Carbon Monoxide Poisoning - D. Penney (CRC, 2008) BBS

CARBON
MONOXIDE
POISONING
8417: “8417_c000” — 2007/9/25 — 20:28 — page i — #1

8417: “8417_c000” — 2007/9/25 — 20:28 — page ii — #2
CARBON
MONOXIDE
POISONING
EDITED BY
DAVID G. PENNEY
Boca Raton London New York
CRC Press is an imprint of the

Taylor & Francis Group, an informa business
8417: “8417_c000” — 2007/9/25 — 20:28 — page iii — #3

8417: “8417_c000” — 2007/9/25 — 20:28 — page iv — #4
Dedication
———————
I wish to dedicate this book first to my mother,
Gertrude Ellen (Goodhew) Penney,
always a source of support and encouragement,
to my grandchildren,
and to all of those victims of carbon monoxide poisoning
who sought but did not find professional help for their suffering.
8417: “8417_c000” — 2007/9/25 — 20:28 — page v — #5

8417: “8417_c000” — 2007/9/25 — 20:28 — page vi — #6
Preface
Carbon Monoxide Poisoning is a new title covering further areas of the expansive
field of carbon monoxide (CO) toxicology that were not covered in the first two books,
Carbon Monoxide and Carbon Monoxide Toxicity, both edited by David G. Penney,
PhD. Both were published by CRC Press, the first in 1996 and the second in 2000.
This book is designed to be complementary to both earlier books, forging into new
areas and following new themes.
The scope of this book is even broader than the earlier two. The first book took
a very scholarly approach, presenting the latest basic and medical science of CO
toxicology in 13 chapters. The contents of that book remain current. The second book
was broader in its approach, and while extending presentations of basic and medical
science, added discussions of human CO exposure under specialized conditions and
in geographic locations other than the United States, in 23 chapters. It also presents
a large body of new data on both acute and chronic CO poisoning. This present,
third book, Carbon Monoxide Poisoning, further extends these presentations both to
new areas such as the law, rehabilitation, personal experience with CO poisoning,
education of the public about CO using the World Wide Web, and so forth, and adds
further new data on chronic CO poisoning.
This book contains some unique features:
1. A critical look at the efficacy of hyperbaric oxygen therapy in decreasing
the damage caused by CO poisoning
2. The use of exciting new scanning techniques in revealing damage from
CO poisoning
3. The introduction of a handheld pulse-oximeter that reads COHb directly
and noninvasively
4. New data showing the persistent health damage that can be caused by
chronic CO poisoning
5. The dangers of CO poisoning possible in motor homes, recreational boats,
and so on
6. The levels of ignorance regarding CO on the part of the general public
Interest in the effects of carbon monoxide on human health has grown rapidly
during the past 20+ years. Governmental agencies, private groups, and the public are
concerned. While an old and familiar poison, CO remains the number one “poison” in
our environment in terms of its “brain-killing” potential, and its potential for overall
immediate and long-term health harm. The public and the medical community need
to obtain quality information about the risks from CO and need the means to identify
and manage victims of CO poisoning successfully. It is hoped that this book, and its
two previous companions, will in some way be of value in meeting these challenges.
8417: “8417_c000” — 2007/9/25 — 20:28 — page vii — #7

8417: “8417_c000” — 2007/9/25 — 20:28 — page viii — #8
Author
David G. Penney, PhD, is a retired professor of physiology, who taught and con-
ducted research on carbon monoxide at the School of Medicine, at Wayne State
University, Detroit, Michigan. He was at one time adjunct professor of occupational
and environmental health in the School of Allied Health Professions at Wayne State
University. He is also a retired director of general surgical research at Providence Hos-
pital in Southfield, Michigan, where for 12 years he directed the scholarly activities
of surgical residents and attending surgeons.
Dr. Penney obtained his BSc degree from Wayne State University in 1963, and
his MSc and PhD degrees from the University of California, Los Angeles, in 1966
and 1969, respectively. Before coming to Wayne State University in 1977, he was
a faculty member at the University of Illinois, Chicago. With his wife, Linda Mae
Penney, the couple have six children.
Dr. Penney’s professional interests have been focused on carbon monoxide for
over 37 years, in both animal models and in humans. His special interests center
around chronic CO poisoning, education of the public about the dangers of CO pois-
oning, the diagnosis and management of CO poisoning victims, and the medicolegal
aspects of CO toxicology.
Dr. Penney has assisted many national and international government and non-
government agencies in matters involving carbon monoxide. He was among the
earliest consultants to the US Environmental Protection Agency (EPA) in setting
CO standards for outside air. He assisted the World Health Organization (WHO) in
the late 1990s in setting similar standards for the world. He has worked with the Aus-
tralian Medical Association (AMA) and with other concerned groups in Australia to
attempt to stem the tide of suicides involving CO. Currently, Dr. Penney assists Under-
writers Laboratory (UL) as a medical expert on CO in establishing standards for CO
alarms and other gas-monitoring equipment, and major gas distributing companies in
educating the public about the dangers of CO poisoning.
Dr. Penney’s published works on CO include over 65 peer-reviewed research
articles, several dozen other articles and abstracts, a number of review articles, book
chapters, and three other books in print. At last count, Dr. Penney had more research
articles and books published on the topic of CO toxicology than anyone else in the
world. He has also published several other books on medical education and on Royal
Oak history, and for some years wrote a column on local history for a hometown
newspaper.
8417: “8417_c000” — 2007/9/25 — 20:28 — page ix — #9

8417: “8417_c000” — 2007/9/25 — 20:28 — page x — #10
Acknowledgments
I wish to thank all the authors, former patients, and all who have contributed to this
book. It has been a long road and at times it seemed impossible. Now it is done.
Thanks to everyone.
I also wish to thank Wayne State University School of Medicine and my Depart-
ment of Physiology chairman, Dr. Joseph Dunbar, for granting me the time off in
2005 to get the book off the ground.
I of course thank CRC – Taylor and Francis Publishers and all their employees
who have been wonderful to work with these past 12 years, in developing my three
books on carbon monoxide.
Finally, I wish to thank my wife Linda for her constant support in developing these
books, hearing my complaints, providing inspiration and also some perspiration in
getting the work done. I of course thank my mother, Gertrude, for her support,
encouragement and even late night help with data entry.
8417: “8417_c000” — 2007/9/25 — 20:28 — page xi — #11

8417: “8417_c000” — 2007/9/25 — 20:28 — page xii — #12
Contributors
Rob Aiers Kosmas Galatsis, Ph.D.

Envirotec (UK) Ltd. Microelectronics Advanced Research
Hampshire, U.K. Corporation
Center on Functional Engineered Nano
Carol L. Armstrong, Ph.D., A.B.P.N. Architectonics
Division of Oncology University of California
The Children’s Hospital of Philadelphia
Los Angeles, California
Department of Neurology
University of Pennsylvania Medical School
James F. Georgis, O.D.
Philadelphia, Pennsylvania
Optometry Clinic
Steve N.M. Collard, B.A., M.E.D. Pueblo, Colorado
Adjunct Professor of Law
Nova Southeastern University James M. Gracey, Ed.D.
Valrico, Florida Colorado Institute for Injury
Rehabilitation, Inc.
David C. Cone, M.D. Denver, Colorado
Division of Emergency Medicine
Yale New Haven Hospital Neil B. Hampson, M.D.
Yale University School of Medicine Center for Hyperbaric Medicine
New Haven, Connecticut Section of Pulmonary and Critical Care
Medicine
Joseph A. Cramer
Virginia Mason Medical Center
Wyoming, Michigan
Seattle, Washington
Jacqueline L. Cunningham, Ph.D.
Department of Psychology Michael F. Hanzlick
The Children’s Hospital of Philadelphia Hanzlick & Associates
Philadelphia, Pennsylvania Highland Ranch, Colorado
Thomas M. Dydek, Ph.D., D.A.B.T., P.E. Alastair W.M. Hay, Ph.D.

Dydek Toxicology Consulting Molecular Epidemiology Unit
Austin, Texas LIGHT Laboratories
School of Medicine
Robert E. Engberg, B.S., P.E.
University of Leeds
Gas Dynamics Corporation
St. Paul, Minnesota Leeds, U.K.
Peter G. Flachsbart, Ph.D., A.I.C.P. Dennis A. Helffenstein, Ph.D.

Department of Urban and Regional Planning Colorado Neuropsychological
University of Hawaii at Manoa Associates
Honolulu, Hawaii Colorado Springs, Colorado
8417: “8417_c000” — 2007/9/25 — 20:28 — page xiii — #13

xiv Contributors
Gunnar Heuser, M.D., Ph.D. Joshua A. Mott, Ph.D.

Neuromed and Neurotox Associates CDC/CCID/NCIRD
Santa Barbara, California Air Pollution and Respiratory Health
Clinical Assistant Professor Branch
University of California, Los Angeles National Center for Environmental
Los Angeles, California Health
Atlanta, Georgia
S. Gregory Hipskind, M.D., Ph.D.
Department of Anthropology David G. Penney, Ph.D.
Western Washington University Wayne State University School of Medicine,
Bellingham, Washington and
Providence Hospital and Medical Centers
Ramona O. Hopkins, Ph.D. (retired)
Psychology Department and Neuroscience St. Augustine, Florida & Beulah, MI
Center
Brigham Young University
Linda M. Penney
Provo, Utah St. Augustine, FL & Beulah, MI
Gary Hutter, Ph.D.

Meridian Engineering & Technology, Inc. Kevin J. Reilly, Jr.
Glenview, Illinois Training Operations & Firefighter
Diversified Security Solutions, Inc.
Richard Karg, B.S., M.S. Saddle Brook, New Jersey
R.J. Karg Associates
Topsham, Maine James W. Rhee, M.D.
Section of Emergency Medicine & Medical
Michael E. King, Ph.D. Toxicology
CDC/CCEHIP/NCEH The University of Chicago
Air Pollution and Respiratory Health Branch Chicago, Illinois
National Center for Environmental Health
Atlanta, Georgia Frank Ricci
New Haven City Fire Department
Jerrold B. Leikin, M.D. New Haven, Connecticut
Rush Medical College
Chicago, Illinois
Carlos D. Scheinkestel, M.D.
and
Monash University
Evanston Northwestern Health Care
Melbourne, Australia
Glenview, Illinois
Jane Brown McCammon, B.S., M.S. Robert E. Schreter, B.S., P.E.

Double Angel Foundation R. Schreter and Associates, Inc.
Broken Circle M Consulting, LLC Roswell, Georgia
Littleton, Colorado
Peter Tikuisis, Ph.D.
Ian L. Millar Human Modeling Group
Monash University Defence Research & Development Canada
Melbourne, Australia Toronto, Ontario, Canada
8417: “8417_c000” — 2007/9/25 — 20:28 — page xiv — #14

Contributors xv
Christian Tomaszewski, M.S., M.D., Stephen P. Willison, B.S., J.D.

F.A.C.E.P., F.A.C.M.T. Willison & Hellman, P.C.
Department of Emergency Medicine Grand Rapids, Michigan
University of Pittsburgh Medical Center
Hannad Medical Corporation Wojtek B. Wlodarski, D.Sc., Ph.D.,
Doha, Qatar M.Sc. E.E.
School of Electrical and Computer Systems
Suzanne R. White, M.D., F.A.C.M.T., Engineering
F.A.C.E.P. Royal Melbourne Institute of Technology
Children’s Hospital of Michigan University
Regional Poison Control Center Melbourne, Victoria, Australia
Department of Emergency Medicine
Wayne State University
School of Medicine
Detroit, Michigan
8417: “8417_c000” — 2007/9/25 — 20:28 — page xv — #15

8417: “8417_c000” — 2007/9/25 — 20:28 — page xvi — #16
Table of Contents
Table of Contents for Carbon Monoxide Toxicity, 2000 . . . . . . . . . . . . . . . . . . . . . . . . . xxi
Table of Contents for Carbon Monoxide, 1996 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxv
Chapter 1
Introduction to and Overview of the Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
David G. Penney
Chapter 2
Exposure to Ambient and Microenvironmental Concentrations of Carbon
Monoxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Peter G. Flachsbart
Chapter 3
Carbon Monoxide Build-Up in Houses and Small Volume Enclosures. . . . . . . . . 43
Robert E. Engberg
Chapter 4
Formation and Movement of Carbon Monoxide into Mobile Homes,
Recreational Vehicles, and Other Enclosures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Robert E. Schreter
Chapter 5
Carbon Monoxide Emissions from Gas Ranges and the Development of a
Field Protocol for Measuring CO Emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Richard Karg
Chapter 6
Investigating Carbon Monoxide-Related Accidents Involving Gas-Burning
Appliances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Michael Hanzlick
Chapter 7
Carbon Monoxide Dangers in the Marine Environment. . . . . . . . . . . . . . . . . . . . . . . . . 157
Jane McCammon
8417: “8417_c000” — 2007/9/25 — 20:28 — page xvii — #17

xviii Contents
Chapter 8
Application of Warnings and Labels for Carbon Monoxide Protection . . . . . . . . . 197
Gary Hutter
Chapter 9
Public Health Surveillance for Carbon Monoxide in the United States:
A Review of National Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Michael E. King and Joshua A. Mott
Chapter 10
Carbon Monoxide Sensors and Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
Kosmas Galatsis and Wojtek Wlodarski
Chapter 11
Marketing of Carbon Monoxide Information and Alarms in Europe and
Beyond: Use of the World Wide Web in Saving Lives . . . . . . . . . . . . . . . . . . . . . . . . . . 271
Rob Aiers
Chapter 12
Investigating Carbon Monoxide Poisonings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
Thomas M. Dydek
Chapter 13
Carbon Monoxide Detectors as Preventive Medicine . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
James W. Rhee and Jerrold B. Leikin
Chapter 14
Misconceptions About Carbon Monoxide. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
David G. Penney
Chapter 15
A Survey Study of Public Perceptions About Carbon Monoxide . . . . . . . . . . . . . . . 325
David G. Penney and Linda M. Penney
Chapter 16
Treatment of Carbon Monoxide Poisoning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
Suzanne R. White
Chapter 17
The Case for the Use of Hyperbaric Oxygen Therapy in Carbon Monoxide
Poisoning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
Christian Tomaszewski
8417: “8417_c000” — 2007/9/25 — 20:28 — page xviii — #18

Contents xix
Chapter 18
Hyperbaric Oxygen for Acute Carbon Monoxide Poisoning: Useful Therapy
or Unfulfilled Promise? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391
Carlos D. Scheinkestel and Ian L. Millar
Chapter 19
A Challenge to the Healthcare Community: The Diagnosis of Carbon
Monoxide Poisoning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
David G. Penney
Chapter 20
Neuroimaging after Carbon Monoxide Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449
Gunnar Heuser
Chapter 21
Recent Advances in Brain SPECT Imaging after Carbon Monoxide Poisoning 457
S. Gregory Hipskind
Chapter 22
Neurocognitive and Affective Sequelae of Carbon Monoxide Poisoning . . . . . . 477
Ramona O. Hopkins
Chapter 23
Neurocognitive and Neurobehavioral Sequelae of Chronic Carbon Monoxide
Poisoning: A Retrospective Study and Case Presentation . . . . . . . . . . . . . . . . . . . . . . . 495
Dennis A. Helffenstein
Chapter 24
Chronic Carbon Monoxide Poisoning: A Case Series . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
David G. Penney
Chapter 25
Functional and Developmental Effects of Carbon Monoxide Toxicity in
Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569
Carol L. Armstrong and Jacqueline Cunningham
Chapter 26
Issues in Rehabilitation and Life Care Planning for Patients with Carbon
Monoxide Poisoning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
James M. Gracey
Chapter 27
Treatment of Carbon Monoxide Poisoning with Yoked
Prism Lenses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 619
James F. Georgis
8417: “8417_c000” — 2007/9/25 — 20:28 — page xix — #19

xx Contents
Chapter 28
Firefighters and Carbon Monoxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 643
Kevin J. Reilly, Jr., Frank Ricci, and David Cone
Chapter 29
The Purpose and the Process of Litigation in a Carbon Monoxide Poisoning
Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655
Stephen P. Willison
Chapter 30
Offering Expert Opinions in a Carbon Monoxide Case . . . . . . . . . . . . . . . . . . . . . . . . . 671
Stephen P. Willison
Chapter 31
Injury Caused by Carbon Monoxide Poisoning: Defining Monetary Damages 683
Steve Collard
Chapter 32
My Carbon Monoxide Poisoning: A Victim’s Story . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725
Joseph A. Cramer
Chapter 33
Noninvasive Measurement of Blood Carboxyhemoglobin with Pulse
CO-Oximetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 739
Neil B. Hampson
Chapter 34
Chronic Carbon Monoxide Exposure: How Much Do We Know About
it?—an Update . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745
Alastair W.M. Hay
Chapter 35
Essential Reference Tables, Graphs, and Other Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . 753
David G. Penney
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 765
8417: “8417_c000” — 2007/9/25 — 20:28 — page xx — #20

Table of Contents for
Carbon Monoxide Toxicity 2000
Chapter 1
History of Carbon Monoxide Toxicology
Dieter Pankow
Chapter 2
Carbon Monoxide in Breath, Blood, and Other Tissues
Hendrik J. Vreman, Ronald J. Wong, and David K Stevenson
Chapter 3
Carbon Monoxide Detectors
Richard Kwor
Chapter 4
The Setting of Health-Based Standards for Ambient Carbon Monoxide and
Their Impact on Atmospheric Levels
James A. Raub
Chapter 5
Effect of Carbon Monoxide on Work and Exercise Capacity in Humans
Milan J. Hazucha
Chapter 6
The Interacting Effects of Altitude and Carbon Monoxide
James J. McGrath
Chapter 7
Interactions Among Carbon Monoxide, Hydrogen Cyanide, Low Oxygen
Hypoxia, Carbon Dioxide, and Inhaled Irritant Gases
David A. Purser
Chapter 8
Carbon Monoxide Poisoning and Its Management in the
United States
Neil B. Hampson
8417: “8417_c000” — 2007/9/25 — 20:28 — page xxi — #21

xxii Contents
Chapter 9
Death by Suicide Involving Carbon Monoxide around the World
Pierre Baume and Michaela Skopek
Chapter 10
Carbon Monoxide as an Unrecognized Cause of Neurasthenia: A History
Albert Donnay
Chapter 11
Update on the Clinical Treatment of Carbon Monoxide Poisoning
Suzanne R. White
Chapter 12
Treatment of Carbon Monoxide Poisoning in France
Monique Mathieu-Nolf and Daniel Mathieu
Chapter 13
Acute Carbon Monoxide Poisonings in Poland - Research and
Clinical Experience
Jerzy A. Sokal and Janusz Pach
Chapter 14
Treatment of Carbon Monoxide Poisoning in the United Kingdom
Martin R. Hamilton-Farrell and John Henry
Chapter 15
Carbon Monoxide Air Pollution and Its Health Impact on the Major Cities
of China
Qing Chen and Lihua Wang
Chapter 16
Use of Scanning Techniques in the Diagnosis of Damage from
Carbon Monoxide
I.S. Saing Choi
Chapter 17
Low-Level Carbon Monoxide and Human Health
Robert D. Morris
Chapter 18
Chronic Carbon Monoxide Poisoning
David G. Penney
Chapter 19
Chronic Carbon Monoxide Exposure: The CO Support Study
Alistair WM. Hay, Susan Jaffer, and Debbie Davis
8417: “8417_c000” — 2007/9/25 — 20:28 — page xxii — #22

Contents xxiii
Chapter 20
Neuropsychological Evaluation of the Carbon Monoxide-Poisoned Patient
Chapter 21
Pediatric Carbon Monoxide Poisoning
Suzanne R. White
Chapter 22
Carbon Monoxide Production, Transport, and Hazard in Building Fires
Frederick W. Mowrer and Vincent Brannigan
Chapter 23
Approaches to Dealing with Carbon Monoxide in the
Living Environment
Thomas H. Greiner and Charles V. Schwab
8417: “8417_c000” — 2007/9/25 — 20:28 — page xxiii — #23

8417: “8417_c000” — 2007/9/25 — 20:28 — page xxiv — #24
Table of Contents for
Carbon Monoxide 1996
Chapter 1
Carbon Monoxide Analysis
Roger L. Wabeke
Chapter 2
Carbon Monoxide Formation Due to Metabolism of Xenobiotics
Dieter Pankow
Chapter 3
Modeling the Uptake and Elimination of Carbon Monoxide
Peter Tikuisis
Chapter 4
Cerebrovascular Effects of Carbon Monoxide
Mark A. Helfaer and Richard J. Traystman
Chapter 5
Pulmonary Changes Induced by the Administration of Carbon Monoxide and
Other Compounds in Smoke
Daniel L. Traber and Darien W Bradford
Chapter 6
Effects of Carbon Monoxide Exposure on Developing Animals and Humans
David G. Penney
Chapter 7
Carbon Monoxide - From Tool to Neurotransmitter
Nanduri R. Prabhakar and Robert S. Fitzgerald
Chapter 8
Toxicity of Carbon Monoxide: Hemoglobin vs. Histotoxic Mechanisms
Claude A. Piantadosi
8417: “8417_c000” — 2007/9/25 — 20:28 — page xxv — #25

xxvi Contents
Chapter 9
Carbon Monoxide-Induced Impairment of Learning, Memory, and
Neuronal Dysfunction
Masayuki Hiramatsu, Tsutomu Kameyama, and Toshitaka Nabeshima
Chapter 10
Behavioral Effects of Carbon Monoxide Exposure: Results
and Mechanisms
Vernon A. Benignus
Chapter 11
Delayed Sequelae in Carbon Monoxide Poisoning and the
Possible Mechanisms
Eric Kindwall
Chapter 12
Treatment of Carbon Monoxide Poisoning
Suzanne R. White
Chapter 13
Options for Treatment of Carbon Monoxide Poisoning, Including Hyperbaric
Oxygen Therapy
Stephen R. Thom
8417: “8417_c000” — 2007/9/25 — 20:28 — page xxvi — #26

1 Introduction to and
Overview of the Field
David G. Penney
CONTENTS
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
I have designed this book to complete the series on carbon monoxide (CO) begun with
Carbon Monoxide, 19961 and continued with Carbon Monoxide Toxicity, 2000.2 This
and the second book are NOT new editions of the first book, as has often been assumed.
While CO may seem a very narrow subject area, it finds its way into many diverse
disciplines and its literature is vast. This third book, Carbon Monoxide Poisoning,
completes the trilogy and should become a standard reference source on CO for years
to come.
The new book covers areas not previously presented, including rehabilitation, edu-
cation of the public using the WWW, litigation involving CO poisoning, economic
loss assessment, and firefighting. There are areas of update, such as the chapter by
Dr. Suzanne White, on diagnosis and management. There are two chapters in which
the authors take opposing views, one stating the case for use of hyperbaric oxygen
therapy (HBOT) by Dr. Christian Tomaszewski and against the use of HBOT by
Dr. Carlos Scheinkestel. One chapter deals with toxicology investigation
(i.e., forensic) procedures. A series of chapters detail the risk of CO poisoning from
kitchen ranges, recreational trailers and motor homes, and recreational powerboats.
Three chapters cover the very important area of neuropsychological evaluation of
adults and children following CO poisoning. The chapter by Dr. Dennis Helffenstein
presents new data on a case series of patients that had sustained chronic CO poison-
ing. I have written a companion chapter to his in which a retrospective review of 61
chronically CO-poisoned patients were symptomatically evaluated (Chapter 24).
Better Education of Physicians: This is essential if CO-poisoned patients are

to be properly diagnosed and treated in the future. Several years ago I over-
heard a prominent emergency room physician say, “the standard of care for
CO poisoning in the U.S. is less than the standard of care.” I estimate that in
80% of CO cases reviewed, one, two, or more mistakes were made in dia-
gnosing and/or treating CO-poisoned patients. This may involve misdiagnosis,
dependence on faulty pulse-oximetry data, administering NBO with the wrong
8417: “8417_c001” — 2007/9/11 — 12:17 — page 1 — #1

2 Carbon Monoxide Poisoning
equipment (e.g., nasal prongs), failure to order HBOT when it was needed and
possible, discharge of patients while still symptomatic, and so forth. Physi-
cians should also be informed of the possible serious permanent health harm
that chronic or lower-level acute CO poisoning can cause if not diagnosed
immediately and treated fully. From my perspective, the CO cases that result
from acute poisoning and those most likely to reach the media are actually
the smaller fraction of all CO poisonings, while the chronic (i.e., occult) CO
poisonings make up by far the largest fraction, and probably result in the most
injuries, but they are the very group that physicians are least trained to properly
deal with.
Better Education of the Public: The public too needs education about the
dangers presented by CO exposure. My chapter presenting the results of surveys
of public perceptions of CO in Michigan and Florida shows this. While almost
everyone knows that CO is a deadly poison, substantial fractions (sometimes
most) of the adult and juvenile population cannot intelligently evaluate the risk
of CO from automobiles, propane radiant heaters, generators, and recreational
powerboats. In some situations people are overly cautious in a given situation,
but in other situations people vastly underestimate the risk of injury and death.
Youth, as opposed to adults, are particularly uninformed.
New Approaches to Treating Acute, Severe CO Poisoning: There appears to be

very little new in treating acute severe CO poisoning. We cannot decide for
sure whether HBOT is more effective in reducing neurologic sequelae, even
though it has been used for approximately 50 years. The pros and cons chapters
on HBOT provide detailed discussions of many aspects of the situation, and
a few ideas about possible new approaches. The bright light in this area is
almost certainly the new generation of pulse-ox devices that read COHb directly
and noninvasively. See Dr. Neil Hampson’s chapter about the testing of this
device.
Requiring Proper Warnings on Combustion Equipment: The lack of proper and

adequate warnings on equipment that do, or under foreseeable conditions might,
emit harmful or lethal amounts of CO remains a real problem. People continue
to die because warnings on combustion devices are not obvious, explicit and
direct, and not on the device itself. Warnings in operating manuals should be
continued and improved, but they alone are insufficient because manuals are
usually separated from the device. Warnings on the device must tell the user
what might occur in using the device a certain way. Statements such as “provide
adequate ventilation” are useless. The warning must specify where and when
not to use the device, and for how long. Warnings in the United States should be
written in both English and Spanish, along with standard prohibition symbols.
Some of the devices this applies to include portable generators, cement saws,
lawn mowers, pressure washers, scissor-lifts, kerosene heaters, propane-radiant
heaters, lamps and cook stoves, powerboats, charcoal grills and hibachis, and
so forth. See Dr. Hutter’s excellent chapter on warnings.
8417: “8417_c001” — 2007/9/11 — 12:17 — page 2 — #2

Introduction to and Overview of the Field 3
Rethinking Work Guidelines for Carbon Monoxide that Reflect the Science:
Another area that needs immediate attention is threshold limit standards for
inhalation of CO. Environmental Protection Agency (EPA) and World Health
Organization (WHO) after extensive study and deliberation some years ago set
the 8-h standard at 9 ppm, or 10 mg/m3 for outside air. On the other hand,
National Institute for Occupational Safety and Health (NIOSH) and Occupa-
tional Safety and Health Administration (OSHA) have set different standards
for the work environment. We know that people with coronary artery dis-
ease, congestive heart failure, asthma, and a state of fetal development are
often members of the workforce, and represent a more sensitive, higher risk
subgroup of the general population. It is also well known that people, even
those with no obvious risk factors, vary widely in their tolerance of CO.
Why then should the standards be so different—9 ppm (EPA, WHO) versus
50 ppm (OSHA) for the same species? This is a 5-1/2 fold difference! I believe
it is time we in the toxicology community re-examine ambient air CO concen-
tration work standards, and make decisions for new standards based only on
the best science.
Realization that Brain Damage Resulting from CO Poisoning Is not Dependent

on COHb and/or Severity of Poisoning: New studies make it clear what many
of us have believed for some time based on experience, that brain damage from
CO poisoning is only very poorly correlated with the severity of the poisoning
by whatever criteria, even the COHb saturation. Some people with very severe
poisoning and/or high initial COHb values make remarkable recoveries, while
some with what appears to be minimal poisoning, and even on occasion, near
normal COHb when measured, incur substantial damage. Clearly, loss of con-
sciousness is not required for the development of neurologic sequelae, although
I sometimes hear the less well-informed say that it is. I believe everyone agrees
that a longer (i.e., soaking) exposure is more detrimental than a short one, but
strangely some still insist that chronic, lower-level CO exposure can cause no
permanent harm. Other markers for brain damage have been proposed such
as acidosis, gait/balance/clumsiness on presentation, and release of cellular
enzymes, but it remains unclear how useful they are.
Carbon Monoxide Disaster Management: Release of CO during certain kinds

of disasters could pose significant problems. Fire almost invariably gives off CO
as an incomplete combustion product, and fire is a larger or a smaller compon-
ent in most disasters, especially those that might be instigated by terrorists.
It is unclear to me whether any overall planning has been done by agen-
cies of the government with respect to CO. Firefighters regularly encounter
CO in the work they do, and have equipment such as self-contained breath-
ing apparatus to deal with it. See the chapter on firefighting by Mr. Reilly,
Mr. Ricci, and Dr. Cone. Other concerns for human health that may not yet be
fully addressed include indoor car, monster truck, and motocross events, work
in coal mines, warehouses, enclosed construction sites, and so forth where
significant CO is generated by combustion devices. Since it is dangerous and
8417: “8417_c001” — 2007/9/11 — 12:17 — page 3 — #3

foolhardy to run a portable generator inside a garage or house, why is it not

dangerous to human health to run several monster trucks with huge gasoline-
fueled engines lacking catalytic converters inside a covered sports stadium? The
recent Sago Mine disaster where a dozen men died slowly, mainly from CO
poisoning, points out the need for adequate emergency equipment that would
allow men to live for extended periods of time in the presence of lethal CO
air concentrations. Autostarters—are they safe? How often will car engines be
started inadvertently (by children, otherwise by mistake) in a closed garage,
leading to injury or death from CO poisoning?
The wonder is that CO has been with man since prehistory, probably since we
first began using fire. Other scourges such as plague, cholera, typhus, smallpox, and
so forth are gone, at least from the developed world, whereas this simple, small
molecule, CO, continues to afflict us, and probably will, at least as long as we are
wedded to the “carbon energy cycle.”
I hope you enjoy this book and will use it with its earlier brothers, Carbon
Monoxide (1996)1 and Carbon Monoxide Toxicity (2000).2
References
1. Penney, D.G., ed. Carbon Monoxide, CRC Press, NY, 1996, 296 pp.
2. Penney, D.G., ed. Carbon Monoxide Toxicity, CRC Press, NY, 2000, 560 pp.
8417: “8417_c001” — 2007/9/11 — 12:17 — page 4 — #4

2 Exposure to Ambient
and Microenvironmental
Concentrations of
Carbon Monoxide
Peter G. Flachsbart
CONTENTS
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2 Standards and Guidelines for Exposure to Ambient Concentrations of
Carbon Monoxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.3 Trends in Carbon Monoxide Emissions and Ambient Air Quality . . . . . . . 8
2.4 Human Exposure to Carbon Monoxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.5 Microenvironmental Exposures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.5.1 Residential Exposures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.5.1.1 Nonfatal Exposures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.5.1.2 Fatal Exposures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.5.2 Occupational Exposures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.5.3 Shopping Center Exposures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.5.4 Recreational Exposures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.5.4.1 Exposures on Recreational Vehicles . . . . . . . . . . . . . . . . . . . . 23
2.5.4.2 Exposures at Indoor Sporting Events . . . . . . . . . . . . . . . . . . . 24
2.5.5 Commuter Exposures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.5.5.1 Defective Exhaust Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.5.5.2 Parking Garages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.5.5.3 Service Stations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.5.5.4 Drive-Up Facilities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.5.5.5 Airbag Deployment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.5.5.6 Motor Vehicle Emission Standards. . . . . . . . . . . . . . . . . . . . . . 30
2.6 Exposure to Methylene Chloride . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.6.1 Nonoccupational Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.6.2 Occupational Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.8 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
8417: “8417_c002” — 2007/9/11 — 12:16 — page 5 — #1

2.1 INTRODUCTION
Carbon monoxide (CO) is a gas commonly produced by incomplete combustion of
fuels containing carbon atoms. Many people use these fuels (i.e., coal, gasoline,
kerosene, natural gas, oil, propane, and wood) around the globe. As a result, CO is
ubiquitous in the atmosphere. However, without sophisticated instruments, a person
is unable to detect CO, because the gas is not irritating and has no color, odor, or
taste. Moreover, the gas is a potential health hazard, because exposure to CO can
starve critical body organs, especially the brain and heart, of oxygen. Once inside the
lungs, CO molecules pass easily into the bloodstream and compete with oxygen for
hemoglobin (Hb) in the red blood cells. About 95% of the absorbed CO readily binds
with Hb to form carboxyhemoglobin (COHb), because the affinity of Hb for CO is
over 200 times stronger than it is for oxygen. Thus, the percentage of total Hb in the
blood that is in the form of COHb is a biomarker of CO exposure.1
The health effects of CO, which are a function of its concentration and the
duration of exposure, range from subtle to severe. They include neurobehavioral,
cardiovascular, and developmental effects, observed at low levels of CO exposure,
to unconsciousness and death, which occur after acute exposure to high CO concen-
trations. Lethal CO exposures are usually linked to CO concentrations greater than
1000 parts per million (ppm) by volume. Coma, convulsions, cardiopulmonary arrest,
and death have been observed when COHb levels reach 50%, although death from
CO poisoning is frequently reported at far lower COHb saturations. The exact COHb
concentrations that trigger acute and chronic health effects in different people differ
widely. Sublethal levels of CO may cause neurological-type symptoms, including
fatigue, headache, nausea, vomiting, deficit in short-term memory, to name a few.
Exposures to these CO concentrations are often misdiagnosed as viral illness, clinical
depression, and so forth. Still lower CO exposures (i.e., those producing less than 10%
COHb) may not be associated under certain circumstances with overt symptoms.2,3
This broad range of effects makes CO relevant to people concerned with ambient
air quality management as well as officials responsible for protecting public health
and safety. Ambient air quality standards are the foundation of air quality management
programs in many countries worldwide. Such standards typically specify maximum
permissible concentrations in ambient air for certain pollutants. To achieve ambi-
ent standards in the United States, the U.S. Environmental Protection Agency (EPA)
implemented progressively tighter tailpipe emission standards for motor vehicles.
As a result, these standards have substantially reduced ambient CO concentrations
in most metropolitan areas of the U.S. and have had other collateral benefits. For
example, the nation had 11,667 fewer deaths from accidental CO poisoning between
1968 and 1998, according to a study by the Centers for Disease Control and Pre-
vention (CDCP).4 Still, an average of 480 U.S. residents died each year during
2001–2002 from nonfire-related unintentional CO poisoning. In addition, an estim-
ated 15,200 persons, that is, people with confirmed or possible nonfire-related CO
exposure or poisoning, were treated annually in U.S. hospital emergency rooms.5 In
fact, more than 50% of all fatal poisonings reported in many countries may be attrib-
utable to CO, because these cases are under-reported or misdiagnosed by medical
professionals.2
8417: “8417_c002” — 2007/9/11 — 12:16 — page 6 — #2

Exposure to Ambient and Microenvironmental Concentrations 7
This chapter explores reasons behind the paradox of declining ambient CO con-
centrations in urban areas of the United States coupled with persistent injuries and
fatalities from CO poisoning. The chapter is organized around the superposition prin-
ciple of CO exposure, which may help to explain this paradox. This principle holds
that CO concentrations at any given point in time and space consist of both ambient
and microenvironmental components. The next section takes a closer look at ambient
CO concentrations in urban areas. The chapter then describes how the development of
portable monitors enabled measurements of personal exposure to CO concentrations
in places where people perform routine daily activities. Since these activities often
occur in specific microenvironments, the chapter then describes typical CO expos-
ures where people live, work, shop, play and commute, and factors that affect these
exposures. The last section offers some concluding thoughts.
2.2 STANDARDS AND GUIDELINES FOR EXPOSURE

TO AMBIENT CONCENTRATIONS OF CARBON
MONOXIDE
The Clean Air Act (CAA) of 1963 was amended by the U.S. Congress in 1970,
1977, and 1990. The 1977 and 1990 versions largely reaffirmed the course set by
the 1970 amendments.6 Under the 1970 CAA amendments, the U.S. EPA established
the National Ambient Air Quality Standards (NAAQS) and set deadlines for their
attainment. The current NAAQS reflect EPA’s scientific judgments about maximum
allowable ambient concentrations and averaging times for certain “criteria” air pol-
lutants including CO. Criteria air pollutants are those that could reasonably endanger
public health or welfare. The 1970 and 1990 CAA amendments also mandated strin-
gent motor vehicle emission standards as a means to achieve the NAAQS for CO and
other air pollutants that have been linked to mobile sources.
Air pollutant concentrations can be expressed either as ppm or as milligrams per
cubic meter (mg/m3 ) of air. Many of the studies reviewed in this chapter refer to the
NAAQS for guidance on allowable limits of CO exposure. The EPA promulgated
identical primary and secondary NAAQS for CO on April 30, 1971. The primary
standards specify a level of air quality sufficient to protect public health and the
secondary standards are intended to protect public welfare. The standards include “an
adequate margin of safety” to reflect scientific uncertainties related to measurement
of the effects of air pollutant exposure in the population. In 1985, EPA rescinded the
secondary standard for CO, but retained two primary standards: 9 ppm (10 mg/m3 )
as an 8-h average and 35 ppm (40 mg/m3 ) as a 1-h average. Each standard may be
exceeded once per year in an air quality control region (AQCR) without violating the
standard.7
The NAAQS for CO are designed to keep COHb levels below 2% in the blood
of 99.9% of nonsmoking healthy adults and people who belong to probable high-risk
groups. Smokers are excluded because they may exhale more CO into the air than
they are inhaling from the ambient environment. The high-risk groups include the
elderly; pregnant women; fetuses; young infants; and those suffering from anemia
or certain other blood, cardiovascular, or respiratory diseases. People at greatest risk
8417: “8417_c002” — 2007/9/11 — 12:16 — page 7 — #3

from exposures to ambient CO levels are those with coronary artery disease. Some
of these people suffer myocardial ischemia as identified by ST-segment depression,
during exercise when their COHb levels ≥ 2.4%.7 The symptoms of this disease are
spasmodic attacks of chest pain (angina pectoris) caused by insufficient oxygen in
the heart muscles. Controlled laboratory studies are needed to observe these health
effects, because the COHb levels are at or near the lower margin of detection of current
instruments.8 Although annual death rates from heart disease have been declining
since 1980, heart disease is still America’s leading cause of death.9 Coronary artery
disease reduces a person’s circulatory capacity, which is particularly critical during
exercise when muscles need more oxygen. Given the widespread prevalence and lack
of awareness of coronary heart disease, Godish3 argues that a significant number of
people still may be at risk from CO exposure, even if ambient CO concentrations do
not exceed the 8-h NAAQS for CO.
The World Health Organization (WHO) guidelines for CO (see below) are also
relevant to this discussion. Relative to the NAAQS for CO in the U.S., these guidelines
have an identical 8-h concentration but a lower 1-h concentration. Unlike the NAAQS,
the guidelines specify maximum concentrations for two shorter time spans (30 min
and 15 min).10,11
Maximum Concentrations Averaging Times
9 ppm (10 mg/m3 ) 8h

25 ppm (30 mg/m3 ) 1h
50 ppm (60 mg/m3 ) 30 min
90 ppm (100 mg/m3 ) 15 min
WHO’s guidelines are intended to prevent blood levels of COHb from exceeding
2.5–3% in nonsmoking populations even when a person engages in relatively heavy
work. Romieu12 reported that average COHb levels are about 1.2–1.5% in the general
population and from 3% to 4% in the blood of cigarette smokers.
2.3 TRENDS IN CARBON MONOXIDE EMISSIONS

AND AMBIENT AIR QUALITY
The CAA amendments have substantially reduced nationwide CO emissions, even
as other socio-economic indicators of growth have increased. For example, between
1970 and 2002, nationwide emissions of CO fell 48%, despite national increases
of 38% in population, 155% in vehicle miles of travel (VMT), and 164% in gross
domestic product.13 The rapid growth of VMT has been attributed to the decent-
ralization of jobs and housing within urban regions during the post World War II
era.14
The CAA amendments have also reduced ambient CO concentrations in urban
areas of the United States. The U.S. EPA determines compliance with the NAAQS
based on measurements of ambient air quality made by a nationwide network of
fixed-site monitoring (FSM) stations. Ambient concentrations of air pollutants are
8417: “8417_c002” — 2007/9/11 — 12:16 — page 8 — #4

typically measured in air “external to buildings, to which the general public has
access.”15 In urban areas, most ambient stations that measure CO concentrations are
located near roadways.16 These stations use nondispersive infrared reference (NDIR)
spectrometry to measure ambient CO concentrations. Monitoring instruments based
on the NDIR method are large, complex and expensive, and require a vibration-free,
air-conditioned facility for the production of accurate and reliable data. The nation-
wide network consists of state and local air monitoring stations (SLAMS), which
send data to EPA’s Aerometric Information Retrieval System (now Air Quality Sys-
tem) within six months of acquisition.17 Several stations within the SLAMS network
belong to a network of national air monitoring stations (NAMS) to enable national
assessments of air quality.
FSM stations typically reveal two peaks in ambient CO concentrations. These
peaks usually coincide with periods of congested rush-hour traffic.7 For that reason,
some exposure analysts consider CO to be a signature air pollutant for mobile sources.
On a nationwide basis, the EPA’s annual emissions inventory revealed that highway
vehicles accounted for 62.8% of the 93.7 million tons of CO emitted from all sources
except fires in 2003.13 Regional inventories show that motor vehicles account for even
higher percentages of all CO emissions released into the ambient air. For example, CO
emissions from motor vehicles ranged from 78% of all CO emissions in Fairbanks,
Alaska, to 96% of all CO emissions in Phoenix, Arizona. EPA classified both cities
as having “serious” levels of ambient CO concentrations in 1999.18
An ambient station is considered to be in violation (i.e., nonattainment) of the
NAAQS for CO, if it records a nonoverlapping average concentration that exceeds
either the 1-h or 8-h standard more than once per calendar year. The historical record
shows that cities have had more difficulty satisfying the 8-h standard than the 1-h
standard. For the 8-h standard, the nonoverlapping average omits other high values
that occur within 8 h of the first value. Also, values of 9.5 ppm, or greater, are counted
as exceeding the 8-h standard due to the standard’s rounding convention. Maximum
8-h average CO concentrations typically exceeded 30 ppm when continuous monitors
were first installed in some U.S. cities in the early 1960s. When EPA promulgated the
NAAQS for CO in 1971, 91.4% of 58 ambient monitors recorded violations of the
8-h standard and 12.1% of 58 stations recorded violations of the 1-h standard.16 In
1996, EPA’s Office of Air Quality Planning and Standards (OAQPS) reported that CO
levels exceeded the NAAQS in seven counties, which had a combined population of
more than 12.7 million people.19
The CAA amendments require states to develop plans to achieve and maintain
ambient air quality that satisfies the NAAQS. To prepare these plans, states inventory
emissions in each AQCR for a baseline year and determine the necessary emission
reductions to achieve the NAAQS. Pursuant to the 1977 CAA amendments, many
states established inspection and maintenance (I/M) programs as required by their
plans for those regions that were in nonattainment of the NAAQS. An AQCR must
satisfy the CO NAAQS for two consecutive years to be considered in attainment by
EPA. States must submit plans to EPA showing how the region will maintain that
attainment for at least 10 years. As of March, 2006, OAQPS reported that there were
38 CO “maintenance areas” in the United States encompassing 89 counties with a
combined population of 46.8 million people.20
8417: “8417_c002” — 2007/9/11 — 12:16 — page 9 — #5

Since 1995, 11 cities have reported violations of the 8-h standard and no monitor
has reported a violation of the 1-h standard.16 Violations appear to persist in areas
with meteorological and/or topographical handicaps. Meteorological handicaps make
it particularly difficult for cities to satisfy the standard, because of the stochastic nature
of ambient air pollutant concentrations. For example, violations of the NAAQS for
CO in Fairbanks, Alaska, have been attributed to stagnant air masses during winter
months. The atmosphere is more stable in Fairbanks during winter, because less
sunlight causes ground-level temperature inversions to occur more frequently. Also,
less air pollutant dispersion occurs in winter, because winds are milder and mountains
surrounding the city hinder horizontal dispersion.16 As of March 2006, the EPA had
classified five American cities (El Paso, Texas; Las Vegas and Reno, Nevada; the Los
Angeles South Coast Air Basin, California; and Missoula, Montana) as urban areas
that were in nonattainment of the NAAQS for CO. They represented eight counties
with a total population of about 15.4 million people.21 Compared to 1996, there was
one more county in nonattainment of the NAAQS for CO by 2006, and the total
population living in such areas had increased by 21.3%.
2.4 HUMAN EXPOSURE TO CARBON MONOXIDE

The study of population exposure is multidisciplinary and the definition of personal
exposure has evolved over time. A recent definition states that exposure is the contact
between an agent and a target at a specified contact boundary, defined as a surface in
space containing at least one exposure point, that is, a point at which contact occurs.
According to this definition, an inhaled CO molecule (the agent) reaches a human
(the target) at the lining of the lung (the contact boundary), where CO exchange takes
place between air and blood.22 Actual studies of CO exposure use small-scale portable
monitors to measure CO concentrations within a few feet of a person’s nasal and oral
cavities. These studies assume that the air surrounding the person is well mixed and
that measured CO concentrations in that air represent the person’s actual exposure
from CO inhalation.
Besides inhalation exposure to CO, metabolic degradation of many drugs, solvents
(e.g., methylene chloride), and other compounds of CO can elevate levels of COHb in
a person’s blood. Because the endogenous production of CO from drugs and solvents
may continue for several hours, it can prolong any cardiovascular stress from COHb.
Moreover, the maximum COHb level from endogenous CO production can last up to
twice as long as COHb levels caused by comparable exposures to exogenous CO.23,24
Hence, this chapter also discusses the literature on exposure to methylene chloride.
Figure 2.1 illustrates an individual’s air pollutant exposure over time. In this
figure, the function Ci (t) describes the CO concentration to which an individual i is
exposed at any point in time t. Ott defined this event as the instantaneous exposure
of an individual.25 The shaded area under the curve represents the accumulation
of instantaneous exposures over some period of time (t1 − t0 ). This area also is
equal to the integral of the air pollutant concentration function, Ci (t), between t0 and
t1 . Ott defined the quantity represented by this area as the integrated exposure. An
exposure analyst can derive the average concentration to which a person is exposed
8417: “8417_c002” — 2007/9/11 — 12:16 — page 10 — #6

Ci (t )
Concentration
t0 t1
Time
FIGURE 2.1 Exposure of person i to air pollutant concentration (C) as a function of time t.
by dividing the person’s integrated exposure by the period of integration (t1 − t0 ).

This average concentration is sometimes referred to as the average exposure. To
compare the average concentration with an established air quality standard, the period
of integration should equal the averaging period of the standard. This concept of
exposure thus combines two parameters, the air pollutant concentration and the time
duration of exposure to the air pollutant. An exposure analyst assumes that these two
parameters are directly proportional to the dosage of CO in the body, as represented
by the level of COHb in the blood stream, and ultimately to health outcomes.
Prior to the use of portable monitoring devices, exposure analysts relied on ambi-
ent data from FSM stations in urban areas to estimate population exposure. Such data
were used to provide crude estimates of population exposure to CO concentrations
that violated the NAAQS. For example, the President’s Council on Environmental
Quality (CEQ) did a crude estimate of population exposure to CO in the late 1970s.26
This method is based on data collected for each county in the United States. The
estimate is derived as follows:

n
TPE = (pi )(di ) (2.1)
i=1
where
TPE = the total population exposure in the United States (person-days)
pi = the resident population of county i (persons)
di = the number of days in a calendar year that violations of the NAAQS for
CO are observed in county i (days)
n = the total number of counties in the United States in a given year
8417: “8417_c002” — 2007/9/11 — 12:16 — page 11 — #7

For 1978, the CEQ estimated that the nation’s total population exposure above
the 1-h NAAQS for CO of 35 ppm was 2.80 billion person-days. This represented
about 3.7% of the total possible exposures of 75.555 billion person-days, which was
derived by multiplying an estimated 207 million U.S. residents in 1978 times 365
days in a year.26
CEQ’s method was a crude estimate of total population exposure, because the
estimate rested on three major assumptions:
1. The CO concentrations measured by the county’s fixed-site monitors were

representative of concentrations to which residents were actually exposed.
If a county had more than one fixed-site monitor, then the monitor with
the worst CO concentration represented the exposure of all residents in the
county.
2. Residents did not leave the county on days that violations of the NAAQS
for CO occurred for that county.
3. There were no violations of the NAAQS for CO in counties that were not
monitored (e.g., rural counties).
Several scientific studies have questioned the validity of the first two assumptions
as discussed below.
The first assumption implies that ambient CO concentrations are spatially homo-
geneous throughout a county. A study in the early 1970s questioned the ability of
FSM stations to accurately represent human exposure to ambient CO concentrations.
Using large Tedlar™ bags filled by a constant flow pump over 5-min periods, Ott
collected 1128 CO concentrations at “breathing height” at outdoor locations in San
Jose, California, on weekdays between October 1970 and March 1971.27 Of 438
samples collected on 21 dates while walking along sidewalks of congested down-
town streets, 60% were above values measured concurrently at the nearest FSM
station. The correlation between the “walking samples” and FSM values was posit-
ive, but low (r = 0.20). On 2 of 7 days, the sidewalk concentrations (13 and 14.2
ppm) averaged over an 8-h period were well above the corresponding concentrations
(4.4 and 6.2 ppm, respectively) reported for the FSM station. Overall, the 8-h average
CO concentrations for the 7 days ranged between 1.4 and 3 times the values observed
simultaneously at the FSM station. The highest values were in late December when
streets were heavily congested with traffic due to Christmas shopping.28
The San Jose study confirmed the hypothesis that FSM stations did not represent
the CO exposures of a person walking in outdoor settings of a major city. Since the
San Jose study measured the CO exposure of only one person, other studies tested the
hypothesis for larger groups of people. Some of these studies focused on commuters,
because CO is a signature air pollutant of motor vehicle tailpipe emissions. For
example, Cortese and Spengler29 recruited 66 nonsmoking volunteers who lived in
different parts of the metropolitan area of Boston, Massachusetts. Each volunteer
carried an Ecolyzer monitor attached to a Simpson recorder for 3–5 days between
October 1974 and February 1975. The study reported that the mean of all commuter
exposures (11.9 ppm) was about twice the concurrent concentration measured at six
FSM stations (6 ppm). Automobile commuters had exposures nearly twice that of
8417: “8417_c002” — 2007/9/11 — 12:16 — page 12 — #8

transit users, and about 1.6 times that of people who did “split-mode” commuting,
which involved both auto and transit.
The first assumption of the CEQ method of exposure estimation also raises
the following question: To what extent do ambient CO levels reflect the CO con-
centrations to which people are exposed indoors? Early studies of human activity
patterns in America30 and other industrialized countries31 have consistently shown
that people spend most of their time indoors. Several studies published in the early
1970s addressed this question.32−34 Each study examined the relationship between
indoor and outdoor CO concentrations. In the absence of indoor sources, these stud-
ies found that indoor CO concentrations of office buildings tended to follow outdoor
CO levels with some degree of time lag and with a tendency not to reach either the
extreme high or low values that were found outdoors. The General Electric study also
reported that CO concentrations were larger indoors than outdoors at heights greater
than 100 ft above the roadway, due to entrapment of CO within the building.33
To simplify the exposure calculation, the CEQ assumed that each person spent
24 h at home, because the method relied on household data provided by the U.S.
Census Bureau. The CEQ method also assumed that people did not travel outside the
area represented by the FSM station. People who live in cities spend a significant
amount of personal time in pursuits away from the home. In a study of metropolitan
Washington, DC, residents in 1968, Chapin found that the hours spent away from
home, on the average, ranged from 6.33 h on Sunday to 10.64 h on Friday.30 In
other words, people spent between 26.4% and 44.3% of their day away from home.
Chapin also reported that people travel an average of 14.2 miles per day, which
suggested that people moved between areas represented by different FSM stations
over the course of the day. By the early 1980s, additional studies raised further
questions as to the ability of FSM stations to represent the CO exposures of the
public.35,36
As stated previously, FSM stations use the NDIR method to measure ambient
CO concentrations. Because NDIR monitors are not portable, they cannot be used
to measure CO exposure as a person performs routine daily activities. The advent
of microelectronics during the 1970s enabled considerable progress to be made in
the development of reliable, compact, mobile air quality monitoring instruments.
The most dramatic of these were the new miniaturized personal exposure monit-
ors (PEMs) as described by Wallace and Ott.37 These instruments could go nearly
anywhere, as they were equipped with batteries and shoulder straps. The utility of
PEMs for measuring personal CO exposure was demonstrated in several early studies.
These included a study of automobile commuters in Los Angeles during the summer
of 1973,38 and field surveys of personal exposure in many commercial settings of sev-
eral California cities between November 1979 and July 1980.39 Continued technical
improvements of PEMs stimulated scholarly interest in how to use them in studies of
personal and population exposure.
Both direct and indirect methods in the use of PEMs have evolved. The direct
method distributes PEMs to ordinary people who record their CO exposures and
activities directly using either a paper diary or an electronic data logger. For example,
Cortese and Spengler used this method in their survey of Boston commuters.29 A
second method relies on indirect estimates of population exposure, based on the
8417: “8417_c002” — 2007/9/11 — 12:16 — page 13 — #9

following model advocated by Fugas40 and Duan.41

n
Ei = ck (tik ) (2.2)
k=1
where
Ei = the total integrated exposure of person i over some time period of interest
(e.g., 24 h)
ck = the air pollutant concentration in microenvironment type k
tik = the amount of time spent by person i in microenvironment type k
n = the number of microenvironment types encountered by person i over the
period of interest.
The total exposure of a population can be determined by summation of the integ-
rated exposures of individuals who are members of that population. The indirect
method assumes that an individual’s total exposure to air pollution is a function of
location and time. It follows that variation in the total exposure of an individual occurs
because air pollution concentrations vary from one location to another and because
time spent in different locations varies substantially from person to person. The indir-
ect method postulates the existence of “microenvironments” in which a person is
exposed to an air pollutant at a given concentration over a fixed period of time. Duan
described a microenvironment as a “chunk of air space with homogeneous pollut-
ant concentration.”41 Because of the potentially large number of microenvironments,
Duan suggested that similar ones should be grouped, either by location (e.g., indoor
or outdoor) or activity performed at a location (e.g., residential or commercial) into
“microenvironment types.”
Under the CAA, the U.S. EPA has authority to perform periodic reviews of the
criteria that support the NAAQS. In the early 1980s, EPA scientists developed a risk-
analysis framework to support its reviews of the NAAQS for CO.42,43 This framework
required estimates of the percentage of an urban population that was exposed to
CO concentrations that exceeded the NAAQS. The need for these estimates was
partly related to EPA’s proposal to change the form of the primary standard from
deterministic to statistical.44 In response to this need, the EPA supported development
of several large-scale population models of CO exposure in the early 1980s. These
models included the Simulation of Human Activity and Pollutant Exposure (SHAPE)
model and the NAAQS Exposure Model (NEM). Subsequently, the EPA supported
development of the probabilistic NEM for CO (pNEM/CO) and the Air Pollutants
Exposure Model (APEX). To provide data for these models, the EPA funded both
direct and indirect studies of urban population exposure. The indirect studies included
measurements of CO in various microenvironments and a nationwide survey of human
activity patterns as described later.
EPA’s direct studies took measurements of the daily CO exposures of the non-
institutionalized, nonsmoking adult populations (ages 18–70 years) living in two
metropolitan areas of the United States. The surveys were performed using port-
able CO exposure monitors equipped with data loggers to reduce the data-collection
8417: “8417_c002” — 2007/9/11 — 12:16 — page 14 — #10

burden on the population. Field surveys of 454 residents of Denver, Colorado, and
714 residents of Washington, DC, occurred during the fall of 1982 and winter of
1983. Study participants carried PEMs equipped with data loggers. They also kept
diaries for 48 h in Denver and for 24 h in Washington. The studies revealed that
ambient CO levels at the composite network of fixed-site monitors were able to
track variation in personal exposures. However, the network overestimated the 8-h
exposures of people with low-level personal exposures and underestimated the 8-h
exposures of people with high-level personal exposures. With respect to the under-
estimates, over 10% of the daily maximum 8-h exposures in Denver exceeded the
NAAQS of 9 ppm, and about 4% did so in Washington. The end-expired breath
CO levels were in excess of 10 ppm, which was roughly equivalent to about 2%
COHb in about 12.5% of the Denver participants and about 10% of the Washing-
ton participants. Recall, that the NAAQS for CO are designed to keep COHb levels
below 2% in the blood of the general public including probable high-risk groups.
During the survey period, the composite CO concentrations at fixed-site monitors
exceeded the 8-h NAAQS for CO (9 ppm) only 3% of the time in Denver, and never
in Washington, DC. Hence, the results of these two surveys raised further doubts
as to the ability of fixed-site monitors to represent the total CO exposure of urban
populations.45
The results of the population exposure studies in Denver and Washington did not
persuade the U.S. EPA to change the form of the NAAQS for CO from determin-
istic to statistical. In September, 1985, the EPA announced that it would retain the
existing primary NAAQS for CO, but that it would rescind the secondary standard
to protect public welfare, because there was no evidence to support it.7 However,
one could argue that the results of these two studies provided some justification for
the tighter CO emission standards that had taken effect for new cars sold outside
of California during the 1980 model year. (Under the CAA, the state of California
has authority to set its own tailpipe emission standards.) For example, Table 2.1
shows results for selected microenvironments of the Denver study. The table shows
that higher CO concentrations were associated with commuting by motor vehicles
(i.e., motorcycle, bus, car, and truck). It also shows that indoor CO concentrations
in excess of the NAAQS were observed in public garages and in service stations or
vehicle repair facilities. In the Washington study, participants who commuted 6 h
or more per week had higher CO exposures than those who commuted fewer hours
per week. Likewise, participants who had occupations that involved motor vehicles
(e.g., professional drivers of trucks, buses, and taxis; automobile mechanics; gar-
age workers; and policemen) had a mean CO exposure (22.1 ppm) that was over
three times higher than the average exposure of those who did not work with motor
vehicles (6.3 ppm).
The Denver population exposure study provided raw data to test the SHAPE
model and the probabilistic version of the NAAQS Exposure Model (pNEM/CO).
The evaluation of SHAPE showed close agreement between the observed and
predicted arithmetic means of the 1-h and 8-h maximum average CO exposures.
However, SHAPE over-predicted low-level exposures and under-predicted high-level
exposures.48 Likewise, an evaluation of pNEM/CO showed relatively close agree-
ment between simulated and observed exposures for CO concentrations near the
8417: “8417_c002” — 2007/9/11 — 12:16 — page 15 — #11

TABLE 2.1
Carbon Monoxide Concentrations of Selected Microenvironments in
Denver, Colorado, 1982–1983 (Listed in Descending Order of Mean
CO Concentration)
Microenvironment n Meana (ppm) Standard Error (ppm)
In-Transit
Motorcycle 22 9.79 1.74
Bus 76 8.52 0.81
Car 3, 632 8.10 0.16
Truck 405 7.03 0.49
Walking 619 3.88 0.27
Bicycling 9 1.34 1.20
Outdoor
Public garages 29 8.20 0.99
Residential garages or carports 22 7.53 1.90
Service stations or vehicle repair facilities 12 3.68 1.10
Parking lots 61 3.45 0.54
Other locations 126 3.17 0.49
School grounds 16 1.99 0.85
Residential grounds 74 1.36 0.26
Sports arenas, amphitheaters 29 0.97 0.52
Parks, golf courses 21 0.69 0.24
Indoor
Public garages 116 13.46 1.68
Service stations or vehicle repair facilities 125 9.17 0.83
Other locations 427 7.40 0.87
Other repair shops 55 5.64 1.03
Shopping malls 58 4.90 0.85
Residential garages 66 4.35 0.87
Restaurants 524 3.71 0.19
Offices 2, 287 3.59 0.002
Auditoriums, sports arenas, concert halls 100 3.37 0.48
Stores 734 3.23 0.21
Health care facilities 351 2.22 0.23
Other public buildings 115 2.15 0.30
Manufacturing facilities 42 2.04 0.39
Homes 21, 543 2.04 0.02
Schools 426 1.64 0.13
Churches 179 1.56 0.25
a An observation was recorded whenever a person changed a microenvironment, and on every
clock hour; thus each observation had an averaging time of 60 min or less.
Source: Johnson, T. A Study of Personal Exposure to Carbon Monoxide in Denver, Colorado,
Report No. EPA-600/4/84-014, Environmental Monitoring Systems Laboratory, U.S. Environ-
mental Protection Agency, Research Triangle Park, NC, 1984 as reported in U.S. EPA. Air
Quality Criteria for Carbon Monoxide, Report No. EPA 600/8-90/045F, Office of Health and
Environmental Assessment, Environmental Criteria and Assessment Office, U.S. Environmental
Protection Agency, Research Triangle Park, NC, 1991.
8417: “8417_c002” — 2007/9/11 — 12:16 — page 16 — #12

average exposure (i.e., within the range of 6–13 ppm for the 1-h standard and within
5.5–7 ppm for the 8-h standard). Like SHAPE, pNEM/CO also over-predicted lower
exposures and under-predicted higher exposures for both standards.49
2.5 MICROENVIRONMENTAL EXPOSURES

This section describes studies of CO exposures in microenvironments that people use
to live, work, shop, play, and commute. In particular, it identifies factors that may
contribute to high-level exposures in these microenvironments. High-level expos-
ures deserve more attention, because population exposure models (i.e., SHAPE and
pNEM/CO) appear to underestimate high-level exposures in these microenviron-
ments, as stated earlier. These microenvironments were also chosen, because they
were identified by the U.S. EPA’s National Human Activity Pattern Survey (NHAPS)
as locations relevant to air pollution exposure. That survey collected 24-h retrospect-
ive diary data on activities and their locations from 9196 respondents interviewed in
the 48 contiguous states from late September 1992 through September 1994. Table 2.2
summarizes the minutes spent on the diary days in six locations for respondents to
the survey.50
2.5.1 RESIDENTIAL EXPOSURES

Exposure to CO in the home is an important component of a person’s total daily
exposure, because an estimated 68.7% of one’s time on average is spent inside a
residence.50 Major sources of CO concentrations inside the home include unvented
or poorly vented furnaces, gas appliances, fireplaces, wood stoves, kerosene space
heaters, and charcoal grills and hibachis. Other CO sources include motor vehicles
inside an attached garage. Studies of exposures to nonfatal concentrations are dis-
cussed first, followed by studies of unintentional deaths caused by high indoor CO
concentrations.
TABLE 2.2
Time Spent in Different Locations by 9196 Participants of The National
Human Activity Pattern Survey (NHAPS), October 1992–September 1994
Location Overall Mean (min) Doer % Doer n Doer Mean (min)
In a residence 990 99.4 9153 996
Office-factory 78 20.0 1925 388
Bar-restaurant 27 23.7 2263 112
Other indoor 158 59.1 5372 267
In an enclosed vehicle 79 83.2 7596 95
Outdoors 109 59.3 5339 184
Source: Modified from Klepeis et al., J. Expo. Anal. Env. Epid., 11, 231, 2001.
8417: “8417_c002” — 2007/9/11 — 12:16 — page 17 — #13

2.5.1.1 Nonfatal Exposures

Many Americans have appliances that emit CO in the home. These homes often use
gas (natural gas and liquid propane) for cooking, heating water, and drying clothes.
Of all participants of the NHAPS study, 38.3% had a gas range or oven at home, and
23.7% said that it had a burning pilot light. The same study showed that some people
still use small appliances to heat a room: fireplace (10%); wood stove (6%); and
kerosene space heater (2%). In the presence of indoor sources such as gas appliances,
indoor CO concentrations often exceed outdoor levels.47 In a 1985 Texas study, CO
concentrations were greater than or equal to the NAAQS for CO of 9 ppm in 12% of
surveyed homes. Residential CO concentrations were high in cases where unvented
gas space heaters were used as the primary heat source.51
According to the Barbecue Industry Association, 44 million American households
owned a charcoal grill in 1989, and an estimated 600 million charcoal-barbecuing
events took place annually.52 An earlier study showed that the air stream from char-
coal grills contained 20–2000 ppm of CO, with 75% of grills emitting 200 ppm and
above.53 Another study reported COHb levels ranging from 6.9% to 17.4% in a family
of four people in northern California who had been exposed to smoke from cooking
indoors on a barbecue grill, which was found by firefighters in the middle of the living
room.54
Based on data for ten counties, a study in Washington state reported features
of unintentional CO poisoning cases that occurred between 1982 and 1993.52 Most
cases occurred when electrical power was interrupted during fall and winter months,
because of either regional storms or unpaid utility bills. Of 509 patients treated with
hyperbaric oxygen, 79 (16%) were exposed to CO emissions when charcoal briquets
were burned for heating or cooking in 32 separate incidents. Non-English speaking
Hispanic whites and Asians were disproportionately represented among the cases. The
COHb levels of these 79 people averaged 21.6% and ranged from 3.0% to 45.8%.
Two studies assessed CO exposure to emissions from unvented portable kerosene
heaters in eight small mobile homes with no gas appliances and low air exchange
rates.55,56 Each home was monitored for an average of 6.5 h per day for 3 days per
week for 4 weeks. For 2 weeks the heater was on, and, for 2 weeks, it was off. When
the heater was turned on, it was in use for an average of 4.5 h. When the heater was in
use, study participants (all nonsmokers) spent most of their time in the family room
or kitchen. Sampling took place in the living area about 1.5–3 m from the heater.
The mean 8-h CO concentrations were 7.4 ppm (1-h peak = 11.5 ppm) when the
heater was on and 1.4 ppm (1-h peak = 1.5 ppm) when it was off. Peaks usually were
observed at the end of the combustion period. The ambient CO level measured 0.5
h prior to heater use ranged from 0 to 8 ppm. When the heater was on, three of the
eight homes had 8-h average CO levels that exceeded the NAAQS, and one home
routinely had levels of 30–50 ppm.
A California study reported CO exposures for a random sample of homes that
used gas appliances during a 48-h period from December 1991 to April 1992.57−59
For periods of 48 h, the median CO concentration was 1.2 ppm (indoors) and 0.8 ppm
(outdoors), and the median of the maximum 8-h average CO concentration was 2.0
ppm (indoors) and 1.4 ppm (outdoors). Of surveyed homes, 13 of 286 homes (4.5%)
8417: “8417_c002” — 2007/9/11 — 12:16 — page 18 — #14

had indoor CO concentrations above the NAAQS of 9 ppm for 8 h, and 8 of 282
homes (2.8%) had outdoor CO concentrations above this standard. Although most of
the exceedances occurred in the Los Angeles basin, these percentages could be low
because the basin was under-represented in the statewide sample. The study did not
translate these percentages into statewide estimates.
The study suggested that a small percentage of California homes would still have
indoor CO problems even if outdoor CO levels at these homes complied with federal
ambient standards. For a common sample of 277 homes, 17 homes (6.1%) had 1-h
maximum concentrations indoors that were at least 5 ppm higher than outdoor levels,
and 10 homes (3.6%) had 8-h maximum CO concentrations indoors that were at least
5 ppm higher than outdoor levels. Using univariate regression analysis, outdoor CO
concentrations explained approximately 55% of the indoor CO variation. Higher net
indoor CO levels (indoor minus outdoor CO concentrations) were traced definitively
to space heating with gas ranges and gas-fired wall furnaces, use of gas ranges with
continuous gas pilot lights, small home volumes, and cigarette smoke. However, sev-
eral other factors also may have contributed to the higher CO levels: malfunctioning
gas furnaces, automobile exhausts leaking into homes from attached garages and car-
ports, improper use of gas appliances (e.g., gas fireplaces), and improper installation
of gas appliances (e.g., forced air unit ducts).59
2.5.1.2 Fatal Exposures

Factors contributing to unintentional deaths from CO poisonings were identified by
studies in California and New Mexico. In California, two studies collected data for
the 1979–1988 period. In the first study, 59 of 444 deaths (13.3%) were caused by
improper use of charcoal grills and hibachis.60 Of the 59 deaths, 54% occurred inside
motor vehicles (e.g., vans, campers) and 46% in residential structures (e.g., homes,
apartments, shacks, tents). Relative to their share of the state’s population, higher
death rates occurred amongAsians, blacks, males, and people aged 20–39. The second
study identified specific factors that contributed to unintentional deaths caused by CO
from several combustion sources (e.g., charcoal grills and hibachis, other heating and
cooking appliances, motor vehicles, small engines, camping equipment).61 In this
study, there was a strong association between alcohol use and CO poisoning from
motor vehicles. Typically, motorists under the influence of alcohol would pull into
their garages, leave the engine running while listening to cassette tapes, and then fall
asleep. Faulty heating equipment used during winter months was implicated in about
50% of all unintentional deaths in both the California study61 and the New Mexico
study.62
The National Center for Health Statistics (NCHS) and the U.S. Consumer Product
Safety Commission (CPSC) estimated that 212 deaths in 1992 could be attributed
to fuel-burning appliances used in the home. Of these deaths, 13 involved use of
gasoline-powered appliances.63 An estimated 3900 CO injury accidents occurred in
1994, of which about 400 were associated with the use of gasoline-powered engines or
tools. In response to the problem, several federal government agencies issued a joint
alert concerning exposure to CO emitted by these sources.64 These sources involved
use of pressure washers, air compressors, concrete-cutting saws, electric generators,
8417: “8417_c002” — 2007/9/11 — 12:16 — page 19 — #15

floor buffers, power trowels, water pumps, and welding equipment. Unintentional
CO poisonings frequently happened indoors even when people took precautions to
ventilate buildings.
Power outages following hurricanes and tropical storms often create demand for
alternate sources of electricity (e.g., portable gasoline generators) to run air con-
ditioners and refrigerators. But these generators can be a significant source of CO
exposures if they are placed in garages or outdoors near windows. The majority of
exposures occur overnight when generators are used to run air conditioners and other
appliances. Hurricanes Katrina and Rita, which struck the U.S. Gulf Coast in the late
summer of 2005, caused 10 deaths from CO poisoning in 18 storm-affected counties
of Alabama and Texas.65 All of the fatalities were caused by gasoline-powered gen-
erators placed either inside the home or in a fully enclosed space outside the home.
Very few homes had functioning CO detectors. In four hurricanes that hit Florida in
2004, some victims of CO poisoning placed generators inside their homes or garages
to protect the devices from weather damage or to prevent theft.66
2.5.2 OCCUPATIONAL EXPOSURES

The NHAPS study reported that 20% of Americans spent nearly 6.5 h per day on
average working inside an office or factory.50 The National Institute for Occupational
Safety and Health (NIOSH) estimated that 3.5 million workers who work in the private
sector potentially are exposed to CO primarily from motor exhaust. The number of
persons potentially exposed to CO in the work environment is greater than that for
any other physical or chemical agent.67 In 1992, there were 900 work-related CO
poisonings resulting in death or illness in private industry according to the U.S.
Bureau of Labor Statistics as cited in a NIOSH report.64
Three risk factors affect industrial occupational exposure: (1) the work environ-
ment is located in a densely populated area that has high background (i.e., ambient)
CO concentrations; (2) the work environment produces CO as a product or by-product
of an industrial process, or the work environment tends to accumulate CO concen-
trations that may result in occupational exposures; and (3) the work environment
involves exposure to methylene chloride (i.e., dichloromethane), which is metabol-
ized to CO in the body. Proximity to fuel combustion of all types elevates CO exposure
for certain occupations: airport employees; auto mechanics; small gasoline-powered
tool operators (e.g., users of chainsaws); charcoal meat grillers; construction work-
ers; crane deck operators; firefighters; forklift operators; parking garage or gasoline
station attendants; policemen; taxi, bus and truck drivers; toll booth and roadside
workers; and warehouse workers.47 Table 2.3 shows results for several occupational
studies (typical CO values and/or ranges), averaging periods, and the measured or
estimated percent COHb levels for nonsmokers, if reported.
The CO exposure of office building workers has received less attention. Flachsbart
and Ott83 developed a “rapid method” for surveying CO concentrations inside several
high-rise buildings. In one case, they observed CO concentrations in excess of 9
ppm, which is the 8-h NAAQS for CO, on four visits to a 15-story office building
in Palo Alto, California. A survey in April 1980, showed that CO concentrations in
the building’s underground garage averaged 40.6 ppm. The CO levels ranged from
8417: “8417_c002” — 2007/9/11 — 12:16 — page 20 — #16

TABLE 2.3
Studies of Occupational Exposures and Dosages
Occupational CO Concentration (ppm) Measured or Estimated
Category and Averaging Period Percent COHb State/Country References
Airport workers 5.0–13.6 (0.25 h) NA Massachusetts, U.S. 68
5–300 (0.1–1.7 h) (INT) NA U.S. 69
Bus drivers 5.8–12.5 TWA (0.5–1 h) NA France 70
Chainsaw/gas NA 9.2–75.6 in 5 farmers U.S. 71
tool operators >200 (<2 min) NA U.S. 64
16.2–24.3 TWA (8 h) > 4 in 10 NS Germany 72
Forklift operators 250–300 (5 h) 5–22 for 4 NS North Carolina, U.S. 73
and workers in NA (4.4 h) 4.2–28.7 for 7 NS North Carolina, U.S. 74
facilities with 370–386 (NA) 21.1 ± 0.7 North Carolina, U.S. 75
forklifts 25–47 TWA (8–12 h) 6.3–13.3 for 4 NS Colorado, U.S. 76
3–34 (8 h) > 3.25 in 5% of NS California, U.S. 77, 78
Garage mechanics 42.6% > 35 (1 h) > 5 in 45% of NS Ontario, Canada 79
Exposure to Ambient and Microenvironmental Concentrations
Traffic/roadway 2.7 (8 h) NA 4 states, U.S. 80

workers 1–4.3 (8 h) NA Denmark 81
5–42 (ENV) <5 Massachusetts, U.S. 82
Note: ENV = short-term environmental measurements; INT = interior of vehicle; NA = not available; NS = non-smokers; and TWA = time-weighted average
Source: Modified from Apte, M. A population-based exposure assessment methodology for carbon monoxide: development of a carbon monoxide passive sampler and
occupational dosimeter, Ph.D. thesis, University of California, Berkeley, 1997 and updated.
8417: “8417_c002” — 2007/9/11 — 12:16 — page 21 — #17

21
10.2 to 18.5 ppm on the first eleven floors of the building, but were only 2–4 ppm on
the top four floors. They attributed these findings to three factors: (1) fans ventilating
the garage had been switched off to reduce electricity costs, because utility rates
had increased sharply due to an international energy crisis; (2) the door connecting
the garage with the main stairwell of the building was kept open; and (3) the first
eleven and top four floors of the building were served by separate ventilation systems.
Once the first two factors were corrected (i.e., the garage fans were switched on and
the connecting door to the garage was closed), Flachsbart and Ott83 reported that
average CO levels in the garage dropped from 40.6 ppm (before) to 7.9 ppm (after),
while typical CO levels in the building fell from 11 ppm to 12 ppm (before) to 1 ppm
to 2 ppm (after).
2.5.3 SHOPPING CENTER EXPOSURES

The NHAPS study estimated that 59.1% of Americans spent nearly 4.5 h per day on
average at other indoor locations, such as shopping malls, stores, schools, churches,
health clubs, laundromats, salons, and parking garages.50 Of these locations, major
shopping centers, in particular, attract and generate relatively large volumes of motor
vehicle traffic that typically circulates at low speeds with frequent stops and starts.
This traffic pattern produces relatively high CO emissions, which can accumulate to
unhealthy levels if emissions occur in enclosed or semienclosed spaces. Since shop-
ping centers also attract large numbers of people, the potential for human exposure
to high CO levels is great.
In the 1980s, the Ala Moana Shopping Center in Honolulu, Hawaii, was an
example of this type of indoor CO exposure problem. First, the center had 155 business
outlets that attracted 40 million people including many tourists each year. Second,
the center had an attached structure with 7800 parking spaces on several decks. CO
emission rates were high, because the posted speed limit for motor vehicles in the
structure was 15 mph for the safety of pedestrians. CO emission rates increase when
vehicle speeds fall below 15 mph. Third, one deck of the parking structure functioned
as a lid on the exhaust emissions of cars at the street level of the structure. Fourth,
many of the 94 outlets at street level kept their doors open during business hours to
attract customers. This allowed CO concentrations from the parking area and internal
driveways to diffuse into many retail outlets.84
Flachsbart and Brown visited 25 of the 94 street-level outlets at the shopping center
every 5 days between November 1981 and March 1982.85 During one visit 4 days
before Christmas Day, they estimated that average 1-h concentrations of CO inside
ten stores exceeded the federal ambient standard of 35 ppm, but not the occupational
standards which were 200 ppm for a 1-h period and 50 ppm for an 8-h period. The
average CO levels during 3-min visits to these ten stores on that date ranged from
a low of 36.3 ppm to a high of 86.7 ppm. An earlier study of the shopping center’s
ground-level parking area found average CO levels ranging from 12 to 37 ppm.86
Although the two studies surveyed different locations of the shopping center, the
lower CO levels observed by the 1981–1982 study of the center could be attributed
to implementation of the federal motor vehicle emission control program during
the 1970s.
8417: “8417_c002” — 2007/9/11 — 12:16 — page 22 — #18

2.5.4 RECREATIONAL EXPOSURES

Recreational exposures occur when people use recreational vehicles and when
they watch or participate in certain indoor sporting events that involve motor
vehicles.
2.5.4.1 Exposures on Recreational Vehicles

Two studies examined personal exposure to CO in the exhaust of recreational vehicles.
In the first study, Simeone sampled CO concentrations in the passenger areas of
large power boats with side-mounted exhausts during routine cruises offshore of
Annapolis, Maryland, and Boston, Massachusetts.87 In Boston harbor, CO concen-
trations averaged 56 ppm during a 60-min cruise and 28 ppm after a 30-min cruise.
For the Chesapeake Bay cruises near Annapolis, average stabilized CO concentra-
tions at the helm ranged from 93 to 170 ppm over 20- to 30-min periods and 272
ppm over 30 min on the rear deck near the transom of the boat. In both studies,
exhaust gas was affected significantly by airflow about the boat under certain head
winds. At head wind speeds of 10–30 knots, turbulent mixing occurred in closer
proximity to the rear of the boats, enabling exhaust gases to migrate freely into
each boat.
In the second study, Snook studied the CO exposure of a second snowmobiler who
tailed a lead snowmobiler on a 2- to 3-miles straight trail over level terrain in Grand
Teton National Park, Wyoming.88 The CO exposure of the second snowmobiler was
measured under stable atmospheric conditions in Tedlar™ bags. The distance between
the two snowmobiles ranged from 25 to 125 ft., and speeds ranged from 10 to 40 mph.
The second snowmobiler’s maximum average centerline exposure was 23.1 ppm,
which occurred at 10 mph and 25 ft. behind the lead snowmobile. Although Snook
reported no averaging times for exposures, one can estimate that these times ranged
from 3 to 18 min from the data given on the snowmobiler’s travel distance and vehicle
speed. In general, the centerline CO concentrations decreased with increasing distance
between snowmobiles and increased with greater speeds, but only for distances greater
than 25 ft. between snowmobiles. At 15 ft. off centerline, average concentrations fell
sharply to levels of 0–7.5 ppm. When the snowmobiler drove alone (self-exposure),
the average concentration minus the background concentration was 1.3–3.0 ppm.
Background concentrations ranged from 0.2 to 0.5 ppm.
At the time of Snook’s study, snowmobile tourism was a booming business
across the nation and in several national parks. During the winter of 1993 and 1994,
over 87,000 tourists traveled by snowmobile in Yellowstone National Park.89 Under
steady-state conditions, a snowmobile may emit from 10 to 20 g of CO per mi, while
a modern U.S. automobile equipped with a catalytic converter emits far less (0.01–
0.04 g of CO per mi) at speeds of 10 to 40 mph. Snook and Davis reported that
there were no federal laws regulating the exhaust from snowmobile engines, and
states were pre-empted from implementing snowmobile emission standards.90 The
typical snowmobile utilized a two-stroke engine, because it is less expensive than a
four-stroke engine and provides a high power-to-weight ratio. However, a two-stroke
engine produces relatively high emissions of CO.
8417: “8417_c002” — 2007/9/11 — 12:16 — page 23 — #19

2.5.4.2 Exposures at Indoor Sporting Events

Significant quantities of CO are emitted by certain types of machines during sporting
events that occur in poorly ventilated indoor arenas. The CO is emitted by several
sources, including ice-resurfacing machines and ice edgers during skating events; gas-
powered radiant heaters used to heat viewing stands; and motor vehicles at motocross,
monster-truck, and tractor-pull competitions. These competitions usually involve
many motor vehicles with no emission controls.
Several studies of CO exposure at indoor sporting events are noteworthy. First,
Kwok reported episodes of CO poisoning among skaters inside four arenas in Ontario,
Canada.91 Mean CO levels ranged from 4 to 81 ppm for periods of about 80 min.
The CO levels in the spectator areas ranged from 90% to 100% of levels on the
ice rinks. The ice resurfacing machines lacked catalytic emission controls. Second,
both Sorensen92 and Miller et al.93 reported CO concentrations greater than 100 ppm
in rinks from the use of gasoline-powered resurfacing machines. High concentra-
tions were attributed to poorly maintained machines and insufficient ventilation in
one rink. Third, based on data collected in the Québec city area, Lévesque et al.94
developed a linear relationship between CO exposure and the CO concentration in
exhaled breath, but could not eliminate other factors affecting the relationship. In a
later study, Lévesque et al.95 measured the alveolar CO of 14 male adult nonsmokers
who played ice hockey, but who were not exposed in occupational settings. Rink CO
concentrations ranged from 0 to 76.2 ppm. The study again found a linear relation-
ship between exposure and absorbed CO, such that for each 10 ppm of CO in the
indoor air, the players absorbed enough CO to raise alveolar CO by 4.1 ppm or about
0.76% COHb.
In the United States, surveys of CO exposure were done at ice arenas in Vermont,
Massachusetts, Wisconsin, and Washington. For a rink in Massachusetts, Lee et al.96
showed that excessive CO concentrations can occur, even with well-maintained equip-
ment and fewer resurfacing operations, if ventilation is inadequate. Average CO levels
were less than 20 ppm over 14 h, with no significant source of outdoor CO. Ventilation
systems could not disperse pollutants emitted and trapped by temperature inversions
and low air circulation at ice level. In another study, Lee et al.97 reported that CO
concentrations measured inside six enclosed rinks of the Boston area during a 2-h
hockey game ranged from 4 to 117 ppm, whereas outdoor levels were about 2–3
ppm. The alveolar CO of hockey players increased by an average 0.53 ppm per 1
ppm CO exposure over 2 h. Fifteen years earlier, Spengler et al.98 found similar CO
concentrations in the Boston area, with levels ranging from 23 to 100 ppm for eight
enclosed rinks.
In 1991, Paulozzi et al.99 reported that 25 people exposed to CO during a Vermont
high-school ice hockey game had mean COHb levels of 8.9%, but did not report
whether any of them were smokers. Although Paulozzi et al. were unable to measure
CO concentrations at the game, Smith et al.100 reported CO levels of 150 ppm (no
averaging time was given) at an indoor ice-hockey rink in Wisconsin. To document
the extent of the problem in Vermont, Paulozzi et al. measured CO concentrations
during eight high-school games in the state, and reported that average CO levels
for the entire game ranged from <5 ppm to 101 ppm with a mean of 35 ppm.101
8417: “8417_c002” — 2007/9/11 — 12:16 — page 24 — #20

Hampson reported a maximum CO level of 354 ppm inside an ice arena in Seattle,
Washington, in March 1996.102 Based on data for 17 persons, whose tobacco use
was not reported, the average COHb level was 8.6% (range of 3.3–13.9%). The
source of the CO was a malfunction in a 20-year-old ice resurfacing machine. The
study also reported that CO may have diffused into an adjacent bingo hall through
an open door. In view of these studies, the State of Minnesota declared in Regulation
No. 4635 that CO measurements taken 20 min after ice resurfacing must be less
than 30 ppm.102
Studies also have been done in sports arenas that allow motor vehicles. Boudreau
et al.103 reported CO levels for three indoor sporting events (i.e., monster-truck com-
petitions, tractor pulls) in Cincinnati, Ohio. The CO measurements were taken before
and during each event at different elevations in the public seating area of each arena
with most readings obtained at the midpoint elevation where most people were seated.
Average CO concentrations over 1–2 h ranged from 13 to 23 ppm (before the event)
to 79 to 140 ppm (during the event). Measured CO levels were lower at higher seating
levels. The ventilation system was operated maximally, and ground-level entrances
were completely open.
High CO concentrations also have been found at motor vehicle competitions
in Canada. Luckurst and Solkoski104 recorded CO concentrations at two tractor-
pull events in Winnipeg, Manitoba. The mean of instantaneous concentrations at 25
locations in the arena ranged from 68 ppm at the start of the first event to 262 ppm by
the end. At the second event, the range was 78–436 ppm. Lévesque et al.105 reported
CO levels at an indoor motocross competition held in a skating rink in the Québec
City region. The May, 1994 event lasted from roughly 8:00 p.m. to midnight. Average
CO concentrations determined at five stations located at different points in the arena
ranged from 19.1 to 38.0 ppm, with the higher levels measured during the second half
of the show. High CO concentrations forced a health official to interrupt the event
seven times to help clear the air. Covariance analysis showed that CO levels were
related to the initial CO concentration, the event duration, engine size, and especially
the number of motorcycles on the track.
2.5.5 COMMUTER EXPOSURES

The NHAPS study estimated that 83.2% of Americans spent 95 min per day on
average inside a motor vehicle.50 Time spent inside a vehicle, as a percentage of
a 24-h day (6.6%), is relatively small compared to the percentage of time spent at
home (69.2%) or at work (26.9%). Still, one might expect commuter CO exposure
to be high. Highway vehicles accounted for 62.8% of the total CO emitted from
all sources except fires in the U.S. in 2003.13 In urban areas, highway vehicles
typically account for even higher percentages of total CO emissions.3 Moreover,
highway vehicles potentially are a major contributor to a person’s total CO exposure,
especially if a person is routinely near motor vehicle emissions.106 Some evid-
ence of this comes from the Denver population exposure study, which indicated
that CO exposure while driving was a major component of a person’s daily CO
exposure.45,48
8417: “8417_c002” — 2007/9/11 — 12:16 — page 25 — #21

In some ways, driving on congested roadways is analogous to swimming in a

polluted river. However, the analogy isn’t perfect, because most commuters have
enclosed vehicles, which may not fully protect their occupants from ambient CO
concentrations. In a Los Angeles study, Petersen and Allen107 reported that interior
CO concentrations averaged 92% of exterior levels. An earlier study by Petersen
and Sabersky found that CO concentrations measured outside vehicles exhibit rapid
variations and high peaks, while simultaneously measured concentrations inside a
vehicle show more gradual variations.38
Several factors affect variation in the CO exposures of commuters as shown
in Table 2.4, which lists several American studies in chronological order that
have been performed on this subject over a 28-year period since 1974. Table 2.4
includes three studies of actual commuters29,45,108 and five studies of commuter
microenvironments.109−113 A Boston study in 1974 revealed variation in the CO
exposure of a bicyclist by street type, as defined by traffic volume and number of
lanes.109 A few years later, Holland found that automobile commuter exposures var-
ied among four cities surveyed during the winter of 1981.110 A later study attributed
Denver’s higher exposures to its colder winter climate and higher altitude.114 Table 2.4
shows that exposures vary by commuting period (morning versus afternoon);29 trip
type (residential versus commuting to an urban area);110 mode of travel (automobile,
bus, and rail);111 season of the year (summer versus winter);108 trip location (rural,
interstate beltway, and urban);112 and year of study.113 Akland et al.45 reported that
actual commuters who spent six or more hours per week in travel had greater expos-
ures than those who spent less than 6 h. Finally, Table 2.4 shows that seven of these
studies found that commuter exposures typically exceeded concurrently measured
ambient concentrations.29,45,108,110−113
2.5.5.1 Defective Exhaust Systems

The need to control CO emissions from motor vehicles can be traced in part to studies
that tied high CO exposures during commuting to defective exhaust systems and cars
without effective emission controls. For example, Amiro115 found that engine and/or
tailpipe emissions of CO leaked into passenger cabins of 9 of 19 vehicles studied.
CO concentrations inside contaminated vehicles ranged up to 400 ppm. Clements
reported that of 645 school buses tested, 7.2% had average CO readings in excess
of 20 ppm, and 5.4% had maximum readings above 50 ppm.116 He estimated that
CO concentrations in excess of 20 ppm existed in school buses ridden by up to 2.1
million U.S. school children on a daily basis in the 1970s. In a two-step study, Ziskind
et al.117 first identified which sustained-use vehicles (buses, taxicabs, and police cars)
had faulty exhaust systems, that is, they leaked CO into passenger cabins. Of 120 rides
taken in faulty vehicles during the second step, the researchers found that 58% had
CO concentrations that exceeded the U.S. NAAQS of 9 ppm for 8 h. In a later study,
Hampson and Norkool reported that 20 of 68 children were treated for accidental
CO poisoning after riding in the back of eight pickup trucks in Seattle, Washington,
between 1986 and 1991.118 Average COHb levels measured in an emergency room
were 18.2% ± 2.4% (mean ± standard error of the mean), and ranged from 1.6% to
37.0%. Seventeen of the twenty children rode under a rigid closed canopy attached to
8417: “8417_c002” — 2007/9/11 — 12:16 — page 26 — #22

TABLE 2.4
Factors Affecting Commuter Carbon Monoxide Exposure.
Mean or Range or Typical
Median CO SD of CO Ambient CO
Travel Concentration Concentration Concentration
Factor Study Period Urban Area Mode (ppm) (ppm) (ppm) References
Type of street Summer ’74 Boston, MA 109
Narrow-busy 2-lanes Bicycle 16.4 (17.3)a 3.3 (4.6)a NA
Wide-busy 4-lanes Bicycle 14.4 (10.9)a 3.5 (1.9)a NA
Narrow-light 2-lanes Bicycle 10.1 (8.8)a 0.88 (1.5)a NA
Commuting period October ’74–February ’75 Boston, MA 6.0 29

Morning Auto 12.8 ±5
Afternoon Auto 13.9 ±6
City January–March ’81 110

Stamford, CT Auto 5.2 ±4.7 4.2–5.5b
Los Angeles, CA Auto 8.5 ±5.8 3.9–5.8
Phoenix, AZ Auto 6.7 ±4.9 2.4–3.9
Denver, CO Auto 10.3 ±7.7 3.1–5.8
Exposure to Ambient and Microenvironmental Concentrations
Trip type January–March ’81 Four citiesc 110

Residential driving Auto 2.9–7.6 ±3.6 to ±5.0 ±2.4 to ±3.9
Urban commute Auto 6.2–16.1 ±4.7 to ±7.7 ±3.9 to ±5.8
Mode of travel January–March ’83 Washington, DC 2.3 111

Auto 9.1–22.3 ±2 to ±9
Bus 3.7–10.2 ±1 to ±7
8417: “8417_c002” — 2007/9/11 — 12:16 — page 27 — #23

Rail 2.2–5.2 0.5 to ±5
(Continued)
27
28
TABLE 2.4
Continued
Mean or Range or Typical
Median CO SD of CO Ambient CO
Travel Concentration Concentration Concentration
Factor Study Period Urban Area Mode (ppm) (ppm) (ppm) References
Weekly travel November ’82–February ’83 Washington, DC 3.2 45
< 6 h/wk All 5.04 NA
≥ 6 h/wk All 7.90 NA
Season May ’87–March ’88 Los Angeles, CA 108

Summer Auto 6.5 ±2.2 2.8
Winter Auto 10.1 ±5.8 4.3
Trip location August–September ’88 Raleigh, NC 112

Rural Auto 4 NA NA
Interstate beltway Auto 11 NA NA
Urban Auto 13 NA 2.8
Year of study San Francisco, CAd 113

February ’80–February ’81 Auto 9.7e ±3.1 1.3
February ’91–February ’92 Auto 4.8e ±1.7 1.31
February ’01–March ’02 Auto 1.5e ±0.8 0.61
Note: a Rush hours: 7–9:30 a.m. and 3–5:30 p.m.; (non-rush hour concentrations)
b Ambient concentration was measured outside the automobile.
c Four cities were Stamford, CT; Los Angeles, CA; Phoenix, AZ; and Denver, CO.
d Surveys occurred in the suburbs of Menlo Park, Palo Alto, and Los Altos.
8417: “8417_c002” — 2007/9/11 — 12:16 — page 28 — #24

e Median of net CO concentrations for 50 trips, where the net value of each trip = average trip CO concentration − background CO concentration.
NA = not available
Carbon Monoxide Poisoning
the bed of the truck, and the other three children rode beneath a tarpulin. Six pickups
had a known leak in their exhaust system: three had a rear-end tailpipe and three had
a side-mounted tailpipe.
2.5.5.2 Parking Garages

Initial studies of CO concentrations in parking garages focused primarily on the
garage itself. Trompeo et al.,119 who surveyed 12 underground garages in Turin,
Italy, reported that CO levels averaged 98 ppm and ranged from 10 to 300 ppm
for 132 observations. Chovin120 measured CO levels of 80–100 ppm on average
in poorly ventilated garages in Paris, France. CO concentrations in parking gar-
ages can reach a peak in the late afternoon when many drivers leave work and start
their cars at the same time. Goldsmith121 and Flachsbart et al.111 found that mass
exits from garages can elevate pollutant levels to very high concentrations. The lat-
ter study observed spikes in CO concentrations (mean = 94.0 ppm based on 3-min
averages, n = 14 cases) inside a test vehicle as it exited the underground garage of
a government office building in Washington, DC, during a winter survey in 1983.
The study also attributed these spikes in CO concentrations to the fact that many
cars were operating under “cold-start” conditions. Once the test vehicle exited the
garage, high CO concentrations inside the vehicle gradually dissipated, but the dis-
sipation rate was slow because the vehicle’s windows were kept closed due to cold
weather.
2.5.5.3 Service Stations

Many motorists typically pump their own fuel at service stations. For safety reasons,
most people shut off the engine of their vehicle after they arrive at the station. How-
ever, some people may be exposed to CO emitted by vehicles entering or exiting the
station. Amendola and Hanes measured CO concentrations in thirteen service stations
and two automobile dealerships in the New England area.122 Results varied by sea-
son, with CO concentrations ranging from 2.2 to 21.6 ppm in warm weather and 16.2
to 110.8 ppm during cold weather. They cited reduced ventilation as another cause of
higher concentrations during winter. Based on a random sample of 100 self-service
stations in four counties of southern California, Wilson et al.123 reported a median
CO concentration (4.3 ppm), which was about twice that of the ambient level (2.0
ppm) for 5-min averaging times.
2.5.5.4 Drive-Up Facilities

Passengers have been exposed to extremely high CO levels when motor vehicles are
idling in a queue. During February and March of 1976, Myronuk124 measured CO
concentrations inside an automobile at space-confined, drive-up facilities in Santa
Clara Valley, California. The in-vehicle CO concentrations ranged from 15 to 95
ppm for 15-min averages, with short-term peaks of 100–400 ppm. High CO averages
could be attributed to older vehicles that lacked catalysts, even though these vehicles
probably had other forms of emission controls and appeared to be well maintained
8417: “8417_c002” — 2007/9/11 — 12:16 — page 29 — #25

and tuned. Background CO levels were only 2–5 ppm and monitoring occurred when
wind speeds were low. Concentrations varied widely if ventilating fans for heater or
air conditioning units were in operation.
2.5.5.5 Airbag Deployment

Airbags are currently a required safety feature on new cars sold in the United
States. On the basis of deployment of four airbags, Wheatley et al.125 reported that
time-weighted-average (TWA) CO concentrations ranged from 174 to 370 ppm. Peak
CO concentrations occurred 2 min after deployment.
2.5.5.6 Motor Vehicle Emission Standards

One of the most significant factors affecting commuter exposure in the United States
has been the federal motor vehicle emission control program.126 The CAA amend-
ments of 1970, 1977, and 1990 established a series of progressively tighter motor
vehicle emission standards and other controls on new cars sold in the United States.
Tighter controls on the tailpipe emissions of new cars had the effect of offsetting
the expansion of motor vehicle travel caused by population growth and urban sprawl
since 1970.127 Prior to the 1968 model year, the CO tailpipe emissions of passenger
cars in the U.S. were 84.0 g per mi. A series of progressively tighter CO emission
standards took effect beginning with the 1968 model year, when a standard of 51.0 g
per mi took effect. By the 1975 model year, CO tailpipe emissions fell to 15.0 g per
mi as a result of the catalytic converter, which became standard equipment on new
cars sold in the U.S. The CO emission standard was then lowered to 3.4 g per mi for
the 1981 model year and then to 1.7 g per mi by the 2003 model year.3 This amounted
to a reduction of nearly 98% in the CO emission rate of new cars between the 1967
and 2003 model years.
Several studies show the effects on CO exposure of this trend. First, Flachsbart did
a meta-analysis of 16 commuter exposure studies that occurred in the U.S. between
1965 and 1992.128 These studies reported typical (mean or median) CO concentrations
for trips, most of which lasted an hour or less. CO concentrations fell from 37 ppm in
1965, as reported by Haagen-Smit129 for a study in Los Angeles, California, to 3 ppm
in 1992 for a study by Lawryk et al.130 in the New Jersey suburbs of New York City.
A more recent study of an arterial highway on the San Francisco Peninsula indicated
that the median CO exposure of 50 trips taken over a 15-month period between 2001
and 2002 was 1.5 ppm.113
If one assumes that the results of these 17 separate studies between 1965 and 2002
are representative of commuter CO exposures in cities nationwide, then exposures
fell 96% over the 37-year period. Notice that this percentage reduction in commuter
exposure (96%) is nearly identical to the percentage reduction in the CO emission
rate for new cars (98%). Also, the time period represented by the CO exposure
reduction (1965–2002) is nearly identical to the period used to estimate the emission
rate reduction (1967–2003). Finally, Mott et al.4 found that unintentional motor-
vehicle-related CO death rates in the United States fell from 4.0 to 0.9 deaths per one
million person-years between 1975, when the catalytic converter was introduced, and
8417: “8417_c002” — 2007/9/11 — 12:16 — page 30 — #26

1998. They estimated that if rates of unintentional CO-related deaths had remained
at pre-1975 levels, then 11,667 more vehicle-related CO deaths might have occurred
by 1998.
It’s quite possible that modern motor vehicle emission control programs have
substantially reduced the sharp peaks in CO exposures observed in older studies of
commuting microenvironments. A longitudinal study of commuting in California
noticed a lack of sharp peaks in CO measurements during surveys of an arterial
highway in 2001–2002 relative to observations of the same highway during previous
surveys in 1980–1981 and 1991–1992. The study attributed this finding to California’s
implementation of (1) an I/M program (i.e., Smog Check) in 1984; and (2) tougher
“durability standards” on emission controls which were adopted in September 1990
and phased-in on new cars sold in California in 1993 and 1994.113
2.6 EXPOSURE TO METHYLENE CHLORIDE

Exposure to halogenated hydrocarbons such as methylene chloride, which can be
metabolized to CO in the body, occurs when the chemicals are found in contaminated
ambient air and groundwater, and in consumer products that contain the chemical as
a solvent, flame-retardant additive, or propellant.
2.6.1 NONOCCUPATIONAL EXPOSURE

Exposure to methylene chloride in the home, for example, primarily occurs through
use of paint and varnish removers. Exposure may also occur through use of aero-
sol propellants such as those found in hair sprays, antiperspirants, air fresheners,
and spray paints. The Agency for Toxic Substances and Disease Registry reported
that some aerosol products may contain up to 50% methylene chloride.24 Ambient
exposure may occur near production and use facilities or near hazardous waste sites
that store methylene chloride. In a study reported in 1983, ambient concentrations
of methylene chloride near organic solvent cleaning and paint and varnish removal
operations ranged from 7.1 to 14.3 parts per billion (ppb) averaged over one year.131
Although methylene chloride readily disperses when released into the air, it may
remain in groundwater for years, and be ingested in drinking water or inhaled when
it volatilizes during showering and laundering.24
Exposure to about 500 ppm methylene chloride for several hours can elevate
COHb levels to 15%. Increases in COHb levels can be detected in the blood of
nonsmokers about 30 min after exposure to methylene chloride. Stewart et al.132
showed that elevated COHb levels were proportional to a series of controlled expos-
ures to methylene chloride. In another controlled experiment, Stewart and Hake133
observed postexposure levels of COHb ranging from 5% to 10% after 3 h use of a
liquid-gel paint remover containing 80% methylene chloride and 20% methanol by
weight. Concurrent exposure to methylene chloride and methanol prolongs the period
of elevated COHb in the body.23,133,134 Peterson reported COHb levels of up to about
10% saturation after inhalation of methylene chloride concentrations ranging from
50 to 500 ppm over 5 days, for 7.5 h per day.135
8417: “8417_c002” — 2007/9/11 — 12:16 — page 31 — #27

2.6.2 OCCUPATIONAL EXPOSURE

Certain occupations expose workers to organic solvents such as methylene chloride.
The solvent is widely used as a degreaser, paint remover, aerosol propellant, and
blowing agent for polyurethane foams. It is used as an extractant for foods and spices,
a grain fumigant, and a low-pressure refrigerant. It also is used in the manufacturing of
synthetic fibers, photographic film, polycarbonate plastics, pharmaceuticals, printed
circuit boards, and inks. More than one million workers have significant potential
for exposure to methylene chloride.24 Moreover, the highest levels of exposure to
methylene chloride often occur in the workplace. To protect worker health, the 8-h
TWA threshold limit value for methylene chloride was set at 50 ppm by the American
Conference of Governmental Industrial Hygienists (ACGIH).136 Exposure at this
concentration leads to COHb levels of about 1.9% in experimental subjects. Exposure
to 500 ppm for several hours may elevate COHb levels as high as 15%. An 8-h
exposure to about 500 mg/m3 (3.5 mg/m3 = 1 ppm) of methylene chloride vapor is
equivalent to an 8-h exposure to 35 ppm of CO.137
Methylene chloride stored in tissue may continue to be metabolized to CO after
several hours of acute exposure. In such cases, COHb levels will continue to rise
and peak as high as 25% about 5–6 h after exposure.24 Shusterman et al.138 reported
an apparent linear elevation of COHb as a function of hours worked by a furniture
refinisher who used paint stripper containing methylene chloride. Ghittori et al.139
reported a significant linear correlation (r = 0.87) between methylene chloride con-
centration in air and CO in alveolar air of nonsmoking and sedentary factory workers
in Italy. Exposure to 600 mg/m3 of methylene chloride for 7.5 h was associated with a
COHb level of 6.8% in eight volunteers. Exposure to methylene chloride also can be
fatal. Leikin et al.140 reported fatalities of two people who were exposed to unknown
concentrations of methylene chloride while they removed paint in an enclosed space.
Death was caused not by CO poisoning, but by solvent-induced narcosis. Before they
died, their COHb levels continued to rise following cessation of exposure despite
treatment with high levels of oxygen.
2.7 CONCLUSIONS
This chapter reviewed several types of CO exposure studies. Some studies measured
CO concentrations in places where people live, work, shop, play, and commute. Other
studies determined the amount of time people spent doing these activities. Still other
studies identified factors that contribute to exposures in excess of federal ambient or
occupational standards. A few conclusions from these studies follows.
The NHAPS found that Americans continue to spend most of their time indoors.
On average over a 24-h period, they spent 86.9% of their time indoors either at home,
work, bar–restaurant, or elsewhere. They spent 5.5% of their time inside a motor
vehicle and only 7.6% of their time outdoors.50 These percentages applied to the
entire national sample of 9196 respondents interviewed in the 1990s, not just to the
“doers” of various activities. If most Americans spend only a few hours outdoors
each day, then indoor and in-vehicle CO exposures are important components of total
human exposure.
8417: “8417_c002” — 2007/9/11 — 12:16 — page 32 — #28

The findings of the NHAPS study underscore the importance of studies that com-
pared indoor and outdoor CO exposures. Some of these studies indicated that indoor
and outdoor CO exposures differed from those predicted from observations at fixed-
site monitors, which are used by the U.S. EPA to determine compliance with the
NAAQS. When ambient CO levels are either high or low on a given day, fixed-site
monitors still reflect the corresponding high or low aggregates of personal exposures
on those days. Otherwise, the stations do not adequately represent the CO exposures
of community residents while they are exposed to motor vehicle exhaust during com-
muting, and to occupational and residential sources of unvented or poorly vented fuel
combustion. The mean COHb level of people exposed to CO from these sources is
greater than their mean COHb level predicted solely from exposure to CO of ambient
origin.
Longitudinal studies show that federal and state motor vehicle emission control
programs have substantially reduced the CO exposures of urban commuters in the
United States. These studies have two implications. First, similar reductions in expos-
ure may exist for other microenvironments, especially those that may be affected
substantially by tailpipe emissions from motor vehicles (e.g., office buildings and
shopping centers attached to parking garages, as well as service stations, waiting
lines at drive-up facilities, and sidewalks along busy streets). Unfortunately, there
appear to be no published trend studies of CO exposure for these other microenviron-
ments. Second, one would expect that results of certain population studies (e.g., the
Denver–Washington study of 1982–1983) may no longer represent the CO exposures
of current urban populations in the United States.
More recent studies have shown that people are still exposed to elevated CO
levels in certain indoor microenvironments (e.g., unventilated parking garages, motor
vehicles with leaky exhaust systems, small homes with unvented or poorly vented gas
stoves and space heaters). Moreover, studies report that high-level CO exposures can
occur when people use unregulated gasoline-powered appliances, engines, and tools
(e.g., chainsaws), even under ventilated conditions. These locations and activities
are not part of the regulated ambient environment. However, these microenviron-
ments typically represent settings of fatal and nonfatal CO poisonings. Studies of
California homes indicated that elevated CO concentrations (>9 ppm) were caused
by several factors, such as attached garages and carports; ranges with continuous
gas pilot lights; and improper use and installation of gas appliances, especially in
small homes. Other studies have found elevated CO concentrations (>9 ppm) when
people ride certain types of recreational vehicles (e.g., snowmobiles, power boats),
gather indoors to barbecue food (sometimes to cope with electrical power outages
after severe storms), and watch sporting events held at indoor arenas. High-level
exposures (>25 ppm) may occur inside arenas when they are used for ice skat-
ing or motocross, monster-truck, and tractor pull competitions. Vehicles in these
competitions often lack any type of emission controls. At some events, ventila-
tion did not sufficiently lower CO concentrations to safe levels (<9 ppm) for the
general public.
Finally, CO exposures are high in certain occupations, such as garage mechanics
and chainsaw/gas tool operators, and for people who operate or work around fork-
lifts. More than one million people have high exposure to methylene chloride in the
8417: “8417_c002” — 2007/9/11 — 12:16 — page 33 — #29

workplace, which is a concern because it can metabolize to CO in the body after

several hours of exposure.
2.8 ACKNOWLEDGMENTS
Portions of this chapter are reprinted with permission from: Chemosphere—Global
Change Science, vol. 1, M.A.K. Khalil, J.P. Pinto, and M.J. Shearer, Eds., P.G. Flachs-
bart, “Human Exposure to Carbon Monoxide from Mobile Sources,” pages 301–329,
copyright (1999), published by Elsevier; Urban Traffic Pollution, D. Schwela and
O. Zali, Eds., P.G. Flachsbart, “Chapter 4: Exposure to Exhaust and Evaporative
Emissions from Motor Vehicles,” pages 89–132, copyright (1999), published by the
WHO; Journal of Exposure Analysis and Environmental Epidemiology, 11(3), Neil E.
Klepeis et al., “The National Human Activity Pattern Survey (NHAPS): a resource for
assessing exposure to environmental pollutants,” pages 231–252, copyright (2001),
published by Nature Publishing Group; A Population-Based Exposure Assessment
Methodology for Carbon Monoxide: Development of a Carbon Monoxide Passive
Sampler and Occupational Dosimeter, copyright (1997) by Michael G. Apte, Ph.D.
Thesis, University of California, Berkeley; and Human Exposure Analysis, W. Ott,
A. Steinemann, and L. Wallace, Eds., Peter G. Flachsbart, “Exposure to Carbon
Monoxide,” pages 113–146, copyright (2007), published by CRC Taylor & Francis.
References
1. Burr, M.L. Combustion products, In Indoor Air Quality Handbook, Spengler, J.D.,
Samet, J.M., and McCarthy, J.F., eds., McGraw-Hill, New York, 2000, chap. 29.
2. Raub, J.A., Mathieu-Nolf, M., Hampson, N.B., and Thom, S.R. Carbon monoxide
poisoning—a public health perspective, Toxicology, 145, 1, 2000.
3. Godish, T. Air Quality, 4th ed., Lewis Publishers, Boca Raton, FL, 2004, 156–159.
4. Mott, J.A., Mitchell, I.W., Alverson, C.J., Macdonald, S.C., Bailey, C.R., Ball,
L.B., Moorman, J.E., Somers, J.H., Mannino, D.M., and Redd, S.C. National vehicle
emissions policies and practices and declining US carbon monoxide-related mortality,
JAMA: J. Am. Med. Assoc., 288, 988, 2002.
5. Vajani, M. Unintentional non-fire-related carbon monoxide exposures - United States,
2001–2003, MMWR: Morb. Mort. Wkly. Rep., 54, 36, January 21, 2005.
6. Fiorino, D. Making Environmental Policy, University of California Press, Berkeley,
CA, 1995, 25–28.
7. U.S. EPA. Air Quality Criteria for Carbon Monoxide, Report No. EPA 600/P-99/001F,
Office of Research and Development, National Center for Environmental Assessment,
U.S. Environmental Protection Agency, Washington, D.C., 2000.
8. Raub, J. Environmental Health Criteria 213: Carbon Monoxide, 2nd ed., World
Health Organization, Geneva, Switzerland, 1999.
9. Arias, E., Anderson, R.N., Kung, H.-C., Murphy, S.L., and Kochanek, K.D. Deaths:
final data for 2001, National Vital Statistics Reports, 52(3), Center for Health
Statistics, Hyattsville, MD, 2003.
10. WHO. Carbon monoxide, In Air Quality Guidelines for Europe, WHO Regional
Publications, European Series No. 23, World Health Organization, Regional Office
for Europe, Copenhagen, 1987, 210–220.
8417: “8417_c002” — 2007/9/11 — 12:16 — page 34 — #30

11. WHO. Updating and revision of the air quality guidelines for Europe, Meeting of
the WHO Working Group “Classical” Air Pollutants, Report No. EUR/ICP/EHAZ
94 05/PB01, World Health Organization, Regional Office for Europe, Copenhagen,
1995.
12. Romieu, I. Epidemiological studies of health effects arising from motor vehicle air
pollution, In Urban Traffic Pollution, Schwela, D. and Zali, O., eds., E & FN Spon,
London, 1999, chap. 2.
13. U.S. EPA. National Air Quality and Emissions Trends Report, 2003 Special Stud-
ies Edition, Report No. EPA 454/R-03-005, Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC,
2003.
14. Muller, P.O. Transportation and urban form: stages in the spatial evolution of the
American metropolis, In The Geography of Urban Transportation, 3rd ed., Hanson,
S. and Giuliano, G., eds., Guilford Press, New York, 2004, chap. 3.
15. U.S. Code of Federal Regulations, 40 CFR 50.1(e).
16. National Research Council. Managing Carbon Monoxide Pollution in Meteorolo-
gical and Topographical Problem Areas, National Academies Press, Washington,
D.C., 2003.
17. Blumenthal, D. The use of real-time air quality data in daily forecasting and decision-
making, EM: The Magazine for Environmental Managers, September, 2005, 18.
18. U.S. EPA. National Air Quality and Emissions Trends Report, 1999, Report No.
EPA/454/R-01-004, Office of Air Quality Planning and Standards, U.S. Environ-
mental Protection Agency, Research Triangle Park, NC, 2001.
19. U.S. EPA. National Air Quality and Emissions Trends Report, 1996, Report No.
EPA/454/R-97-013, Office of Air Quality Planning and Standards, U.S. Environ-
mental Protection Agency, Research Triangle Park, NC, 1998.
20. See: http://www.epa.gov/oar/oaqps/greenbk/cmsum.
21. See: http://www.epa.gov/oar/oaqps/greenbk/cntc.
22. Zartarian, V.G., Ott, W.R., and Duan, N. A quantitative definition of exposure and
related concepts, J. Expo. Anal. Environ. Epid., 7, 411, 1997.
23. Wilcosky, T. and Simonsen, N.R. Solvent exposure and cardiovascular disease, Am.
J. Ind. Med., 19, 569, 1991.
24. Agency for Toxic Substances and Disease Registry. Methylene chloride toxicity, Am.
Fam. Phy., 47, 1159, 1993.
25. Ott, W. Concepts of human exposure to air pollution, Environ. Int., 7, 179, 1982.
26. Council on Environmental Quality. Environmental Quality - 1980: The Eleventh
Annual Report of the Council on Environmental Quality, U.S. Government Printing
Office, Washington, D.C., 1980, 27–40.
27. Ott, W. An urban survey technique for measuring the spatial variation of carbon
monoxide concentrations in cities, Ph.D. thesis, Stanford University, Stanford, CA,
1971.
28. Ott, W. and Eliassen, R. A survey technique for determining the representativeness of
urban air monitoring stations with respect to carbon monoxide, J. Air Pollut. Control
Assoc., 23, 685, 1973.
29. Cortese, A.D. and Spengler, J.D. Ability of fixed monitoring stations to represent
personal carbon monoxide exposure, J. Air Pollut. Control Assoc., 26, 1144, 1976.
30. Chapin, F.S., Jr. Human Activity Patterns in the City, Wiley-Interscience,
New York, 1974.
31. Szalai, A., ed. The Use of Time: Daily Activities of Urban and Suburban Populations
in Twelve Countries, Mouton and Co., The Hague, The Netherlands, 1972.
8417: “8417_c002” — 2007/9/11 — 12:16 — page 35 — #31

32. Yocom, J.E., Clink, W.L., and Cote, W.A. Indoor/outdoor air quality relationships,
J. Air Pollut. Control Assoc., 21, 251, 1971.
33. General Electric Company. Indoor–Outdoor Carbon Monoxide Pollution Study,
Report No. NTIS-PB220428, National Technical Information Service, Springfield,
VA, 1972.
34. Godin, G., Wright, G., and Shephard, R. Urban exposure to carbon monoxide, Arch.
Environ. Health, 25, 305, 1972.
35. Meyer, B. Indoor Air Quality, Addison-Wesley, Reading, MA, 1983, 159–160.
36. Spengler, J. and Soczeck, M. Evidence for improved ambient air quality and the need
for personal exposure research, Environ. Sci. Technol., 18, 272A, 1984.
37. Wallace, L.A. and Ott, W.R. Personal monitors: a state-of-the-art survey, J. Air Pollut.
Control Assoc., 32, 601, 1982.
38. Petersen, G.A. and Sabersky, R.H. Measurement of pollutants inside an automobile,
J. Air Pollut. Control Assoc., 25, 1028, 1975.
39. Ott, W.R. and Flachsbart, P.G. Measurement of carbon monoxide concentrations in
indoor and outdoor locations using personal exposure monitors, Environ. Int., 8,
295, 1982.
40. Fugas, M. Assessment of total exposure to an air pollutant, Paper No. 38-5 in Proc.
Int. Conf. Environ. Sensing and Assess., Vol. 2, Las Vegas, NV, 1975, 1.
41. Duan, N. Models for human exposure to air pollution, Environ. Int., 8, 305, 1982.
42. Padgett, J. and Richmond, H. The process of establishing and revising national
ambient air quality standards, J. Air Pollut. Control Assoc., 33, 13, 1983.
43. Jordan, B.C., Richmond, H.M., and McCurdy, T. The use of scientific information in
setting ambient air standards, Environ. Health Perspectives, 52, 233, 1983.
44. Federal Register. Carbon monoxide; proposed revisions to the national ambient air
quality standards: proposed rule, 45, 55066, August 18, 1980.
45. Akland, G.G., Hartwell, T.D., Johnson, T.R., and Whitmore, R.W. Monitoring human
exposure to carbon monoxide in Washington, D.C., and Denver, Colorado, during the
winter of 1982–1983, Environ. Sci. Technol., 19, 911, 1985.
46. Johnson, T. A Study of Personal Exposure to Carbon Monoxide in Denver, Colorado,
Report No. EPA-600/4/84-014, Environmental Monitoring Systems Laboratory, U.S.
Environmental Protection Agency, Research Triangle Park, NC, 1984.
47. U.S. EPA. Air Quality Criteria for Carbon Monoxide, Report No. EPA 600/8-90/045F,
Office of Health and Environmental Assessment, Environmental Criteria and Assess-
ment Office, U.S. Environmental Protection Agency, Research Triangle Park, NC,
1991.
48. Ott, W.R., Thomas, J., Mage, D.T., and Wallace, L. Validation of the simulation of
human activity and pollutant exposure (SHAPE) model using paired days from the
Denver, CO, carbon monoxide field study, Atmos. Environ., 22, 2101, 1988.
49. Law, P.L., Lioy, P.J., Zelenka, M.P., Huber, A.H., and McCurdy, T.R. Evaluation
of a probabilistic exposure model applied to carbon monoxide (pNEM/CO) using
Denver personal exposure monitoring data, J. Air Waste Manage. Assoc., 47, 491,
1997.
50. Klepeis, N.E., Nelson, W.C., Ott, W.R., Robinson, J.P., Tsang, A.M., Switzer, P.,
Behar, J.V., Hern, S.C., and Engelmann, W.H. The national human activity pat-
tern survey (NHAPS): a resource for assessing exposure to environmental pollutants,
J. Expo. Anal. Environ. Epid., 11, 231, 2001.
51. Koontz, M.D. and Nagda, N.L. A Topical Report on a Field Monitoring Study of
Homes with Unvented Gas Space Heaters: Volume III, Methodology and Results,
Contract No. 5083-251-0941, Gas Research Institute, Chicago, 1988.
8417: “8417_c002” — 2007/9/11 — 12:16 — page 36 — #32

52. Hampson, N.B., Kramer, C.C., Dunford, R.G., and Norkool, D.M. Carbon monoxide
poisoning from indoor burning of charcoal briquets, JAMA: J. Am. Med. Assoc., 271,
52, 1994.
53. Yates, M.W. A preliminary study of carbon monoxide gas in the home, J. Environ.
Health, 29, 413, 1967.
54. Gasman, J.D., Varon, J., and Gardner, J.P. Revenge of the barbecue grill: carbon
monoxide poisoning, West. J. Med., 153, 656, 1990.
55. Mumford, J.L., Williams, R.W., Walsh, D.B., Burton, R.M., Svendsgaard, D.J.,
Chuang, J.C., Houk, V.S., and Lewtas, J. Indoor air pollutants from unvented kerosene
heater emissions in mobile homes: studies on particles, semivolatile organics, carbon
monoxide, and mutagenicity, Environ. Sci. Technol., 25, 1732, 1991.
56. Williams, R., Walsh, D., White, J., Jackson, M., and Mumford, J. Effect on carbon
monoxide levels in mobile homes using unvented kerosene heaters for residential
heating, Indoor Environ., 1, 272, 1992.
57. Wilson, A.L., Colome, S.D., and Tian, Y. California Residential Indoor Air Quality
Study, Volume 1 - Methodology and Descriptive Statistics, Integrated Environmental
Services, Irvine, CA, 1993.
58. Wilson, A.L., Colome, S.D., and Tian, Y. California Residential Indoor Air Quality
Study, Volume 1 - Methodology and Descriptive Statistics, Appendix, Integrated
Environmental Services, Irvine, CA, 1993.
59. Colome, S.D., Wilson, A.L., and Tian, Y. California Residential Indoor Air Qual-
ity Study, Volume II - Carbon Monoxide and Air Exchange Rate: A Univariate &
Multivariate Analysis, Integrated Environmental Services, Irvine, CA, 1994.
60. Liu, K., Girman, J.R., Hayward, S.B., Shusterman, D., and Chang, Y. Unintentional
carbon monoxide deaths in California from charcoal grills and hibachis, J. Expo. Anal.
Environ. Epid., 3, 143, 1993.
61. Girman, J.R., Chang, Y.-L., Hayward, S.B., and Liu, K.-S. Causes of unintentional
deaths from carbon monoxide poisonings in California, West. J. Med., 168, 158, 1998.
62. Yoon, S.S., Macdonald, S.C., and Parrish, R.G. Deaths from unintentional carbon
monoxide poisoning and potential for prevention with carbon monoxide detectors,
JAMA: J. Am. Med. Assoc., 279, 685, 1998.
63. National Center for Health Statistics and U.S. Consumer Product Safety Commission.
Death certificate file, U.S. Department of Health and Human Services, Public Health
Service, Centers for Disease Control and Prevention, Atlanta, GA, 1992.
64. National Institute for Occupational Safety and Health. Preventing Carbon Monoxide
Poisoning from Small Gasoline-Powered Engines and Tools [NIOSH alert], Report
No. DHHS (NIOSH) 96–118, Public Health Service, U.S. Department of Health and
Human Services, Cincinnati, OH, 1996.
65. Tucker, M., Eichold, B., Lofgren, J.P., Holmes, I., Irvin, D., Villanacci, J., Ryan, J.,
et al. Carbon monoxide poisonings after two major hurricanes—Alabama and Texas,
August – October 2005, MMWR: Morb. Mort. Wkly. Rep., 55, 236, 2006.
66. Sniffen, J.C., Cooper, T.W., Johnson, D., Blackmore, C., Patel, P., Harduar-Morano,
L., Sanderson, R., et al. Carbon monoxide poisoning from hurricane-associated use
of portable generators - FL, 2004, MMWR: Morb. Mort. Wkly. Rep., 54, 697, 2005.
67. Pedersen, D.H., and Sieber, W.K. National Occupational Exposure Survey, Volume
III: Analysis of Management Interview Responses, Report No. PB90-193640, U.S.
Department of Health and Human Services, National Institute for Occupational Safety
and Health, National Technical Information Service, Springfield, VA, 1988.
68. Bellin, P. and Spengler, J.D. Indoor and outdoor carbon monoxide measurements at
an airport, J. Air Pollut. Control Assoc., 30, 392, 1980.
8417: “8417_c002” — 2007/9/11 — 12:16 — page 37 — #33

69. McCammon, C.S., Halperin, W.F., and Lemin, R.A. Carbon monoxide exposure from
aircraft fueling vehicles, Arch. Environ. Health, 36, 136, 1981.
70. Limasset, J., Diebold, F., and Hubert, G. Assessment of bus drivers’ exposure to the
pollutants of urban traffic, Sci. Total Environ., 134, 39, 1993.
71. Kahler, M., Kuhse, W., and Wintermeyer, L.A. Unintentional carbon monoxide pois-
oning from indoor use of pressure washers - Iowa, January 1992–January 1993,
MMWR: Morb. Mort. Wkly. Rep., 42, 777, 1993.
72. Bünger, J., Bombosch, F., Mesecke, U., and Hallier, E. Monitoring and analysis
of occupational exposure to chain saw exhausts, Am. Ind. Hygiene Assoc. J., 58,
747, 1997.
73. Baucom, C.D., Freeman, J.I., and MacCormack, J.M. Carbon monoxide poisoning in
a garment-manufacturing plant - North Carolina, MMWR: Morb. Mort. Wkly. Rep.,
36, 543, 1987.
74. Fawcett, T.A., Moon, R.E., Francia, P.J., Mebane, G.Y., Theil, D.R., and Piantadosi,
C.A. Warehouse workers headache: carbon monoxide poisoning from propane-fueled
forklifts, J. Occup. Med., 34, 12, 1992.
75. Ely, E.W., Moorehead, B., and Haponik, E.F. Warehouse workers’ headache: emer-
gency evaluation and management of 30 patients with carbon monoxide poisoning,
Am. J. Med., 98, 145, 1995.
76. McCammon, J.B., McKenzie, L.E., Heinzman, M. Carbon monoxide poisoning
related to the indoor use of propane-fueled forklifts in Colorado workplaces, Appl.
Occup. Environ. Hyg., 11, 192, 1996.
77. Apte, M. A population-based exposure assessment methodology for carbon monox-
ide: development of a carbon monoxide passive sampler and occupational dosimeter,
Ph.D. thesis, University of California, Berkeley, 1997.
78. Apte, M.G., Cox, D.D., Hammond, S.K., and Gundel, L.A. A new carbon monoxide
occupational dosimeter: results from a worker exposure assessment survey, J. Expo.
Anal. Environ. Epid., 9, 546, 1999.
79. Gourdeau, P., Parent, M., and Soulard, A. Exposition à l’oxyde de carbone dans les
garages d’automobiles: évaluation chez les mécaniciens, Canada J. of Public Health,
86, 414, 1995.
80. Boeniger, M.F. Description of Facilities and Employee Exposure to Carbon
Monoxide at Four Different Bridge and Tunnel Toll Authorities in the United
States, Report No. IWS-45-02, National Institute for Occupational Safety and
Health, Division of Surveillance, Field Studies, Hazard Evaluations, Cincinnati,
OH, 1995.
81. Raaschou-Nielsen, O., Nielsen, M.L., and Gehl, J. Traffic-related air pollution: expos-
ure and health effects in Copenhagen street cleaners and cemetery workers, Arch.
82. Kamei, M. and Yanagisawa, Y. Estimation of CO exposure of road construction
workers in tunnel, Ind. Health, 35, 119, 1997.
83. Flachsbart, P.G. and Ott, W.R. A rapid method for surveying CO concentrations in
high-rise buildings, Environ. Int., 12, 255, 1986.
84. Flachsbart, P.G. and Brown, D.E. Surveys of Personal Exposure to Vehicle Exhaust
in Honolulu Microenvironments., Department of Urban and Regional Planning,
University of Hawaii at Manoa, Honolulu, HI, 1985.
85. Flachsbart, P.G. and Brown, D.E. Employee exposure to motor vehicle exhaust at a
Honolulu shopping center, J. Archit. Plan. Res., 6, 19, 1989.
86. Bach, W. and Lennon, K. Air pollution and health at Ala Moana Shopping Center in
Honolulu, Hawaii Med. J., 31, 104, 1972.
8417: “8417_c002” — 2007/9/11 — 12:16 — page 38 — #34

87. Simeone, L.F. The Intrusion of Engine Exhaust into the Passenger Areas of Recre-
ational Power Boats, Report No. DOT-VNTSC-CG-91-1, John A. Volpe National
Transportation Systems Center, Cambridge, MA, 1991.
88. Snook, L.M. An investigation of driver exposure to carbon monoxide while traveling
in the wake of a snowmobile, Ph.D. thesis, University of Tennessee, Knoxville,
1996.
89. Wilkinson, T. Snowed under, National Parks, 69, 32, 1995.
90. Snook, L.M. and Davis, W.T. An investigation of driver exposure to carbon monoxide
while traveling in the wake of a snowmobile, Paper No. 97-RP143.02 presented at the
90th Annual Meeting of the Air & Waste Management Association, Toronto, Ontario,
Canada, 1997.
91. Kwok, P.W. Reduction of carbon monoxide in indoor skating arenas, Environ. Health
Rev., 25, 60, 1981.
92. Sorensen, A.J. The importance of monitoring carbon monoxide levels in indoor ice
skating rinks, J. Am. Coll. Health, 34, 185, 1986.
93. Miller, R.K., Ryan, M.C., and Bilowus, P. Carbon monoxide poisoning in indoor ice
skating arenas, Va. Med., 116, 74, 1989.
94. Lévesque, B., Dewailly, E., Lavoie, R., Prud’Homme, D., and Allaire, S. Carbon
monoxide in indoor ice skating rinks: evaluation of absorption by adult hockey players,
Am. J. Public Health, 80, 594, 1990.
95. Lévesque, B., Lavoie, R., Dewailly, E., Prud’Homme, D., and Allaire, S. An experi-
ment to evaluate carbon monoxide absorption by hockey players in ice skating rinks,
Vet. Hum. Toxicol., 33, 5, 1991.
96. Lee, K., Yanagisawa, Y., and Spengler, J.D. Carbon monoxide and nitrogen dioxide
levels in an indoor ice skating rink with mitigation methods, Air Waste, 43, 769, 1993.
97. Lee, K., Yanagisawa, Y., Spengler, J.D., and Nakai, S. Carbon monoxide and nitrogen
dioxide exposures in indoor ice skating rinks, J. Sports Sci., 12, 279, 1994.
98. Spengler, J.D., Stone, K.R., and Lilley, F.W. High carbon monoxide levels measured
in enclosed skating rinks, J. Air Pollut. Control Assoc., 28, 776, 1978.
99. Paulozzi, L.J., Satink, F., and Spengler, R.F. A carbon monoxide mass poisoning at
an ice arena in Vermont, Am. J. Public Health, 81, 222, 1991.
100. Smith, W., Anderson, T., Anderson, H.A., and Remington, P.L. Nitrogen dioxide and
carbon monoxide intoxication in an indoor ice arena - Wisconsin, MMWR: Morb.
Mort. Wkly. Rep., 41, 383, 1992.
101. Paulozzi, L.J., Spengler, R.F., Vogt, R.L., and Carney, J.K. A survey of carbon monox-
ide and nitrogen dioxide in indoor ice areas in Vermont, J. Environ. Health, 56, 23,
1993.
102. Hampson, N.B. Carbon monoxide poisoning at an indoor ice arena and bingo hall -
Seattle, 1996 [reprint of MMWR: Morb. Mort. Wkly. Rep., 45, 265, 1996] JAMA: J.
Am. Med. Assoc., 275, 1468, 1996.
103. Boudreau, D.R., Spadafora, M.P., Wolf, L.R., and Siegel, E. Carbon monoxide levels
during indoor sporting events - Cincinnati, 1992–1993 [reprint of MMWR: Morb.
Mort. Wkly. Rep., 43, 21] J. Am. Med. Assoc., 271, 419, 1994.
104. Luckurst, D.G. and Solkoski, G. Carbon monoxide levels in indoor “tractor-pull
events”—Manitoba, Can. Dis. Wkly. Rep., 16, 79, 1990.
105. Lévesque, B., Allaire, S., Prud’Homme, D., Rhainds, M., Lebel, G., Bellemarre,
D., and Dupuis, K. Indoor motocross competitions: air quality evaluation, Am. Ind.
Hygiene Assoc. J., 58, 286, 1997.
106. Colvile, R.N., Hutchinson, E.J., Mindell, J.S., and Warren, R.F. The transport sector
as a source of air pollution, Atmos. Environ., 35, 1537, 2001.
8417: “8417_c002” — 2007/9/11 — 12:16 — page 39 — #35

107. Petersen, W. and Allen, R. Carbon monoxide exposures to Los Angeles area
commuters, J. Air Pollut. Control Assoc., 32, 826, 1982.
108. Shikiya, D., Liu, C., Kahn, M., Juarros, J., and Barcikowski, W. In-vehicle Air
Toxics Characterization Study in the South Coast Air Basin, South Coast Air Quality
Management District, El Monte, CA, 1989.
109. Kleiner, B.C. and Spengler, J.D. Carbon monoxide exposures of Boston bicyclists, J.
Air Pollut. Control Assoc., 26, 147, 1976.
110. Holland, D.M. Carbon monoxide levels in microenvironment types of four U.S. cities,
Environ. Int., 9, 369, 1983.
111. Flachsbart, P.G., Mack, G.A., Howes, J.E., and Rodes, C.E. Carbon monoxide
exposures of Washington commuters, J. Air Pollut. Control Assoc., 37, 135, 1987.
112. Chan, C.-C., Ozkaynak, H., Spengler, J.D., and Sheldon, L. Driver exposure to
volatile organic compounds, CO, ozone, and NO2 under different driving conditions,
Environ. Sci. Technol., 25, 964, 1991.
113. Flachsbart, P., Ott, W., and Switzer, P. Long-term trends in exposure to carbon
monoxide on a California arterial highway, Paper No. 69237 presented at the 96th
Annual Meeting of the Air & Waste Management Association, San Diego, CA,
2003.
114. Ott, WR., Mage, D.T., and Thomas, J. Comparison of microenvironmental CO con-
centrations in two cities for human exposure modeling, J. Expo. Anal. Environ. Epid.,
2, 249, 1992.
115. Amiro, A. CO presents public health problems, J. Environ. Health, 32, 83, 1969.
116. Clements, J. School Bus Carbon Monoxide Intrusion, Report No. DOT-HS-803-705,
National Highway Traffic Safety Administration, U.S. Department of Transportation,
Washington, D.C., 1978.
117. Ziskind, R., Rogozen, M., Carlin, T., and Drago, R. Carbon monoxide intrusion into
sustained-use vehicles, Environ. Int., 5, 109, 1981.
118. Hampson, N.B. and Norkool, D.M. Carbon monoxide poisoning in children riding in
the back of pickup trucks, JAMA: J. Am. Med. Assoc., 267, 538, 1992.
119. Trompeo, G., Turletti, G., and Giarusso, O. Concentrations of carbon monoxide in
underground garages, Rassegna di Medicina Industriale, 33, 392, 1964.
120. Chovin, P. CO, analysis of exhaust gas investigations in Paris, Environ. Res., 1, 198,
1967.
121. Goldsmith, J. Contribution of motor vehicle exhaust, industry, and cigarette smoking
to community CO exposures, Ann. NY Acad. Sci., 15, 122, 1970.
122. Amendola, A. and Hanes, N. Characterization of indoor carbon monoxide levels
produced by the automobile, In Indoor Air Volume 4, Chemical Characterization
and Personal Exposure, NTIS Report No. PB85-104214, Berglund, B. et al., eds.,
Swedish Council for Building Research, Stockholm, 1984, 97–102.
123. Wilson, A.L., Colome, S.D., and Tian, Y. Air Toxics Microenvironment Exposure
and Monitoring Study, South Coast Air Quality Management District, El Monte, CA,
1991.
124. Myronuk, D. Augmented ingestion of carbon monoxide and sulfur oxides by occu-
pants of vehicles while idling in drive-up facility lines, Water Air Soil Pollut., 7, 203,
1977.
125. Wheatley, A.D., Sadhra, S., and Beach, J.R. Exposure to toxic gas and particle phase
pollutants evolved during deployment of airbags in vehicles, Indoor Environ., 6, 134,
1997.
126. Flachsbart, P.G. Human exposure to carbon monoxide from mobile sources, In Carbon
Monoxide, Khalil, M.A.K., Pinto, J.P., and Shearer, M.J., eds. [special issue of papers
8417: “8417_c002” — 2007/9/11 — 12:16 — page 40 — #36

from an international conference, December, 1997, Portland, OR] Chemosphere:

Global Change Science, 1, 301, 1999.
127. Transportation Research Board. Expanding Metropolitan Highways: Implications for
Air Quality and Energy Use, National Academy Press, Washington, D.C., 1995.
128. Flachsbart, P.G. Long-term trends in United States highway emissions, ambient con-
centrations, and in-vehicle exposure to carbon monoxide in traffic, J. Expo. Anal.
Environ. Epid., 5, 473, 1995.
129. Haagen-Smit, A.J. Carbon monoxide levels in city driving, Arch. Environ. Health,
12, 548, 1966.
130. Lawryk, N.J., Lioy, P.J., and Weisel, C.P. Exposure to volatile organic com-
pounds in the passenger compartment of automobiles during periods of normal and
malfunctioning operation, J. Expo. Anal. Environ. Epid., 5, 511, 1995.
131. Systems Applications, Inc. Human Exposure to Atmospheric Concentrations of
Selected Chemicals, Vol. II, Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency, Research Triangle Park, NC, 1983.
132. Stewart, R.D., Fisher, T.N., Hosko, M.J., Peterson, J.E., Baretta, E.D., and Dodd, H.
C. Experimental human exposure to methylene chloride, Arch. Environ. Health, 25,
342, 1972.
133. Stewart, R.D. and Hake, C.L. Paint-remover hazard, JAMA: J. Am. Med. Assoc., 235,
398, 1976.
134. Buie, S.E., Pratt, D.S., and May, J.J. Diffuse pulmonary injury following paint
remover exposure, Am. J. Med., 81, 702, 1986.
135. Peterson, J.E. Modeling the uptake, metabolism and excretion of dichloromethane by
man, Am. Ind. Hyg. Assoc. J., 39, 41, 1978.
136. ACGIH. Threshold Limit Values for Chemical Substances and PhysicalAgents: Biolo-
gical Exposure Indices, American Conference of Governmental Industrial Hygienists,
Cincinnati, OH, 1996.
137. U.S. EPA. Health Assessment Document for Dichloromethane Methylene Chloride,
Report No. EPA/600/8-82/004F, Office of Health and Environmental Assessment,
U.S. Environmental Protection Agency, Research Triangle Park, NC, 1985.
138. Shusterman, D., Quinlan, P., Lowengart, R., and Cone, J. Methylene chloride intoxic-
ation in a furniture refinisher: a comparison of exposure estimates utilizing workplace
air sampling and blood carboxyhemoglobin measurements, J. Occup. Med., 32, 451,
1990.
139. Ghittori, S., Marraccini, P., Franco, G., and Imbriani, M. Methylene chloride exposure
in industrial workers, Am. Ind. Hyg. Assoc. J., 54, 27, 1993.
140. Leikin, J.B., Kaufman, D., Lipscomb, J.W., Burda, A.M., and Hryhorczuk, D.O.
Methylene chloride: report of five exposures and two deaths, Am. J. Emerg. Med., 8,
534, 1990.
8417: “8417_c002” — 2007/9/11 — 12:16 — page 41 — #37

8417: “8417_c002” — 2007/9/11 — 12:16 — page 42 — #38
3 Carbon Monoxide
Build-Up in Houses and
Small Volume Enclosures
Robert E. Engberg
CONTENTS
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.2 Carbon Monoxide Safety and Monitoring Devices . . . . . . . . . . . . . . . . . . . . . . . 44
3.3 Metabolic Uptake of Oxygen and Room Air Ventilation for Humans . . . 45
3.4 The Tracer Gas Model for Predicting CO Accumulation in Structures . . 47
3.5 Ventilation Requirements for Humans in Small Structures . . . . . . . . . . . . . . . 48
3.6 Oxygen Depletion Sensor Applications to Small Burners . . . . . . . . . . . . . . . . 48
3.7 Combustion Equations for Gas Burners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.8 Carbon Monoxide Accumulation Due to Small Engine Exhaust Gases . . 53
3.9 OSHA and Industry Standards for CO Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.10 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
3.1 INTRODUCTION
The problem of carbon monoxide (CO) production within habitable enclosures (i.e.,
breathing spaces) has been a chronic problem since the use of hydrocarbon fuels
became widespread. We know that CO generation from these fuel sources is com-
monly caused by the lack of, or starvation of, adequate oxygen during the combustion
process whether in a heating appliance, engine, or lighting fixture. The creation of
CO becomes a problem when it enters an occupied space used by living creatures,
either by accident or by design.
Humans cannot sense such occult emissions until the blood saturation with CO
begins to affect coordination and mental state. CO is transparent and colorless to
the occupant of the space, as well as being odorless and nonirritative. When CO is
discovered in significant concentration in a breathing space, the occupants may suffer
sickness and even death. The presence of CO is not usually suspected in an enclosed
space until dangerous levels are reached.
CO generation from burning hydrocarbon fuels in combustion devices is due most
often to improper fuel–air combustion ratios. This can be caused by oxygen starvation
43
8417: “8417_c003” — 2007/9/11 — 12:16 — page 43 — #1

at the burner and by improper setting of the fuel–air ratio at the appliance. Testing
of combustion devices installed in small enclosures reveals that 4–5 air changes per
hour (ACH) is necessary to allow the device to operate safely and not generate a CO
concentration of greater than 100 parts per million (ppm). Experience shows that the
fuel–air ratio of the burner can be easily upset by lack of inspection and maintenance of
the combustion device. Primary air adjustment shutters are commonly obstructed by
the collection of dust and lint, while combustion air passages can become restricted
or exhaust ventilation ducts plugged. This causes the production of CO in closed
spaces.
It was not until the 1980s that the use of burner safety devices were legislated and
used within the United States to attack the problem of CO deaths and injuries in closed
breathing spaces. During the past 10 years, manufacturers of small portable heating
appliances have designed and installed burner safety devices to affect shutdown under
ambient conditions that might produce dangerous levels of CO.
The most common and simple safety device built into fuel consuming appliances
has been the oxygen depletion sensor (ODS). It is incorporated into the flame safety
system of most burners. A typical ODS consists of a thermocouple that senses flame
outage and acts to shut down the fuel valve and the pilot flame. This combination
makes up the typical flame safety and ODS assembly on a fuel-burning device. A
second option has been to incorporate a CO sensor into devices such as gas engines.
The past decades have brought national and state codes and standards to bear in
an effort to eliminate CO hazards in fuel burning devices, and in building construc-
tion standards and techniques. The American National Standards Institute (ANSI),
the National Fire Protection Association (NFPA), the Consumer Product Safety Com-
mission (CPSC), Underwriters Laboratory (UL), Canadian Gas Association (CGA)
and the Occupational Safety and Health Administration (OSHA) are at the forefront
as agencies that write and enforce CO safety device requirements in commercial
products and building construction. One example of a standard is the requirement
for minimum exhaust concentration limits generated within a test enclosure that is
heated with a liquefied petroleum (LP) gas fueled, infrared (IR) camp type heaters
(ANSI and CGA).
Technology has greatly advanced in CO monitoring techniques and the meas-
urement of environmental conditions. We can now estimate the oxygen uptake
of a human at a given activity level in a closed space, in the presence of CO
being produced by a malfunctioning heater operating under specified air exchange
and combustion rates. We will discuss each of these techniques in the following
sections.
3.2 CARBON MONOXIDE SAFETY AND

MONITORING DEVICES
It has become commonplace to install CO sensors in breathing spaces that have
the potential of accumulating CO due to fuel combustion by heating appliances and
engines. CO alarms are available at most hardware and home supply stores. The law
requires CO monitors in most new house and apartment construction. For the UL
8417: “8417_c003” — 2007/9/11 — 12:16 — page 44 — #2

Carbon Monoxide Build-Up in Houses and Small Volume Enclosures 45
standards governing the set points of common CO alarms see Chapter 28 in this
book. The way UL has set up the standards, lower CO concentrations take longer to
produce audible alarm while higher CO concentrations trip the alarm in less time.
The intent is to roughly model the rate of CO uptake by humans. Other types of CO
monitors are designed for marine engine rooms, industrial environments, and so forth.
Monitoring and data logging CO sensors are available for personal use, where
walk around monitoring of an indoor environment may be needed. These units are
available in pocket-sized and handheld models that record data from the environment
at preset time intervals and store the data for downloading to computer for later display
and interpretation.
As stated above, the ODS has become popular for the safe shutdown of burner
systems. The device has a calibrated shutdown feature that gauges ambient oxygen
concentration by changing flame shape. The predetermined pilot flame will degrade
in shape with the reduction of ambient oxygen. When this happens, the flame moves
off the thermocouple, causing thermocouple output current to decline, which trips a
gas valve, shutting down the burner. This principle was developed in Europe. The
ODS has been a common safety device required on gas heating appliances since
the 1970s.
The pilot flame is a common safety device used with gas fueled appliances. It has
been used successfully for many decades in the United States and Europe. It assures
the ignition of any gas fumes or vapors around the burner and serves to ignite the
burner. Its continued operation is also monitored through the use of a thermocouple
that senses pilot flame heat. Pilot flames may be ignited either by a match or the ever-
popular piezoelectric spark igniter. The latter has become an added safety feature on
gas burning appliances, since no open continuously burning flame is present, which
might present a fire hazard and also consumes fuel over time.
Exhaust stack gas analyzers are another type of CO sensor used to monitor boiler
burner combustion efficiency. These handheld analyzers and monitors can be used to
data-log exhaust stack conditions in terms of CO, oxygen, and carbon dioxide. They
may also monitor other burner characteristics such as temperature, draft, excess air,
and unburned hydrocarbons. These units also allow burner overall efficiency to be
computed in a continuous fashion.
Placing a thermocouple near the burner screen on an LP-fueled camp heaters is
another safety device. One characteristic of some infrared heaters is the variation
of burner screen temperature with respect to declining oxygen levels in a heated
breathing space. The manufacturer may chose to design the screen-thermocouple
such that it will shut down the burner with decreasing oxygen levels. This will occur
before burner CO production is excessive. The details of this option are discussed
later in this chapter.
3.3 METABOLIC UPTAKE OF OXYGEN AND ROOM

AIR VENTILATION FOR HUMANS
A major concern for humans occupying a small breathing space is the provision for
adequate ventilation of that space for creature comfort, health, and safety. One design
8417: “8417_c003” — 2007/9/11 — 12:16 — page 45 — #3

question is just how much oxygen is consumed by the human body and how much
ventilation is needed. To do that we must estimate the oxygen uptake of the human
body. Coupled with this concern will be determining the oxygen depletion within the
space that can be expected or allowed. For a description of the metabolic variables
necessary for this calculation see Reference 1. All of these variables and concerns are
best addressed by an example of a typical breathing space and occupant model. Such
an example might be two hunters sleeping in a camp tent.
The parameters that will be assumed for this calculation are as follows:
1. A tent will be of a cabin-type design, with nominal floor dimensions of

7.5 × 9.5 × 6.0 ft. (w × l × h), respectively, or 235 cu. ft., approximately.
The tent will have been treated with a water repellent coating.
2. Two men sleeping and each weighing 200 pounds, no outdoor wind, and
all tent windows and opening flaps closed.
3. A propane fueled, infrared heater set at a low heat setting of 8000 BTU/h,
using commercial propane with a higher heating value (HHV) of
2550 BTU/cu. ft.2
4. Ambient air having 20.9% oxygen, and a design lower limit of 18.5%
oxygen, where CO begins to be generated when using a typical camp
heater.
The mathematical variables used in the following equations and calculations are
defined as follows with their associated units denoted in [… ..]:
ACH = Vm /Ve , where Vm = minimum rate of fresh airflow per person,

[cu. ft./h] and Ve = volume of the enclosure [cu. ft.].
M = metabolic rate (met) for a human is given as, 1 met = 18.4 [BTU/h *
sq. ft.].
And M = 567(0.23∗ RQ+0.77)∗ QO2 /Ad, where RQ = 0.83 for light and sedent-
ary activity; QO2 = volumetric rate of oxygen consumption at conditions (STPD) of
32◦ F, 14.7 psi (std. atm.) [cu. ft./h], and for a sleeping man the value to be used is
M = 13 BTU/h*sq. ft. Now, solving for the oxygen uptake:
QO2 = 13∗ 19.6/567∗ (0.23∗ 0.83 + 0.77) = 0.47 cu. ft./h.
So, for two men, the oxygen uptake in the tent will be 0.94 cu. ft./h.
Now, from our defined limits of minimum oxygen reduction in this space, we have:
QO2 avail = (0.209 − 0.185)∗ V = 0.024∗ 235 = 5.64 cu. ft. of oxygen available. We
next consider the oxygen uptake by the LP, infrared camp heater set at the “low” heat
position, with a fuel flow as defined above. Using the HHV for propane vapor as
2550 BTU/cu. ft. (Reference 2, p. 24) and the stoichiometric combustion of propane,
(volume of oxygen reacting for each volume of LP vapor is 5), we have:
VO2 /VLP = 5
Therefore, when operating at the ‘low’ heat setting of 8000 BTU/h, we get: VLP =
8000/2550 = 3.137 cu. ft./h of propane used.
8417: “8417_c003” — 2007/9/11 — 12:16 — page 46 — #4

Next, the volume of oxygen utilized by the heater becomes VO2 = 5∗ 3.137 =
15.69 cu. ft./h. The time (t) needed to bring the oxygen level down to the critical
minimum level at the onset of CO production will be: t = QO2 avail /(VO2 +2∗QO2 ) =
5.64/(15.69 + 0.94) = 5.64/16.62 = 0.339 h or 20.4 min. We can expect that
after this amount of time spent in the tent, CO will begin to be generated by the
heater assuming that no air exchange, or infiltration, has occurred during this period
of time.
The next job is to estimate the amount of infiltration air (ACH) that will be
needed to maintain the oxygen level within the tent above 18.5%, remembering that
the heater and the two men are competing for the available oxygen. We first write
the equation that: QO2 (consumed) = QO2 (infiltrated) = 15.69 + 0.94) = 16.63
cu. ft./h = (0.209 − 0.185)∗ Vair , so, Vair = 16.63/0.024 = 692.7 cu. ft. of air
exchange. Thus, ACH = Vm /Ve = 692.7/235 = 2.95 air changes per hour are
needed to maintain an 18.5% level of oxygen in the tent. This value for ACH is in
good agreement with experimental values for ACH required for LP heaters to not
accumulate CO in closed breathing spaces.3
3.4 THE TRACER GAS MODEL FOR PREDICTING CO

ACCUMULATION IN STRUCTURES
This approach allows one to predict ACH for any enclosure. It is assumed that the
tracer gas is distributed uniformly in the initial charge of fugitive gas used for this
measurement. It is also assumed that air infiltration into the structure is uniformly
distributed throughout the space. The typical tracer gas used for this measurement is
sulfur hexafluoride (SF6 ). Its concentration (C) is described by the equation
Ct = Co e−kt and solving for kt, ln (Ct /Co ) = −kt
The latter equation indicates that a semilog plot of the quantity ln (Ct /Co ) versus
time will be linear with time and that the ACH, (k) will be equal to the slope of this
line. Since this line is expected to be linear, linear regression can be used to fit a line
to the data and the expression describing how well the line fits the data is R2 , where
R is the correlation coefficient. Assume that if testing yields an R2 greater than 0.9,
then the test is acceptable.
Now, if the ACH (k) is known, then the source strength (S) for CO generation of
the heater can be computed from a simple mass balance of CO in the chamber. The
equation predicting the CO generation rate is
V ∗ k[Ct(i+1) − Ct(i) ∗ e−k{ti+1 −ti } ]

St(i+1) = ,
[1 − e−k{ti+1 −ti } ]
where V = the tent volume; St(i+1) is the generation rate of CO at time ti+1 , k = ACH,
Ct(i+1) is the concentration of CO at time ti+1 and Ct(i) is the concentration of CO at
time ti . The above equation was derived based on the assumption that the air in the
8417: “8417_c003” — 2007/9/11 — 12:16 — page 47 — #5

chamber is well mixed and that CO is not absorbed or adsorbed into the tent material
or its contents.
3.5 VENTILATION REQUIREMENTS FOR HUMANS IN

SMALL STRUCTURES
The ventilation rate for a small enclosure is much the same as for larger spaces that are
designed for occupation by humans. The American Society of Heating, Refrigeration,
and Air Conditioning Engineers (ASHRAE) standard 62 specifies guidelines for the
minimum ventilation airflow rate for various habitable structures. A flow rate of 15
scfm per person is a common requirement for residential structures. This flow rate
will provide enough air to maintain the normal oxygen level for an average adult and
will remove most pollutants as they accumulate in typical spaces. To reiterate, the
oxygen depletion due to its uptake by two adults using a small camp heater within
a tent space requires a minimum of about 3–6 ACH to supply enough oxygen to
allow proper metabolism and to prevent the dangerous accumulation of CO. One
must look at the geometry of the tent to predict whether the ACH will be effective for
the occupants.
Please make reference to the ASHRAE Fundamentals Handbook 1 that describes
the dynamics of ventilation in a breathing space. It asks designers to incorporate cross
ventilation such that lower air levels are systematically exchanged at the prescribed
rate. This prevents vent gases or other pollutants from settling down and accumulating
in sleeping areas. The method of cross ventilation that incorporates a low level air
entry opening with an upper level exhaust or discharge opening of equal area is the
optimum design, as it promotes thermal differences and relative ambient winds to
power air exchange within the space. The designer should use, as a rule of thumb,
about 1 sq. in. per 1000 BTU/h of burner capacity. The definition of a confined space
is considered to be 50 cu. ft. for each 1000 BTU/h of input burner rating. No single
vent opening dimension should be less than 3 in.
This ventilation flow rate will translate to an ACH in the tent computation to be:
ACH = 15 cu. ft./min*60 min/h/235 cu. ft. = 3.83. We may compare this value
with the calculated ventilation needed to maintain the minimum oxygen level for
combustion in the above example of 2.95 ACH. Clearly, a ventilation rate to avoid
CO buildup was about 6.0 ACH as shown in Reference 3. This resulted in less than
100 ppm CO.
3.6 OXYGEN DEPLETION SENSOR APPLICATIONS TO

SMALL BURNERS
An ODS system is simply a pilot burner whose flame impinges on a thermocouple
that in turn powers a gas safety valve electromagnet. What is unique about an ODS
is the tripping of the gas safety valve at a specified level of oxygen depletion. The
tent enclosure computation above specified the oxygen safety valve trip level was
18.5%, as compared to a normal ambient oxygen concentration of 20.9%. Codes and
8417: “8417_c003” — 2007/9/11 — 12:16 — page 48 — #6

standards recognize that it is not healthy or safe for humans to function in oxygen
reduced atmospheres (<18%) for more than 8 h.
The theory of ODS operation is based on flame geometry changing during a
reduction in ambient oxygen (Figure 3.1). Change in position of the pilot flame with
respect to the thermocouple allows the thermocouple tip to cool to a point that will
cause the electromagnet to trip the safety valve, shutting off the heater’s burner. Safety
shutdown is typically designed to occur at 18.5% ambient oxygen. This oxygen level
is above the threshold that will cause excessive CO to be produced within the burner.
Typical ODS flame geometry changes are shown in Figure 3.1.
The ODS was developed and used extensively in Europe for 30 years prior to it
being adopted in the United States in 1980. It was first mandated here on ventless,
Pilot burner
Thermo couple
Ignition plug
Gas
Pilot burner
Thermo couple
Ignition plug
Pilot burner
Thermo couple
Ignition plug Gas
FIGURE 3.1 Operation of an oxygen depletion sensor. Upper panel: Normal operation—
20.9% oxygen. Pilot flame touches tip of thermocouple, generating the thermoelectricity needed
to hold the safety valve open. Middle panel: Oxygen level dropping—19% oxygen. The flame
begins to lift off the precision pilot burner. The thermocouple begins to cool. Lower panel:
Safety shutdown—18% oxygen. The unstable pilot flame moves away from the thermocouple
to stop generating the electricity needed to hold the spring-loaded safety valve open. The heater
shuts down. (Source: Adapted from a sales brochure, no copyright, ca. 1996.)
8417: “8417_c003” — 2007/9/11 — 12:16 — page 49 — #7

gas space heaters. In the late 1990s, this safety device was beginning to be applied to
gas heaters used by the hunters, campers, and the recreational industry generally for
satisfying the demand for safe, small, infrared, and catalytic heaters. Many infrared
camp heater designs use LP gas to fuel the main burner at relatively high vapor
pressures, that is, typically 7–15 psig. Pressure at the burner orifice or gas tip is
controlled by a gas regulator on the heater control panel, that is, the heat setting
control knob. This knob usually adjusts the regulator to about 7 psig in the low heat
range and to about 15 psig in the heater’s high heat range. These regulator pressures
are problematic for the ODS pilot burner, because it usually operates at about 3–5
inches water column (in. w.c.), or approximately 0.11–0.18 psig. One solution for
this problem is to add a second stage regulator that reduces the high pressure going
to the main burner regulator to a lower pressure.
Normally, some pilot flame shielding is required to deflect radiated heat from
the main burner and also to shield the small ODS flame from wind and drafts, thus
stabilizing the ODS flame’s play on the thermocouple. The approximate heat output
of a typical ODS flame at 3.5 in. w.c. pressure, using a 0.00665 in. diameter, 60◦
orifice angle, will be 175 BTU/h. This is the maximum amount of gas that may pass
through the pilot burner without having a code required safety valve to shut off the
pilot gas flow when flame outage occurs. It may be desirable to relax this orifice
diameter in order to fit a commercially available drill bit size, for example, 0.006 in.
Using this bit and the 175 BTU/h code limitation, we can use a 5.3 in. w.c. orifice
pressure. This will produce a desired maximum gas flow and satisfy the safety code.
The burner system designer may now determine the most economical arrangement
of gas safety shutdown piping and equipment for his product. Also, the heater designer
has the choice of commercially available vertical, angled, and horizontally mounted
ODS designs in order to best accommodate his design preferences. Some other ODS
variants have been used by Copreci, an Italian manufacturer.
3.7 COMBUSTION EQUATIONS FOR GAS BURNERS

The reader is advised to refer to Reference 4 for a complete summary of formulae to
determine the products of combustion for gaseous fuels such as propane and butane
that are typically used to fuel space and construction heaters for small volume enclos-
ures. This reference also details the combustion analysis that relates the amount of
combustion air needed for the stoichiometric combustion of various gaseous fuels.
Such an equation for the stoichiometric or theoretical oxidation of propane will be
Balanced equation: C3 H8 + 5O2 = 3CO2 + 4H2 O

Molecular weights: 44.094 + 5∗ 32 = 3∗ 44.01 + 4∗ 18.016
Relative combining weights: 1 lb + 3.629 lb = 2.995 lb + 1.634 lb
Combustion data for typical hydrocarbon fuels are given in Table 3.1.2 Addi-
tional combustion data and LP-gas properties are found in reference 4. It should
be remembered that commercial propane (i.e., LP-gas) contains a number of other
hydrocarbons that are limited by the HD5 or HD-5, P grade specification. That is,
8417: “8417_c003” — 2007/9/11 — 12:16 — page 50 — #8

TABLE 3.1
Combustion Data for Hydrocarbons
Cubic Feel Required Products of Combustion Pounds Required for Products of Combustion Ultimate Carbon
per Cu. Ft. of Gas or ln Cu. FL. formed by Combustion 1 lb. in Pounds from Burning Dioxide ln Flue
Vapor for Combustion Burning 1 cu. ft. of Gas or Vapor One Pound Gas or Vapor Products Percent
Chemical Carbon Water Carbon Water
Formula Air Oxygen Dioxide Vapor Nitrogen Air Oxygen Dioxide Vapor Nitrogen
Methane CH4 9.55 2.0 1.0 2.0 7.55 17.24 3.98 2.74 2.24 13.26 11.7
Ethane C2 H4 16.70 3.5 2.0 3.0 13.20 16.13 3.73 2.92 1.79 12.40 13.1
Propane C2 H3 23.86 5.0 3.0 4.0 18.86 15.71 3.63 2.99 1.63 12.08 13.7
Butane C4 H10 31.02 6.5 4.0 5.0 24.52 15.49 3.58 3.03 1.55 11.91 14.0
Pentane C5 H12 38.19 8.0 5.0 6.0 30.19 15.35 3.54 3.05 1.50 11.80 14.2
Ethylene C2 H4 14.32 3.0 2.0 2.0 11.31 14.80 3.42 3.13 1.28 11.38 15.0
Propylene C3 H6 21.48 4.5 3.0 3.0 16.98 14.80 3.42 3.13 1.28 11.38 15.0
Butylene C4 H8 28.58 6.0 4.0 4.0 22.58 14.80 3.42 3.13 1.28 11.38 15.0
Carbon Monoxide Build-Up in Houses and Small Volume Enclosures
Source: Handbook Propane–Butane Gases, 4th ed., Chilton Company, Los Angeles, CA, PP. 22,23 1962.
8417: “8417_c003” — 2007/9/11 — 12:16 — page 51 — #9

51
TABLE 3.2
Physical and Chemical Properties of LP-Gasa
Propane Butane
Formula C3 H8 C4 H10
Boiling point (◦ F) −44 15
Specific gravity of gas (air = 1.00) 1.50 2.01
Specific gravity of liquid (water = 1.00) 0.504 0.582
lb per gallon of liquid at 60◦ F 4.20 4.81
BTU per gallon of gas at 60◦ F 91,502 102,032
BTU per lb. of gas 21,548 21,221
BTU per cu. ft. of gas at 60◦ F 2488 3280
cu. ft. of vapor (at 60◦ F) gallon 36.38 31.26
cu. ft. of vapor (at 60◦ F) lb 8.66 6.51
Latent heat of vaporization at boiling point BTU per gallon 773 808
Combustion data
cu. ft. air required to burn
1 cu. ft. gas 23.86 31.02
Flash point (◦ F) −156 N.A.
Ignition temperature in air (◦ F) 920–1120 900–1000
Maximum flame
temperature in air (◦ F) 3,595 3,615
Limits of flammability
percentage of gas in air mixture
At lower limit (%) 2.15 1.55
At upper limit (%) 9.6 8.6
Octane number
(ISO-Octane = 100) Over 100 92
a Commercial quality. Figures shown in this chart represent average values.
Source: LP-Gas Serviceman’s Manual Engineered Controls International, Inc. (Reg O), Elon
College, NC, pp. 2,3, 1962, Revised March, 2000.
propane must constitute a minimum of 90.0% of the gas, propylene can be at most
5.0%, and the butanes and heavier aliphatics must not exceed 2.5%. For a summary of
the properties of commercial propane and butane, the reader is referred to Table 3.2.5
The LP-gas is to have a vapor pressure at 100◦ F of greater than 175, but not exceeding
208 psig (Table 3.3). Other reference sources are 4 and 6.
It is sometimes necessary to compute CO in combustion products in a specific
manner to comply with industry standards as defined on an free-air basis. Instruments
can determine the amount of CO on an air-free basis by first measuring the amount of
oxygen and CO present in the sample, then using the following equation to calculate
CO on a free-air basis:
CO air-free = 20.9∗ CO/(20.9 − O2 %)
8417: “8417_c003” — 2007/9/11 — 12:16 — page 52 — #10

TABLE 3.3
Vapor Pressures of LP-Gasesa
Temperature Approximate Pressure (PSIG)
(◦ F) (◦ C) Propane Butane
−40 −40 3.6
−30 −34 8
−20 −29 13.5
−10 −23 23.3
0 −18 28
10 −12 37
20 −7 47
30 −1 58
40 4 72 3.0
50 10 86 6.9
60 16 102 12
70 21 127 17
80 27 140 23
90 32 165 29
100 38 196 36
110 43 220 45
a Conversion formula:
Degrees C = (◦ F − 32) × 5/9

Degrees F = (9/5 ×◦ C) + 32
Source: LP-Gas Serviceman’s Manual, Engineered Controls
International, Inc. (Reg O), Elon College, NC, pp. 2,3, 1962,
Revised March, 2000.
This computation compensates for the amount of excess air provided by the burner.
Excess air from a burner dilutes the products of combustion and causes a test for CO
to be understated. A CO air-free measurement eliminates the excess air dilution.
3.8 CARBON MONOXIDE ACCUMULATION DUE TO

SMALL ENGINE EXHAUST GASES
Some serious injuries and deaths have occurred because the general public is not
fully aware of the latent danger in operating internal combustion engines in confined
breathing spaces or in poorly ventilated structures. An analysis of the combustion
products or exhaust gases from engines fueled by hydrocarbons shows they gener-
ally produce CO in great abundance as compared to burners in various heaters and
lighting appliances. As a comparison, a gas burning appliance will be “red tagged” in
many municipalities if its flue products contain an excess of 400 ppm of CO. Some
states have limited the legal level of CO in the automobile engine exhaust at 1.2%.
Precatalytic CO production in typical gasoline engines is 12,000 ppm CO or more.
A poorly tuned gasoline engine may produce 30 times the CO that a misadjusted gas
8417: “8417_c003” — 2007/9/11 — 12:16 — page 53 — #11

burning appliance will typically generate. In most cases where gasoline engines are
operated, engine exhaust is vented directly into the habitable space as opposed to an
exhaust stack as with gas burning appliances. Use of a breathing space shared by
a running gasoline engine subjects the people to lethal concentrations of CO very
quickly and unexpectedly.
The consequences of operating a small gasoline engine in a closed space are seen
in Reference 7. For example, a 5 hp, four-cycle, gasoline-fueled engine will generate
CO at a rate of 1.72 cfm. A simple calculation reveals that for most small structures,
a significantly elevated CO concentration will exist there after just a few moments of
engine operation. Thus, gasoline-powered engine operation in small rooms becomes
hazardous to the occupants very quickly and requires excessive ventilation, that is,
many ACH. In fact, the number of ACH necessary to limit CO buildup in such a
space may be in excess of 20 just to limit the CO concentration to 400 ppm. The rule
of thumb is that small gasoline-fueled engines should NEVER be operated in closed
breathing spaces.
3.9 OSHA AND INDUSTRY STANDARDS FOR CO

EXPOSURE
In times past, a 100 ppm CO exposure level was acceptable for humans over extended
periods of time, that is, the 8-h workday. The history of acceptable CO standards for
human exposure has been one of a gradual decline in that number over the past 100
years. National Institute for Occupational Safety and Health (NIOSH) and OSHA
now use a level of 35 ppm, time-weighted-average (TWA), over 8 h. The concern has
been to keep blood carboxyhemoglobin saturation to less than 5% for both chronic
exposure in the workplace and for workers with cardiovascular or pulmonary disease.
Other approval and standards organizations such as ASHRAE defer in their Stand-
ard 62-89 to the Environmental Protection Agency (EPA) standard for outdoor air
quality. EPA sets the ambient limit at 9 ppm, averaged over 8 h. EPA also allows
exposure to 35 ppm CO or lower, averaged over 1 h. Finally, the ANSI guide num-
ber Z21.1, allows the ambient atmosphere surrounding an unvented space heater to
achieve a maximum of 200 ppm CO when measured on an air-free basis. On the same
basis, a maximum flue gas concentration of 400 ppm is allowed and an unvented gas
oven may produce a maximum of 800 ppm. See Section 3.7 for the equation used to
calculate CO on a free-air basis.
Many state statutes limit flue gas products of combustion for a heating appliance
to no more than 400 ppm at anytime in the equipment setup or operating range. Many
state and municipal inspectors will “red tag” any fuel burning appliance that is found
to exceed that CO limit.
3.10 CONCLUSION
The aspects of CO accumulation in small structures introduced by using consumer
products such as LP-gas heaters, lanterns, and small gasoline engines motivates a
special series of cautions, warnings, designs, and code compliance considerations.
8417: “8417_c003” — 2007/9/11 — 12:16 — page 54 — #12

These concerns must be addressed by the consumers, designers, and certainly by the
engineers creating products and structures for the stream of commerce. The number
of mishaps and associated litigation speaks for the issue of properly operating fuel-
burning products in small structures such as tents, fishing shanties, campers, garages,
and so forth to name just a few.
The first aspect of operating these appliances in small breathing spaces must
be that of safety for the consumer, that is, public. The chemical and physiological
effects of CO have made it a serious design factor for engineers, manufacturers, and
consumers alike.
References
1. Berglund, L.G. ASHRAE Handbook-Fundamentals, American Society of Heating,
Refrigerating and Air Conditioning Engineers, Inc., Atlanta, GA, pp. 8.6, 8.7, Chapter
8, 2001.
2. Denny, L.C., Luxon L.L. and Hall, B.E. Handbook Propane–Butane Gases, 4th ed.,
Chilton Company, Los Angeles, CA, pp. 22, 23, 1962.
3. Tucholski, D.R. Technical Feasibility of a CO Shutdown System for Tank-Top Heaters,
U.S. Consumer Products Safety Commission (CPSC), Washington, D.C., p. 53, 2005.
4. Reed, R.J. North American Combustion Handbook, 3rd ed., Vol. 1, North American
Manufacturing Company, Cleveland, OH, p. 49, 1986.
5. LP-Gas Serviceman’s Manual, Engineered Controls International, Inc. (RegO), Elon
College, NC, 1962, Revised March, 2000.
6. Segeler, C.G. Editor-in-Chief; Ringler, M.D., Assistant Editor; Kafka, E.M., Technical
Assistant, Gas Engineers Handbook, The Industrial Press, New York, NY, pp. 6–33,
1966.
7. Earnest, G.S., Mickelsen, R.L., McCammon, J.B. and O’Brien, D.M. Carbon monox-
ide poisoning from small, gasoline-powered, internal combustion engines: just what
is a well-ventilated area? Am. Ind. Hyg. Assoc. J., 58, 787–791, 1997.
8417: “8417_c003” — 2007/9/11 — 12:16 — page 55 — #13

8417: “8417_c003” — 2007/9/11 — 12:16 — page 56 — #14
4 Formation and
Movement of Carbon
Monoxide into Mobile
Homes, Recreational
Vehicles, and Other
Enclosures
Robert E. Schreter
CONTENTS
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.2 Accident Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.3 Codes and Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.4 Potential Sources of Carbon Monoxide within Motor Homes and Other
Enclosures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.4.1 Internal Combustion Engines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.4.1.1 Large Vehicle Propulsion Engines . . . . . . . . . . . . . . . . . . . . . . 62
4.4.1.2 Auxiliary Power Generators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.4.2 Interior Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.4.2.1 Ranges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.4.2.2 Ovens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.4.2.3 Gas Refrigerators (LP Gas) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.4.2.4 Domestic Water Heaters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.4.2.5 Gas or Oil-Fired Space Heaters . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.4.2.6 Clothes Dryers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.4.2.7 Fireplaces (LP Gas) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
4.4.2.8 Other Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
4.5 Primer on the Mechanism of Formation of Carbon Monoxide from
Fuel-Fired Appliances found in Mobile Homes . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
4.5.1 The Physical and Chemical Properties of Carbon Monoxide . . . . . 67
4.5.2 The Combustion Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
57
8417: “8417_c004” — 2007/9/11 — 12:16 — page 57 — #1

4.5.2.1 Basic Requirements to Initiate and Perpetuate the

Combustion Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.5.3 Thermochemical Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
4.5.4 Excess Air Combustion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.5.5 Excess Fuel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.6 Application of Combustion Theory to Actual Carbon Monoxide
Poisoning Situations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.6.1 Introduction to the Flow of Exhaust Gases . . . . . . . . . . . . . . . . . . . . . . . . 72
4.6.2 Total Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.6.3 Stagnation Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
4.6.4 Thermal Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
4.6.5 Air Flow Around and Through Enclosures . . . . . . . . . . . . . . . . . . . . . . . . 77
4.6.6 Wind Rose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
4.6.7 Scoops, Deflectors, and Openings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
4.6.8 Resonance, Cyclical Flow into and out of an Enclosure. . . . . . . . . . 81
4.6.9 Flow into and out of Living Quarters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.6.10 Flue and Chimney Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.6.11 Back-drafting of Flue Gases into the Enclosure . . . . . . . . . . . . . . . . . . 83
4.6.12 Dynamics of Generator Tailpipe Exhaust . . . . . . . . . . . . . . . . . . . . . . . . . 83
4.6.13 Exhaust Flow Through Bent or Displaced Exhaust Pipes . . . . . . . . 85
4.6.14 Missing Exhaust Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.7 Accumulation of Carbon Monoxide within an Enclosure. . . . . . . . . . . . . . . . . 87
4.7.1 Rate of Dissipation of Carbon Monoxide from an Enclosure . . . . 89
4.8 Test Methods for Determining the Location of Leakage into an
Enclosure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4.8.1 Smoke Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
4.8.2 Basic Considerations during Smoke Testing . . . . . . . . . . . . . . . . . . . . . . 91
4.8.3 Positive Pressure Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
4.8.4 Negative Pressure Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
4.8.5 Typical Smoke Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.9 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
4.1 INTRODUCTION
This chapter deals with nonfire related carbon monoxide (CO) poisoning accidents
that occur in mobile homes, live-in trailers, portable offices and other mobile living
or work enclosures containing devices that can produce CO or other toxic gases. The
flow mechanisms and equations presented apply equally to all stationary buildings
subject to wind and thermal activity.
I will briefly discuss American National Standards Institute (ANSI) and National
Fire Protection Association (NFPA) codes that apply to recreational vehicles, motor
homes and the like and stress the code requirements involved in sealing the living
compartment against toxic gases and vapors produced externally.
This chapter addresses the individual sources that generate CO within the living
compartment and sources that are part of the mobile home but external to the
living compartment, as well as those produced in the vicinity. It discusses each
8417: “8417_c004” — 2007/9/11 — 12:16 — page 58 — #2

Formation and Movement of Carbon Monoxide 59
of the appliances, the mechanisms by which they produce CO and the expected
concentration of CO associated with them. It compares the probability of their entering
into life threatening asphyxiations. It also discusses other sources of CO emanating
from other vehicles or generators in the vicinity.
I will discuss the physical properties and behavior of CO and the mechanism of its
formation in various types of appliances and internal combustion engines. It discusses
the combustion process, thermochemical equations, and how excess air, excess fuel,
mixing of air and fuel and available time for the completion of combustion affect its
production.
I will introduce the concept of total pressure, including wind and thermal pressure
and how these pressures are the driving forces that result in CO penetrating leaks into
the living compartment.
The chapter covers the application of combustion theory and aerodynamics and
how they apply to CO accidents. It discusses how wind and the aerodynamic shapes
and openings encountered in mobile homes and other enclosures affect air and gas
flow movement into, through, and out of enclosures.
There is a general discussion of exhaust systems for internal combustion engines
and how they play a role in life threatening CO asphyxiations. It discusses how missing
or damaged engine generator tailpipes are major contributors to fatal accidents.
This chapter presents methods used to evaluate the rate of accumulation of CO
within enclosures. Computational methods used to determine CO concentration at any
point in time during the inflow are presented. It also presents computational methods
to determine the dissipation of CO during periods where the generating source is shut
down or its flow rate curtailed. The methods presented predict the partial pressure
within an enclosure and how it may be used in conjunction with the Coburn–Forster–
Kane equation to predict carboxyhemoglobin (COHb) in a victim’s blood (see other
books in this series).
Finally, the chapter discusses methods that may be used to determine the locations
of leaks into the living compartment and the magnitude of inflow into the enclosure.
4.2 ACCIDENT STATISTICS

Between 1995 and 1999, an estimated 10,200 people reported to hospital emer-
gency rooms each year for nonfire-related CO poisoning injuries associated with
consumer products.1 There were more than 500 unintentional nonfire related deaths
and injuries resulting from CO-poisoning accidents associated with the use of con-
sumer products. An average of 22 deaths per year was associated with engine-driven
generators and engine-driven products, which includes deaths associated with mobile
homes. Between 1993 and 1997 deaths from motor vehicle exhausts, averaged 534
annually. Many CO accidents are not recognized, and because of data collection prob-
lems, many cases go unreported. Nonfatal injuries mostly go unreported, therefore,
there are few accurate statistics.
4.3 CODES AND STANDARDS

TheANSI 119.2,2 and NFPA501c3 have written and promulgated standards applicable
to CO in mobile homes and recreational vehicles. The standards cover fire and life
8417: “8417_c004” — 2007/9/11 — 12:16 — page 59 — #3

safety criteria for recreational vehicles to provide reasonable protection from loss of
life from fire and explosion.
In particular, the standard sets down requirements for fuel burning devices used
in connection with recreational vehicles. They pay special attention to the loca-
tion of internal combustion engines and their exhaust ports relative to leakage into
the living compartment of the vehicle. In essence, the codes state that exhaust
gases shall be directed away from the vehicle and that the exhaust system shall
be installed in a manner by which the exhaust pipes are not unduly subject to road
damage.
They also state that all fuel-burning devices, except ranges and ovens, shall be
designed and installed to provide for the complete separation of the combustion system
from the interior atmosphere of the recreational vehicle. Throughout this chapter,
the interior of the recreational vehicle would be referred to interchangeably as the
living compartment or living enclosure. Essentially, this means that the combustion
air intake, the burner, the combustion chamber, heat transfer elements, and exhaust
piping shall be outside the living compartment. It further states that exhaust or flue
outlets shall not terminate under the recreational vehicle.
The codes further state that internal combustion engines (motor generators),
engine driven pumps or the like shall be mounted in compartments that are vapor
tight to the interior of the vehicle (living compartment). The wording is somewhat
confusing in that of necessity the engine compartment cannot be sealed because the
engine requires combustion and cooling air, therefore, cannot be sealed. However,
the meaning of the code is quite clear that the living compartment must be sealed
to prevent exhaust gas penetrating the enclosure. In addition, the code states that all
penetrations into the living compartment shall be sealed. Furthermore, fuel lines and
exhaust systems shall not penetrate the living compartment.
All recreational vehicles that are equipped with or are equipped to accommodate
a future installation of an internal combustion engine shall be equipped with a listed
CO detector or alarm.
Readers interested in detail information regarding codes or standards are referred
to ANSI 119.2 and NFPA 501 c.2,3
4.4 POTENTIAL SOURCES OF CARBON MONOXIDE

WITHIN MOTOR HOMES AND OTHER
ENCLOSURES
Motor homes contain a variety of appliances utilizing combustion that can generate
CO as well as other toxic gases. The probability of creating serious injury or death
is a function of the amount of CO produced by the appliance and the concentration
ultimately trapped in the enclosure.
The likely devices and appliances fall into one of two classes depending on level
of CO that can be expected
1. Internal combustion engines produce levels of CO ranging from 10,000 to

70,000 ppm or more; CO concentration drops by 95% or more when the
8417: “8417_c004” — 2007/9/11 — 12:16 — page 60 — #4

exhaust system is equipped with a catalytic converter. Without a catalytic

converter, the probability of life threatening accidents occurring is very
high if exhaust gases leak into occupied enclosures.
2. Ranges, stoves, ovens, water heaters, and so forth produce much lower
levels of CO, generally in the range of a few ppm up to a few hundred
ppm. The probability of these devices causing life-threatening accidents
is considerably lower than with internal combustion engines. However,
these concentrations are capable of causing brain damage and other health-
related injuries and cannot be dismissed out of hand.
In any CO asphyxiation incident, regardless of the CO level, it is a good practice

to inspect all devices that could potentially generate CO. Measure their output, and
catalog emissions even though they may appear to be unrelated to the accident and
the environment in which the accident occurred.
4.4.1 INTERNAL COMBUSTION ENGINES

Internal Combustion Engines present the greatest hazard from the standpoint of
CO poisonings. As stated above, CO concentrations of 10,000–70,000 ppm are not
uncommon. CO at these magnitudes, even when diluted 30 or 40 to 1 can be lethal,
can quickly result in death.
Internal combustion engines produce high levels of CO because of the brief period
available to complete combustion. As an example, a single cylinder engine running
at 2400 rpm must fire 20 times per second. It must draw in a charge of air, meter and
inject an appropriate quantity of fuel through a crude carburetor, and mix the air and
fuel in 0.0236 s. The mixture must then be ignited, the flame propagate throughout the
cylinder, and finally complete the power stroke in 0.014 s. In this brief period of time,
it is virtually impossible to perfectly mix each fuel molecule with the appropriate
amount of air and burn it to completion in about 0.04 s.
Consequently, incomplete combustion is the inevitable result. In addition, the
boundary of the metal cylinder, the cylinder head, and piston are all relatively cold
surfaces, therefore cause quenching and additional incomplete products of combus-
tion. As a result, the small internal combustion engine is naturally inefficient and
produces high levels of CO and other products of incomplete combustion.
Modern automobile and truck engines, although greatly improved, still produce
high levels of CO, but because of catalytic exhaust systems now deliver much lower
levels of CO to the atmosphere.
Of the fatal CO motor home accidents that I have investigated, the majority can be
traced to the motor generator and its exhaust system. The majority occurred after the
exhaust system was damaged, displaced, or missing. The damage to or the absence
of the tailpipe allowed high levels of CO and other toxic exhaust gases to be directed
under the motor home where they accumulated and found their way into the living
compartment.
It should be noted that despite numerous accidents many generator exhaust sys-
tems are still located in unprotected areas where they are subject to damage from
direct impact from road hazards.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 61 — #5

E H
B F G
D
I
J K
C
FIGURE 4.1 Air flow over and through an enclosure.
4.4.1.1 Large Vehicle Propulsion Engines

Older motor home engines, built before mandated emission standards, commonly
produced in excess of 4% (40,000 ppm) CO. Modern engines still produce high
levels of CO; however, modern catalytic exhaust systems have reduced CO discharge
drastically in order to meet the current CO emission standard.
Other than intentional suicide, and operating the motor home engine for heating
purposes in quiescent zones, the main propulsion engine is rarely the cause of fatal
asphyxiations. The propulsion engine is rarely running long enough, when the vehicle
is not in motion, to accumulate a high enough concentration of CO. When in motion,
the position of the exhaust pipe encourages rapid dilution and dissipation of CO,
discharging it away from the vehicle. There is one exception, and that is in the area
directly behind the vehicle. Large moving objects create a suction wake directly
behind the vehicle, which encourages exhaust gases to accumulate, much like the
vortex G, shown in Figure 4.1. Codes recognize this problem and require special
sealing to avoid leaks into the living compartment in those areas. In addition, codes
require that openings like windows must be permanently sealed to avoid opening.
Main propulsion engine exhaust pipes are generally large and structurally strong,
with substantial means of attachment to the vehicle. Although they are frequently in
unprotected areas, the pipes and supports are strong and can sustain impacts that would
damage smaller lighter generator exhaust pipes. Although this does not preclude an
accident, it greatly reduces the probability.
A motor home if left with the engine idling, for a prolonged period of time,
especially, if parked in a protected zone or in a depression, may permit exhaust gases
to accumulate underneath. The accumulation of exhaust gases in the vicinity of the
discharge end of the tailpipe thwarts the natural dilution of exhaust gases, thereby
exposing leakage areas to highly concentrated exhaust gases. The temperature of
undiluted exhaust gas is high thereby creating a thermal head that facilitates infiltration
into the living compartment through breaches in the vehicle floor.
4.4.1.2 Auxiliary Power Generators

The majority of standby power generators, used in motor homes, are driven by single
cylinder, four-stroke engines. Some of the larger generators use larger two-cylinder,
8417: “8417_c004” — 2007/9/11 — 12:16 — page 62 — #6

four-stroke cycle engines. The range of sizes run from 2.5 kW (7 bhp) to 7 kW
(14 bhp). They are available for gasoline, liquefied petroleum (LP), liquid or vapor
state fuels, and occasionally diesel fuel.
Codes require that all motor-driven generators be mounted in compartments that
are outside of the living compartment. The generators are generally located in a ventil-
ated compartment along the periphery of the motorhome. The generator compartment
cannot be sealed and must be open to the outside air to maintain a constant flow of
cooling and combustion air to the engine.
The code requires that all generators be mounted in compartments that must be out-
side of the living enclosure. Typically, the generators are equipped with exhaust sys-
tems that consist of a muffler, tailpipe, and tailpipe support brackets. The exhaust pipe,
approximately 1 in. in diameter, usually made of 16 gauge stainless steel, is routed
under the motorhome and generally protrudes out the side or back of, and beneath the
lower skirt of the motorhome. Engines are provided with mufflers to reduce exhaust
noise levels, which are typically 67–70 dBA (sound pressure level at 10 ft.). On some
“Quiet” models sound levels as low as 60 dBA are available. It should be noted that
even at 60 dBA the noise signature is loud enough that a defect in the exhaust sys-
tem can readily be detected by a person, used to the sound, standing in the vicinity
of the generator. This provides a safety feature to persons that check the generator
in use.
In general, low horsepower engines are inefficient and produce very high levels of
CO. Exhaust gas composition may range from 4.0% to 7.0%, (40,000–70,000 ppm) of
CO with associated carbon dioxide levels ranging from 6% to 10% (60,000–100,000
ppm). Exhaust gases are discharged at a temperature of 400–800◦ F depending on the
generator load.
Of all appliances, motor generators present the greatest danger to human life and
health. The high CO concentration (40,000–70,000 ppm) in the exhaust gas, even
when the engine is in good working order, creates a serious poisoning hazard. By
comparison, with a few exceptions, other CO-producing devices commonly found in
motor homes, produce CO levels up to 400–500 ppm and only then when they are
not working properly. These lower CO levels may still present a health hazard, but
of a far lower magnitude than a generator.
Because generators present such a high exhaust gas hazard, it is required that a
special standard of care be exercised to ensure that the generator engine and all its
associated components be maintained in first class condition, and be protected against
accidental damage.
Generator engines are quite reliable when properly maintained. However, the
exhaust gas is highly toxic regardless of the operating condition of the engine. Essen-
tially, nothing can be done to relieve the fundamental CO hazard. With regard to life
safety, the exhaust system is the most critical component of the motor generator and
should be protected at all cost.
In general, the exhaust pipe and muffler are located under the engine with the
tailpipe projecting below the skirt of the motor home. In the majority of cases, the
discharge of the tailpipe is located perpendicular to the side of the motor home and
projects out a minimum distance of 1 in. beyond the side of the motor home. For
additional protection, codes require that the exhaust pipe may not be located closer
8417: “8417_c004” — 2007/9/11 — 12:16 — page 63 — #7

than 6 in. from any air intake. This assumes that the pipe discharges into clean air to
dilute the exhaust products.
The tailpipe, by virtue of its location below the vehicle, is in an unprotected
position and is subject to impact from curbs, road hazards, uneven terrain, and flying
debris. Most frequently, tailpipe support brackets are weak and will not withstand
substantial impact. The pipe is normally a thin wall stainless steel tube that is readily
subject to impact damage. My experience shows that the majority of fatal accidents
are caused by damaged, out of position, or missing tailpipes.
4.4.2 INTERIOR DEVICES

4.4.2.1 Ranges
The majority of cooking stoves used in motor homes or trailers burn LP as a fuel.
Propane, butane, or a combination thereof are commonplace. Exhaust gases from the
burners are vented directly into the living compartment of the motor home. Gravity
vents may be provided or a fume hood may be located above the range. A fume
hood collects range exhaust products of combustion along with cooking vapors and
discharges the gases through a register located on the outside wall or roof of the motor
home. The registers are normally equipped with flap type check valves to prevent back
flow into the motor home from wind movement.
Some fume hoods used over stoves filter out grease and smoke, but return com-
bustion products to the living compartment. These systems do not remove CO, carbon
dioxide, or other toxic gases, but simply circulate them within the enclosure. Exhaust
fan capacities are typically in the range of 50–125 acfm. All exhaust systems require
similar quantities of make-up air to prevent back drafting.
Tests conducted by the author show that CO levels, taken just above the visible
flame, range from 15 to 70 ppm; properly adjusted burners have blue flames. Badly
adjusted or damaged burners, frequently typified by yellow tipped flames, have CO
levels of 100–770 ppm (this figure may vary by location). Burner pilot lights can
contribute an additional 28–56 ppm.
Ranges used for cooking rarely are hazardous because they are used for a limited
time. When used for prolonged cooking they are generally used in conjunction with
an exhaust hood. Ranges can pose a hazard if used in a small, tight compartment
without sufficient air infiltration or when operated for prolonged periods of time.
Ranges should never be used for space heating.
4.4.2.2 Ovens
Ovens also discharge products of combustion and cooking vapors directly into the
living compartment. Ovens having blue flames can produce 35–768 ppm of CO and
their pilots can add up to an additional 960 ppm.4
Some ovens use a gravity flue to discharge oven gases outside the living com-
partment. When used they should be equipped with backflow protection. Because
of the relatively low temperature of exhaust gases and the short stack length, the
flue draft developed is limited. Back drafting may present a problem when dryers or
8417: “8417_c004” — 2007/9/11 — 12:16 — page 64 — #8

exhaust fans are used simultaneously within the same enclosure. Under back-drafting
conditions, CO can increase to dangerous levels. During back drafting, CO levels far
exceed the norm expected of the appliance.
Like cooking ranges, ovens are used for relatively short periods of time and are
generally vented. However, when used in tightly sealed enclosures without sufficient
infiltration air, they can present a hazard. Ovens should never be used for space or
supplemental heating.
4.4.2.3 Gas Refrigerators (LP Gas)

The majority of motor homes and trailers use gas or combination gas-electric
refrigerators. The combustion unit for a gas refrigerator is contained in a sep-
arate metal enclosure that is located outside the living quarters. Combustion
air is drawn into the combustion tube, mixed with vaporized propane, ignited
and burned and flue products passed through a heat exchanger and discharged
to the outside atmosphere. Thus, combustion gases are isolated from the living
compartment.
Of all appliances, gas refrigerators show a low frequency of CO poisonings.
Negative living compartment pressure does not pose a problem.
4.4.2.4 Domestic Water Heaters

The combustion unit for a gas fired water heater is contained in a separate metal
enclosure that is located outside the living quarters. Combustion air is drawn into the
combustion tube, mixed with vaporized propane, ignited, burned, and flue products
passed through a heat exchanger. They are discharged to atmosphere outside the
living compartment. Make-up and hot water connections that penetrate the living
compartment must be sealed to prevent external gases from penetrating the liv-
ing compartment. Negative living compartment pressure does not pose a problem
except through leaking seals around pipes or electrical connections entering/leaving
the living compartment. The risk of dangerous amounts of CO entering the living
compartment, except at seals, is small.
4.4.2.5 Gas or Oil-Fired Space Heaters

The combustion unit of the space heater must have an air intake and discharge that
are external to the living compartment. Hot combustion gases are contained inside
a stainless steel heat exchanger that separates them from the living compartment.
Pressurized air flow across the outside of the heat exchanger picks up and carries hot
air into the living compartment. The risk of dangerous levels of CO penetrating the
living compartment is small.
4.4.2.6 Clothes Dryers

Clothes dryers when fuel-fired have the combustion unit outside the living com-
partment. The combustion unit is much like those used in approved space heaters,
8417: “8417_c004” — 2007/9/11 — 12:16 — page 65 — #9

water heaters, and refrigerators. Clothes dryers present an additional but indir-
ect hazard, because the dryer exhaust can lower living compartment pressure to
the point where back drafting by other appliances occurs. Lower living compart-
ment pressure also increases the probability of drawing external exhaust gases
into the living compartment. If a clothes dryer is used it is imperative that a suf-
ficient make up air supply be provided to prevent back drafting into the living
compartment.
4.4.2.7 Fireplaces (LP Gas)

Fireplaces and even wood burning stoves have become available for use in mobile
homes. Because their flames are exposed inside the living compartment, anything that
could result in back drafting will cause exhaust gases to be discharged directly into
the living compartment. Gas fireplaces, even in conventional homes, are notorious
for discharging CO into living areas. Fireplaces are subject to back drafting any
time that living compartment pressure is reduced below atmospheric pressure (see
Figure 4.2).
4.4.2.8 Other Sources

The use of portable kerosene, gasoline, and propane or natural gas heaters is not
recommended. Regardless of the type, they discharge their products of combustion
into the living compartment. Even when operating properly they require an adequate
supply of combustion air, which may not be present in a tight enclosure. Heaters of
this type are portable and can be easily upset, presenting an additional fire and toxic
gas hazard. Many of these devices are position sensitive and cannot function in other
than a vertical attitude. Their use is prohibited by ANSI 119.2 in that the products
of combustion are discharged into the living compartment. There is a history of
numerous asphyxiation accidents related to this type heater.
Fresh air
Leakage air
Dryer exhaust
Exhaust gas
Dryer exhaust back flow
Clothes dryer
FIGURE 4.2 Back-drafting in a fireplace flue.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 66 — #10

4.5 PRIMER ON THE MECHANISM OF FORMATION

OF CARBON MONOXIDE FROM FUEL-FIRED
APPLIANCES FOUND IN MOBILE HOMES
4.5.1 THE PHYSICAL AND CHEMICAL PROPERTIES OF CARBON
MONOXIDE
To understand how CO is generated and transported into a motor home, it is necessary
to understand its physical properties, where and how it is formed, and how dangerous
concentrations are transported into living compartments.
CO is a clear gas that is odorless, tasteless and has a density slightly less than air.
It is generally a product of incomplete combustion of carbon or hydrocarbon fuels.
Its physical properties are seen in Table 4.1.5
The common understanding is that CO is lighter than air, as it is in the pure
state. At standard temperature and pressure, it has a density of 0.0779 lb/cu. ft.; air is
0.0807. The difference in density creates a natural buoyancy. However, under most
circumstances where we find it, CO is a product of incomplete combustion. Therefore,
it is not a pure gas, but contained in a mixture of exhaust gases containing nitrogen,
carbon dioxide, and water vapor in proportions dictated by Equation 4.1. Because of
the density of these gases, the combined mixture has an essentially neutral buoyancy.
For illustrative purposes, assume that a motor generator is fueled by octane,
C8 H18 . Its exhaust products are 6% CO and 8% CO2 . Under these conditions the
exhaust gas density is 0.0805 lb/cu. ft. at STP (standard temperature and pressure 60◦ F
and 29.92 mm Hg), which has essentially neutral buoyancy when compared to air
TABLE 4.1
Physical Constants for Carbon Monoxide∗
Name Symbol Value Units
Chemical symbol CO
Density, STP rho 0.0779 lb/cu.ft.
Molecular weight MW 28.011
Specific gravity (air) SG 0.967
Specific gas constant R 55.19 ft. lbf /lbm R◦
Specific heat p = c Cp 0.248 Btu/lb◦ F
Specific heat v = c Cv 0.177 Btu/lb◦ F
Specific heat ratio k 1.4 —
Specific volume nu 13.557 cu. ft./lb
Speed of sound C 1154.8 ft./s
Heat of combustion 310 Btu/cu. ft.
Heat of combustion 4340 Btu/lb
Flammability in air lower LFL 12.5 Vol% of total volume
Flammability in air upper UFL 74 Vol% of total volume
Flame temperature in air 3542 ◦F
Auto ignition temperature 1228 ◦F
∗Also see Chapter 35.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 67 — #11

0.0807,lb/cu. ft. at STP. Therefore, an exhaust gas mixture, with neutral temperature
differential, will not have natural buoyancy at STP.
Any buoyancy characteristics associated with combustion products will be due
solely to a temperature differential between the exhaust gas stream and the surrounding
air. For example, a hot jet stream of exhaust at 800◦ F in air will be buoyant and have
a resultant direction and magnitude that will be the vector sum of the vertical thermal
component and jet vector of the initial gas stream.
Ahot gas stream jetted into ambient air naturally entrains large volumes of ambient
air, thereby lowering its temperature, thus its buoyancy. As the stream gains volume
it slows down. Any computations should take into account the momentum exchange,
the increasing volume of the stream and diminishing temperature.
4.5.2 THE COMBUSTION PROCESS

Under most circumstances, CO is an undesirable product of combustion. It is formed
as a result of incomplete or inefficient burning of carbonaceous substances. By defin-
ition, combustion is the rapid oxidation of fuel accompanied by the liberation of heat
and light. The combustion process, even in its simplest form, say burning methane in
air, is not fully understood.
Equation 4.2 shows a methane and excess air mixture burning to completion,
forming products of combustion containing water, carbon dioxide, and excess air.
Unlike the simplified end-points equation shown, there can be up to 31 intermediate
steps in the oxidation of methane. If at any time during the combustion process the
reaction is chilled or in any way interrupted, the intermediate reactions will freeze
leaving one or more intermediate partial products of combustion. One of the more
common partial products is CO. If temperatures remain high and oxygen is present
some or all the intermediate products may be oxidized to completion. If the temper-
ature drops below the critical reaction temperature or if oxygen is not present, the
intermediate products will become stable and be represented in the exhaust gases.
The most common causes for incomplete combustion are quenching, incorrect fuel
air ratios and inadequate mixing resulting in stratification, and burning in the fuel rich
or reducing fuel air region.
The combustion process involves mixing of multiple gas products, mass transport,
chemistry, and heat transfer. For the purpose of this discussion, many of the more
complex processes will be simplified and we will deal, as much as possible, with
the reactants and end products of combustion as defined by the thermo-chemical
equations. It should be understood that the equations shown are representative of
those containing intermediate reaction, but only at the beginning and the end.
4.5.2.1 Basic Requirements to Initiate and Perpetuate the

Combustion Process
Fuel must be present in a state that is burnable. Liquid and solid fuels will not burn.
Fuel must be in a gaseous state for it to combine with oxygen and burn.
Liquid fuels have to be atomized to small particles so that on exposure to high
temperatures, the pilot flame, they quickly gasify so they can burn. Similarly, solid
8417: “8417_c004” — 2007/9/11 — 12:16 — page 68 — #12

fuels must be pulverized so that on exposure to high temperature, the pilot flame,
they quickly gasify and burn. The fuel must also be present in sufficient quantity to
be within the flammable limits (Table 4.1), and the initial flame size must exceed the
minimum kernel diameter in order to propagate. An oxidizer, air or oxygen, must be
present and intimately mixed with the fuel. The mixture must be within flammable
limits, at least in a localized area, where ignition is to occur.
There must be a source of ignition, of sufficient energy to raise the temperature
of a critical mass of the fuel mixture above the ignition temperature to initiate the
reaction. If the reaction is to perpetuate itself, the heat generated must be at least
equal to the net heat loss from the flame and the heat required to heat the incoming
mixture to ignition temperature. If the heat generated is equal to or greater than those
requirements, the reaction will propagate.
4.5.3 THERMOCHEMICAL EQUATIONS

Combustion like any other chemical reaction may be represented by thermochemical
reaction equations. The equations show the fuel and air input and the products of
combustion and heat liberated. The typical equation shows the initial ingredients and
the final products of combustion, but does not show the intermediate reactions. For
those interested in the detailed combustion process, see various sources.6–8
In general, this presentation will concern itself with the end-products of rich com-
bustion, primarily CO and carbon dioxide. There are times when quenching, chilled
flames, and blow off create intermediate toxic products of combustion. Essentially,
we will concern ourselves with three thermochemical equations, Equation 4.1 for rich
or excess fuel (i.e., deficiency of air), Equation 4.2 is for an excess of air, and Equation
4.3 is for secondary oxidation of CO formed in Equation 4.1. The term “E” represents
the equivalence ratio or the ratio of actual to stoichiometric air. Stoichiometric air
being that amount of oxygen required for perfect combustion.
Thermochemical Equation for Methane Excess Fuel
CH4 + 2 E O2 + 7.52 E N2 = (4 E − 3) CO2 + (4 − 4 E) CO + 2H2 O

+ 7.52 E N2 + Heat (1040 BTU/ft 3 ) (4.1)
Thermochemical Equation for Methane and Excess Air.
CH4 + 2 E O2 + 7.52 E N2 = CO2 + 2H2 O + 2(E − 1)O2

+ 7.52 E N2 + Heat (E∗ 1040 BTU/ft 3 ) (4.2)
Thermochemical Equation for Carbon Monoxide.
CO + O = CO2 + Heat (323.5 BTU/ft 3 ) (4.3)
Equation 4.2 is presented in graphic form (see Figure 4.3). This chart gives a
quick visualization of how combustion proceeds to either excess fuel or excess air.
An equivalence ratio of 1, the abscissa, occurs at stoichiometric or perfect combustion.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 69 — #13

DEN*100
CO2
Exhaust gas percent wet methane CH4
CO
18
O2
16
14
12
Percent (Wet)
10
0
0 1 2 3 4 5 6
Equivalence ratio
FIGURE 4.3 Exhaust gases from the combustion of methane.
Some notable observations about Figure 4.3.

1. The CO2 curve peaks at an equivalence ratio of 1, and falls off on the
reducing and oxidizing side of stoichiometric.
2. Maximum combustion efficiency is achieved when carbon dioxide is
maximized.
3. Therefore, the CO2 curve is a good indicator of the quality of combustion.
4. Since CO2 is readily measured, it can be used, in conjunction with a
determination of the presence of oxygen to assess the equivalence ratio.
5. Flame temperature also follows the CO2 curve (flame temperature actually
peaks slightly on the reducing side of E = 1 but for general purposes the
difference can be disregarded).
6. These observations are helpful in determining if a flame is reducing and
generating CO.
The chart shown is for methane. There are two charts for each fuel: (1) wet gas
analysis, where the water of combustion remains in the exhaust gas, and (2) dry gas
analysis, assuming that the water of combustion has condensed. Figure 4.3 represents
a wet gas analysis. The use of wet gas analysis is appropriate for combustion systems
where the exhaust gas temperature is above the dew point. The majority of charts
found in the literature are set-up on a dry basis. This was done when the most common
combustion analyzer was the Orsat. It being a wet absorption analyzer caused the gas
temperature to drop below the dew point, removing all the water, and thereby yielding
the products of combustion on a dry basis. Care must be taken when using a chart
to determine if analysis is on a wet or dry basis. Section 4.10 shows a similar graph
drawn on a dry basis. This approach may be more recognizable and is used more
frequently by those involved in general combustion work.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 70 — #14

The maximum CO2 for each fuel will be different and may be calculated from
Equation 4.1. Burners are normally adjusted for near peak CO2 on the slightly oxidiz-
ing side to insure complete combustion. The majority of good combustion equipment
operates with an equivalence ratio in the range of 1.0–1.4. Electronic instruments can
be calibrated to read either wet or dry percent. Therefore, it is important to know how
the instrument is calibrated.
4.5.4 EXCESS AIR COMBUSTION

Excess air combustion occurs at any time that the quantity of combustion air exceeds
the stoichiometric requirements. This can occur on an overall basis, where fuel and air
are uniformly mixed throughout the combustion chamber, or on a local basis within
the combustion envelope, usually classified as stratified mixing. Stratified mixing can
result in rich and lean zones occurring within the same flame envelope. One of the
fundamental problems that occurs in most combustion systems is that it is difficult to
mate every fuel molecule with the correct number of oxygen molecules. Failure to
do so can result in rich excess fuel zones and lean excess air zones within the same
combustion envelope.
When there is too much air for a given amount of fuel, the flame temperature
drops because of heating the additional excess air. If the quantity of excess air is
sufficient then the flame temperature can drop below the critical reaction temperat-
ure and flameout can occur. However, prior to flameout, flame quenching starts and
partial products of combustion such as CO and various aldehydes appear and become
stable in the exhaust gases. If the quenching is local, and the partial products are sub-
sequently reheated to a temperature in excess of the critical reaction temperature then
the partial products may go to completion in the subsequent burn. Figure 4.4 shows a
curve of flame temperature (◦ F) versus excess air (%). As excess air, increases flame
Temperature versus excess air

4000
Temperature
3500
3000
Temperature (F°)
2500
2000
1500
1000
500
0
0 500 1000 1500 2000 2500
Excess air (% )
FIGURE 4.4 Flame temperature versus excess air (%), natural gas. (R. Schreter and
Associates, 1997).
8417: “8417_c004” — 2007/9/11 — 12:16 — page 71 — #15

temperature decreases. As the temperature approaches 1200◦ F, the critical reaction

temperature, partial products of combustion and unburned fuel emerge. At very high
levels of excess air, CO and formaldehyde can be generated.
4.5.5 EXCESS FUEL

Excess fuel combustion occurs at any time that the quantity of fuel exceeds the
stoichiometric requirements. As with excess air, this can occur on an overall basis
or on a local basis within the combustion envelope. As depicted in Figure 4.3, CO
appears at any value of E < 1. Small amounts of CO may occur even at E = 1.05,
slightly excess air, because of difficulty mixing all the fuel with all the air. The rise
of CO is very rapid on the excess fuel side as can be seen in Figure 4.3.
Rich reactions contribute to the formation of CO in appliances when they are
deficient in combustion air. The same type of reaction occurs when back drafting
occurs. During back drafting, the reversal of flow upsets the mixing process and
flame stability, resulting in rich combustion, blow off, quenching and formation of
partial products of combustion.
4.6 APPLICATION OF COMBUSTION THEORY TO

ACTUAL CARBON MONOXIDE POISONING
SITUATIONS
Aerodynamics and gas dynamics are the primary factors driving exhaust gas leak-
age into or out of the living compartment of a motor home, or for that matter any
enclosure. The following theory is intended for use with motor homes; however, the
same approach may be used in any low-pressure flow situation that might occur in a
residential or commercial building.
4.6.1 INTRODUCTION TO THE FLOW OF EXHAUST GASES

The flow rate of air, CO or a mixture of exhaust gases through leaks in the enclosure
boundary is of critical importance to the concentration of toxic gases within the enclos-
ure. The physical conditions governing low pressure gas flow usually encountered
in motor homes greatly simplifies calculations, because the compressibility of the
gases may be disregarded. The range of total pressure and differential pressure is low,
further simplifying the flow equation. Flow is mainly through orifice type openings
and friction can generally be disregarded. A discharge coefficient appropriate for the
type of opening is required. Discharge coefficients may be found in most sources on
fluid flow. Because differential pressures across orifices or leakage holes are low, a
simplified velocity equation may be used as expressed in Equation 4.4.
Gas Velocity at Low Pressures.

T0
U= 2∗gc ∗R∗ ∗ (P0 − P1 ) (4.4)
P0
8417: “8417_c004” — 2007/9/11 — 12:16 — page 72 — #16

The continuity equation may then be used to determine the flow rate (Equation 4.5).
Gas Flow
Q = Cd∗A∗U (4.5)
For very low pressures where (P0 − P1 /P1 ) ≤ 0.01, a much simplified flow
equation may be used.
Low Pressure Gas Flow.

δg
Q1−2 = 1658.5∗Cd∗a∗ (h0 − h1 ) ∗ (4.6)
δa
where
a is area (sq. inches)

A is area (sq. ft.)
δa is density of air at STP
δg is density of gas at flowing conditions
F is frictional loss in units consistent with pressure
gc is the constant 32.2
h0 is upstream pressure (inches of water)
h1 is downstream pressure (inches of water)
Cd is discharge coefficient
P0 is upstream pressure (lb/sq. ft. abs)
P1 is downstream pressure (lb/sq. ft. abs)
Q1−2 is flow rate (cu. ft./h)
R is universal gas constant (ft lb/lb R0 )
T0 is upstream absolute temperature (R0 )
U is velocity (ft./s)
4.6.2 TOTAL PRESSURE

The pressures, ht and Pt , represent total pressures that are used in connection with
moving air streams. Total pressure includes static pressure and a velocity component.
Moving gas streams have both potential and kinetic energy components. In low-
pressure gas flow, they are generally thought of as total, static, and velocity pressures
(Equation 4.7). Total pressure is the sum of the static and velocity pressures and may
be represented by any consistent set of units.
Total Pressure.
ht = hs + hv + f (4.7)
8417: “8417_c004” — 2007/9/11 — 12:16 — page 73 — #17

where
f is frictional loss [in. water column (w.c.)]

hs is static pressure (in. w.c.)
ht is total pressure (in. w.c.)
hv is velocity pressure (in. w.c.)
An example will be helpful to illustrate the concept of total pressure and the
interchange of static and velocity pressures. Assume a large tank has been pressurized
to a static pressure of hs . Because the tank is large and the outflow is small, there is
negligible velocity within the tank. Therefore within the tank hv = 0. From Equation
4.7, ht = hs . As the fluid approaches and enters the pipe, it must accelerate and
gain velocity. The static pressure diminishes and the velocity pressure increases (see
plane a). Concurrently, because of turbulence there is a permanent frictional head loss
that causes the total pressure, ht to decline slightly (see point e). In the straight section
of pipe, plane (a to b), fluid reaches maximum velocity, requiring a corresponding loss
in static pressure (see plane a). The fluid then enters a divergent nozzle (b to c), where
the velocity diminishes, restoring static pressure. Throughout the pressure interchange
(a to c), there is a gradual loss of total pressure due to friction and turbulence (e to f).
As the fluid emerges from the nozzle, it continues to expand and then it impacts
the wall. The remaining velocity is converted back into static pressure with an accom-
panying friction loss (f to g). The static pressure is now equal to the total pressure
since velocity is now zero. The regaining of velocity pressure into static pressure,
hs , is referred to as stagnation pressure (see Figure 4.5).
Wall
Flow h
Tank
g
e Total pressure f
Friction
Static pressure
Velocity pressure
D a b c d
FIGURE 4.5 Total velocity and static pressure versus position.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 74 — #18

After striking the wall at point h, the static pressure being higher than the ambient
pressure causes the gas to accelerate in a direction perpendicular to the original jet,
again causing another interchange between total, velocity, and static pressure. This
concept is extremely important and will be used throughout discussions of moving
gas streams.
4.6.3 STAGNATION PRESSURE

When a moving gas stream impacts a stationary barrier or wall the stream velocity
diminishes or approaches zero. The kinetic energy of the moving stream is transformed
into potential energy in the form of pressure. The change into potential energy raises
the local static pressure, which is termed stagnation pressure.
Stagnation pressure plays an important role in the inflow and outflow of gases
in a motor home or like enclosure. The stagnation pressure can be due to a moving
body, such as a motor home impacting stationary air, or to a moving air stream or
wind impacting a stationary body. The magnitude of the stagnation pressure may be
calculated from the velocity component and the angle of incidents to the plane of
the impact surface. In low-pressure air flow calculations it is traditional to refer to
velocity in ft./min and stagnation pressure in terms of inches of water column (see
Equation 4.8).
Stagnation Pressure.
2
Um
h = ∗ δ ∗ sin φ (4.8)
1096.5
Where
δ is density of fluid at flowing temperature and pressure (lb/ft3 )

h is stagnation pressure (in. w.c.)
φ is angle of incidence of Vm to plane of stagnation degrees
Um is velocity of stream (ft./min)
A useful chart, Figure 4.6, shows the magnitude of stagnation pressure for air at
various wind speeds impacting a stationary surface. To put this in perspective consider
the following example: a 20 mph wind impacts a stationary wall at an angle of 90◦ .
From Figure 4.6, the stagnation pressure will be 0.20 in. w.c. Assume that there is a
single 2 ft. × 4 ft. closed double hung window in the wall. Assume that the window
has a 1/64 inch crack along all moving surfaces. The total leakage area will be
1
A= × 12 × (2 + 2 + 2 + 4 + 4) = 2.625 sq.in
64
Using Equation 4.6 we can calculate that the air flow through the window crack
will be 1572 acfh, which is a substantial leakage.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 75 — #19

Stagnation pressure (in. water column)

1.40
Stagnation pressure in. water column

1.20
1.00
0.80
0.60
0.40
0.20
0.00
0 10 20 30 40 50 60
Wind speed MPH
FIGURE 4.6 Stagnation pressure.
4.6.4 THERMAL PRESSURE

Any time there is a temperature gradient within or between the inside and outside
of an enclosure, thermal pressure will develop at the top of the enclosure. Thermal
pressure is a function of temperature difference, density of the fluid or fluids, and
vertical height from the inlet to outlet in feet. The usual units for thermal pressure are
inches of w.c. Equation 4.9 represents the equation for thermal pressure:
Thermal pressure.

p∗MW ∗H 1 1
ht = ∗ − (4.9)
55.77 T2 T1
Where
ht is thermal pressure head (in. w.c.)

H is height of the enclosure (ft.)
MW is molecular weight of gas
p is atmospheric pressure (psia)
T2 is absolute atmospheric temperature (Rankine degrees)
T1 is absolute enclosure temperature (Rankine degrees)
Thermal pressure can be treated as any other pressure to determine flow across
an opening (see Equation 4.6). Thermal pressure is important in the flow of exhaust
because of the temperature gradient between the hot exhaust gas and that of the
surrounding atmosphere. To put it into perspective, a temperature gradient of 100◦ F
in a 6-ft. high enclosure will create a thermal pressure of 0.0237 in. w.c. That thermal
pressure would create a flow of 204 scfh per square inch of leakage area.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 76 — #20

4.6.5 AIR FLOW AROUND AND THROUGH ENCLOSURES

Aside from fans, blowers, exhausters, or similar power gas moving devices, thermal
head, and wind are the principal cause of flow of gases into or out of enclosures such
as motor homes or buildings. Any time that wind moves across or around any object,
pressure gradients form around it. It is these pressure gradients acting on openings
in the enclosure that create flow into and out of the enclosure. Figure 4.1 shows an
elevation view of a simplified enclosure (B) exposed to wind flow. The wind blowing
from left to right. The horizontal lines (A) represent stream tubes of air. As the
lower stream tubes approach the vertical wall, their horizontal velocity approaches
zero, converting kinetic energy to static pressure (stagnation pressure). This builds a
positive pressure region along the upwind face. The static pressure gradient on the
upwind face is greatest at the bottom where virtually all kinetic energy or velocity
pressure is converted to static pressure (C). At higher levels there is flow moving
up the wall, therefore static pressure will be somewhat lower due to transition into
velocity. Total pressure remains essentially constant across the face with the exception
that there is a frictional loss due to turbulence.
Air within the positive pressure region (M) flows into the lower pressure region
(D) and accelerates to join the higher velocity stream tubes flowing across the top of
the object at (B). As the velocity increases, static pressure diminishes forming a low
pressure region (F) along the top of the object. This effect may be likened to the low
pressure created as air flows across the top of an airfoil. As the stream tubes flow past,
the downstream edge of the object they deflect down due to the low-pressure vortex
(G) on the downstream side of the body. Surface drag from the stream tubes acting on
the low pressure region (G) cause a clockwise rotation to the low-pressure wake (G).
At very low Reynolds numbers vortex (G) will remain stable in the cavity formed
by the ground and the downstream wall of the object. At Reynolds numbers above
50, which is most common in air flow, vortex (G) becomes unstable and periodically
moves downstream, as represented by vortex (H). Then a new vortex is established and
grows in size at (G) until it separates. This process is referred to as vortex shedding.
While vortex shedding introduces instability downstream for the object, the pressure
in the immediate downstream area while variable in magnitude remains essentially
negative.
When openings exist in the enclosure, as shown in Figure 4.1, positive pressure
on the windward side causes flow into the enclosure through openings (I). As a result
of the inflow, internal pressure increases. The internal pressure being higher than that
on the downwind side causes a pressure differential that results in flow through any
openings (K). The negative pressure (G) further increases the pressure differential
resulting in a further increase of flow. If the enclosure is divided into rooms or other
subdivisions, flow will work its way through openings in the partitions, traveling
from high to low pressure. Flow rate will be governed by pressure differential and
area of leakage paths, in accordance with Equation 4.6. If multiple leaks are involved,
a series of simultaneous equations may be written to determine the net flow through
the enclosure.
It is apparent that if CO was discharged on the upstream side, then it would be
induced into the interior of the enclosure.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 77 — #21

4.6.6 WIND ROSE

When a CO-poisoning incident occurs and external wind conditions appear to be a
factor, it is desirable to be able to determine the wind effect over the time span of
the accident. The wind may be at a constant velocity and direction during the entire
incident or it may vary widely in direction and velocity.
The wind intensity and direction are a vector quantity that may constantly change
with time. Because of the highly variable nature of the wind, it is helpful to develop
a graphic image, as well as a mathematical model of the wind effect relative to time.
There are three factors that are pertinent: velocity, direction, and duration which
add to the difficulty in graphing the event. National Oceanographic and Atmospheric
Administration (NOAA)9 hourly weather reports provide all the data necessary to
evaluate the role of weather on an incident. The reports give wind direction, speed,
temperature, and a number of other factors, for each hourly segment. The NOAA
weather service developed a chart called a wind rose used to assess average wind
direction and speed for given airport locations for the purposes of runway direction.
Using this basic concept as a starting point, I developed a chart that is useful for
forensic investigation of wind-related incidents. Initially, the wind direction, wind
speed and the time for each hourly segment is logged into a chart (Table 4.2). The
data are then plotted on a 360◦ polar graph in which wind direction, degrees, and
intensity (knots) is entered on the polar graph. Entries are shown as 1-h segments
for any given direction. The length of each vector represents the wind speed for
that particular vector. Where there are multiple vectors or any given heading, each
alternating sector is colored for easy differentiation. I normally place time in, {0800},
showing the actual time for each sector so that I do not have to refer to the database for
that information. Plotting wind impact surfaces on the chart also aids in visualization.
The rose (Figure 4.7) shows that during a 24-h period there were two prominent
groupings of wind. The one of interest spans from about 2200 on 2/13 to 0800 on 2/14,
having a wind direction of from 70◦ to 180◦ , which is predominantly out of the east.
In this case, the front of the building is represented by the line from 30◦ to about 210◦ .
The concern was for stagnation pressure on the “East East South” face of the building
within the time frame of the accident. Study of the data (Table 4.2) showed that during
the critical hours winds out of the east created substantial stagnation pressure, thereby,
pressurizing not only the face of the building but all the rooms having openings on
that side of the building. In the case represented by this wind rose there were a number
of large louvers opening into a boiler room that contained the boiler that generated
CO. As a result of the wind, the room was pressurized driving CO into other sections
of the building, which contained the living quarters.
The wind rose also shows that more intense winds out of the southwest occurred
later in the day; however, they occurred outside of the critical time and were of no
consequence.
4.6.7 SCOOPS, DEFLECTORS, AND OPENINGS

It is commonplace to have ventilators and exhaust fans on or in the roof of motor
homes and most portable trailers and offices. These ventilators frequently project into
8417: “8417_c004” — 2007/9/11 — 12:16 — page 78 — #22

TABLE 4.2
Weather Data Giving Wind Direction, Wind Speed, and Temperature
Wind Direction Wind Speed Temperature
Date Time (degree) (knots) (◦ F)
2/13/2004 2200 90 4 33.8
2/13/2004 2300 90 8 35.6
2/13/2004 2400 180 3 32.0
2/14/2004 100 126 1 30.2
2/14/2004 200 290 3 30.2
2/14/2004 300 150 4 30.2
2/14/2004 400 0 0 37.4
2/14/2004 500 180 3 33.8
2/14/2004 600 0 0 33.8
2/14/2004 700 0 0 30.2
2/14/2004 800 140 4 33.8
2/14/2004 900 0 0 37.4
2/14/2004 1000 0 0 44.6
2/14/2004 1100 0 0 48.2
2/14/2004 1200 240 9 53.6
2/14/2004 1300 260 11 55.4
2/14/2004 1400 240 9 55.4
2/14/2004 1500 220 7 57.2
2/14/2004 1600 220 6 55.4
2/14/2004 1700 230 14 55.4
2/14/2004 1800 230 10 53.6
North
360° Front face of building
Wind speed
3 (knots)/ring
270° 90°
(2300)
(2200)
(0100)
(0800)
(0300)
(0500)
(2400)
180°
FIGURE 4.7 Wind rose for 02/13/04 (2200 hrs) to 02/14/04 (1800 hrs).
8417: “8417_c004” — 2007/9/11 — 12:16 — page 79 — #23

A B
C E
FIGURE 4.8 Ventilator acting as a deflector.
A B G
− H
+ −
C + −
+ − −
+
+
F
D
E
FIGURE 4.9 Ventilator acting as a scoop.
the wind stream and induce flow during motion of the vehicle, or in windy conditions
with the vehicle being stationary. Figure 4.8 and 4.9 show pressure gradients and
direction of flow into or out of an enclosure, depending on the direction of wind flow
across the ventilator.
Figure 4.8 shows wind flow (A) from left to right, that is deflected upward as
it encounters the cover sloping up to the right. The increased velocity at B lowers
the static pressure resulting in a low-pressure region at D, downstream of the cover.
Assuming neutral pressure within the enclosure (C), flow from inside the enclosure
(E) will move from the neutral pressure zone (C) to the negative pressure zone (D),
downstream of the deflecting cover. Knowing the relative wind velocity, its angle of
incidence the shape and configuration of the ventilator, and the area of the opening
the flow rate through the enclosure can be calculated using Equation 4.6.
Figure 4.9 shows airflow (A) from left to right that impacts the underside of the
cover (B), causing partial stagnation of stream (C), resulting in a positive pressure
region (D) on the underside of the cover and at the mouth of the ventilator opening.
This assumes that the interior of the enclosure is at neutral pressure (E). The pressure
differential plus the remaining velocity pressure will result in air flowing into the
enclosure. The flow rate entering the enclosure can be calculated using Equation 4.6.
If the pressure within the enclosure is other than neutral, then the differential pressure
(h0 −h1 ) must be adjusted accordingly.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 80 — #24

4.6.8 RESONANCE, CYCLICAL FLOW INTO AND OUT OF AN

ENCLOSURE
Airflow across any opening leading into a closed chamber can result in cyclical pres-
sure pulsations within the enclosure. The cyclical pressure pulsations result in a
cyclical inflow and outflow of gas from the enclosure. The pulsations, depending on
the shape of the opening can result in a net inflow or outflow from the enclosure.
This phenomenon is well known to passengers riding inside an automobile when the
rear windows are partially open. The cyclical low frequency pressure surges are com-
monly referred to as buffeting. The pressure surges can go unnoticed or be significant
to the extent of being painful to the ears.
A similar effect can be experienced inside a motor home or other enclosure when
airflow across an open window or ventilator results in buffeting. The frequency of the
oscillation is a function of the volume of the enclosure and the physical characteristics
of the opening. The technical term for this type of oscillation is helmholtz resonance.
The equation that governs this phenomenon is
Helmholtz resonance.

1 c2∗ S
fo = ∗ ∗ (4.10)
2 π lc∗ V
Where
c is speed of sound, at flowing temperature and pressure of the gas (ft./s)

fo is resonant Frequency (Hz)
lc is effective length of the throat opening
S is cross-sectional area of the throat opening (sq. ft.)
V is volume of the enclosure (cu. ft.)
An example will illustrate the significance of Helmholtz resonance. Assume a

trailer 8
× 30
× 7
has a volume of a 1680 cu. ft. Assume it has a single partially open
window having an open area of 0.44 sq. ft. and a 0.5
throat, thickness of the pane,

at the opening. Also assume that the fluid medium is air at 60◦ F. The frequency of
the oscillation will be 7.1 Hz (cps). This frequency is well below the range of human
hearing; therefore, the oscillation will go undetected. However, it will still induce
large volumes of outside fluid during the negative cycle. It should be noted that low
frequency pulsations, if of sufficient amplitude, could be detected by feeling. The
inducted fluid will then mix with the contents of the enclosure. On the positive cycle,
a similar volume of mixture will be expelled. This of course assumes that the opening
is bidirectional in configuration. If the atmosphere outside the window contains a
significant percentage of CO, then each half cycle will induce CO into the interior.
Over a period of time, the interior concentration of CO will increase and could become
significant. See Section 4.7 for a discussion of the rate of accumulation.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 81 — #25

4.6.9 FLOW INTO AND OUT OF LIVING QUARTERS

After the foregoing discussion it is apparent that flow into or out of an enclosure
occurs only if there is a pressure imbalance and an opening that will allow flow. The
pressure imbalance can be due to a mechanical device like a pump or fan; thermal
activity creating pressure imbalance due to density variations; or wind pressure acting
on the enclosure. The leakage areas may be the obvious ones such as windows, doors,
ventilators, or unsealed openings around pipes or electrical penetrations into the living
compartment. Leaks can also occur at joints in the walls and floors, through cracks
or labyrinth openings that are not apparent to an observer.
In CO poisonings, the percent CO concentration in the enclosure is of prime
importance. CO may be generated within the enclosure or outside and leak into the
enclosure, or a combination of the two. Flow into or out of the enclosure is critical
because it can increase the concentration within or it can dilute the concentration by
bringing in fresh air and washing out CO during the air exchange process.
4.6.10 FLUE AND CHIMNEY DYNAMICS

A flue or chimney is used to remove the products of combustion from inside a building
or enclosure. It operates on the same principle as thermal pressure except that the
pressure is negative when measured at the base of the flue. In addition, because flow
is confined within a solid boundary, the chimney or flue, friction within the duct must
be considered.
Equation 4.11 may be used to calculate the suction created by a flue or chim-
ney. Careful inspection of the equation shows a similarity with that of Equation 4.9,
except that there is an additional back end representing the fluid friction inside the
chimney.
Flue draft

p∗MW∗H 1 1 .09∗H∗U 2 p∗MW
h= ∗ − − 1+ ∗ (4.11)
55.77 Tf Ta D∗2∗g 10.729∗Ta
Where
D is flue diameter (ft.)

g is a constant = 32.2
h is flue suction (in. w.c.)
H is height of the flue or chimney (ft.)
MW Molecular Weight of the flowing gas
p is absolute pressure (psia)
Ta is absolute temperature of ambient air (R◦ )
Tf is absolute temperature of flowing gas (R◦ )
U is the velocity of the flue gas (ft/sec)
8417: “8417_c004” — 2007/9/11 — 12:16 — page 82 — #26

To solve Equation 4.11 it is necessary to assume a flue gas velocity. An acceptable

range to start with is 14.0–16.5 ft./s
4.6.11 BACK-dRAFTING OF FLUE GASES INTO THE ENCLOSURE

Back drafting is one of the common root causes for CO poisoning. Back drafting
is a process whereby the normal direction of exhaust flow is reversed because the
suction inside the enclosure is greater than the thermal suction created as a result
of the temperature difference across the flue. Back drafting can occur in any closed
or semiclosed enclosure. The most common cause is a result of an exhaust fan, or
similar air removal device, overpowering the flue suction. The exhaust device may
be an exhaust fan, a bathroom or kitchen exhauster, clothes dryer or even another
chimney of greater height and temperature difference.
Figure 4.2 shows an example of back drafting through a chimney as a result of a
clothes dryer operating. The room is relatively tight which limits leakage air into the
room. The exhaust fan in the clothes dryer has a capacity, which exceeds the leakage
rate of air into the room, thereby lowering the pressure within the room. If the suction
pressure within the room is equal to or greater than the flue suction, the flow up the
chimney stalls, resulting in a reversal of flow in the chimney.
Ambient air is drawn down the chimney, providing combustion air to the flame
from above. This changes the normal direction of airflow across the flame holder,
disturbing mixing and flame stability. The flame instead of being sucked up the
stack, is blown toward the inside of the enclosure. The exhaust gases then change
direction and are discharged into the enclosure. Additionally, the flow reversal
disturbs the normal combustion process and its flame stability, causing the flame
to blow off, increasing the production of partial products of combustion. Figure
4.2 shows a back drafting chimney as a result of a suction created by a clothes
dryer.
Back drafting has become much more common as homes and buildings becoming
tighter (less air leaks) and the prevalence of air exhausting appliances like clothes
dryers and exhaust fans, roof exhausters and multiple chimneys becomes more com-
mon. The way to avoid the problem is to install appliances that have their own outside
air intake, which are frequently power driven. Another method would be to install a
general outside air intake.
Motor homes, trailers and portable buildings, because of their requirement to seal
the living compartment tend to have fewer air leaks, thereby, become prone to back
drafting when exhaust devices are used.
4.6.12 DYNAMICS OF GENERATOR TAILPIPE EXHAUST

Generator tailpipe exhaust, by code, must always be external to the living compart-
ment. As described earlier, the tailpipe is generally perpendicular to the outside wall
of a motor home, is below, and protrudes beyond its periphery. For the purpose of
perspective, consider that a 4 kW single cylinder, four-cycle gasoline-fueled engine
(Octane, C8 H18 , will be used as the formula for gasoline) generator will be used.
At a 3 kW load the exhaust volume will be about 2030 acfh at 800◦ F. At the 1.1 in.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 83 — #27

diameter exhaust pipe, typical for that size generator, exit velocity will be 85 ft./s.
When the exhaust jet enters stationary ambient air, the surface drag of the jet stream
moves ambient air along with it. Ambient air is entrained into the outer portion of
the jet stream causing the outer layer to lose velocity and through drag eventually to
slow the core stream. The velocity profile across the jet will be Gaussian as shown
in Figure 4.10. As cold air is entrained into the hot jet the average temperature of
the jet diminishes. Normally it would be expected that the hot jet would have a tend-
ency to rise, due to its buoyancy. However, in the early stages of the jet, the first
20–30 diameters, the horizontal velocity is high compared to the vertical velocity
resulting from thermal buoyancy. Therefore, the vertical movement of the jet will be
small.
In the example shown in Figure 4.10 the vertical rise at 36 diameters is approxim-
ately 1.5 diameters, which is not significant. When the horizontal jet velocity slows,
the thermal rise becomes more significant.
From Figure 4.11, the entrainment at 36 diameters (3 ft.) will be in the order
17:1 and could be as high as 30:1. The example used a 17:1 entrainment ratio.
Gaussian velocity
Entrainment of distribution Thermal rise
ambient gas
l.45 D
0D 4D 8D 12 D 16 D 20 D 24 D 36 D
Exhaust pipe Jet profile
FIGURE 4.10 Expanding jet.
Entrainment ratio versus distance

100
Velocity fps; entrainment ratio
80
60
40 Velocity
20 Entrainment ratio
−20
0 5 10 15 20
Distance in nozzle diameters
FIGURE 4.11 Jet velocity and entrainment versus jet distance in diameters.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 84 — #28

Wall
Trapped exhaust gas
Direction of flow
of exhaust gas
Motor home
Entrained exhaust
Tailpipe
Exhaust discharge
FIGURE 4.12 Entrainment in a confined area.
The initial CO concentration would be diluted by a factor of 17:1. The CO con-

centration in the vicinity of the jet at 3 ft., for an initial CO concentration of 40,000
ppm, would be approximately 2350 ppm. Further dilution would occur at greater
distances downstream.
If the exhaust jet is discharged into a region that has limited air circulation or that
traps or confines the exhaust, then the jet will be immersed in a cloud of exhaust gases
(see Figure 4.12). Initially, the exhaust jet entrains ambient air diluting the exhaust
concentration. As the jet strikes the wall, its velocity drops to zero and the exhaust
gases stagnate. Because of the confinement, the exhaust gases accumulate between
the road, the wall, and the motor home. The entraining action of the jet instead of
drawing in fresh air and diluting the CO level in the jet, it entrains and re-circulates
exhaust gas, gradually concentrating it, up to the maximum concentration in the
exhaust jet.
On the basis of this behavior, it is important to insure that the exhaust system is
in good condition and discharges outside the motor home and into an area, that is
unrestricted where exhaust gases can readily dissipate.
4.6.13 EXHAUST FLOW THROUGH BENT OR DISPLACED

EXHAUST PIPES
Because exhaust systems are usually located in exposed locations they are prone to
impact and may become damaged, bent or rotated out of position. When this happens,
they may discharge exhaust gases under the motor home or possibly discharge it
against the inside of the skirt or other interior surfaces. If the jet discharges under the
vehicle, there is a tendency for the exhaust to become stagnant there. As a result, the
exhaust jet entrains the locally stagnant exhaust and the CO does not dissipate. If
the jet strikes another object, the jet loses its velocity, therefore its entraining ability.
The exhaust becomes stagnant and does not dilute or dissipate. The slow moving or
stagnant cloud of highly concentrated CO remains under the motor home and inside
the generator compartment.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 85 — #29

Without entrainment the highly concentrated CO cloud remains at high temperat-

ure, therefore is highly buoyant, creating thermal pressure on any leaks that may exist
into the living compartment. In addition, any suction pressure within the enclosure,
resulting from wind or other effects, will add to the entrainment of the offending
gases. The concentrated pocket of exhaust under the floor of the motor home is in a
location where leaks into the living quarters are more likely.
If the exhaust jet impinges on the skirt or other surface at an angle, the jet stream
will attach to and flow along the wall, (Coanda effect or wall jet attraction, Henry
Coanda, French aviator, ca. 1918). The flattened jet is under negative pressure, caus-
ing it to entrain local ambient gases. Since this occurs in a confined area under the
motor home, the gas being entrained is primarily exhaust gases, resulting in a high
concentration of CO. The concentrated gases are in a location that is prone to have
leaks into the living compartment.
4.6.14 MISSING EXHAUST SYSTEMS

If the exhaust system, either the tailpipe or the muffler has been broken off, the remain-
ing vertical spud directs the exhaust jet toward the ground. The jet characteristics are
similar to those previously described. The hot exhaust gas jet entrains ambient gas
and the mixed jet then impacts the ground.
The hot jet after striking the ground flows outward from the point of impact
and because of its buoyancy then flows upward. Thereafter, a recirculation pattern
develops (Figure 4.13). The exhaust cloud remains in position, continually being
regenerated. The suction in the exhaust jet entrains local gases, which are primarily
exhaust gas and fresh air dilution is minimal. The re-entrained exhaust jet builds
up a stationary, highly concentrated CO cloud under the generator and motor home.
Winds acting perpendicular to the exhaust stream could move the cloud to a safer
location depending on the direction and magnitude of the wind stream and terrain or
boundaries in the vicinity of the generator.
The stationary cloud has a relatively high temperature and therefore creates
thermal pressure on the underside of the motor home and generator compartment
where the probability of leaks into the living compartment is most likely.
Generator compartment
Muffler spud
Generator platform Hot exhaust gas
Pavement
Recirculation vortex
FIGURE 4.13 Jet entrainment and recirculation under a motor generator without a tail pipe.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 86 — #30

4.7 ACCUMULATION OF CARBON MONOXIDE

WITHIN AN ENCLOSURE
The extent of disability or death of individuals exposed to CO and other toxic exhaust
gas products of combustion is directly related to the partial pressure of the toxic gases
within the enclosure, the duration of exposure, as well as the physical condition and
rate of exertion of the individual and presence of other gases such as carbon dioxide.
This discussion does not consider the medical aspects of such poisoning and refers
the reader to other authoritative texts on the subject (other chapters in this book,
and10,11 ).
Earlier I described the sources of exhaust gases that are produced either within
the enclosure or the volume flow rate of exhaust gases leaking into the enclosure from
exterior sources. This section describes the procedures for determining the exhaust
gas concentration at any specific time in the CO accumulation cycle. This information
in conjunction with the Coburn–Forster–Kane equation may be used to evaluate the
individual exposure to CO and their resultant COHb.
Equation 4.14 evaluates the exponential increase of CO concentration within an
enclosure that is initially filled with clean uncontaminated air. The program assumes
that perfect mixing will occur within the enclosure. This is a reasonable assumption
because the contaminating fluid is either jetted into the enclosure or is drawn in as
a result of pressure differential, or is due to a thermal plume or jet with a sizable
thermal head, all of which create mixing. In addition, the enclosures are generally
small and are not subdivided into tight rooms or enclosures, that might create isolation
or stratification of the leakage gases.
Equation 4.14 allows for simultaneous introduction of exhaust gas, CO and also
leakage of uncontaminated outside air. This same equation may be used to evaluate the
concentration of other component gases contained in the exhaust gas by substituting
the gas concentration in ppm, that is, QCO2 for QCO in ppm.
Volume flow of carbon monoxide in incoming exhaust (CFM).
QCO = QX ∗ CO/1000000 (4.12)
Total volume of exhaust plus leakage air into the enclosure (CFM).
QMm = QLm + QX (4.13)
Concentration of carbon monoxide at time t.

100∗ QCO ∗ (QMm )∗ tn
Cn,m = 1−e V (4.14)
QMm
where
CO is CO concentration, ppm, in incoming exhaust

Cn,m is CO concentration within the enclosure at time tn and QMm , ppm
e is a constant = 2.7183
8417: “8417_c004” — 2007/9/11 — 12:16 — page 87 — #31

m is step iterator (cfm)

n is iterator of time (min.)
Qco is volume flow of CO in the incoming exhaust gas (cfm)
QL m is leakage of air entering the enclosure. Iterated by m (cfm)
QMm is total volume of exhaust plus leakage air entering the enclosure (cfm)
QX is volume of exhaust entering the enclosure (cfm)
tn is the iterated value of time in min
V is net volume of the enclosure (cu. ft.)
Example: Assume a motor home having a 1680 cu. ft. volume, initially containing
clean air, and with two openings, one at the top and one near the floor. Assume there is
a 150 cfm roof exhauster located in the ceiling. Assume another motor home is parked
parallel to this one whose motor generator is discharging exhaust gases containing
50,000 ppm CO in the direction of the first motor home, near the lower opening.
Because of dilution and the size opening, assume that 100 cfm of diluted exhaust
containing 2000 ppm of CO is drawn into the lower opening of the first motor home.
Simultaneously, 50 cfm of clean air, (150 fan-100 exhaust), is drawn in through
the upper opening. Determine the CO concentration with respect to time. Plot CO
concentration when the total volume of exhaust plus leakage, QM m = 100, 125, 150,
and 175, 200 cfm.
Figure 4.14 shows the CO concentration in ppm within the enclosure with respect
to time. Iterative computations were done for Qmx = 100 to 200 cfm. The middle
dash line (m = 2) represents a flow of 150 cfm, when m = 150, which is equal to the
exhaust fan volume of 150 cfm. The series of curves shows how CO concentration
changes with higher and lower leakage flow rates, QL, into the enclosure.
CO concentration (ppm) at time tn

2000
C ppm0,n
C ppm1,n
CO concentration (ppm)
1500 C ppm2,n
C ppm3,n
1000 C ppm4,n
500
0
0 20 40 60 80 100
tn
Time (min)
FIGURE 4.14 Accumulation of carbon monoxide within an enclosure.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 88 — #32

Based on this information it is possible to predict the CO concentration at any

period of time or to integrate the area under the curve, and to divide by time to
determine the average concentration for any time span.
Some important points may be made regarding this set of curves:
1. The solid curve, representing zero air leakage, shows that the CO concen-
tration in the enclosure reaches a maximum concentration in the incoming
exhaust gas after about 80 min.
2. Each successive curve maxes out at a lower concentration, all at about 80
min or less.
3. The CO concentration rises to a level dangerous to humans within 6 or
7 min.
The CO concentration, for the example given, rises very rapidly for the first 20
min, reaching about 1100 ppm. Thereafter, the rate of rise diminishes and levels
off at a maximum CO concentration of 1300 ppm at about 50 min. Thereafter the
concentration stabilizes and remains constant.
This information derived from this calculation can then be used in conjunction
with the Coburn, Forster, Kane (CFK) equation, to determine blood COHb rise for
any given period of time.
4.7.1 RATE OF DISSIPATION OF CARBON MONOXIDE FROM AN

ENCLOSURE
Earlier I discussed the rate of accumulation of CO within an enclosure. This section
directs its attention to the dissipation of the accumulated CO when the rate of influx
of CO, QCO , decreases or is zero, while a fan or exhauster continues bringing in
fresh air.
Dissipation of carbon monoxide within an enclosure.

βn,m = ppmx ∗ e−((QCO + Qx) ∗ tn /V ) (4.15)
where
βn,m is CO concentration in the enclosure at the end of time tn (ppm).

n is time iterator (min)
ppmx is CO concentration at the start of the reduction
QCO is volume flow rate of CO entering the enclosure, if any (cu. ft./min)
Qx is volume flow rate of clean air entering the enclosure (cu. ft./min)
t is time (min)
V is volume of enclosure (cu. ft.)
Reconsider the previous example with the following modifications: Assume that
after 20 min of operation the motor generator is turned off, but all other conditions
remain the same. Determine the rate of dissipation of CO within the enclosure with
8417: “8417_c004” — 2007/9/11 — 12:16 — page 89 — #33

Dissipation of CO versus time

1500
1200 β 2, n
Co concentration (ppm)
900
600
300
0
0 20 40 60 80 100
tn
Time (min)
FIGURE 4.15 Dissipation of carbon monoxide from an enclosure.
respect to time. The exhaust fan, still running, takes up 150 cu. ft./min of fresh air
which dilutes the 1100 ppm CO concentration within the enclosure. Equation 4.15
may be used to calculate the CO concentration after any time period. Figure 4.15
shows the dissipation curve starting at 1100 ppm at time t = 0 which correlates with
Figure 4.14 at 20 min.
Calculation of the type shown in section 4.7 and 4.7.1 can be used to calculate
CO concentration where the source of CO cycles on and off, as would occur with a
thermostatically controlled furnace. Depending on the on-off cycle time, the resultant
graph would appear as a saw tooth curve. Integration of the area under that curve
divided by time would yield the average concentration over any selected time span.
If this information is to be used in the determination of blood COHb, a similar
saw tooth curve would have to be produced accounting for the rise in COHb during
periods where the CO partial pressure of the atmosphere exceeds the partial pressure
of CO in the lungs. Also half-life calculations would have to be done to account for
decreasing COHb where the partial pressure of CO in the enclosure is below that of
the lungs during off time.
4.8 TEST METHODS FOR DETERMINING THE

LOCATION OF LEAKAGE INTO AN ENCLOSURE
In cases where CO intrusion into the living compartment occurs, it becomes important
to determine the location and relative size of the opening or breach in the living
compartment. NFPA3 and ANSI2 safety codes require that the living compartment
of a mobile home or trailer equipped with a motor generator must be sealed against
8417: “8417_c004” — 2007/9/11 — 12:16 — page 90 — #34

intrusion of exhaust gases. Therefore, any opening or potential leakage path into the
living compartment that is not sealed is of significance.
Since exhaust gases, CO, carbon dioxide and nitrogen, are all colorless, odorless,
and tasteless, it is not possible to detect their leakage using normal sensory perceptions.
Scientific instruments, that detect the presence of the offending gas, may be used as
long as it is possible to replicate original conditions regarding flow and generation of
CO. Replication, particularly when wind or atmospheric conditions play an active role
is usually very difficult. Additives to the gas stream may be used to enhance sensory
perception as long as they do not affect the results. Smoke testing has proven to be
simple, inexpensive and accurately shows the magnitude and position of leaks.
4.8.1 SMOKE TESTING

Injecting smoke into an area where exhaust gases were known to accumulate is a
practical way to make leakage into the living compartment visible, without appre-
ciably changing the physical characteristics of the leakage stream. Visual contrast of
smoke is good, as are photographic characteristics. Smoke makes it easy to detect
the location of small and large leaks as well as to assess their magnitude. It should
be pointed out that this is not an attempt to replicate the original exhaust gas stream.
It is intended to show leakage paths and give an order of magnitude to the leakage
stream.
Commercial smoke-producing devices are readily available, which are inexpens-
ive and safe to operate. These include candles of various sizes as well as other smoke
generators. Candles will generate smoke over periods from 30 s to 3 min, in volumes
from 4,000 to 40,000 cu. ft. Smoke generators can provide a continuous stream of
high quality smoke where required.
Colored smoke is available in candle form. Colored smoke gives excellent contrast
especially in bland backgrounds, however, colored smoke is apt to cause permanent
stains. Most manufacturers state that their smoke is not toxic. My experience has
been that breathing smoke, even for a short duration, leaves an aftertaste, and lung
congestion that is objectionable, and that persists for many hours. The use of a good
high quality respirator seems to alleviate that problem.
Smoke particle size in ranges from 2 µm (0.00004 in.) to 60 µm (0.0024 in.)
therefore, has little effect on changing the flow characteristics of a gas stream. The
smoke is stable and persistent, having a settling time of about 1.5 h, which is more
than enough for most tests. The smoke, because of its small size will penetrate very
small openings. Smoke can be released in open air or injected into the intake of a
portable blower, when a pressurized stream is required.
The internet is a good source for smoke candles or generators. Smoke candles, or
the fluid used for a smoke generators are not permitted as carry on or shipped baggage
on commercial aircraft.
4.8.2 BASIC CONSIDERATIONS DURING SMOKE TESTING

In order to have either an inflow or an outflow of gases there must be a pressure
differential between the inside and outside of the enclosure. The simplest way to
8417: “8417_c004” — 2007/9/11 — 12:16 — page 91 — #35

accomplish the pressure differential is to increase or decrease the enclosure internal

pressure while allowing the ambient or outside pressure to remain at atmospheric
conditions. Negative internal pressure replicates the conditions that would normally
draw CO into the living compartment, therefore tends to replicates the actual accident
conditions, assuming that the leakage areas have not been changed. Under some
atmospheric conditions, generally high ambient temperatures, a thermal head may
develop as a result of temperature differentials occurring during testing and must be
overcome by additional pressure differential. Tests can be done by pressurizing or
by creating a negative pressure within the enclosure. The specific type of test will be
dictated by the circumstances.
4.8.3 POSITIVE PRESSURE TESTING

Positive pressure testing requires a means for increasing internal pressure within the
motor home or enclosure. This requires the installation of a turbo blower or other
device that is not normally found in this type of enclosure. Additionally it requires
that exhaust fans, ventilators, dryers, and similar openings be sealed during the test.
Observation of small quantities of leaking smoke may be difficult to see in ambient or
bright light conditions. Windy conditions will tend to dilute all but very large leakage
streams. The diluted streams are much more difficult to photograph in bright sunlight
where it may be impossible to control front lighting. Many of these problems can be
eliminated if the testing can be done inside a building where wind and lighting can
be controlled.
4.8.4 NEGATIVE PRESSURE TESTING

Negative pressure testing is done by lowering the pressure inside the enclosure while
providing a reservoir of dense smoke on the outside of the enclosure. It is desirable to
locate the source of the smoke in the area where leaks are suspected. The advantage
of this type of testing is that most motor homes, trailers or portable buildings already
have exhaust ventilators, bathroom exhausters, fume hoods and dryers, that can be
used to lower the enclosure pressure. Additionally, it is easier to see the smoke in
interior controlled lighting conditions, where front and side lighting is easily achieved.
Most tests can be performed without modification to the living enclosure. This is a
distinct advantage in production testing but it is especially valuable in forensic cases
where approval for modifications would have to be obtained, in advance of testing,
from opposing counsel.
Most CO intrusion incidents are from auxiliary generator engine exhaust that is
normally discharged outside the living enclosure. As a generalization, motor homes
develop exterior leaks in the floors and transition area of walls to floors, around wheel
wells, and around conduits or pipes that penetrate the living enclosure. Frequently
the leakage path is convoluted and difficult to detect using a light source or by trying
to push a flexible object through the opening. Smoke testing is an easy and efficient
way to detect these types of openings.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 92 — #36

4.8.5 TYPICAL SMOKE TESTING

Smoke testing will give a good visual picture of the location and magnitude of leaks
into a living compartment. Still and video photographs provide a means of recording
real time results of the test. If photographs are judiciously taken they can be used to
determine relative location and relative magnitude of leaks. Photos and videos are
invaluable in forensic cases since they provide visual proof that leakage paths are
present in the living compartment, which is in direct violation of NFPA and ANSI
safety codes.
Pictures are worth a thousand words, but only if it they are clear and depict the
objects of interest accurately without distortion. As an example, the photographs in
Figures 4.16 through Figure 4.22 tell a story of how and where CO leaked into the
motor home. They show a replication, using smoke as a visible medium, of exhaust
gas entering a motor home. The desired effect in this case was to show how large
volumes of CO entered the motor home in a very short period of time. The photographs
clearly show how the accumulation of smoke (the smoke representing CO) occurred.
At the start of the test in the example shown, the interior of the motor home was
clear without any obstruction. Figure 4.16 shows a view looking toward the back
of the motor home. This shows that the internal atmosphere is clear. An additional
photograph taken looking forward through the motor home and out the windshield
shows that that section of the motor home is also free of haze (Figure 4.17).
Figure 4.18 shows the first wisps of smoke appearing through the upper drawer of
the end table. This photograph was taken after 1 min of elapsed time (ET) following
initiation of smoke. This leak occurred at the left rear corner of the motor home.
This was particularly important because that is where the generator was located. The
smoke candle was located directly under the generator. The smoke traveled through
a convoluted passage that could not be detected during inspection.
FIGURE 4.16 Looking aft before testing.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 93 — #37

FIGURE 4.17 Looking forward before testing.
FIGURE 4.18 Left rear corner, first smoke 1 min elapsed time.
From preliminary testing it was suspected that a large leak existed in the middle-
area of the motor home. Thus attention was focused on that area. Figure 4.19 shows
a view looking aft from the middle-point of the motor home. A large discharge of
smoke was detected coming from the lower part of a storage cabinet. The interior
atmosphere has become partially obscured at 2 min ET. Figure 4.20, 25 s later, shows
obscuration has all but eliminated all detail in the previous photograph. Figure 4.21
shows total obscuration of the cabinet shown on the right of the previous photograph.
This series of six photographs shows the location of the two leaks, and it shows that
the mid point leak is the larger of the two. It shows that the rate of leakage is very
8417: “8417_c004” — 2007/9/11 — 12:16 — page 94 — #38

FIGURE 4.19 Looking aft second leak area, 2 min elapsed time.
FIGURE 4.20 Looking aft second leak 2 min 25 s elapsed time.
high in the interior part of the vehicle. The photographs show that the leaks were
very large and smoke permeated the interior of the vehicle in a very short time. Total
obscuration took place in three minutes. Timing with a stop watch, backed up by the
time record from the digital and video cameras, used in conjunction with the known
volume of the vehicle, made it possible to approximate the leakage rate of smoke into
the interior of the vehicle.
I have found that smoke testing is an efficient means of determining the location
of leaks and also their relative magnitude. It has proven to be an invaluable asset in
forensic cases.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 95 — #39

FIGURE 4.21 Second leak, totally obscured cabinet, 3 min elapsed time.
Exhaust gas percent dry methane CH4

20
18
16
14
Percent (dry)
DEN*100
12
CO2
10
CO
8
O2
6
4
2
0
0 1 2 3 4 5 6
Equivalence ratio
FIGURE 4.22 Exhaust gas percent dry methane.
4.9 CONCLUSION
Using the techniques presented in this paper, the source and magnitude of CO leaks
into mobile homes and other enclosures may be determined. Photographs of the smoke
intrusion provide a visual demonstration of the presence of leaks.
References
1. Mah, J.C. Non-Fire Carbon Monoxide Deaths and injuries Associated with the Use
of Consumer Products U.S. Consumer Products Safety Commission, October, 2000.
2. American National Standards Institute, Washington, DC.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 96 — #40

3. National Fire Protection Association, Quincy, MA.

4. Environmental Protection Agency. Air Quality Criteria for Carbon Monoxide, EPA
600/P-00/001B, Tables 3–7, 1999, Cincinnati, OH.
5. Boltz R. E. and Tuve G. L. Handbook of Tables for Applied Engineering Science, 2nd
edition, CRC Press, Boca Raton, FL, 1987.
6. Williams, G. C. et al., The Combustion of Methane in a Jet Mixed Reactor, Twelfth
Symposium on Combustion, Pittsburg, PA. 1968.
7. Jones, J.C. Combustion Sciences: Principles and Practices, Millennium Books,
Newton, Australia, 1933.
8. Linan, L. and Williams, F. Fundamental Aspects of Combustion, Oxford University
Press, Oxford, UK, 1993.
9. National Oceanographic and Atmospheric Administration, Washington, DC
10. Penney, D.G. Carbon Monoxide, p. 296, CRC Press, NY, 1996.
11. Penney, D.G. Carbon Monoxide Toxicity, p. 560, CRC Press, NY, 2000.
8417: “8417_c004” — 2007/9/11 — 12:16 — page 97 — #41

8417: “8417_c004” — 2007/9/11 — 12:16 — page 98 — #42
5 Carbon Monoxide
Emissions from Gas
Ranges and the
Development of a Field
Protocol for Measuring
CO Emissions
Richard Karg
CONTENTS
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

5.1.1 Objectives of Karg Field Protocol Development. . . . . . . . . . . . . . . . . . 102
5.1.2 ANSI Standard Z21.1-1993, Household Cooking Gas Appliances 102
5.1.3 As-Measured and Air-Free Carbon Monoxide Measurement . . . . . 103
5.1.4 Single-Zone Mass Balance Model and Room CO Concentrations 106
5.1.5 Field Testing, Laboratory Research, and Karg Field Protocol
Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
5.1.5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
5.1.5.2 Range Top Burners, Field Testing . . . . . . . . . . . . . . . . . . . . . . . 109
5.1.5.3 Range Top Burners, Laboratory Testing . . . . . . . . . . . . . . . . 112
5.1.5.4 Oven Bake Burners, Field Testing . . . . . . . . . . . . . . . . . . . . . . 115
5.2 Karg Field Protocol for Measuring Carbon Monoxide Emissions from
Gas Ranges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
5.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
5.2.2 Visual Inspection and Customer
Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
5.2.2.1 Range Top Inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
5.2.2.2 Oven Inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
5.2.3 Measurement of Emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
5.2.3.1 Safety During Measurement of Emissions. . . . . . . . . . . . . . 124
5.2.3.2 Preparation for Burner Testing . . . . . . . . . . . . . . . . . . . . . . . . . . 124
5.2.3.3 Range Top Burner Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
99
8417: “8417_c005” — 2007/9/11 — 12:16 — page 99 — #1

5.2.3.4 Oven Bake Burner Testing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

5.2.3.5 Burner or Range Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
5.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
5.4 Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
5.1 INTRODUCTION
Indoor air quality studies and carbon monoxide (CO) emissions testing from combus-
tion appliances have pointed to a number of factors affecting indoor CO levels: source
and source-use characteristics, building features, ventilation rates, air mixing within
and between rooms, the existence and effectiveness of contaminant removal systems,
and outdoor concentrations.1 Gas ranges are used by a significant percentage of the
population and, as an unvented combustion appliance, can contribute to hazardous
concentrations of CO in indoor air when they malfunction or are misused.
The U.S. Consumer Products Safety Commission (CPSC) estimates that 130
people died in 2001 from nonfire, nonauto related CO poisonings; 75 of these deaths
were attributed to heating systems and 10 to gas ranges/ovens.2 From 1990 to 2001 the
CPSC estimated between 6 and 14 people died each year from CO emitted from gas
ranges/ovens.2−5 Data for the years 1994–1998 show an additional 600–900 per year
suffered nonfatal poisonings.3 It is obvious from these data that gas ranges/ovens can
malfunction and emit hazardous levels of CO. To reduce injuries and deaths from CO
poisonings, gas range top burners and ovens should be tested regularly and maintained
properly.
Testing combustion appliances for CO emissions during servicing or home
weatherization is not currently a standardized practice in the United States, although
the low-cost electronic combustion analyzers available today make it easier to test
CO emissions in the field. The lack of standard field testing procedures for measuring
CO emissions, the uncertainty of the accuracy of field measurements, and the dearth
of technicians trained to consider the complex nature of CO diagnostics have led to
widely variable emission measurement protocols or to a total lack of field testing.6−10
Heating systems are more likely than gas cooking appliances to receive regular
servicing.5 Most natural gas utilities recommend servicing of gas heating systems
every 1–3 years. Gas ranges/ovens are seldom, if ever, serviced by a trained tech-
nician. Tilkalky suggested: “. . . routine service adjustments on the 10 range set
resulted in statistically significant decreases in CO and NO2 following servicing.
This finding implies that some servicing of the gas range may be desirable and
have an impact on indoor levels of NO2 and CO, and should be investigated fur-
ther on a larger data set to determine if this initial indication holds true for a larger
population.”6 Furthermore, if a homeowner desires servicing of a cooking appliance,
anecdotal reports suggest that they might have difficulty finding a trained and willing
technician.11,12
From 1995 to 1998 the administrative office for the Ohio low-income weatheri-
zation program, funded by the U.S. Department of Energy (DOE), conducted a survey
of the 50 state and District of Columbia low-income weatherization programs to
determine the CO emission testing methods used by these programs on all combustion
8417: “8417_c005” — 2007/9/11 — 12:16 — page 100 — #2

Carbon Monoxide Emissions from Gas Ranges and the Development of a Field 101
appliance types. Many of the questions used on these Ohio surveys were related to
gas range tops and ovens. Tables 5.1 and 5.2 include oven and range top data from the
1998 Ohio survey.13 The wide variation and arbitrary nature of these testing methods
was alarming.
This lack of consistency among testing methods within this DOE program
demonstrated the need for a standardized CO emissions testing protocol for tech-
nicians in the field. In addition, there were other reasons to develop a field protocol,
TABLE 5.1
Gas Range Top Burner Testing Data from Ohio Survey of carbon monoxide
Protocols, 51 State Low-Income Weatherization Programs, 199813
Do you have a CO emissions Yes = 37 No = 14
protocol? n = 51
Pan on burner for test? n = 31 Yes = 4 No = 27
Sampling probe height Range = 6–36, Mean = 13.8,
above burner (inches)? n = 28 SD = 8.2, Median = 12,
Favored = 12
Highest acceptable CO
reading (ppm)? n = 26 As-measured, n = 25 Air-free, n = 1
Range = 9–300 Value = 100
Mean = 58
SD = 62
Median = 50
Favored = 100
Burner warm up time before
test (minutes)? n = 25 Range = 0.25–20, Mean = 5.4,
SD = 3.5, Median = 5, Favored = 5
TABLE 5.2
Gas Oven Testing Data from Ohio Survey of CO Protocols, 51 State Low-
Income Weatherization Programs, 199813
Do you have a CO emissions protocol? n = 51 Yes = 37 No = 14

Oven door open or closed? n = 36 Open = 10 Closed = 26
Oven control on broil or bake? n = 42 Broil = 19 Bake = 23
Highest acceptable CO reading (ppm)? n = 32 As-measured, n = 28 Air-free, n = 4
Range = 9–400 Range =
Mean = 110 35–400
SD = 103 Mean = 158
Median = 100 SD = 164
Favored = 50 and 100 Median = 100
Favored = 100
Burner warm up time before test? n = 34 Range = “at startup” to “steady-state”, Favored =
5 or 10 min
8417: “8417_c005” — 2007/9/11 — 12:16 — page 101 — #3

including (1) the use of gas ovens/ranges was widespread, (2) these appliances were
often used as space-heating devices dwellings were becoming tighter, and (3) gas
ranges and ovens were not regularly cleaned and tuned.14
The author was asked by the U.S. DOE’s Low-Income Weatherization Programs
of the Chicago Region to develop the Field Protocol for Gas Range Carbon Monoxide
Emissions Testing in 2001, hereafter referred to as the Karg Field Protocol.15,16
5.1.1 OBJECTIVES OF KARG FIELD PROTOCOL DEVELOPMENT

The primary objective of developing the Karg Field Protocol was to identify range
top and oven burners that are high emitters of CO. As residences are made tighter
to save energy, a given CO emission for a gas range can become more hazardous to
the occupants. This is discussed in Section 5.1.4 with the use of the single-zone mass
balance model. Of course, the possibility of an increased hazard to occupants is a
concern to the management and technicians working with weatherization programs.
Secondly, the Karg Field Protocol was developed to bring standardization to
range testing among 51 DOE low-income weatherization programs. The Ohio survey
of emissions testing methods clearly demonstrated this need and made it obvious that
there was no existing consensus.
The final objective was to minimize arbitrary or biased evaluation by field
technicians.
5.1.2 ANSI STANDARD Z21.1-1993, HOUSEHOLD COOKING

GAS APPLIANCES
Two important questions had to be answered while developing the Karg Field
Protocol: What basis should be used to develop the threshold level of CO emissions
at burners to ensure an acceptable ambient level for the occupants of a residence?
Second, at what time after burner start-up should readings be recorded and what
method of measurement should be used to accurately reflect the long-term CO emis-
sions and thus, the potential hazard of a burner’s emissions to occupants. This second
question is discussed in Section 5.1.5.4.
Manufacturers of gas ranges must comply with American National Standards
Institute document Household Cooking Gas Appliances (ANSI Z21.1, 1993).17 This
standard states “An appliance shall not produce a concentration of CO in excess of
0.08% (800 ppm) in an air-free sample of the flue gases when the appliance is tested in
a room having approximately a normal oxygen supply.”17 The test procedure required
by this standard is lengthy, but simply stated: containers filled with five pounds of
water are placed on each of the four range top burners. All the range top burners and
the oven burner are started simultaneously. After 5 min of operation, the CO air-free
is measured in a collection hood above the range. This air-free concentration must be
800 ppm or less.
In 1996, the Gas Research Institute published a topical report entitled Critique
of ANSI.Z21.1 Standard for CO Emissions from Gas-Fired Ovens and Ranges.8
This Battelle Laboratory appraisal of the 800 ppm emission level, originally set
in 1925, found “The underlying basis for and allowable limit set by the original
8417: “8417_c005” — 2007/9/11 — 12:16 — page 102 — #4

CO standard were valid and conservative in 1925, and remain so today. . .”8 The
report continues with “Gas ovens/ranges do not pose a public safety or health threat
with regard to CO emissions, a performance characteristic that can be reliably val-
idated either in the lab, factory, or field by using the current Z21.1 measurement
protocol, and its specified limit of 800 ppm CO, O2 -free.”8 Although not all research-
ers agree with this conclusion,10 the gas range Karg Field Protocol used this American
National Standards Institute (ANSI) level of 800 ppm CO air-free as the basis for the
development of CO emission thresholds.
It would have been problematic to select a basis for the Karg Field Protocol other
than ANSI Z21.1, since it is the current standard by which manufacturers design and
build gas ranges and ovens. A field protocol based on values more or less stringent
than ANSI Z21.1 would have limited credibility within the industry and would have
compromised field technicians. Therefore the Karg Field Protocol was designed to
determine if a gas range complies with ANSI Z21.1.
5.1.3 AS-MEASURED AND AIR-FREE CARBON MONOXIDE

MEASUREMENT
Gas ranges are probably the most common unvented gas appliances in use in North
America. Because they are unvented, they have the potential of contributing to unac-
ceptable ambient indoor levels of CO. Fortunately, in most residential situations,
they are not operated long enough to adversely affect occupants. However, there are
cases where gas oven/ranges can emit enough CO to cause health hazards or trigger
CO alarms.18 Examples include: (1) user operation of a gas oven/range as a space-
heating appliance;14 (2) user alteration of the oven, such as lining the oven bottom
with aluminum foil, inadvertently covering the secondary air ports;5,19 (3) malfunc-
tioning equipment, including excessive gas pressure, closed air shutters, damaged
orifices, and warped oven flame spreaders;16 and (4) baking a large roast or fowl.
As the awareness of CO hazards increases, more technicians are using electronic
instruments for measurement of CO emissions. These electronic devices are con-
nected to a flexible plastic tube with a hollow metal probe at the other end. When
measuring CO emissions from a vented appliance, the metal probe is inserted into
a drilled hole in the metal flue pipe. A vacuum pump in the measurement device
pulls a combustion gas sample through the metal probe and plastic tube, into the CO
measurement cell, and finally expels the gases through an exhaust port.
Measurement of CO emissions from a gas oven is performed with a similar tech-
nique, however, rather than inserting the probe into a drilled hole in a vent pipe, the
metal probe is inserted into the oven-vent port that is typically located at the rear of
the range top.
There are two scales with which to measure CO: one is “as-measured” and the
other is “air-free.” As-measured CO is determined from a sample of combustion
gases with no regard for the amount of excess air (oxygen, O2 ) diluting the CO
concentrations. Excess air is the amount of oxygen in the combustion gases in excess
of the exact amount needed for perfect combustion. When combustion is perfect, just
the right amounts of fuel and oxygen are supplied to the combustion process so that
all the oxygen is utilized, a process called stochiometric combustion.
8417: “8417_c005” — 2007/9/11 — 12:16 — page 103 — #5

The basic problem with the as-measured method is this: As the amount of excess air
increases, the as-measured CO value falls for a given source strength of CO.20 In other
words, the amount of excess air in an emissions sample can, by dilution, significantly
decrease the as-measured value. This can cause a technician to mistakenly think that
a hazardous burner is working properly.
On the other hand, air-free measurement of CO takes account of the amount of
excess air in an emissions sample, incorporating an adjustment to the as-measured
CO ppm value, thus simulating air-free (oxygen-free) conditions in the sample. To
do this, a reading of oxygen percentage is taken from the emissions sample along
with the CO as-measured ppm. This can be done with an emissions meter having
the capability of measuring CO and oxygen percentage. Most emissions meters that
measure oxygen and CO as-measured have an integral electronic chip that calculates
the air-free level from as-measured CO ppm and oxygen percentage. CO air-free
and the firing rate of the burner can be used to determine the CO emission rate; CO
as-measured cannot. The use of the CO air-free measurement is generally supported
within the scientific community because it is a normalized value.21
Understanding as-measured and air-free CO measurements can be aided with an
analogy. If four drops of blue food color are placed in a small glass of water and another
four drops in a full pitcher of water, the water in the glass is darker shade of blue
than the water in the pitcher. When a sample from each water vessel is analyzed, the
concentration of food color in the pitcher is lower than in the glass; this is analogous
to CO as-measured. On the other hand, CO air-free is analogous to calculating the
number of drops of food color in each vessel by analyzing the concentration by
measuring (similar to CO as-measured) and by performing calculations to determine
the amount of food coloring in each vessel. Of course, the resulting calculation would
reveal four drops of food color in each vessel, a determination that could not be made
if only the concentration were measured. Finally, either of these vessels of water,
when dumped into a 5 gallon container of water, will color the 5 gallons of water to
nearly the same degree.
Either of these equivalent equations can be used for determining air-free CO in
an emissions sample with values for CO as-measured and oxygen percentage:

20.9
COAFppm = · COppm (5.1a)
20.9 − O2
or

COppm
COAFppm = (5.1b)
1 − 4.78 (O2 )
where
COAFppm = Carbon monoxide, air-free ppm
COppm = As-measured combustion gas carbon monoxide, ppm
O2 = Oxygen in combustion gas, as a percentage
From any combustion emission sample, the highest percentage of oxygen pos-
sible is 20.9%—that of the earth’s atmosphere—and the lowest is zero, resulting
8417: “8417_c005” — 2007/9/11 — 12:16 — page 104 — #6

(a) (b)
CO as-measured = 43 ppm CO as-measured = 522 ppm

O2 = 20% O2 = 10%
CO air-free = 1000 ppm CO air-free = 1000 ppm
Oven A Oven B
FIGURE 5.1 (a) Oven A has more dilution air than (b) oven B, so the CO as-measured value
is lower, but the CO air-free values are the same for each oven, so each oven will pollute the
kitchen air to the same degree.
from stoichiometric or oxygen-starved combustion. Whereas the oxygen content of

combustion gases from natural gas or propane furnaces or boilers is usually within a
range of 4.0–10.0%, the percentage of oxygen for a natural gas or propane range top
burner or oven is much higher, normally within a range of 13.6–20.1% and averaging
about 19.8%.16 This higher range of oxygen in the combustion gases leads to more
diluted and lower as-measured readings.
The two ovens pictured in Figure 5.1 demonstrate the inherent problem with
as-measured values. For Oven A, the CO as-measured level is 43 ppm; safe by some
standards. For Oven B, the CO as-measured level is 522 ppm; considered unsafe by
most standards. However, each oven is emitting the same level of CO air-free—1000
ppm, which is unsafe by all standards. (Refer to Equation 5.1a for these calculations.)
In other words, the source strength of CO from each oven as indicated by the CO
air-free measurement is the same; each oven will lead to the same ambient level of CO
in the kitchen air. A technician might pass Oven A and fail Oven B, not understanding
that each is contributing equally to hazardous levels of ambient indoor CO. Measuring
CO air-free gives the technician a better idea of the impact the CO emissions has on
indoor air concentrations.
As indicated above, the higher the percentage of oxygen in the combustion gases,
the lower the CO as-measured readings will be. If excess air is zero, as-measured
and air-free measurements will be equal. As excess air increases, CO as-measured
readings decrease, but CO air-free readings do not change. The high excess air percent-
ages for gas ovens and range top burners increases the importance of taking air-free
measurements, because the high excess air can significantly lower as-measured
readings, thus masking the true CO source strength.
It is common to find an oxygen content of 19.8% in range top burner combustion
gases. If this value is plugged into Equation 5.1a, the resulting CO air-free level
8417: “8417_c005” — 2007/9/11 — 12:16 — page 105 — #7

is 20 times the as-measured value. This is a rough general guideline relating CO

as-measured to CO air-free values for gas range top burners.
5.1.4 SINGLE-ZONE MASS BALANCE MODEL AND ROOM CO

CONCENTRATIONS
Vented appliances are only a hazard when they spill or backdraft. On the other hand,
gas ranges vent into the living space, so it is important to understand the relationship
between the source strength of CO emissions and the resulting concentrations of CO
in the ambient indoor air. The single-zone mass balance model equation is helpful for
demonstrating this relationship.

COAFppm · Vg · Gr · 1 − (1/2.713ACHt )
COppm = (5.2)
ACH · v
where
COppm = As-measured indoor ambient carbon monoxide, ppm

COAFppm = Air-free carbon monoxide, ppm
Vg = ft 3 of flue gas per ft3 of fuel gas (8.5 ft3 for natural gas, 21.8 ft3 for
propane)
Gr = Gas flow rate, in ft3 /h = input rate (Btu/h) / heat value of fuel (Btu/ft3 )
—an average input rate for an oven is 18,000 Btu/h, for a range-top
burner 9,000 Btu/h
—an average heat value for natural gas is 1000 Btu/ft3 , for
propane 2500 Btu/ft3
2.713 = Napierian logarithmic base
ACH = Number of natural air changes per hour of room or house
t = Time interval, hours
v = Volume of room or house, ft3
For example, if a natural gas range with a total input rate of 54,000 Btu/h is
emitting CO air-free at a constant rate of 800 ppm in a house of 8000 ft3 having an air
exchange rate per hour (ACH) of 1.5, the ambient indoor CO concentration will be
29 ppm after 2 h (Refer to Equation 5.3). The rate of 54,000 Btu/h is the sum of the
input of four range top burners and the oven operating at the same time, a possibility
if a user were operating the gas range as a space heater.

800AFppm · 8.5 · 54 · 1 − (1/2.7130.5·2 )
COppm = (5.3)
1.5 · 8000
COppm =29
8417: “8417_c005” — 2007/9/11 — 12:16 — page 106 — #8

This assumes the ambient CO is evenly dispersed throughout the house. In reality,
the concentrations are likely to be higher in the kitchen than in remote sections of the
living space.
If the structure is tightened, reducing the ACH from 1.5 to 0.5, the ambient indoor
CO level will increase. As shown in the Equation 5.4, at the end of the 2-h period the
average ambient CO concentration will be 58 ppm, rather than 29 ppm. At the end of
4 h, the room concentration will be 80 ppm.

800AFppm · 8.5 · 54 · 1 − (1/2.7130.5·2 )
COppm = (5.4)
0.5 · 8000
COppm =58
It is clear that tightening the house (lowering the ACH) without also lowering
range CO emissions or operating a vented range hood during range operation can
create a more hazardous indoor condition. This has important implications for the
weatherization of residential buildings.
Figure 5.2 shows the above examples along with two other house leakage rates—
0.35 to 2.0 ACH. Notice that as the leakage rate decreases the indoor levels of CO
increase and the number of hours before reaching a stabilized level increases. For
example, at 2.0 ACH the stabilized level of 23 ppm is reached after 2 h. In contrast,
at 0.5 ACH the stabilized level of 91 ppm is reached after 10 h. This situation might
occur if occupants were operating all the burners for space heating with the oven door
140.00
0.35 ACH
120.00
Ambient CO level, ppm
100.00
0.5 ACH
0.35 ACH
80.00 0.5 ACH
1.0 ACH
60.00
1.0 ACH 2.0 ACH
40.00
2.0 ACH
20.00
0.00
0 1 2 3 4 5 6 7 8 9 10
Hours of CO production, 800 ppm air-free
FIGURE 5.2 Room ambient CO concentrations at various house leakage rates over 10-h
period. CO production is at a constant rate of 800 ppm air-free from a total burner input of
54,000 Btu/h in an open-plan house of 8,000 ft3 .
8417: “8417_c005” — 2007/9/11 — 12:16 — page 107 — #9

open, thus the duty cycle of the oven would be 100% because the oven thermostat
would never be satisfied.
Field measurements of 18 gas ovens under the author’s direction in 2001 resulted
in 8 of these tested units emitting in excess of 800 ppm CO air-free after 15 min of
operation. Although the average gas oven has a firing rate of approximately 18,000
Btu/h—only one-third the value of 54,000 Btu/h input rate used in the total range
input rate above—these failed ovens emitting in excess of the 800 ppm air-free CO
emissions could cause hazardous ambient CO levels if one of these ovens were used
to bake a large turkey or as a space heater. One notable oven emitted CO at a rate of
2270 ppm air-free. In the same example house as above, with a 0.35 ACH and the
oven operating at 18,000 Btu/h, this CO emission level would result in an ambient
CO of 62 ppm after 2 h and 93 ppm after 4 h.
Other field and laboratory studies have revealed that lining the bottom of an oven
with aluminum foil to catch spills can lead to higher CO emissions if the foil obstructs
the secondary air openings at the oven bottom, as noted in the U.S. CPSC testing.5
This practice interrupts the oven venting path and becomes even more hazardous if
the oven is operated with its door open.5 “The worst case situation of a consumer
using a gas oven as a space heater with an aluminum foil, protective liner, produced
the highest CO emission rates. These high emission rates can push CO ambient air
concentrations to dangerous levels.”5
5.1.5 FIELD TESTING, LABORATORY RESEARCH, AND KARG FIELD

PROTOCOL DEVELOPMENT
The Karg Field Protocol for Gas Range Carbon Monoxide Emissions Testing
was completed by the author in 2001. Funding and support for this protocol
research and development was provide primarily by the U.S. DOE and the Chicago
Region low-income weatherization programs of Illinois, Indiana, Iowa, Michigan,
Minnesota, Missouri, Ohio, and Wisconsin. Funding and support for laboratory work
was provided by the Gas Research Institute of Des Plaines, Illinois, through GARD
Analytics of Park Ridge, Illinois and WEC Consulting of Potomac, Maryland. A
group of technical advisors was assembled for the project that included college fac-
ulty, scientists from government organizations, indoor air quality experts, technical
consultants, and private laboratories. Tim Lenahan of the Ohio Office of Energy
Efficiency was the contract manager for the project.
5.1.5.1 Introduction
At the foundation of the development of this Karg Field Protocol was research in
the field and laboratory. Eighteen oven bake burners, and 81 range top burners were
tested in the mid-coast Maine area, and 19 range top burner tests were done at the
Gas Research Institute laboratory facility. Of the 21 oven bake burners tested, 4 were
natural gas and 17 were propane (LPG). All 21 of the gas ranges tested in the field
were in primary residences or summer homes. Laboratory range top burner testing
was done over a 3-day period to determine the best method for measuring range top
burner CO emissions. No oven testing was done at the laboratory.
8417: “8417_c005” — 2007/9/11 — 12:16 — page 108 — #10

Field testing required approximately 3 h in each house. During emissions testing,

the houses were closed up with all windows and exterior doors shut. Range hood
fans and other exhaust fans in the houses were not operated during the testing, unless
ambient CO levels reached 35 ppm. All test equipment was calibrated with CO span
gas at the beginning and end of each house test regimen. After each burner test the
instrument was rinsed for 1–2 min. All oven emissions testing was done with two
test instruments inserted into the oven vent at the back of the range top. During each
30-min oven test, these instruments were checked for correspondence.
The oldest range tested had been in service 53 years; the newest only 1 year. Ten
of the ranges had standing pilots and eleven had spark ignitions. Gas pressure was
measured in ten of the ranges. Natural gas measured in three ranges was between
4 and 5 in. of water column and propane gas pressure measured in eight ranges was
between 7 and 12 in. of water column.
The actual firing rate of each burner was not measured, but label firing rates were
recorded for 16 ranges. The mean and median range top burner label firing rate were
8,563 and 9,000 Btu/h and the mean and median oven bake burner label firing rate
were 17,677 and 18,000 Btu/h.
Ambient CO levels were continuously monitored during the field testing for
the safety of house occupants and the field analyst. After range top burner test-
ing, the highest ambient CO level recorded was 7 ppm; all other readings were
zero. After oven bake burner testing, the highest level was CO level recorded was
78 ppm and the lowest was zero. The mean, median, and standard deviation were
7.8 ppm, 1.5 ppm, and 19.21 ppm for 16 field tests for which this information was
recorded.
Although the tested ranges were not adjusted after the initial emissions test and
then tested again, other research has found evidence that adjustment of a gas range
by a knowledgeable technician can lower CO emissions.6 Ranges should be cleaned,
tested for CO emissions, and adjusted, if necessary. Anecdotal evidence suggests that
gas ranges are seldom adjusted and rarely tested for CO emissions.
5.1.5.2 Range Top Burners, Field Testing

Range top burner emission rates are affected by disruption of the burner flame. Air
currents and flame impingement greatly affect CO output. In addition, sampling
location also affects measured CO and oxygen concentrations. To reduce variability
and increase correspondence between measured CO output and actual CO output, a
number of devices and methods have been devised. Range top burner testing in the
field was done using five different devices and methods to determine which method
was the most uniform, repeatable, and best simulated actual burner use in a home.
These devices and methods include (1) the CO Hot Pot™ (Figure 5.3), (2) a device
similar to the CO Hot Pot™, but with a stainless steel kettle within the cylinder,
(3) a device similar to the CO Hot Pot™, but with no bowl or water kettle within the
cylinder, (4) a free-standing aluminum kettle on the burner, and (5) testing the burner
CO emissions over an open burner with no special device. The first of three devices
were designed and fabricated by the author to introduce standardization and increase
the accuracy of field testing. The last two methods were found to be in common use
8417: “8417_c005” — 2007/9/11 — 12:16 — page 109 — #11

(a) (b)
FIGURE 5.3 CO Hot Pot™ for measuring CO emissions from range top burners. Notice the
stainless steel bowl at the bottom of the 8-in. diameter by 12-in. high cylinder. (a) Front view
and (b) Top view of the CO Hot Pot™.
for testing range top burners by technicians working in the U.S. DOE low-income
weatherization program.13
The CO Hot Pot™ was used on each range top burner for a five-minute test.
This simple device (Figure 5.3) is an 8-in. round by 12-in. high galvanized sheet
metal cylinder with a 6-in. round stainless steel bowl fastened concentrically at the
cylinder’s bottom.22 This bowl is intended to simulate a pot or pan on the range burner.
The Y-shaped brace at the top of the cylinder strengthens the top of the cylinder and
supports an eye-bolt that holds the tip of the electronic instrument probe in place after
it is inserted through a hole drilled in the side of the cylinder. A metal handle fastened
to the other side of the cylinder facilitates easy placement of the cylinder on the range
top burner grate. The cylinder contains the emissions and protects the combustion
from lateral drafts that can significantly increase CO emissions as a result of flame
impingement. This device was also used during range top burner testing in the Gas
Research Institute laboratory.
A second device that was field and laboratory tested was a cylinder similar to the
CO Hot Pot™, but with a stainless steel water kettle in the place of the bowl. The spout
of this kettle passed through a snug opening in the cylinder so that evaporating water
from the kettle could escape without affecting the emissions readings taken inside the
cylinder just above the kettle. This device had the same Y-brace and eye-bolt to hold
the instrument probe. The kettle was always filled with four pounds of 60◦ F water
just before a test.
The third device tested was an open cylinder that was also used for range top
burner testing, a replica of the CO Hot Pot™, but without the stainless steel bowl at
the bottom.
In addition to these three devices, the fourth method tested in the field and
laboratory was a free-standing aluminum water kettle, without a cylinder around
it, that was used on each range top burner. The size of this water kettle was similar to
the stainless steel one housed in a cylinder. This aluminum kettle was always filled
with four pounds of 60◦ F water before a test.
Finally, the fifth method tested was the CO instrument probe held at a precise
position over the open flame of the range top burner; no cylindrical device or water
kettle.
8417: “8417_c005” — 2007/9/11 — 12:16 — page 110 — #12

CO air-free ppm
Left front range top burner, unit10 CO as-measured ppm
700 35
CO as-measured (ppm)
600 30
CO air-free (ppm)
500 25
400 20
300 15
200 10
100 5
0 0
0 1 2 3 4 5
Time after startup (minutes)
FIGURE 5.4 A typical emission profile of a 5-min range top burner test with CO Hot Pot™
showing CO as-measured and air-free ppm.
When each of these devices and methods were used on a range top burner, the
end of the instrument probe was always positioned exactly ten inches above the top
of the burner grate. Each test was conducted for 5 min while emissions at the burner
were monitored and recorded. All burner measurements for CO as-measured, CO
air-free, oxygen (O2 ) percentage, and temperature were logged every 0.25 min and
recorded to a computer spreadsheet for analysis. Range top burner testing was always
performed before oven testing to prevent the oven from preheating the range top area.
Figure 5.4 shows typical profiles of CO as-measured and air-free ppm using the
CO Hot Pot™ for a 5-min test. It was common for the emissions to peak within
the first 2 min and then begin to fall as the burner parts warm up and steady state
is reached. Range top burner emissions are less problematic to measure than oven
emissions because the burner does not turn on and off (duty cycle) in order to hold
a constant temperature. Notice that the CO air-free and as-measured emissions track
each other in parallel, although air-free is approximately 17–20 times greater than
as-measured. This is typical. The parallel characteristic of these two measurements
is a result of a fairly consistent percentage of oxygen in the emission gases.
Of the 81 range top burners tested during the field research, only 1 failed the
Karg Field Protocol threshold of 35 ppm CO as-measured, using the CO Hot Pot™
at 5 min. This failed burner had an emissions rate of 38 ppm CO as-measured and
882 ppm CO air-free (20% oxygen).
Oxygen percentages were within a fairly narrow range when using the CO Hot
Pot™—between 17.9 and 20.8, with a mean of 19.8. Oxygen percentages this high
can significantly affect the relationship between CO as-measured and air-free values.
For example, for the failed range top burner mentioned above, the multiplier to get
from CO as-measured to CO air-free is 23.2 (see Equations 5.1a and 5.1b). The high
oxygen levels typically found in emissions from gas range top and oven burners lead
to reduced levels of certainty when taking measurements in the field.
Although one study by the Gas Research Institute found “The data sets reveal
mean CO emissions on the low setting to be much larger than the values obtained
8417: “8417_c005” — 2007/9/11 — 12:16 — page 111 — #13

when the burners were set on high or medium,”6 the author’s testing in the field and
the laboratory was done with the range burner setting on high. The highest setting was
selected for the Karg Field Protocol because this setting; (1) most closely matches
the Btu/h rating on the range name plate; (2) is the easiest setting for the technician
to find; and (3) is the only repeatable setting on most range top burners, allowing the
technician to retest at the same setting—Btu/h firing rate—after a burner is adjusted.
5.1.5.3 Range Top Burners, Laboratory Testing

Laboratory chamber testing was done over a 3-day period at the Gas Research Institute
facility to determine the best method of measuring CO emissions from range top
burners. The CO Hot Pot™ and four other methods—all described in the previous
section—were tested to determine which of the five methods would best predict the
ambient emission CO levels in the test chamber. The CO Hot Pot™ was tested seven
different times; the four other methods were each tested three times. The CO Hot
Pot™ was tested more often because by the third day of testing, it already showed
evidence of being the most promising of the five methods.
During the chamber testing of range top burners, readings of CO as-measured, CO
air-free, and oxygen were recorded with hand-held equipment at the burner with each
of the five methods (devices) while the chamber ambient CO concentrations were
recorded with stationary laboratory equipment. Both burner and chamber emissions
were recorded every 0.25 min during the test periods. The equipment was set up on a
range burner in the chamber, the technician left the chamber, the burner and emissions
recording equipment was then activated from outside of the chamber. During most of
the tests, a small fan was operating on the floor of the chamber to help mix emissions
with the chamber air; care was taken to ensure air did not flow across the range top.
The burner was operated for 30 min and then turned off from outside the chamber.
The decay of the emissions was measured for another 15 min so that the rate of air
changes per hour within the chamber could be calculated after each 30-min emissions
test. For most of the tests, the chamber was then purged with an exhaust fan operating
for 15 min within the chamber before the next test began. All tests were performed
with the burners at the highest setting. (See Figure 5.5)
The burner input rate is usually printed on the range nameplate for both natural gas
and propane fuels. ANSI Standard Z21.1-1993 for manufacturers states “When oper-
ated for 5 min, starting with all parts of the appliance at room temperature, the burner
adjustment shall be within + 5% of the capacities specified (on the nameplate).”17
Metering gas consumption during laboratory testing for this study showed that the
metered consumption in Btu/h of five range top burners on two ranges varied from
rated consumption by +15% to –29%. The metered burner consumption closest to
rated consumption was –12%. These variations were significantly greater than the
5% required by the ANSI Standard. More field research is needed to determine the
typical variation between nominal nameplate ratings and actual burner Btu/h.
Figure 5.6 shows the chamber concentrations of CO during three tests using the
CO Hot Pot™ over the 30-min test, the 15-min decay, and the 15-min purge. At the
end of the 30-min test, the chamber concentrations were within less than 1 ppm of
CO for the three tests on the same burner.
8417: “8417_c005” — 2007/9/11 — 12:16 — page 112 — #14

FIGURE 5.5 Chamber testing with CO Hot Pot™ and Testo instrument sending emissions
data directly to a computer outside the chamber. Notice the 12-in. square opening in the wall
at the upper right and the floor fan to spread combustion emissions within chamber. Photo was
taken through a sealed glass observation port.
10
Test 3, Range C
Test 10, Range C
9
Test 15, Range C Burner turned off
Decay allowed to
8
continue without
purging chamber
5
Chamber purged
4
0 15 30 45
FIGURE 5.6 Test chamber CO concentrations for three range top burner tests using CO Hot
Pot™. See Figure 5.7 for corresponding burner CO emission rates.
8417: “8417_c005” — 2007/9/11 — 12:16 — page 113 — #15

Test3, Range C, CO air-free

Test10, Range C, CO air-free
350 Test15, Range C, CO air-free 60
Test3, Range C, CO as-measured
CO air-free ppm Test10, Range C, CO as-measured
300 (scale at left axis)
Test15, Range C, CO as-measured 50
250
40
CO air-free (ppm)
200
30
150 CO as-measured ppm
(scale at rightaxis)
20
100
50 10
Initial burner warmup
0 0
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
FIGURE 5.7 Burner CO emission rates for three range top burner tests using CO Hot Pot™.
See Figure 5.6 for corresponding test chamber concentration.
Figure 5.7 shows the corresponding CO emissions measured at the burner for
these three 30-min tests. At the end of the three tests, the CO as-measured values
were 21, 21, and 19 ppm and the CO air-free values were 231, 231, and 199 ppm.
At 6 min after burner startup, the test time selected for the Karg Field Protocol, the
CO as-measured values for the three tests were 15, 14, and 10 ppm and the CO
air-free values were 261, 225, and 161 ppm. The higher chamber CO concentrations
and emission rate at the burner for test 3 shown in Figures 5.6 and 5.7 might have
resulted from the CO Hot Pot™ being positioned slightly off center over the burner
(laboratory notes and a photograph support this misalignment), although further tests
were not conducted to determine if an off-center position over the burner would affect
emissions.
After field and laboratory testing, the CO Hot Pot™ was selected over the other
four test methods for a number of reasons.
1. The cylinder of this device protects the CO emissions testing from the
unpredictable affect of lateral air movement, making testing in the field
more consistent and repeatable. During a test in the laboratory chamber
with the free-standing water kettle, the chamber air conditioner inad-
vertently turned on, blowing air across the range top and causing the
ambient CO concentration in the chamber to rise to just under 15 ppm.
Similar tests with this free-standing water kettle on the same burner
without the affect of the air conditioner fan resulted in a chamber CO just
above 3 ppm. This minor laboratory accident demonstrated the import-
ance of preventing air flow across burners during emissions measurement
and supports the use of the cylindrical protective housing of the CO
8417: “8417_c005” — 2007/9/11 — 12:16 — page 114 — #16

Hot Pot™. Of course, emissions testing with the instrument probe merely
held above an oven flame is subject to the same, if not more, variab-
ility from lateral air movement than when using a free-standing water
kettle.
2. Burner emissions measurements with the CO Hot Pot™ were found to level
off to steady-state conditions 5–6 min after burner startup in almost all tests
in the laboratory and field. The other four methods were less consistent in
performance.
3. Some of the other measurement methods where either too complicated to
be practical in the field, or they demonstrated design flaws. The meth-
ods not using a cylindrical housing made the emissions testing subject to
more error resulting from lateral drafts, as described above. The cylin-
der with a water kettle inside and the free-standing water kettle required
filling with water and yielded variable results because of the changing water
temperature. Finally, the open cylinder with no bowl or water kettle con-
centrically fastened at the bottom of the cylinder created a strong enough
draft during some tests to cause the flames to lift off the burner, resulting
in artificially high emissions readings.
4. The oxygen percentages in the emission gases when using the CO Hot
Pot™ were usually close to 19.85. The field testing mean was 19.84% and
the laboratory testing mean was 19.3. This rather tight range for oxygen
percentage allowed the adoption of a relatively simple method for the Karg
Field Protocol range top burner testing, using the CO Hot Pot™ for CO as-
measured readings rather than the more complex CO air-free measurement.
The Karg Field Protocol calls for CO as-measured emissions measured with
the CO Hot Pot™ of 35 ppm or less. For an oxygen percentage of 19.85%,
this equates to a CO air-free threshold of just less than 700 ppm, 100 ppm
less than the ANSI Standard Z21.1 level of 800 CO air-free ppm to which
manufacturers must adhere. The ANSI Standard requires measurement at
5 min, whereas the Karg Field Protocol calls for measurement at 6 min.
The author’s field and laboratory testing indicated that range top burner CO
emissions measured with the CO Hot Pot™ are more likely to have settled
into steady-state conditions at 6 min than at 5 min. The later emission
measurement of the Karg Field Protocol is likely to yield results closer
to the longer-term emissions of a range top burner than ANSI Standard
test a 5 min.
5.1.5.4 Oven Bake Burners, Field Testing

Oven testing was performed on bake burners only, even if a separate broil burner
was installed and working in a tested oven. Oven testing was done at a setting of
350◦ F with the oven door closed, utensils removed from within and under the oven,
and foil lining removed from the oven floor (none of the ovens floors were lined
with foil when the field analyst arrived to test). Ovens were not cleaned before test-
ing. Testing was done for thirty minutes while emissions at the oven burner and
ambient kitchen air were monitored and recorded. All field measurements for CO
8417: “8417_c005” — 2007/9/11 — 12:16 — page 115 — #17

FIGURE 5.8 Redundant testing of CO emissions during field test. Two instrument
probes are inserted into oven vent in a manner ensuring that room air will not dilute the
sample.
as-measured, CO air-free, oxygen percentage, and temperature were logged very

0.25 min and recorded to a computer spreadsheet for analysis. The oven burner
emissions were measured at the oven vent, which in all cases was at the back of
the range top surface. Care was taken to measure emissions without the diluting
affect of room air. Oven temperature was measured at the end of the emissions
probe and with a separate oven thermometer placed in the oven. All emissions
measurements were measured with two electronic instruments that were checked
during the 30-min test period for correspondence. (See Figure 5.8) All instruments
were calibrated before and after each test series at each site with 250 ppm CO
calibration gas.
Tikalsky et al. used 400◦ F for their field testing while Tsongas suggested a
350◦ F setting for his oven field protocol.6,23,24 The oven temperature setting used
for the project field testing was 350◦ F and burner setting selected for the Karg Field
Protocol was also 350◦ F. This temperature was selected because it is a common bak-
ing temperature and the oven requires less time to reach steady-state conditions than
for a higher setting. The actual measured oven temperature at the end of the 30-min
field tests for the 18 ovens ranged from 290◦ F to 375◦ F, with a mean of 347◦ F, a
median of 340◦ F and a standard deviation of 20.7 F◦ .
Separate broil burners were not included in the final protocol because it was
thought that the additional time that would be required would not be worth the reduced
risk to the range user. It was assumed that broil burners are not used as often as bake
burners and, when they are used, they are not operated as long.25
Of the eighteen valid oven field tests that were conducted in 2001, ten passed
the Karg Field Protocol and eight failed. The highest CO air-free ppm measured at
the 15-min Karg Field Protocol mark was 1907, the lowest 244, the mean 932, the
8417: “8417_c005” — 2007/9/11 — 12:16 — page 116 — #18

median 769, and the standard deviation 588. Oven oxygen percentage ranged from a
high of 20.9 to a low of 10.4 with a mean of 18.3.
Two profiles for oven test data are shown in Figure 5.9 (oven number 15, which
passed) and Figure 5.10 (oven number 11, which failed). Each figure shows CO
air-free ppm (left Y-axis), CO as-measured ppm (right Y-axis), oxygen percentage
times 20 (right Y-axis), temperature in ◦ F (right Y-axis), and time in minutes (X-axis).
Notice in both figures how CO emissions peak just after start-up, then fall, and
eventually settle into a steady-state, saw-tooth pattern as a result of the burner duty
cycle.
For each of these profiled field tests, the steady-state condition was reached
approximately 11 min after startup. The longest time required to reach steady-state
for the 18 oven field tests was 19.5 min, the shortest was just over 3 min, and the
mean and median were 10. As a result of these observations, the timing for the oven
bake burner CO emissions test for the Karg Field Protocol was set at 15 min from
Oven setting at 350°F
CO as-measured, temperature, and O2

700 600
CO air-free ppm
CO as-measured ppm
600 O2 percent times 20 500
Temperature °F
CO air-free (ppm)
500
400
400
300
300
200
200
100 100
0 0
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Time from startup (min)
FIGURE 5.9 Oven number 15 passed the developed oven bake burner Karg Field Protocol.
Oven setting at 350°F

CO as-measured, temperature, and O2
3500 600
CO air-free ppm
CO as-measured ppm
3000 O2 percent times 20 500
Temperature °F
CO air-free (ppm)
2500
400
2000
300
1500
200
1000
500 100
0 0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Timeafter startup, (min)
FIGURE 5.10 Oven number 11 failed the developed oven bake burner Karg Field Protocol.
8417: “8417_c005” — 2007/9/11 — 12:16 — page 117 — #19

start-up or later. Only one of the 18 field tested ovens required more than 15 min to
reach steady-state.
As mentioned in Section 5.1.2, two important questions had to be answered before
developing the Karg Field Protocol: First, what basis should be used to develop
the threshold level of CO emissions at the burners to ensure an acceptable ambient
level for the occupants of a residence? This is discussed in Section 5.1.2. Second,
at what time after burner startup should readings be recorded and what method of
measurement should be used to accurately reflect the long-term CO emissions and
thus, the potential hazard of a burner’s emissions to occupants?
The Karg Field Protocol time and method of measurement for ovens CO emissions
was designed not only to reflect the ANSI Standard, but also to accurately predict the
emission rate after the oven bake burner settled into a steady-state pattern.15
To gauge the level of accuracy, the emissions data for the eighteen ovens were used
to compare the Karg Field Protocol method of measuring CO air-free ppm emissions
at 15 min with the CO air-free emissions averaged from 15 to 30 min. The field
testing process collected data every 0.25 min during a 30-min test of each oven. This
analysis found that the oven Karg Field Protocol value had a coefficient of linear
correlation to the CO air-free emissions averaged from 15 to 30 min of R = 0.95 and
R2 = 0.9. (See Figure 5.11). In addition to the Karg Field Protocol, correlation was
also checked with protocols developed by ANSI, Tsongas, and Building Performance
Institute (BPI). Of the four, the Karg Field Protocol gave the highest correlation. As
a result of designing the Karg Field Protocol to give close correspondence from CO
air-free values averaged over 15–30 min, the protocol values averaged 102.4% of
the longer term CO air-free emissions, whereas the ANSI protocol was found to be
71.6%. The Tsongas and BPI protocols could not be checked for this correspondence
because they use a CO as-measured test method. These data for these four protocol
comparisons to the author’s CO air-free values averaged over 15–30 min for 18 oven
field tests are shown in Figure 5.12.
These oven field data were also analyzed to determine the correlation between
the CO air-free ppm emissions at 5 min and the CO air-free emissions averaged
Karg field protocol, oven measurement

2000
CO air-free averaged from
15–30 min, ppm
1500
R = 0.95
R2 = 0.90
1000
500
0
0 500 1000 1500 2000 2500
Karg protocol, CO air-free at 15 min, ppm
FIGURE 5.11 Relationship in 18 tested ovens between Karg Field Protocol oven CO air-free
measurement and CO air-free emissions averaged from 15 to 30 min.
8417: “8417_c005” — 2007/9/11 — 12:16 — page 118 — #20

Carbon Monoxide Emissions from Gas Ranges and the Development of a Field
FIGURE 5.12 Data for comparison of CO air-free emissions averaged from 15 to 30 min from 18 oven field tests by the author to 4 oven test protocols.
8417: “8417_c005” — 2007/9/11 — 12:16 — page 119 — #21

Columns c, f, i, and k are from the author’s oven field test data measured according to the corresponding protocol. A brief description of each protocol CO
measurement method is included in the last row.
119
ANSI Z21.1 Protocol

2000
15–30 min (ppm) 1500
R = 0.51
1000
R2 = 0.26
500
0
0 500 1000 1500 2000 2500 3000
ANSI protocol, CO air-free at 5 minutes (ppm)
FIGURE 5.13 Relationship in 18 tested ovens between ANSI Z21.1 standard measurement
of CO emissions at 5 min and CO air-free emissions averaged over 15–30 min.
from 15 to 30 min. This was done because the ANSI Standard Z21.1 with which gas
range manufacturers must comply, requires CO emissions be measured at 5 min
after startup; these emissions must be 800 ppm air-free or less.17 This analysis
found that the 18 tested ovens at the 5-min ANSI Standard measurement require-
ment had a coefficient of linear correlation to the CO air-free emissions averaged
from 15 to 30 min of R = 0.51 and R2 = 0.26. (Refer to Figure 5.13) These field data
demonstrate that measuring CO air-free from a bake burner at 5 min after start-up is
not a good predictor of longer-term CO air-free emissions and that it may underes-
timate steady-state, long-term emissions. This significant finding indicates that more
research is needed to determine the match between the ANSI Standard method and
actual field CO emissions from oven bake burners in the field.
The oven field test data were analyzed to determine the comparability with two
other published oven CO emission field protocols. The first, published in 1995 by
Tsongas calls for the first CO as-measured ppm peak to be recorded after startup.23
If this peak is 100 ppm or greater at an oven temperature setting of 350◦ F, the oven
fails the protocol. Of the 18 ovens tested in the field on 2001, only 2 passed this
protocol with 70 and 39 CO as-measured ppm, with the average time for the first
peak in CO as-measured emissions of 2.25 min. The Tsongas protocol was tested
with the field data from the 18 ovens to determine how well it predicted the longer
term oven emissions. (Results are illustrated in Figure 5.14) The Tsongas protocol
did not correlate well with longer term CO air-free emissions. The Tsongas protocol
CO as-measured value of 100 ppm appears to be significantly more stringent than the
ANSI Z21.1 Standard. (Referring back to the oven emission profiles in Figures 5.9
and 5.10 will help demonstrate this poor correlation.)
Finally, the oven field test data were analyzed to determine how well the BPI
oven protocol predicts the longer term oven emissions.26 (Results are illustrated in
Figure 5.15) The BPI protocol (being used across the U.S. by BPI certified technicians)
calls for oven bake burner CO as-measured ppm testing at 5 min at the maximum
oven temperature setting, short of broil. If CO as-measured emissions from ovens are
from 100 to 300 ppm, or less, a CO alarm must be installed and a recommendation
for oven service must be made to the customer. If CO oven emissions are more than
300 ppm, the unit must be serviced before any weatherization work is performed.
8417: “8417_c005” — 2007/9/11 — 12:16 — page 120 — #22


Tsongas oven measurement protocol
2000
15–30 min (ppm)

1500
1000
R = 0.53
R2 = 0.28
500
0
0 200 400 600 800 1000 1200 1400 1600
Tsongas protocol, CO as-measured at first peak (ppm)
FIGURE 5.14 Relationship in 18 tested ovens between CO as-measured ppm emissions at

the first peak after start-up and CO air-free emissions averaged over 15–30 min.
BPI oven measurement protocol

2000
15–30 min (ppm)
1500
R = 0.56
1000
R2 = 0.31
500
0
0 200 300 400 500 600 700 800
BPI Protocol, CO as-measured at 5 min (ppm)
FIGURE 5.15 Relationship in 18 tested ovens between CO as-measured ppm emissions at 5

min after start-up and CO air-free emissions averaged over 15–30 min.
Of the 18 ovens tested in the field in 2001, 13 passed this BPI protocol (300 ppm).
Eight of these BPI passes also passed the Karg Field Protocol and five failed. Of
five ovens that failed the BPI protocol, three failed the Karg Field Protocol and two
passed. Although the 18 oven field tests were done at an oven setting of 350◦ F and the
BPI protocol calls for the highest setting, the field test data at 5 min is valid. This is
because at 5 min the oven set at 350◦ F is still moving toward a steady-state condition,
just as it would be at a higher temperature setting.
Of the four oven protocols compared with the field data of eighteen oven tests, the
Karg Field Protocol appears to most accurately predict the averaged emissions from
15 to 30 min. There is no evidence to suggest that the steady-state, saw-tooth pattern
of oven CO emissions over 15–30 min will significantly change after the 30-min
mark, but because the author’s testing stopped at 30 min, he cannot demonstrate this.
Because CO emissions from a gas range burner are seldom, if ever, hazardous unless
the burner operates for a long period, this relationship between protocol readings
and longer term emissions is important. Any protocol that does not correlate closely
with longer term, steady-state emissions will not adequately reflect a CO hazard to
occupants.
8417: “8417_c005” — 2007/9/11 — 12:16 — page 121 — #23

5.2 KARG FIELD PROTOCOL FOR MEASURING

CARBON MONOXIDE EMISSIONS FROM GAS
RANGES
5.2.1 INTRODUCTION
The purpose of this protocol is to guide the field analyst through a systematic procedure
of gas range testing to determine whether a gas range burner is emitting unacceptable
levels of CO.
The burner limits for this protocol are 35 ppm CO as-measured for range top
burners and 800 ppm CO air-free for oven bake burners. Oven broil burners are not
required to be tested.
This method covers residential grade floor-mounted gas ranges, drop-in range top
burners, and built-in ovens only. If drop-in range top burners or built-in ovens are
encountered, follow the appropriate sections of this protocol for these appliances.
This protocol is not intended for use with (1) outdoor gas grills, (2) ovens in catalytic
cleaning mode, (3) ovens vented into flues or chimneys, or (4) range/ovens with a
downward vented exhaust fan while the fan is operating.
This protocol is not intended to determine whether gas ranges operate acceptably
during misuse, such as when a range is used for space heating.
Accurately measuring CO emissions in the field is difficult due to the complex
nature of combustion and dilution airflow patterns. Use of this protocol can increase
the accuracy of measurements to, perhaps, +30%.8,10 This means that the protocol
will sometimes result in false failures and false passes.
Because there is a broad variety of gas ranges in the field, there is the possibility
that range characteristics not addressed in this protocol will be encountered.
Funding and support for this protocol research and development was provided
in 2001 primarily by the U.S. DOE and the Chicago Region low-income weather-
ization programs of Illinois, Indiana, Iowa, Michigan, Minnesota, Missouri, Ohio,
and Wisconsin. Funding and support for laboratory work was provided by the Gas
Research Institute of Des Plaines, Illinois, through GARD Analytics of Park Ridge,
Illinois and WEC Consulting of Potomac, Maryland. Tim Lenahan of the Ohio Office
of Energy Efficiency was the contract manager. This section of this chapter includes
an abridged version of the Field Protocol and is not subject to the copyright of the
publisher of this book.
5.2.2 VISUAL INSPECTION AND CUSTOMER

EDUCATION
5.2.2.1 Range Top Inspection
Inspect the range top burner area for cleanliness. If the burners or burner area are
dirty enough to adversely affect the combustion process, inform the customer that the
range should be cleaned to reduce the possibility of unacceptable emissions. Do not
test for CO emissions until the problem is corrected. Inspect the burners for proper
alignment and seating.
8417: “8417_c005” — 2007/9/11 — 12:16 — page 122 — #24

All cooking vessel support grates should (1) be in place, (2) fit properly in the
burner well, and (3) be in the configuration the manufacturer intended, with no broken
parts.
If any of the grates are missing or in unsatisfactory condition, the customer should
not use the affected range burner until the substandard or missing grate is replaced.
If a grate cannot be repaired or replaced, the decision whether to replace the range
should be made.
If the range top burners are ignited with a standing pilot light, verify that the pilot
flame is present, is about 5/16 in length, and is soft blue in color (not yellow).
5.2.2.2 Oven Inspection

Inspect the oven for cleanliness. If the burners or oven area are dirty enough to
adversely affect the combustion process, inform the customer that the range should
be cleaned to reduce the possibility of unacceptable emissions. Do not test for CO
emissions until the problem is corrected.
Check the oven for obstruction of the oven-floor vents. These vent holes must not
be blocked by anything in the oven, such as aluminum foil, or anything stored below
the oven. These vent openings must never be obstructed because they are a vital part
of the oven combustion and venting systems.
Check for air blockage at the bottom of the range and drawer and/or broiler
compartment under the oven. Dust, lint, pet hair, rugs, or any other obstruction
blocking free airflow to the oven bake burner must be removed. Check the oven
bake-burner spreader plate (burner baffle). Most bake burners (the one at the bottom
of the oven compartment) have a flame spreader plate just under the oven floor bottom
and above the bake burner flame. A warped or detached spreader plate can result in
flame impingement and quenching (cooling) of the gas flame, causing increased pro-
duction of carbon monoxide. Many spreader plates are intentionally bent into curved
or angular shapes, or dimpled, to add strength. Inspect carefully with a flashlight and
inspection mirror to determine if the spreader plate has distorted from its original
shape or has detached from the oven bottom. Ignite the bake burner to inspect the
flame. The flame should not extend beyond the edge of the spreader plate. Also,
inspect for carbon buildup on the spreader plate and the oven bottom. Any carbon
build-up can be an indication of incomplete combustion caused by flame quenching
or a fuel-rich gas mixture.
If the range also has a broil burner at the top of the oven compartment, check its
flame for proper size and color.
If the oven bake burner is ignited with a standing pilot light, verify that the pilot
flame is present, that it is about 5/16 in length, and is soft blue in color (not yellow).
Inspect gas range installation for code compliance by referring to the latest edition
of the National Fuel Gas Code (NFPA 54).
Verify that the range is set up for the appropriate supply gas, either natural gas
or propane (LPG). Although a mismatch between supply gas type and range set up
is not a common occurrence, each range should be checked. Natural gas (methane)
contains 1000 Btu/ft3 , while propane contains 2500 Btu/ft3 . Gas ranges using natural
gas usually operate at a gas pressure of 3.5–5 in. of water, while propane operates
8417: “8417_c005” — 2007/9/11 — 12:16 — page 123 — #25

at 10–11 in. of water. Because of these characteristics, natural gas requires a larger
orifice size than propane at each burner, and a lower gas pressure.
If a range is set up for natural gas but has propane piped to it, it will be over-firing,
probably creating unacceptable levels of CO. A gas range in this condition must not be
used until the problem is corrected. Symptoms of this problem include noisy, yellow,
and large flames rising above the burner support grates on the range top burners;
carbon (smoke) emissions; or unacceptable carbon monoxide emissions.
If a range is set up for propane but has natural gas piped to it, it will be under-firing.
In this case, the customer might complain of the long period required to boil water or
the extended time required for baking. This condition is usually not hazardous, but it
should be corrected.
If the flexible gas line connector can be inspected without moving the range,
make sure the flexible connector is (1) not brass, (2) is not a two-piece con-
nector, and (3) has no pre-1973 rings (in some cases, the date can be found on
the flare nuts rather than the date rings). Do not move the range for the sole pur-
pose of inspecting the flexible connector; this movement might crack or otherwise
damage it.
Check for gas leaks at the range top burner area, oven area, and any accessible gas
lines with an appropriate combustible gas detector. Check for propane leaks below
connections (propane falls) and for natural gas leaks above connections (natural gas
rises). If any gas leaks are found, specify repair. Shut off the gas to the appliance and
do not proceed with testing until the leak is repaired.
If the gas range fails any of these items above (1) specify repair of the gas range or
(2) specify replacement of the gas range, depending on the character of the problem.
If the range does not fail, proceed, with the measurement of emissions.
5.2.3 MEASUREMENT OF EMISSIONS

5.2.3.1 Safety During Measurement of Emissions
While performing the emissions testing, monitor CO concentrations in the kitchen.
If indoor air concentrations rise above 35 ppm, shut down the burner(s), discontinue
testing, and open windows and/or doors. Be cautious not to burn hands or other
body parts on hot test equipment or the range. Also, be mindful not to damage test
equipment by open flames or hot surfaces. Do not damage customer’s counters, floors,
carpeting, or furniture with hot equipment or open flames.
This protocol calls for range top burners to warm up for at least 6 min before
recording emissions. Make sure that the open flame is not left unattended during
this warm-up period. If the analyst wishes to attend to other tasks during the burner
warm-up period, ask the customer to watch the burners during warm up.
5.2.3.2 Preparation for Burner Testing

Always calibrate the emissions measurement instrument according the manufacturer’s
recommendations. Before using the instrument, make sure that the most recent cal-
ibration is valid (check for the calibration label on the instrument). If the calibration
8417: “8417_c005” — 2007/9/11 — 12:16 — page 124 — #26

period has expired, calibrate the instrument before use. Zero the instrument according
to the manufacturer’s recommendations before testing.
5.2.3.3 Range Top Burner Testing

Test range top burners before testing the oven. Remove all objects from the range top.
The range top burners are to be tested in order of right rear (RR), left rear (LR), right
front (RF), and left front (LF). Test each range top burner with the CO Hot Pot™.
This protocol with its limit of as-measured CO per burner is based on the use of the
CO Hot Pot™, exactly as designed.22 The use of any other device or a variation of
the CO Hot Pot™ is likely to yield different results.
If room air from a fan or open window or door is blowing across the range top
burners, ask the customer to turn off or redirect the fan, or close the window or door.
The natural flow of combustion gases upward from the burner must not be disrupted
during the emissions testing process.
Center the CO Hot Pot™ on the burner grate. Prepare the emission measurement
instrument for the test. Ignite the burner and turn to the highest setting. Start timing
device.
Insert the probe of the emission measurement device into the hole on the side of
the CO Hot Pot™ and through to the center of the cylinder.
At 6 min after ignition, watch the instrument CO readings for 2 min. Record high
and low readings for this period. Average the high and low readings to get the 2-min
average CO as-measured.
The CO measured at the burner averaged over the 2 min must be 35 ppm or less,
as-measured, using the CO Hot Pot™.
Determine whether the burner passes or fails the limit. Test each of the top burners
in the order specified above.
5.2.3.4 Oven Bake Burner Testing

Test the oven bake burner only. If the oven has a separate broil burner, do not test it.
Clear the oven of all pots, pans, or other objects. Clear area below oven of
all objects. Leave oven shelves in place. If the vent holes on the oven bottom are
obstructed with foil, catch pans, or anything else, ask the customer to remove the
obstructions.
Ignite the burner, with the temperature setting at 350◦ F. The oven burner may not
ignite immediately; this is normal for some electronic ignition systems. Bake burners
with standing pilots usually ignite faster. Start timing device.
Insert the probe of the emission measurement instrument into the oven vent sleeve
at the back of the range top. Make sure the open end of the instrument probe is fully
inserted into the oven vent opening at its center. Do not allow dilution air to mix with
the sampled combustion by-products. Ensure that grease or other build-up does not
inadvertently block the instrument probe tip.
After beginning the oven test, do not open the oven door. If the oven door is
opened after the testing period begins, wait at least 5 min or to the end of the 15-min
warm up time, whichever is longer, before taking emissions readings.
8417: “8417_c005” — 2007/9/11 — 12:16 — page 125 — #27

It is not necessary to turn on the emissions measurement instrument at the begin-

ning of the warm-up, but it must be ready to take readings after 15 min of oven
warm-up time.
After 15 min of burner warm up, watch the emission measurement instrument
for the minimum and then maximum CO as-measured values. The corresponding
CO air-free must be calculated and averaged for these minimum and maximum CO
as-measured readings.
The step-by-step details of measurement: After 15 min of warm up, watch for the
minimum CO as-measured value (not the minimum CO air-free value). Record this
minimum CO as-measured value and the corresponding oxygen percentage (if your
instrument automatically calculates CO air-free, record this value at the minimum
CO as-measured value). Continue to watch the instrument until you detect the next
maximum CO as-measured value. Record this maximum value and the corresponding
oxygen percentage (if your instrument automatically calculates CO air-free, record
this value at the maximum CO as-measured value). Calculate the CO air-free emission
rates for the minimum and maximum CO as-measured readings from Equation 5.1a
(for natural gas and propane). Average the CO air-free emission rates for the minimum
and maximum CO as-measured readings. Some emissions measurement instruments
calculate CO air-free automatically. If this is the case, this equation need not be used.
Averaged CO air-free must be 800 ppm or less, averaged from the CO air-free
values corresponding to the CO as-measured minimum and maximum occurring after
fifteen minutes of warm-up, with oven set to 350◦ F.
Determine whether the burner passes or fails the limit.
5.2.3.5 Burner or Range Failure

If a failed burner can be adjusted in a way that reduces the CO emissions to below
those set by the levels of this protocol, then the range passes the protocol after the
field analyst retests the range to ensure that the burner(s) now passes limits of the
protocol.
If the failed burner(s) cannot be tuned or replaced to pass the protocol levels or
the gas range construction does not allow for adjustment or parts replacement, the
gas range should be replaced.
5.3 CONCLUSIONS
The Ohio surveys of the national weatherization program technicians demonstrated
a significant lack of uniformity among gas range field testing protocols.13 This
diversity of methods demonstrated disagreement about testing protocol and a dearth
of understanding regarding reliable field test data and methods. Using the Karg
Field Protocol for Gas Range Carbon Monoxide Emissions Testing developed by
the author would increase accuracy and standardization and repeatability to field CO
testing.15,16
The Karg Field Protocol demonstrates a fairly close correlation between field
measurements of oven CO emissions and longer-term emissions averaged from
8417: “8417_c005” — 2007/9/11 — 12:16 — page 126 — #28

15 to 30 min. The 1996 Battelle study to determine the validity of the assump-
tions of the ANSI Standard Z21.1-1993, Household Cooking Gas Appliances8 did
not examine the correlation of CO burner emissions at the 5-min mark of the ANSI
test with the longer term oven emissions. The evidence from the author’s field testing
suggests poor correlation between the data for ovens at the ANSI Z21.1 prescribed
5-min mark and the longer term, steady-state emissions averaged from 15 to 30 min.
The author’s data suggests the longer term, steady-state emissions averaged from
15 to 30 min gives more representative results. This significant finding indicates that
more research is needed to determine a valid oven testing method.
The field research for the development of the Karg Field Protocol reveals that
oven burners were more hazardous than range top burners. While just over 1% of
the range top burners tested in the field failed the Karg Field Protocol, 40% of the
over burners failed.15 Furthermore, during normal household use, oven burners are
usually operated longer than range top burners, increasing the potential hazard of a
high-emission oven burner. This increased oven failure rate in the test sample and
longer run time for ovens indicates it is more important to field test oven burners than
range top burners.
5.4 ACKNOWLEDGMENT
I would like to thank Tom Greiner, Ph.D. for reviewing and editing this chapter.
References
1. Persily, A.K., Carbon Monoxide Dispersion in Residential Buildings: Literature
Review and Technical Analysis, NISTIR 5906, National Institute of Standards and
Technology, Gaithersburg, MD, October 1996.
2. U.S. Consumer Product Safety Commission, Non-Fire Carbon Monoxide Deaths
Associated with the Use of Consumer Products: 2001 Annual Estimates, CPSC,
Bethesda, MD, May, 2004.
3. U.S. Consumer Product Safety Commission, Non-Fire Carbon Monoxide Deaths and
Injuries Associated with the Use of Consumer Products: Annual Estimates, CPSC,
Bethesda, MD, June, 1999.
4. U.S. Consumer Product Safety Commission, Non-Fire Carbon Monoxide Deaths and
Injuries Associated with the Use of Consumer Products: Annual Estimates, CPSC,
Bethesda, MD, October, 2000.
5. U.S. Consumer Product Safety Commission, Summary of Carbon Monoxide Emis-
sions Test Results of Gas Ranges with Self-Cleaning Ovens, Memorandum to Ronald
Jordan from Dwayne Davis and Christopher Brown, February 6, 2001.
6. Tikalky, S. et al., Gas Range/Oven Emissions Impact Analysis, Gas Research Institute,
Chicago, IL, December, 1987.
7. Reuther, J.J., Billick, I.H., Wiersma, S., Misconception, reality, and uncertainty
regarding co-emissions from unvented gas appliances, paper given to Karg in 2001
by Billick.
8. Reuther, J.J., Critique of ANSI Z21.1 Standard for CO Emissions from Gas-Fired
Ovens/Ranges, GRI-96/0270, September, 1996 for Gas Research Institute by Battelle,
Columbus, OH, 1996.
8417: “8417_c005” — 2007/9/11 — 12:16 — page 127 — #29

9. Greiner, T.H., The case of the CO leak: solving the mysteries of carbon monoxide
exposures, Home Energy, 21–28, November/December 1997.
10. Greiner, T.H., Comments concerning carbon monoxide emissions from gas-fired
ovens and ranges: with special reference to Battelle’s 1996 Critique of ANSI Z21.1,
prepared for Affordable Comfort Conference, Chicago, IL, available from Iowa State
University Extension, Ames, IA, 1998.
11. Andrews, T., Ohio Office of Energy Efficiency, personal communication, June
6, 2006.
12. Lenahan, T., Ohio Office of Energy Efficiency, personal communication, June
6, 2006.
13. Andrews, T., Survey Results of State Weatherization Programs Concerning Carbon
Monoxide Standards and Testing for Heating and Cooking Appliances, 1998, Ohio
Office of Energy Efficiency, Columbus, OH, 1998.
14. Heckerling, P.S. et al., Predictors of occult carbon monoxide poisoning in patients
with headache and dizziness, Annals of Internal Medicine, 107, 174–176, 1987.
15. Karg, R.J., Field Protocol for Gas Range Carbon Monoxide Emissions Testing,
Chicago Regional Diagnostics Working Group and U.S. Department of Energy,
Washington, D.C., 2001. Unpublished. Available at www.karg.com/papers.htm.
16. Karg. R.J., CO testing for the real world, Home Energy, 30–33, January/February
2002.
17. American National Standards Institute, Household Cooking Gas Appliances, Standard
Z21.1, American Gas Association, Cleveland, OH, 1993.
18. Greiner, T.H. and Schwab, C.V., Approaches to dealing with carbon monoxide in the
living environment, Carbon Monoxide Toxicity, Penney, D.G., (ed.), CRC Press, New
York, 2000, chap. 23.
19. Hedrick, R. and Billick, I., Gas Range/Oven Testing at the GRI Research House, slide
presentation to GRI Industry Environmental Council, Rosemont, IL, March 15, 2000.
20. Karg, R.J., Air-free measurement of carbon monoxide emissions from gas ranges:
analysis and suggested field procedure in Affordable Comfort ’98 Selected Readings,
Madison, WI, May, 1998.
21. Reuther, J.J., Interlaboratory Program to Validate a Protocol for the Measurement
of NO2 Emissions from Rangetop Burners, GRI-94/0458, December 1994 for Gas
Research Institute by Battelle, Columbus, OH, 1994.
22. Karg, R.J., CO Hot Pot™ Assembly Instructions, www.karg.com/COhotpot.htm.
23. Tsongas, G., Carbon monoxide from ovens: a serious IAQ problem, Home Energy,
18–21, September/October 1995.
24. Tsongas, G., Field monitoring of elevated CO production from residential gas ovens,
in Proceedings: Indoor Air Quality ’94, St. Louis, MO, American Society of Heating,
Refrigerating, and Air Conditioning Engineers, Inc., Atlanta, GA, 1994.
25. Hedrick, R., personal communication, June 20, 2001.
26. Building Performance Institute (BPI), Technical Standards for Certified Building
Analyst1, BPI, Malta, NY, 2005.
8417: “8417_c005” — 2007/9/11 — 12:16 — page 128 — #30

6 Investigating Carbon
Monoxide-Related
Accidents Involving
Gas-Burning Appliances
Michael Hanzlick
CONTENTS
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

6.2 Basic Gas Appliance Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
6.3 The Venting System and Air for Combustion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
6.3.1 The Draft Diverter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
6.3.1.1 Safety Problems Associated with the Draft Diverter . . . 133
6.4 Gas Appliance Safety Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
6.5 Gas Appliance Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
6.5.1 Household Cooking Gas Appliances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
6.5.2 Clothes Dryers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
6.5.3 Space Heating Appliances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
6.5.3.1 Central Furnaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
6.5.3.2 Room Heaters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
6.5.3.3 Wall Furnaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
6.5.3.4 Low-Pressure Boilers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
6.5.3.5 Water Heaters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
6.5.3.6 Special Design Appliances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
6.6 Gas Appliance Certification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
6.7 Gas Appliance Installation Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
6.8 National Fuel Gas Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
6.9 Manufacturers Installation Instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
6.9.1 Proper Fuel Input Adjustments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
6.10 Fuel Gases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
6.11 Basic Principles of Combustion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
6.11.1 Products of Complete Combustion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
6.11.2 Products of Incomplete Combustion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
6.12 CPSC Estimates of Carbon Monoxide Deaths . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
6.13 The Heating/Gas Appliance System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
129
8417: “8417_c006” — 2007/9/11 — 12:16 — page 129 — #1

6.14 Failure Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146

6.15 Investigation of Root Cause . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
6.15.1 Initial Discovery and Investigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
6.15.2 Follow-Up Inspections and Investigation . . . . . . . . . . . . . . . . . . . . . . . . . 149
6.15.3 Determination of Root Cause . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
6.15.4 Estimation of Ambient Carbon Monoxide Levels in Occupied
Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
6.16 Case Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
6.16.1 Investigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
6.16.2 Failure Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
6.16.3 Case Study Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
6.1 INTRODUCTION
The gas appliance industry is one of the oldest and most stable industries in the
United States, having its origin in the early 1800s.1 The industry has prided itself
on its record related to the safe and efficient operation of gas appliances and has
consistently worked to improve the safety of gas appliances by incorporating design
changes when appropriate.
Of major concern in the industry is the production of carbon monoxide (CO) during
appliance operation. CO is a colorless, odorless, highly toxic gas. While there are
many sources for CO, it can be produced in gas burning appliances by the incomplete
combustion (burning) of a carbon-based fuel, such as natural gas or propane. In
the gas appliance industry, CO has been called the “silent killer” because its victims
are normally unaware of any operational problems with the gas appliance or that they
have been poisoned.
Modern gas appliances, those manufactured within the last 50 years, are designed
and safety certified to perform their intended function without producing dangerous
levels of CO when installed, operated, and maintained in accordance with the man-
ufacturer’s instructions. However, there are many key factors that are not controlled
by the appliance manufacturer that can cause an otherwise safe appliance to begin to
produce CO.
6.2 BASIC GAS APPLIANCE DESIGN

The efficient and safe operation of a gas-fired appliance is dependent on a number
of factors some of which are controlled by the overall design and construction of the
appliance. Some of the most basic of these factors are (see Figure 6.1)
• The proper amount of fuel must be mixed with the proper amount of oxygen
and its temperature raised to the ignition point so that combustion (burning
of the fuel to produce heat) takes place.
• The actual mixing of the fuel and air takes place within the gas burner
assembly contained within the appliance.
8417: “8417_c006” — 2007/9/11 — 12:16 — page 130 — #2

Investigating Carbon Monoxide-Related Accidents 131
Products of combustion
mixed with dilution air
Warm air plenum exits to outside atmosphere
Vent system
Warm air
to heated area
Draft diverter
Dilution air
Enters at draft diverter
relief opening
Combustion air
Hot products of
Mixes with fuel and is
combustion
combusted in the gas burner
(flue gases)
assembly
Fuel
Gas burner assembly Natural gas
or
propane lgas
Cool return air Blower compartment

from heated area Air blower
FIGURE 6.1 Typical gas-fixed appliance (furnace).
• The initial ignition of the fuel is often accomplished by the use of an

electric ignition device or by a small constant burning flame referred to
as a pilot light or pilot burner.
• The actual burning of the fuel and production of heat takes place in a
confined combustion chamber within the appliance.
• As the fuel is combusted or burned, hot products of combustion (flue
gases) are formed and travel through the appliance.
• Heat energy is removed from the hot products of combustion in the heat
exchanger and used to provide useful functions (heat air to heat a home,
cook food, dry clothes, or even produce light.)
• Finally, the flue gases exit the appliance and are dispersed into the
atmosphere through a venting system.
6.3 THE VENTING SYSTEM AND AIR FOR

COMBUSTION
As the appliance operates, products of combustion (flue gases) must be removed from
the appliance and new air for combustion supplied. New air for combustion is mixed
with the fuel and combusted at the main burner. This process establishes a “natural”
8417: “8417_c006” — 2007/9/11 — 12:16 — page 131 — #3

draft through the appliance due to the buoyancy of the hot flue gases, and establishes
the rate at which heat is extracted from the appliance through its heat exchanger. This
effect also acts to establish the overall operating efficiency of the appliance. In order
for an appliance to operate at an efficient and safe level on a consistent basis, this
natural draft through an appliance must be maintained at as near a constant level as
possible.
Products of combustion for a properly adjusted gas appliance contain amounts of
carbon dioxide, nitrogen, excess oxygen, and water vapor.2 If the appliance is in poor
condition, poorly maintained or installed incorrectly, products of combustion can
also contain dangerous levels of CO. The products of combustion and flue gases are
removed from the appliance and vented to the outside atmosphere through a venting
system.
The venting system that an appliance is connected to can have a dramatic effect
on the draft through an appliance. The draft rate through a vent system can increase
or decrease depending on a number of variables some of which include
• Height of the vent terminal above the appliance

• Size of the vent system
• The difference between inside and outside temperatures (stack effect)
• Adequate supply of new makeup air for combustion to the appliance
• Outside wind conditions
• Obstructions near the vent terminal
The venting system must not be allowed to restrict or block the natural draft of
combustion products through an appliance. If the natural draft through an appliance
is blocked, incomplete combustion resulting in the production of CO and possible
fire hazards are created.
A continuous source of new combustion air containing oxygen must be supplied
to the appliance to maintain proper combustion and operation of the appliance. If the
supply of new combustion air is restricted incomplete combustion resulting in the
production of CO can result.
6.3.1 THE DRAFT DIVERTER

The draft diverter, or draft hood, is one method utilized to “couple” the gas appliance
to the venting system and neutralize the effects that a vent system may have on the
operation of the appliance. The draft diverter is a device that is built into an appliance
and is designed to
• Provide a path for the escape of the flue gases from the appliance in the
event of no draft, back draft, or stoppage beyond the draft hood in the vent
system.
• Prevent a back draft from entering an appliance.
• Neutralize the effect of stack action of the gas vent upon the operation of
the appliance.
8417: “8417_c006” — 2007/9/11 — 12:16 — page 132 — #4

In its most basic form, a draft diverter is in the shape of a “box.” It has an inlet for
the products of combustion to enter, an outlet at the top connecting to the vent system
and a “relief” opening on one side that is open to the area around the appliance.
A draft diverter may contain one or more baffles to control how the products of
combustion flow through the draft diverter. In operation, products of combustion
from the appliance enter the draft diverter through the inlet, collect in the diverter
box, and exit the draft diverter through the outlet to the venting system. During
normal operation “dilution” air, air from the area around the appliance, enters the
draft diverter through the relief opening and mixes with the products of combustion.
This process of adding dilution air through the relief opening acts to allow the flow rate
through the vent system to increase or decrease, depending on vent system design and
outside atmospheric conditions, while maintaining a constant natural draft through
the appliance.
The design of the draft diverter also allows products of combustion to exit
the appliance through the relief opening in the event that the vent system should
become blocked or if a down draft or reverse draft should develop in the venting
system.
6.3.1.1 Safety Problems Associated with the Draft Diverter

While a draft diverter provides an important safety function of protecting the appliance
from the effects of the venting system, the basic design of the draft diverter presents a
serious safety concern in itself. Specifically, a blocked vent or a down draft condition
in the vent will cause products of combustion to “spill” out of the draft diverter relief
opening into the area around the appliance. This is normally not a problem if the
blockage or down draft condition is caused by a temporary condition and the spillage
only occurs for short periods of time.
However, if the spillage condition is allowed to continue for a prolonged period
of time a serious safety condition will develop. Products of combustion will be drawn
back into the appliance, replacing the normal supply of combustion air, causing the
appliance to have “incomplete” combustion resulting in the production of CO. This
cycle, if allowed to continue, can generate sufficient amounts of CO in the area around
the appliance to cause a serious threat to human life.
Spillage of products of combustion from the draft diverter can be caused by
• Blockage or restriction in the vent system

• A poorly designed, maintained or installed vent system
• Back draft in the vent system
An inadequate supply of combustion air can also cause spillage. As oxygen is

depleted in the combustion process, in the area around the appliance, new air con-
taining oxygen for combustion will be drawn to the appliance from the path of least
resistance. If the appliance is installed in a confined area, the path of least resistance
can be through the venting system, which creates a back draft. This back draft causes
the products of combustion to exit the appliance through the relief opening of the
draft diverter.
8417: “8417_c006” — 2007/9/11 — 12:16 — page 133 — #5

6.4 GAS APPLIANCE SAFETY DEVICES

All modern day appliances are equipped with basic, and sometimes very complicated,
safety devices in an attempt to ensure the safe operation of the appliance. Some of
these safety controls include
• Thermocouple used on constant burning pilot systems—A device that
utilizes the heat from the constant burning pilot flame to generate an electric
current to hold open a gas valve piped in series with the main gas burner.
If the constant burning pilot flame is extinguished the gas valve closes and
shuts off all gas to the appliance. The appliance user must manually relight
the constant burning pilot flame. This type of safety is the most basic and
is commonly found on all appliance types.
• Electronic ignition systems—Electronic ignitions systems are either spark
ignition systems or some form of a hot wire or carbon bar system. These
systems eliminate the constant burning pilot flame and either ignites a small
pilot flame or the main burner flame directly. The presence of a flame must
be “proven” electronically within a very short time frame. If the flame
fails to ignite all gas to the appliance is shut off and the appliance will not
operate. This type of system can be very complicated and most require a
trained service representative to repair.
• High limit control—An electrical switch that senses the operating tem-
perature of the appliance. The control is electrically connected to the
appliance so that if the switch senses an over temperature condition the
switch will act to shut down the operation of the appliance. Normally,
the switch will automatically reset itself after the appliance cools down
and allows the appliance to operate without any intervention by the user.
• Flame rollout switch—A flame rollout switch is an electric temperature
sensitive switch that monitors the temperature in the area around the main
burner area. If a malfunction occurs that would cause the main burner flame
to burn outside of the burner compartment, the rollout switch will act to
shut down the operation of the appliance. In some cases this switch must be
manually reset after it senses a malfunction. Although the technology for
flame rollout switches has existed since at least the early 1950s, appliance
manufacturers did not begin to routinely install flame rollout switches on
appliances until the mid 1980s.
• Vent safety shutoff system—A vent safety shut off system normally con-
sists of an electrical switch that senses the temperature of the air in the draft
diverter relief opening area. During normal operation the temperature in
this area is at, or near, room temperature. If the vent system malfunc-
tions causing products of combustion to spill out of the relief opening, the
temperature at this location will rapidly rise to an elevated temperature
because hot flue gases are exiting the appliance through the relief opening.
The switch is electrically connected to the appliance so that if the switch
senses an over temperature condition the switch will act to shut down the
operation of the appliance. Normally, the switch will automatically reset
itself after the appliance cools down and allows the appliance to operate
8417: “8417_c006” — 2007/9/11 — 12:16 — page 134 — #6

without any intervention by the user. Although the technology for vent
safety shut off systems has existed since at least the early 1950s, appli-
ance manufacturers did not begin to routinely install this safety device on
appliances until the mid 1980s.
6.5 GAS APPLIANCE TYPES

Although all modern gas appliances share the same basic design concepts, not all
appliances are designed to operate alike or provide the same function to the consumer.
The most common types of gas appliances are:
6.5.1 HOUSEHOLD COOKING GAS APPLIANCES

Commonly referred to as gas stoves or gas ranges. Most household cooking appliances
contain small top burners for cooking food in pots or pans, an oven for baking food
and a broiler for broiling or grilling food. Household cooking appliances may be free
standing, built-in or a combination of the two.
Household cooking appliances are normally not equipped with a draft diverter and
are not connected to a venting system. Instead, household cooking appliances usually
exhaust the products of combustion directly into the area just above the appliance.
This is the same area that is normally occupied by the user of the cooking appliance.
For this reason, household cooking appliances are of special concern when
considering the overall safety of operation of the appliance.
6.5.2 CLOTHES DRYERS

Primarily used in the family living environment to dry clothes after washing. The gas
fired clothes dryer looks very much like its electric counter part.
Clothes dryers are not equipped with a draft diverter. Clothes dryers depend on
an internal blower to circulate the hot products of combustion directly around and
through the wet clothes to achieve the drying action. This same blower then forces
the very wet cooler products of combustion from the clothes dryer through the dryer
vent to the outside atmosphere.
Clothes dryers consume an enormous amount of air from the living environment
during the drying operation. For this reason it is critical that clothes dryers (both
electric and gas) not be installed in the same room as other gas fired appliances.
6.5.3 SPACE HEATING APPLIANCES

There are four basic types of space eating appliances:
6.5.3.1 Central Furnaces

The most typical gas appliance utilized to heat a home. A central furnace is a self-
contained gas-burning appliance for heating air by the transfer of heat through a heat
exchanger. The warm air is then circulated to the home through a heat distribution
system (heating ducts.) A central furnace is normally equipped with an air blower
8417: “8417_c006” — 2007/9/11 — 12:16 — page 135 — #7

that provides the primary means for circulating the air. There are three variations of
the central furnace
• Upflow—The blower is located at the bottom of the furnace and air flows
up through the appliance.
• Downflow—The blower is located on the top of the furnace and air flows
down through the appliance.
• Horizontal—The furnace appears to be laying on its side; the blower is
located at one end of the furnace and air flows horizontally through the
appliance.
Central furnaces are often equipped with a cooling coil in the warm air outlet of
the furnace. During the hot summer months the furnace blower is utilized to circulate
cool air to the home. Central furnaces are normally equipped with a draft diverter and
are connected to a venting system.
6.5.3.2 Room Heaters

Aself-contained, freestanding gas appliance utilized to heat a small area. Like a central
furnace, room heaters heat air by the transfer of heat through a heat exchanger. The
warm air is then allowed to circulate into the area to be heated. They may be equipped
with a small fan or blower to assist the circulation of the warm air through the area
being heated. Unlike a central furnace, a room heater is not connected to a heat
distribution system.
Room heaters are normally equipped with a draft diverter and most room heaters
(vented) are connected to venting system. However, some room heaters (unvented)
are designed to exhaust their products of combustion directly into the room being
heated and are not connected to a venting system.
6.5.3.3 Wall Furnaces

A self-contained gas appliance utilized to heat a small area. Wall furnaces are either
recessed into a wall of the room being heated or installed flush on the wall. Like a
central furnace, wall furnaces heat air by the transfer of heat through a heat exchanger.
The warm air is then allowed to circulate into the area to be heated.
Wall furnaces may be equipped with a small fan or blower to assist the circulation
of the warm air through the area being heated. Unlike a central furnace, a room heater
is not connected to a heat distribution system. Some wall furnaces are designed to
provide heat to two adjacent rooms. In this case the wall furnace is located on the
common wall between the two rooms and a small opening from the back of the wall
furnace allows heat to flow into the second area. Wall furnaces are normally equipped
with a draft diverter and are connected to a venting system.
6.5.3.4 Low-Pressure Boilers

A self-contained gas burning appliance for supplying hot water for space heating. The
boiler heats water by the transfer of heat through a heat exchanger. The hot water is
8417: “8417_c006” — 2007/9/11 — 12:16 — page 136 — #8

then circulated to the home through a piping distribution system and utilizes radiators
to heat specific areas inside the home.
A hot water heating system is normally equipped with one or more circulator
pumps that circulate the hot water from the boiler to the areas being heated (zones.)
Each area, or zone, may operate independently of the other zones. Low-pressure
boilers normally maintain the water contained in the boiler at a constant temperature,
usually around 180˚ F. Low-pressure boilers are normally equipped with a draft
diverter and are connected to a venting system.
6.5.3.5 Water Heaters

A water heater is a closed vessel in which water is heated through a heat exchanger
by the hot products of combustion. The most common type of water heater is the
automatic storage type that heats and then stores water, within the appliance, for
delivery to the user on demand. Typical automatic storage type water heaters have
30 gallon, 40 gallon, or 50 gallon water capacities and deliver water between 100˚ F
and 120˚ F. They are normally equipped with a draft diverter and are connected to a
venting system.
Water heaters are normally installed in the same area of the home as the central
heating appliance. When this is the case, the water heater and the central heating
appliance will usually share the same venting system. Each appliance is connected
to the “common” vent system through its own vent connector. The vent connector
consists of a short section of vent pipe that connects the outlet of each appliance to
the inlet of the common vent system. In this type of installation, any failure of the
common portion of the venting system will affect both appliances connected to it.
6.5.3.6 Special Design Appliances

While most gas-fired appliances are equipped with draft diverters and are connected
to a venting system, there are two special designs that eliminate the draft diverter and
in some cases eliminate the need for a typical venting system.
Direct Vent—A system consisting of an appliance, combustion air, and flue gas
connections between the appliance and the outside atmosphere. The appliance, flue
gas connections, and vent termination cap are supplied by the appliance manufacturer
and constructed so that all air for combustion is obtained from the outside atmosphere
and all flue gases are discharged to the outside atmosphere.
In this type of system there is no draft diverter. The appliance, normally a wall
furnace, is installed on an outside wall of the structure, and the combustion and flue
gas connections are directly to and from the appliance through the wall to the outside
atmosphere.
Induced Draft—A system that utilizes a small air blower to induce or force the
products of combustion through the venting system to the outside atmosphere. The
blower takes the place of the draft diverter and forces the products of combustion
through the system.
This type of system is more efficient than the type of system that utilizes a draft
diverter and is normally utilized on mid and high efficiency central space heating
8417: “8417_c006” — 2007/9/11 — 12:16 — page 137 — #9

systems. Both of these types of special design appliances are safer designs because
they eliminate most of the problems associated with the draft diverter and the vent
system.
6.6 GAS APPLIANCE CERTIFICATION

In the United States, most building codes in effect today require that the building
code authority approve any gas appliance before it is installed. Normally any gas
appliance that has been tested and certified (listed) by an approved testing agency
will be considered “approved” by the building code authority.
Most gas appliance manufacturers in the United States voluntarily have their
products tested and certified through a certification program provided by the Canadian
Standards Association (CSA)3 and the Z21/83 Committee.
The CSA and the Z21/83 Committee are jointly accredited by the American
National Standards Institute (ANSI) to develop the Z21 series of standards for
gas burning appliances and related accessories. These standards cover construction
requirements, safe operation, performance aspects, and laboratory test methods. CSA
provides the administrative services required to operate the program, and the Z21/83
Committee, along with its Technical Advisory Groups, develop the Z21 series of
standards in accordance with ANSI approved procedures.
CSA is accredited by ANSI as a Nationally Recognized Testing Laboratory and
is considered to be an approved testing agency as required under most building
codes.
For gas appliances manufactured in the United States, the CSA Star indicates that
the appliance has been tested under the applicable Z21 ANSI Standard and is certified
to meet the requirements of that standard. (see Figure 6.2)
Before the CSA, the American Gas Association (AGA) was recognized as the
approved testing agency and used the same star certification seal as an indication
that the appliance was certified to meet the requirements of the applicable Z21 stand-
ard. This seal, in some form, has been found on certified appliances manufactured
since 1925.
ESIG N
D
ER D
C
TIFIE ®
FIGURE 6.2 CSA Star.
8417: “8417_c006” — 2007/9/11 — 12:16 — page 138 — #10

6.7 GAS APPLIANCE INSTALLATION REQUIREMENTS

The installation, adjustment, operation, and repair of gas burning appliances are gov-
erned by national building codes adopted by the building authority having jurisdiction
in a local community such as a city or county. There are currently four major national
building code groups in the United States.
• Building Officials and Code Administrators International, Inc. (BOCA)

Primarily used in the Northeastern region of the United States. BOCA
publishes the National Mechanical Code and the National Plumbing Code.
• International Conference of Building Officials (ICBO) Primarily used
in the Western half of the United States. ICBO publishes the Uniform
Mechanical Code.
• Southern Building Code Congress International, Inc. (SBCCI) Primarily
used in the Southeastern region of the United States. SBCCI publishes the
Standard Mechanical Code and the Standard Plumbing Code.
• International Association of Plumbing and Mechanical Officials (IAPMO)
Primarily used in the Western half of United States. IAPMO publishes the
Uniform Mechanical Code (jointly with ICBO) and the Uniform Plumbing
Code.
In 1972 the Council of American Building Officials (CABO) was established

by BOCA, ICBO, and SBCCI to compile and standardize the code requirements
published by BOCA, ICBO, and SBCCI. The International Code Council, Inc. (ICC)
was organized on December 9, 1994 as successor to CABO. The ICC publishes the
International Fuel Gas Code, the International Mechanical Code, and the International
Plumbing Code.
While each national building code group has developed its own set of codes that
apply to the installation of gas appliances, the codes are very similar to one another
with only minor differences between them. In some cases, the local building authority
may supplement sections of a national code with more stringent requirements.
6.8 NATIONAL FUEL GAS CODE

The National Fire Protection Association (NFPA) was organized in 1896 to improve
the methods of fire protection and to obtain and circulate information on these
subjects.4 The NFPA in its efforts has developed a set of National Standard Codes
that includes the National Fuel Gas Code—NFPA 54. The National Fuel Gas Code—
NFPA 54 was developed by gas industry experts, and is considered to be the industry
standard for the installation of gas piping and gas appliances.
6.9 MANUFACTURERS INSTALLATION

INSTRUCTIONS
The appliance manufacturer supplies each gas-burning appliance with detailed
installation instructions. All of the national building codes referenced above require
8417: “8417_c006” — 2007/9/11 — 12:16 — page 139 — #11

that the appliance be installed in accordance with the manufacturer’s installation

instructions.
Most manufacturer’s installation instructions require the installation to comply
with the National Fuel Gas Code—NFPA 54 and provide instructions for
• Proper gas appliance location
• Adequate gas piping supply
• Adequate combustion and ventilation air supply
• Proper venting system
• Proper fuel input adjustment
All of these topics are covered in detail in the National Fuel Gas Code—
NFPA 54.
6.9.1 PROPER FUEL INPUT ADJUSTMENTS

Even though all of the items listed in the installation instructions are critical for a
safe installation, the most common installation requirement that is often overlooked
is the proper fuel input adjustment. The fuel input rating for a gas-burning appliance
can be found on the appliance rating plate attached to the appliance. This rating is the
amount of fuel, expressed in British Thermal Units (BTUs), which the appliance will
consume or burn during each hour of operation at sea level (zero feet in elevation.)
Gas-burning appliances are preadjusted to burn this amount of fuel by the manu-
facturer before shipment to the distributor. The appliance installer will often assume
that the fuel input is correct because it was pre-adjusted by the manufacturer. While
this practice may be valid at sea level, at higher elevations above sea level it is not
valid.
A cubic foot of fuel (natural gas or propane gas) requires a specific amount of
oxygen (supplied in the air) for proper and safe combustion. As the elevation above
sea level increases, the amount of oxygen in the air decreases. Because an appliance is
physically designed to burn a specific amount of gas with a specific amount of oxygen
supplied in the combustion air, the fuel input rate to the appliance must be adjusted
down, or derated, to account for the lower levels of oxygen at increased elevations.
As stated in most appliance installation instructions, at elevations above 2000 ft.
the fuel input rate must be reduced by 4% for each 1000 ft. above sea level. For
example, a central furnace rated at 100,000 BTU/h as listed on its rating plate, would
only be rated at 80,000 BTU/h when installed at 5,000 ft. above sea level (Denver,
Colorado.) If the installer of this furnace did not check and adjust the fuel input rate
for the increased elevation, the appliance would be burning a too rich mixture of fuel
and air with the increased probability of producing dangerous amounts of CO.
6.10 FUEL GASES

There are four major groups of fuel gases; natural gases, liquefied petroleum (LP)
gases including propane and butane, manufactured gases, and mixed gases. The most
common type fuels for household use in the United States are natural gas followed
by propane and then to a much lesser extent butane. Natural gas, propane gas, and
8417: “8417_c006” — 2007/9/11 — 12:16 — page 140 — #12

butane gas are all classified as hydrocarbon gases. A hydrocarbon gas is a gas that is
made up of both hydrogen and carbon atoms, thus the term hydrocarbon.
Natural gas is a naturally occurring mixture of hydrocarbon gases as well as small
amounts of nonhydrocarbon gases found in porous formations beneath the earth’s
surface. The principle component of most natural gases found in the United States is
methane. Lesser amounts of ethane, propane, butane, and other gases are also found
in natural gas.
The major source of propane and butane in the United States is crude petroleum.
Another important source of propane and butane is from natural gas through a process
known as absorption. These fuels are absorbed from natural gas in oil and then refined
into propane, butane, and natural gasoline.
The boiling point of natural gas at atmospheric pressure is −260◦ F.5 In order to
liquefy natural gas it is necessary to cool it to its boiling point. As the pressure of
natural gas is increased the boiling point also increases. The critical temperature of
natural gas is about −116◦ F at about 673 pounds per square inch (PSI) of pressure.
Simply put, this indicates that natural gas will always be in a vapor state whenever
the temperature of the gas is above −116◦ F regardless of the pressure. A natural gas
utility company normally provides natural gas to the end user through a distribution
system of underground pipes.
The boiling point of propane is −44◦ F at atmospheric pressure.6 When propane
is enclosed in a container at room temperature it will reach a state in which some
of the propane will be in a vapor state and some of the propane will liquefy. The
propane vapor causes the pressure inside the container to increase. The pressure inside
the container will depend on the temperature (see Table 6.1). This characteristic of
propane allows the fuel to be stored very easily in a storage tank as a liquid and then
vaporized and used as a fuel in a gas-fired appliance. Propane gas is most often found
on portable appliances (outdoor grills), on recreational vehicles, and in rural areas as
a heating fuel.
The boiling point of butane is 32◦ F at atmospheric pressure. Like propane, when
butane is enclosed in a container at room temperature it will reach a state in which some
TABLE 6.1
Vapor Pressures of Propane and Butane Gases
Propane Vapor Butane Vapor
Temperature (◦ F) Pressure (psi) Pressure (psi)
−40 1.3
−20 10.7
0 23.5
20 40.8
40 63.3 3.0
60 92.5 11.5
70 109.3 16.5
80 128.1 22
100 172.3 37
8417: “8417_c006” — 2007/9/11 — 12:16 — page 141 — #13

of the butane will be in a vapor state and some of the butane will liquefy. Because the
boiling point of butane is 32◦ F it will always remain a liquid at temperatures below
32◦ F. This characteristic makes it very difficult to use butane as a heating fuel in
locations where the storage temperature is below 32◦ F.
Another important characteristic shared by natural gas, propane gas, and butane
gas is how the number of atoms affects the gases. Referring to Figure 6.3, note
that methane has fewer carbon and hydrogen atoms than ethane, propane has more
atoms than ethane, and butane has more atoms than propane. As the number of atoms
increase the actual weight or density of the gas increases. Also as the number of atoms
increases the heating value of the gas increases.
The weight of a gas is normally expressed as a ratio of the weight of the gas
compared to the weight of the same volume of dry air. This ratio is referred to as
the specific gravity of the gas. A specific gravity greater than 1 indicates the gas is
heavier than air and if the gas is released into the air it will sink to the lowest level.
A specific gravity less than 1 indicates the gas is lighter than air and when released
into the air it will rise. Referring to Table 6.2, note that natural gas is lighter than air.
H H H
H C H H C C H
H H H
Methane CH4 Ethane C2H6
H H H H H H H
H C C C H H C C C C H
H H H H H H H
Propane C3H8 Butane C4H10
FIGURE 6.3 Carbon atoms and hydrogen atoms linked together; the aliphatic series.
TABLE 6.2
Specific Gravity and Heating Values of Several Hydrocarbon Gases
Natural gas Propane Butane
Specific gravity
(Air = 1.00) 0.58 − 0.79 1.53 2.00
Heating value
BTU/cu. ft. 900 − 1200 2516 3280
8417: “8417_c006” — 2007/9/11 — 12:16 — page 142 — #14

Also propane and butane are both heavier than air. This characteristic makes natural
gas a safer fuel because if released into the air it will rapidly dissipate into the
atmosphere while propane and butane will tend to collect and pool at floor level.
The heating value of a gas is normally expressed in British Thermal Units per
cubic foot (BTU/cu. ft.) One (1) BTU is the amount of heat required to raise one
(1) pound of water one (1) degree Fahrenheit (◦ F.) Referring to Table 6.2, note that
butane has a higher heating value than propane and propane has a higher heating value
than natural gas. This characteristic makes it critical that the gas-fired appliance be
properly adjusted to burn the type of fuel that is supplied. A gas-fired appliance that
has been adjusted to burn propane fuel will not operate safely on natural gas just as
a gas-fired appliance adjusted to burn natural gas will not operate safely on propane.
6.11 BASIC PRINCIPLES OF COMBUSTION

Combustion, or burning, of a fuel is defined as the rapid oxidation of the fuel. It
is the process of oxygen acting with the fuel to produce heat and light very rapidly.
Some people have defined combustion as the controlled explosion of a mixture of fuel
and oxygen to produce controlled amounts of heat or light. For combustion to take
place the proper amount of fuel must be mixed with the proper amount of oxygen
(air.) Heat must be added to raise the mixture to its ignition point. Once ignition
takes place the combustion process will continue as long as the fuel and air supply is
maintained.
About 10 cu. ft. of air containing normal amounts of oxygen is required to burn
1000 BTU of either natural gas or propane gas at optimal conditions. However, the
limits of flammability, or explosive limits, of natural gas is between about 4% and
14% gas in air and for propane between about 2.4% and 9.6% gas in air. If the fuel
mixture is below the explosive limit the mixture is too lean to burn; and, if the mixture
is above the explosive limit the mixture is too rich to burn. The ignition point of natural
gas and propane gas is normally between about 900◦ F and about 1000◦ F.
6.11.1 PRODUCTS OF COMPLETE COMBUSTION

When a hydrocarbon fuel gas, like natural gas, is burned in a mixture of air containing
a sufficient amount of oxygen (O2 ) the normal products of combustion are carbon
dioxide (CO2 ) and water vapor (H2 O.) Included with the combustion products is the
nitrogen (N2 ) that was contained in the combustion air. Nitrogen, because it is an
inert gas and does not burn, passes through as part of the flue gases. The total volume
of the flue gases contains (see Figure 6.4) CO2 and H2 O (products of combustion),
N2 (carry through from the combustion air), and excess air. Heat as shown refers to
ignition of the gas mixture. Excess air is air that was drawn into the appliance with
the combustion air but was not needed for combustion.
When burning a fuel such as natural gas or propane in a gas-designed appliance,
the volume of gas burned on a per hour basis remains constant. Therefore, the volume
of the products of combustion, CO2 and H2 O, will remain constant. However, the
total volume of the flue gases will vary depending on the amount of excess air that
8417: “8417_c006” — 2007/9/11 — 12:16 — page 143 — #15

100 cu.ft.
carbon dioxide
Products
of
1000 cu.ft. combustion
combustion air 200 cu.ft.
(Oxygen + Nitrogen) water vapor
100 cu.ft.
natural gas
250 cu.ft.
excess air 800 cu.ft.
nitrogen
HEAT
250 cu.ft.
excess air
1350 cu.ft. (fuel + combustion air + excess air) 1350 cu.ft. flue gases
FIGURE 6.4 Products of complete combustion.
is drawn into the appliance. Appliance manufacturers utilize excess air as a built-in
safety factor when designing the combustion system in an appliance to ensure that
complete combustion results.
CO is not a product of complete combustion.
6.11.2 PRODUCTS OF INCOMPLETE COMBUSTION

Sometimes when a hydrocarbon fuel gas is burned incomplete combustion results.
When this occurs, CO is one byproduct of the incomplete combustion in addition to the
normal byproducts of complete combustion, carbon dioxide (CO2 ) and water vapor
(H2 O). The most common cause of incomplete combustion in a gas-fired appliance
is an inadequate supply of combustion air.
Another less common cause of incomplete combustion is flame cooling, and to a
lesser extent, insufficient mixing of the fuel and combustion air. Flame cooling results
when the flame impinges on a cool surface, such as a heat exchanger wall. This causes
the flame to cool to below the ignition point of the fuel. When this occurs the mixture is
not fully combusted and CO, as well as other byproducts, is an unwanted by product.
CO is actually a hydrocarbon fuel similar to propane or butane. The heating value
of CO is about 322 BTU/cu. ft. and the specific gravity is 0.9672, making it slightly
lighter than air. CO has no odor.
Although CO has no detectable odor, there is another byproduct of incomplete
combustion that does. Just as CO is the byproduct of the carbon component in the fuel,
there are also byproducts of the hydrogen component, namely aldehydes, hydrogen,
and other hydroxylated compounds. Aldehydes usually have a sharp penetrating odor
that acts to sting the nose and eyes. Very small amounts of aldehydes in air are
detectable by the normal human nose.
The presence of aldehydes indicates the presence of CO. However, the lack of a
detectable odor does not necessarily indicate that combustion is complete since CO
and hydrogen are both odorless.
8417: “8417_c006” — 2007/9/11 — 12:16 — page 144 — #16

TABLE 6.3
CPSC Estimated Annual Carbon Monoxide Poisoning Deaths 1994–2002
(Nonfire Related)
Percent of
1994–1998 1999 2000 2001 2002 Total Total (%)
Unspecified gas
heating 25 3 7 6 15 56 7.3
LP gas heating 46 22 29 26 50 173 22.6
Natural gas heating 35 19 37 28 22 141 18.4
Gas water heaters 7 1 3 0 1 12 1.6
Gas cooking ranges/
ovens 7 6 11 10 3 37 4.8
All other causes 80 58 51 60 97 346 45.2
Total 200 109 138 130 188 765 100.0
6.12 CPSC ESTIMATES OF CARBON MONOXIDE

DEATHS
The Consumer Product Safety Commission (CPSC) has tracked deaths in the
United States due to CO from the late 1970s. In a report7 titled “Non-Fire Carbon
Monoxide Deaths Associated With The Use Of Consumer Products 2002 Annual
Estimates”, published on July 12, 2005, the Commission has estimated the number
of deaths that has occurred each year from 1994 due to gas-fired appliances (see
Table 6.3).
As demonstrated in Table 6.3, the majority of deaths were caused by natural
gas and LP gas (propane and butane) heating systems followed by gas-fired cooking
ranges/ovens and to a lesser degree gas-fired water heaters.
As noted in the CPSC report, these are estimates of documented deaths and do
not include reported and nonreported injuries and illness caused by CO or deaths that
were misdiagnosed as non-CO deaths.
Table 6.3 also indicates, regardless of any advances in appliance design, the
number of deaths each year attributed to gas-fired heating systems has remained
fairly constant over the past 10 years. This supports the concept that the manner in
which the gas appliance is installed and how the appliance is serviced and maintained
has a huge influence on the safe operation of the appliance. Or, simply stated, the
operating condition of the heating system as a whole has just as much influence, if
not more, on the safe operation of the gas-fired appliance.
6.13 THE HEATING/GAS APPLIANCE SYSTEM

When evaluating the operation of a gas-fired appliance, it is critical to consider the
operation of the appliance including all of the external factors that impact the safe
operation of the appliance. The gas-fired appliance and all of the external components
impacting the operation of the appliance make up the heating system or gas appliance
8417: “8417_c006” — 2007/9/11 — 12:16 — page 145 — #17

system. As a minimum, the following external components or factors, along with the
gas-fired appliance make up the system
• The fuel supply— As discussed earlier, it is critical to adjust the appliance

for the type of fuel that is supplied. In addition to the fuel type, the fuel line
pressure and quality of the fuel is also important. If the fuel line pressure is
too low improper mixing of the fuel and combustion air could result causing
incomplete combustion. Conversely, too high of a fuel line pressure could
cause the appliance to malfunction resulting in a fire hazard. The quality
of the fuel, moisture content, and other contaminants could also cause the
appliance to malfunction.
• The combustion and ventilation air supply—For an appliance to operate
as designed, it is critical that a sufficient amount of air for combustion
is supplied to the appliance. A deficient supply of combustion can cause
the venting system to malfunction resulting in a down draft through the
vent system rendering the vent system inoperative. This causes all of the
products of combustion to vent out of the appliance through the draft
diverter relief opening. As discussed under the section “Safety Problems
Associated With The Draft Diverter,” this condition can, and often does,
result in the production of CO. It is just as important to ensure that a
sufficient amount of ventilation air is also supplied to the appliance. An
adequate amount of ventilation air will ensure that the draft effect in the
venting system, and the appliance itself, will be maintained. Ventilation air
is used to help cool the area around the appliance and, more importantly,
provide air for dilution at the draft diverter as needed.
• The venting system—The proper design, installation, and operation of the
venting system is the single most important factor in the safe operation
of a gas-fired appliance. Normally, if the venting system is functioning as
intended, any CO produced by the appliance will be vented to the outside
atmosphere. If the vent system fails, it is probable that the appliance will
produce CO at some point in its operation.
• The service personnel—Even if the gas-fired appliance system is prop-
erly installed and adjusted, it is critical that the operation of the system be
checked annually and repaired as necessary. Normally it requires a min-
imum of two failures for CO to find its way into the living occupied area of
a home. There must be source of CO (failure 1) and there must be a path-
way into the occupied area (failure 2). Proper servicing of the system will
help to ensure there are no failures during the useful life of the appliance.
6.14 FAILURE MODES

As noted above, it normally requires at least two failures of the system to result in CO
reaching the occupied area. For example, the first failure would cause the appliance
to produce CO and the second failure would cause the CO to travel to the occupied
area rather than being vented to the outside atmosphere.
8417: “8417_c006” — 2007/9/11 — 12:16 — page 146 — #18

The following failure modes can result in incomplete combustion causing the
appliance to produce CO:
• Appliance burning too much fuel resulting in a rich mixture

• Incorrect combustion air adjustment on the burner causing the flame to lift
and impinge on a cooler surface
• Restricted internal flue passageway resulting in poor draft through the
appliance
• Insufficient combustion air supply
Normally the appliance will appear to operate safely under the above listed failures
provided that the venting system is operating to vent the products of combustion to
the outside atmosphere.
The following failure modes can cause the venting system to malfunction:
• Improper design, improper size, or insufficient vent height above the roof
line.
• Missing or damaged vent terminal cap at the vent terminal. A damaged or
missing vent cap will allow wind to enter the vent system causing a back
draft or no draft. A damaged or missing vent cap will also allow birds and
other animals to enter the vent in an attempt to nest.
• Insufficient combustion air supply can cause the vent to downdraft, or
reverse flow. The lack of combustion air causes the pressure inside the
room containing the appliance to be reduced. A downdraft through the vent
system occurs when the pressure inside the room containing the appliance
is decreased to a point below the outside atmospheric pressure.
Note that an insufficient supply of air can cause two failures in the system, incom-
plete combustion resulting in the production of CO and the failure of the vent system.
Other failure modes that can cause two failures in the system include
• Excessive leakage in the return air system of a forced air furnace. Excessive
leakage causes air from around the furnace to be pulled into the return air
system of the furnace. Depending on the severity of the leakage this effect
can cause the vent system to fail. In severe cases it can cause the burner
flame to be pulled out of the furnace causing not only a CO hazard but a fire
hazard as well. Once the furnace starts producing CO a direct path into the
heated area is provided by the furnace heat distribution system pumping
the CO into all areas of the heated space.
• Cracked heat exchanger. A cracked heat exchanger can interfere with the
draft effect through the appliance causing the appliance to produce CO.
Depending on the location and size of the crack, products of combus-
tion containing CO can enter the heated air stream rather than exiting the
appliance through the venting system.
Some appliances, such as a gas-fired cooking appliance (range/oven) and unvented

room heaters, are not connected to a vent system. These appliances vent the products
of combustion directly into the area above the appliance. For example, when the top
8417: “8417_c006” — 2007/9/11 — 12:16 — page 147 — #19

burners of a gas range are operated the products of combustion are vented directly into
the area occupied by the user of the appliance. When the oven is used the products of
combustion from the oven are vented from the appliance at a location at the back of
the cook top directly into the area occupied by the user. In these types of appliances
it is critical that the appliance be installed, operated, and maintained so that CO is
not produced. Any failure causing incomplete combustion can result in a hazardous
condition.
6.15 INVESTIGATION OF ROOT CAUSE

Gas appliances are designed and performance certified using ANSI standards
developed by the Z21/83 Committee. For example, gas-fired central furnaces are
certified under ANSI Z21.478 to produce CO levels less than 400 parts per million
(ppm) in an air free (no excess air) sample of the flue gases when tested in an atmo-
sphere having a normal oxygen supply. While ANSI standards allow some amounts of
CO, most gas appliances when installed in accordance with national and local codes
and properly maintained will produce only trace amounts of CO (normally less than
10 ppm.)
When investigating an incident it may or may not be evident that CO poisoning
was the cause of the illness or death. However, even in those cases where it is obvious
from the beginning that CO poisoning was the cause, it may not be as obvious as to
what the root cause or source of the CO poisoning was. Many times police and fire
investigators, and sometimes well trained heating technicians, base a conclusion on
an obvious observable factor while missing the root cause of the accident. Most of
the time, in CO accidents involving gas appliances, there are a number of factors that
contribute to the cause. Many of these factors may have been present for a number
of years before an observable incident has occurred.
A malfunction in any one of the components of the heating system could cause
the appliance to produce CO. When this happens, the appliance may seem to operate
in a safe manner because the CO is being vented to the outside atmosphere through
the venting system. However, if the venting system should fail for some reason or if
a second malfunction should occur (i.e., faulty heat exchanger), dangerous levels of
CO could be circulated to the occupied areas of the structure.
Besides the obvious causes of failure, other less obvious causes are faulty installa-
tions, poorly maintained heating system components, and sometimes poorly designed
features within the appliance itself. Many times a deteriorating condition is noticeable
before a serious failure occurs. Rusted and discolored vent pipes, odors, unexplained
gas pilot outages, water vapors condensing on windows during cold periods, and com-
plaints of poor heating are all indicators of a malfunctioning heating system which,
if left uncorrected, could result in a CO poisoning.
Many times the people who first respond to an accident disturb key evidence
while evaluating the scene and attempting to make the location safe. Typical actions
taken by response personnel include moving boxes or furniture that may be blocking
combustion air openings and removing panels from gas fired appliances to shut off
the gas. In order to minimize any destruction of key evidence it is imperative that a
8417: “8417_c006” — 2007/9/11 — 12:16 — page 148 — #20

qualified expert who specializes in forensic investigations be contacted and allowed

to inspect the site as soon as possible.
When investigating the CO accident the following factors must be evaluated
before a probable root cause can be determined. Extensive photographs and video-
taping of the scene should be completed to document the conditions as found. Every
effort should be made to preserve key evidence for evaluation. The conditions
as found (initial discovery) must not be changed or corrected until the investiga-
tion is completed. In most cases the gas appliance and venting system components
should be removed after the follow-up inspection and investigation and preserved as
evidence.
While it is preferred that a forensic expert be involved from the initial discovery
through the completion of the investigation, this is normally not the case. What usually
occurs is the forensic expert is called in at some point after the initial discovery and
investigation is completed. From a forensic viewpoint it is important to involve an
expert, at least on a consulting basis, as soon as possible after the incident.
6.15.1 INITIAL DISCOVERY AND INVESTIGATION

• First discovery—Who discovered the victim? How did they discover the
victim? What happened to cause them to find the victim?
• Ambient conditions in the space as found—What were the conditions when
the victim was first discovered? Was the room cool, warm or hot? Was there
water moisture on the windows? Was there an odor in the room and if so
what did it smell like?
• First response—Who responded? What did the police and fire departments
find? Did the local building code department respond? Did the local gas
supplier or gas utility respond? Did they take photographs? What condi-
tions did they find? Did they turn off the appliance? If so, how? Were
any tests conducted on the appliances? Did they prepare a report? What
conclusions did they reach?
• Make safe—Did the local authorities call in someone to inspect the gas
appliances to make safe? Was the building evacuated? If so, why? What
actions were taken to make the building safe? Was the suspect gas appliance
repaired or altered after the incident? If so, how and by whom?
6.15.2 FOLLOW-UP INSPECTIONS AND INVESTIGATION

• Weather data—What was the local weather conditions preceding the
incident? Outside temperature patterns, wind speed and direction, snow
cover, rain, sleet, and so forth all may have a bearing on the operation of
the venting system and combustion air and ventilation system. Weather
data should be collected from local sources such as the state university and
local newspapers.
• Building orientation and construction—What direction is the building
facing? Make a sketch of the building and rooms showing location of
the victims, heating appliances, other gas burning appliances, fireplaces,
8417: “8417_c006” — 2007/9/11 — 12:16 — page 149 — #21

and other air consuming devices such as exhaust fans and exhaust hoods.
Record all dimensions, including ceiling height and window and door loca-
tions. Note the age of the construction and construction type (frame, brick,
etc.) and the tightness of the building (weather stripped, caulked, etc.).
Mark locations of gas vents and sources of combustion air including size
and type of air grills and air louvers.
• Overall condition and appearance of the building and gas appliances—
Does it appear that the building and appliances are in good condition? Is
there an existing hazardous condition? Is there a hazard tag on the gas appli-
ance? If so, when was it placed and by who? What action is recommended
to correct the hazard?
• Maintenance history—Have there been any previous problems reported?
What types of repairs have been completed in the past? When were the
problems reported and when were the repairs completed? Who discovered
the problems and who made the repairs?
• Evaluate appliance installation—Is the appliance installed in compliance
with the manufacturer’s installation instructions? Does the installation
comply with national standards (NFPA 54) and local building codes?
Record locations of fuel lines, gas pressure regulators including gas pres-
sure regulator vent lines and propane tanks if fired on propane. Sketch
the existing appliance installation and venting system. Identify sources of
combustion and ventilation air and any other information that may be per-
tinent to the operation of the appliance. Note the condition of the appliance
and venting system.
• Collect appliance data—Record all nameplate data for all appliances.
Include serial numbers and model numbers on all automatic gas control
valves. Also note any special devices or controls that may be installed
on the appliances. Examples of special devices may include automatic
vent dampers, heat reclaimers, and other retrofitted energy conservation
devices.
• Conduct combustion tests—Test the gas appliance(s) to determine what
their normal operating characteristics are. Test all gas appliances in the
structure. Repeat all tests under different conditions to duplicate condi-
tions that existed at the time of the incident. At a minimum, test for CO at
the flue outlet (inlet to draft diverter) and measure stack temperature. Also
it is necessary to measure oxygen or carbon dioxide content to be able to
calculate CO on an air free basis. Note burner ignition and operating char-
acteristics (yellow tip, blue flame, primary air adjustments, etc.). Ambient
CO levels in the appliance area, as well as in the area where the victim was
found, should be recorded and monitored.
• Measure the fuel input rate—Is the appliance properly adjusted for the
elevation? The nameplate rating for certified appliances must be derated
at the rate of 4% for each 1000 ft. in elevation. The actual heat (fuel)
input should be determined and recorded. For propane-fired appliances it
is necessary to measure the burner orifice sizes and compute the actual
input. For natural gas fired appliances it is necessary to measure the actual
8417: “8417_c006” — 2007/9/11 — 12:16 — page 150 — #22

input by “clocking” the gas meter. It is necessary to contact the local gas
utility for actual heating (BTU) and specific gravity values.
• Miscellaneous factors—Are there any other factors or conditions that might
have a bearing on the cause of this incidence? Such factors may include
faulty maintenance by a third party or failure of a third party to provide a
warning when they had knowledge of a hazardous condition.
6.15.3 DETERMINATION OF ROOT CAUSE

Although the cause of the incident may seem obvious at first, guard against premature
conclusions. Only after all of the items and information discussed above have been
collected and analyzed is it possible to determine the root cause or causes of the CO
poisoning. As stated earlier, with modern gas- fired appliances there are usually two
or more failures that finally result in a tragic accident.
The key is to collect as much data as soon after the incident as possible and
evaluate that data to reach a determination of probable cause.
6.15.4 ESTIMATION OF AMBIENT CARBON MONOXIDE LEVELS IN

OCCUPIED AREAS
As a last step in evaluating a CO incident in a structure, it may be feasible to estimate
the ambient CO levels inside the structure caused by the failures identified in the
investigation. The factors that must be identified before an estimate of the ambient
CO level can be calculated are
• Type of fuel, heating value, and total volume of combustion products

produced in cubic feet per 1000 BTU, dry basis (water vapor removed.)
• Actual fuel input rate, BTU/h.
• Combustion data indicating amount of CO produced by the appliance on
an air free basis (excess air removed.)
• Type of construction and estimated air exchange rate in the structure,
expressed as air changes per hour.
• Estimate of the operating cycle based on appliance usage and weather data.
• Failure mode indicating how much of the produced CO actually enters the
structure.
After all of the above listed factors have been identified, an estimation of the
ambient CO level within the structure can be calculated. As an illustrated example
consider the following case study.
6.16 CASE STUDY

During the early morning hours on January 12, two adults, male and female, were
transported from their home to a local hospital emergency room suffering from
severe headaches, nausea, and confusion. Blood tests indicated carboxyhemoglobin
8417: “8417_c006” — 2007/9/11 — 12:16 — page 151 — #23

(COHb) levels in the adult male of 32% and in the female of 35%. Neither individual
smoked.
As a result of this incident the couple’s single level home was inspected and an
investigation conducted in an effort to locate the cause of the CO exposure.
6.16.1 INVESTIGATION
• Incident location
1. Northern New Mexico
2. 5000 ft. elevation
• Home description
1. Determined from local records and inspection
a. Frame construction completed in 1972
b. Single level built over a crawl space
c. 1680 sq. ft. in area
d. 9 ft. ceilings
e. 15,120 cu. ft. internal volume
• Type appliance
1. Determined by inspection
a. Forced air furnace, upflow
b. Rated fuel input rate
i. 125,000 BTU/h nameplate
ii. 100,000 BTU/h at local conditions
c. Atmospheric venting system with draft diverter
d. Furnace installed in confined space utility closet.
e. No combustion air supplied to utility room (operation
depends on normal infiltration of air from the outside for
combustion air.)
• Fuel type—Natural gas supplied from the Texas Panhandle area.
1. Published data
a. Heating value
i. 1013 BTU/cu. ft. at sea level
ii. 858 BTU/cu.ft. at local conditions
b. Total volume of combustion products—8.545 cu. ft. dry
basis
• Actual fuel input rate
1. Determined by measurement
a. 122.55 cu. ft. per h natural gas
b. 105,149 BTU/h (5.1% over fired)
• Combustion data at steady state conditions
1. Determined by measurement
a. % O2 = 12.6
b. CO, ppm with diluted with excess air = 1650
c. CO, ppm excess air free = 4125
8417: “8417_c006” — 2007/9/11 — 12:16 — page 152 — #24

• Air changes per hour and type construction

1. Determined by measurement and calculation utilizing pub-
lished American Society of Heating, Refrigerating And
Air-Conditioning Engineers (ASHRAE) tables and formula.9
a. Estimated air exchange rate
i. 0.75 air changes per hour at design conditions
ii. 0.56 air changes at average temperature difference
(January)
b. Average to tight
• Weather conditions for time period
1. Local data sources
a. January time period
i. Average outdoor temperature 29.9◦ F
b. Indoor temperature 72◦ F
c. Average temperature difference 42.1◦ F
d. Design outdoor temperature −1◦ F
• Furnace operating cycle
1. Calculation based on furnace size and average weather data for
time period
a. Duty cycle 57.7%
b. Burner cycles per hour 4
6.16.2 FAILURE MODE

• Determined by investigation
1. Production of CO by furnace (source)
a. Over fired (contributing cause of CO)
b. Insufficient combustion air supply
c. 4125 ppm CO air free basis
2. Total vent system failure caused by insufficient combustion air
resulting in vent draft reversal
3. 100% spillage of combustion products into area around furnace
4. Excessive leakage of air into return air system of furnace at fur-
nace base and junction of furnace to return air plenum connection
(pathway into heated area)
5. 100% of combustion products circulated into occupied area
6.16.3 CASE STUDY CONCLUSIONS

Figures 6.5 and 6.6 were generated by applying the collected data to standard formulas
to calculate estimated ambient indoor CO levels.10−13 These plots indicate that after
about 7 h of total elapsed time from start, the ambient CO level inside the structure
has increased from 0 to around 250 ppm. The plots also indicate that the ambient
8417: “8417_c006” — 2007/9/11 — 12:16 — page 153 — #25

300
250
Ambient Co level (ppm)
200
150
100
50
0
Time 0–7 h
FIGURE 6.5 Estimated ambient carbon monoxide levels from starts to 7 h of elapsed time.
265.0
260.0
Ambient CO level (ppm)
255.0
250.0
245.0
240.0
235.0
230.0
Time 7–14 h
FIGURE 6.6 Estimated ambient carbon monoxide levels from 7 to 14 h of elapsed time.
level of CO within the structure stabilized after about 10 h from start and then varied
between 245 and 260 ppm as the furnace cycled on and off.
The Environmental Protection Agency (EPA)14 has published maximum allow-
able exposure limits for CO of 9 ppm for an 8-h exposure or 35 ppm for a 1-h
exposure. The EPA has established that this exposure limit not be exceeded more
than once per year. Exposure above these levels is considered harmful to public
health.
The estimated ambient CO levels demonstrated in this case study exceed the EPA
recommended maximum allowable CO exposure limits for humans and indicate that
the occupants were exposed to an extremely hazardous risk of CO poisoning.
8417: “8417_c006” — 2007/9/11 — 12:16 — page 154 — #26

References
1. American Gas Association, Gaseous Fuels, Properties, Behavior, and Utilization,
1954.
2. American Gas Association, Fundamentals of Combustion Revised, Catalogue No.
XH9601.
3. Canadian Standards Association, CSA in the USA, http://www.csa-international.org/
testing_certification_us/
4. National Fire Protection Association, National Fuel Gas Code Handbook, 1999,
Lemoff, T.C., ed., Inc., NFPA No.: F9-54HB99.
5. Gas Engineers Handbook, Fuel Gas Engineering Practices, 1977, Industrial Press,
Inc., NY, Segeler, C.G., editor First ed.—5th Printing.
6. Engineered Controls International, Inc. LP-Gas Serviceman’s Manual, L-545, 1962.
7. Consumer Product Safety Commission, Non-fire carbon monoxide deaths associated
with the use of consumer products 2002 annual estimates, July 12, 2005.
8. American National Standards Institute, Inc., ANSI Z21.47, American National
Standard for Gas-Fired Central Furnaces. 23rd ed., 1987.
9. American Society of Heating, Refrigerating and Air-Conditioning Engineers, 2005
ASHRAE Fundamentals Handbook.
10. Calspan, Investigation of Safety Standards for Flame-Fired Furnaces, Hot Water Heat-
ers, Clothes Dryers and Ranges, Report No. YG-5569-D-3, July, 1975, Contract No.
CPSC-C-74-131.
11. Calspan, Safety Devices For Gas-Fired Appliances, Report No. 6608-D-1, May, 1980,
Contract No. CPSC-C-79-1007.
12. Gas Research Institute, Environmental and Safety Research, September 1996, Topical
Report, Critique of ANSI Z21.1 Standard for CO emissions from gas-fired ovens and
ranges, Prepared by J.J. Reuther, Battelle.
13. Thomas H. Greiner, Ph.D., P.E., Associate Professor, Iowa State University, Com-
ments concerning carbon monoxide emissions from gas-fired ovens and ranges with
special reference to Battelle’s 1996 Critique of ANSI Z21.1 (1)
14. Environmental Protection Agency, National Ambient Air Quality Standards
(NAAQS) for pollutants considered harmful to public health and the environment,
http://epa.gov/air/criteria.html.
8417: “8417_c006” — 2007/9/11 — 12:16 — page 155 — #27

8417: “8417_c006” — 2007/9/11 — 12:16 — page 156 — #28
7 Carbon Monoxide
Dangers in the Marine
Environment
Jane McCammon
CONTENTS
7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158

7.2 Identifying Marine CO Poisonings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
7.2.1 Identification of Marine CO Poisonings at Lake Powell. . . . . . . . . . 160
7.2.2 Identification of Marine CO Poisonings Nationwide . . . . . . . . . . . . . 160
7.2.3 Additional Carbon Monoxide Poisoning Case Identification . . . . . 161
7.3 Descriptive Characteristics of Known Marine Carbon Monoxide
Poisonings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
7.3.1 Demographics, Distribution, and Outcome of Poisonings . . . . . . . . 161
7.3.2 Location of the Victim and Source of Carbon Monoxide . . . . . . . . 162
7.4 Medical Characteristics of the Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
7.4.1 Carboxyhemoglobin Measurement Limitations . . . . . . . . . . . . . . . . . . . 164
7.4.2 Boat-Related Carbon Monoxide Poisonings Resulting
in Death . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
7.4.3 Carbon Monoxide Exposure as a Drowning Risk Factor . . . . . . . . . 165
7.4.4 Non-Fatal Boat-Related Carbon Monoxide Poisonings . . . . . . . . . . 166
7.5 Boats and CO Sources Related to Marine Poisonings, with Case Study
Presentations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
7.5.1 Ski Boats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
7.5.2 Cabin Cruisers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
7.5.3 Houseboats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
7.6 Airborne Carbon Monoxide Concentrations Measured on and Near
Boats. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
7.6.1 Methods for Measuring, Analyzing, and Evaluating CO in Air . . 175
7.6.2 Carbon Monoxide Concentrations Associated with Boat-Related
CO Poisonings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
7.6.2.1 Ski Boats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
7.6.2.2 Cabin Cruisers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
7.6.2.3 Houseboats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
7.6.2.4 Carbon Monoxide Exposure in Areas of Congested
Boat Traffic (Lake Havasu Study). . . . . . . . . . . . . . . . . . . . . . . 179
157
8417: “8417_c007” — 2007/9/11 — 12:16 — page 157 — #1

7.7 Prevention Efforts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180

7.7.1 U.S. Coast Guard Regulation of Boat Manufacturing/Recall
Authority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
7.7.2 Environmental Protection Agency Regulation of Marine
Engine Emissions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
7.7.3 State Legislative Action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
7.7.4 Effectiveness of Marine Carbon Monoxide Detector/Alarms in
Prevention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
7.7.4.1 Case-Based Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
7.7.4.2 Marine Carbon Monoxide Detector Evaluation . . . . . . . . 183
7.7.5 Engineering Control Research and Development . . . . . . . . . . . . . . . . . 184
7.7.6 Innovations in EMS Medical Management . . . . . . . . . . . . . . . . . . . . . . . 187
7.8 Summary and Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
7.1 INTRODUCTION
Although carbon monoxide (CO) poisoning has been recognized since the earliest
medical writings, CO exposure continues to be a major cause of death and illness.
Many fatal and nonfatal exposures go undetected and unreported. The development of
effective prevention programs is severely hampered by the absence of a centralized,
coordinated surveillance system.
CO poisonings in the marine environment began receiving attention as early as
1984, with poisonings occurring inside living quarters (also referred to as the boat
cabin) being recognized early on.1 In 1990 and 1991, the United States Coast Guard
(USCG), responsible for regulating recreational boat manufacture, investigated issues
related to intrusion of engine exhaust into boat cabins.2,3 CO poisoning associated
with occupancy of cabins of recreational boats was first described in the scientific
literature in 1995.4 Subsequently, the American Boat and Yacht Council (ABYC—a
nonprofit membership organization that develops safety standards for boat building
and repair) developed standards for CO detection systems for boats manufactured
after July 31, 1998.5
However, CO poisonings occurring outside the boat cabin were another matter.
Reports of outdoor CO poisonings in any setting are scant.6−9 Published reports of
boat-related CO poisoning outside of a boat cabin, in which the boat occupant was
swimming behind a motorboat10 and behind a houseboat11 appeared in 1998.
The absence of data on outdoor CO poisonings confused National Park Service
(NPS) Emergency Medical Services (EMS) providers on Lake Powell (within the
Glen Canyon National Recreation Area—GCNRA) who had noticed a number of
poisoning incidents involving boat occupants losing consciousness outdoors, begin-
ning in the early 1990s. Through the early years of these poisonings, park officials
sought evidence of other outdoor CO poisonings, but failed to find documentation of
this problem elsewhere in the United States. When the first verified fatal marine CO
poisoning occurred at Lake Powell in 1994, NPS EMS personnel had treated a total
of 29 nonfatal CO poisonings associated with boats.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 158 — #2

Carbon Monoxide Dangers in the Marine Environment 159
On an August evening in 1994, NPS officials at the GCNRA dispatch center

received an emergency call from boaters requesting assistance regarding the possible
drowning of a 12-year old boy. Three boys had been swimming near the rear of a
houseboat while family members were inside the air-conditioned living quarters of
the boat. One of the boys became ill, and parents took him inside the boat to rest,
attributing his illness to the hot weather. The other two boys continued swimming.
The victim (the second boy) was on the swim platform at the time and complained
to his companion about his legs feeling funny. The third boy climbed onboard and
went inside the boat, complaining to his mother that he, too, did not feel well. The
victim remained on the swim platform at the rear of the boat while others attended
to his companion. Nearly an hour later, the group discovered that the victim was
missing and began to search for him, concurrently summoning NPS assistance. NPS
divers recovered the boy’s body from 21 ft. of water more than 2 h later. During
the course of their investigation of this drowning, NPS discovered that the onboard
gasoline-powered generator had been operating when the boys were swimming, and
communicated this information to the medical examiner. As a result, a carboxyhemo-
globin (COHb) analysis was requested as part of the autopsy. The analysis, repeated
due to the high initial concentration, revealed a COHb of 61% documenting that
the boy’s drowning was secondary to CO poisoning, the source of which was the
generator.12
Although NPS requested assistance from the USCG after this boy’s death, and
again in 1998 when two CO-related drownings occurred within 12 days of each other
behind houseboats of a similar design, active response was lacking. This changed
in August 2000, when two brothers, Dillon and Logan Dixey (aged 8 and 11 years),
swimming at the rear of their family houseboat on Lake Powell, died as a result
of CO poisoning. Their deaths were quick, occurring within minutes of expos-
ure to generator exhaust under the swim deck, and were strikingly similar to five
previously recognized fatalities on this lake behind houseboats of the same design.
The brothers’ deaths, combined with the growing body of evidence at Lake Powell,
triggered an interagency, multidisciplinary investigation, the goal of which was to
identify effective prevention strategies at this lake.13 The interagency investiga-
tion rapidly grew to cover similar poisonings nationwide, with a number of local,
state, and federal governmental agencies, individual boat and equipment manufactur-
ers, boat rental companies, trade organizations, nonprofit organizations representing
consumers, and individual boaters joining the effort to identify poisonings and
interventions.
7.2 IDENTIFYING MARINE CO POISONINGS

Data discussed in this chapter were derived from a variety of sources, and vary in level
of documentation and detail available from the source. The most important thing to
remember is that these are not all the poisonings that have occurred, but rather are the
poisonings we know about from the sources discussed below. Case data are influenced
by reporting bias (such cases were not recognized early on, recognition of cases has
improved only slightly, awareness of reporting requirements is scant, and many cases
8417: “8417_c007” — 2007/9/11 — 12:16 — page 159 — #3

were reported because of increased news coverage); and misdiagnosis of disease and
death etiology (many deaths were and are classified as drownings that were actually
CO poisonings first).
7.2.1 IDENTIFICATION OF MARINE CO POISONINGS AT LAKE

POWELL
Todate, nearly 1/3 of all recognized cases occurred at Lake Powell, and were iden-
tified through extensive case-finding research conducted jointly by the Centers for
Disease Control and Prevention’s (CDC), National Institute for Occupational Safety
and Health (NIOSH), the Medical Advisor for Prehospital Care at GCNRA, NPS
GCNRA staff, and the US Department of the Interior.
The isolation of the lake, and related centralization of NPS emergency response
and patient transport, resulted in early identification and thorough documentation of
fatal and nonfatal marine poisonings by Glen Canyon NPS medical and law enforce-
ment personnel. Computerized and hard copy records were abstracted to identify and
describe Lake Powell CO poisonings.14 These included: NPS GCNRA law enforce-
ment dispatch logs; NPS EMS response sheets; Page Hospital Emergency Department
treatment records; hospital discharge data; and medical examiner/coroner autopsy
reports.
7.2.2 IDENTIFICATION OF MARINE CO POISONINGS NATIONWIDE

In the United States, data related to boating accidents are collected through the USCG
Boating Accident Report Database (BARD), which provides vital information for
USCG regulation of the manufacture and recall of recreational boats and boating-
related equipment.
Data used to compile the BARD statistics come from two sources: (1) Boating
Accident Report data forwarded electronically to the USCG by recognized reporting
authorities, typically the state boating law administrator in each state; and (2) reports
of USCG investigations of fatal boating accidents that occurred on waters under
Federal jurisdiction.15
To obtain the most accurate data, BARD relies as much as possible on recreational
boating accident investigations conducted by local law enforcement personnel and
submitted to the USCG electronically by the recognized reporting authority. In the
absence of investigational data, information is collected from accident reports filed by
recreational boat operators, who are required to report boating accidents that involve
the vessel or its equipment.16
Included in the USCG database are recreational boat-related or equipment-related
accidents in which:
1. A person dies; or
2. A person is injured and requires medical treatment beyond first aid, that is,
treatment at a medical facility or by a medical professional other than at
the accident scene; or
8417: “8417_c007” — 2007/9/11 — 12:16 — page 160 — #4

3. Damage to vessels and other property totals $2000 or more or there is a

complete loss of any vessel; or
4. Aperson disappears from the vessel under circumstances that indicate death
or injury.
CO poisonings were specifically added to the listed USCG reporting guidelines

in 2001, in response to confusion expressed among the reporting authorities as to
whether recreational marine CO poisonings met the reporting requirements. Thus, a
fraction of marine CO poisonings were reported to the Coast Guard prior to 2001.
7.2.3 ADDITIONAL CARBON MONOXIDE POISONING CASE

IDENTIFICATION
Intense national news coverage of the poisonings at Lake Powell, as well as technical
presentations and other information dissemination, resulted in the identification of
additional cases by researchers closely identified with this issue. The pattern that
emerged from collection of related records from these independently identified cases
indicated a broader problem, and pointed to emerging problems nationwide related
to the design of houseboats and swim platforms on other types of motorboats.
These randomly identified cases were added to those identified in the Lake Powell
investigation and those contained in BARD to comprise a document now known as
the National Case Listing.17 Data discussed below are derived from the January 2006
listing and analysis of the supporting database.
7.3 DESCRIPTIVE CHARACTERISTICS OF KNOWN

MARINE CARBON MONOXIDE POISONINGS
7.3.1 DEMOGRAPHICS, DISTRIBUTION, AND OUTCOME OF
POISONINGS
The National Case Listing contains detailed information about 607 US boat-related
CO poisonings requiring medical treatment or emergency response occurring on water
bodies in or adjacent to 32 states. Further information is described in publications by
the CDC.13,18,19
One-hundred-twenty-two (20%) of the 607 people died as a result of the poisoning.
Ninety-eight percent of the poisonings occurred between 1990 and 2005. Age was
known for 432 (71%) of the cases: 42% of the 432 cases were children (aged 18 or
younger). Gender was known for 393 (64%) of the cases, which was fairly evenly
divided between males (204 or 52%) and females (189 or 48%). Figures 7.1 through
7.3 show the distribution of cases by year, month, and state.
The decline in cases in 2005, shown in Figure 7.1, is primarily an effect of the
time lapse between the occurrence of the incident and reporting or verification of the
case. (As of this writing, we have preliminary information regarding 13 additional
cases for 2005.) Monthly distribution of marine CO poisonings is opposite that of
8417: “8417_c007” — 2007/9/11 — 12:16 — page 161 — #5

Number of boat-related CO poisonings by year

Total poisonings identified =607 Date of chart: 1/2006
80
70
60
50
40
30
20
10
0
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
unk
Non–fatal Fatal
FIGURE 7.1 Distribution of boat-related carbon monoxide poisonings by year.
Distribution of boat-related CO poisonings by month
120
100
Number of people poisoned
80
60
40
20
0
EC
T
K
P
R
AR
V
B
N
AY
N
N
SE
O
JU
FE
AP
JU
AU
JA
D
O
M
U
M
FIGURE 7.2 Distribution of boat-related carbon monoxide poisonings by month.
other poisonings (higher in the summer instead of higher in the winter) and reflects
the pattern of boat use.
7.3.2 LOCATION OF THE VICTIM AND SOURCE OF CARBON

MONOXIDE
Location (on the boat) was known for 546 of the cases. Two of every three victims
(353/546) were documented to be inside the boat’s living quarters when poisoned.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 162 — #6

1 13 1
15
7 7 3
1 4 1 6
3
30 2 15
19* 1
5
1 5 44
45 30
10
29
Lake 2
powell 22* 2
176 4 12
Location
(29%) 20
*Excludes 1 unspecified
lake 39
17
powell
cases
FIGURE 7.3 Distribution of boat-related carbon monoxide poisonings by state.
These poisonings in the boat cabin occurred in 88 incidents primarily involving

multiple victims (as many as 17 victims at a time). In contrast, 190 people were
poisoned outside the boat cabin, in 149 incidents which were more likely to involve
one to three victims. Thus, the higher number of incidents of poisoning occurred
outside the boat.
The specific location of the victim was documented for 100 of the CO-related
deaths, with 56 of the victims exposed to CO outside the boat’s living quarters, 44
exposed inside the living quarters or within a full boat canopy. Location on the boat
was characterized for 446 of the 485 survivors of marine CO poisonings: 309 were
exposed to CO inside the boat cabin in 69 incidents; 134 were exposed outside a
boat cabin in 103 incidents; and 3 people (1 incident) were characterized as hav-
ing been at various locations inside and outside. Again, although more poisoning
survivors were inside boats when poisoned, more incidents occurred outside boat
cabins.
The specific source of CO exposure was documented for 487 of the 607 pois-
onings. Exhaust from onboard gasoline-powered marine generators used to produce
electricity for onboard appliances (air-conditioners, televisions, refrigerators, etc.)
was the most common source, linked to 302 of the 480 poisonings. Gasoline-powered
propulsion engine exhaust was linked to 168 poisonings. Twelve people were exposed
to exhaust from both types of engines (generators and propulsion engines) when they
were poisoned. Source was undifferentiated (e.g., the record said “engine exhaust” or
“exhaust fumes” but did not clearly define which engine) or unspecified for 120 cases.
At this point, there are no poisonings known to have been associated with marine
diesel-powered engines of either type.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 163 — #7

7.4 MEDICAL CHARACTERISTICS OF THE CASES

Medical characteristics of marine CO poisonings to be discussed here include COHb
as an indicator of exposure, death and loss of consciousness (LOC) as indicators of
exposure severity, and Glasgow Coma Scale (GCS—a scale that assesses the degree
of brain function) as a measure of victim status when EMS personnel arrived on the
scene.
7.4.1 CARBOXYHEMOGLOBIN MEASUREMENT LIMITATIONS

Many boat-related CO poisonings occur while the victim is swimming near the boat.
In such incidents, the victim loses consciousness, sinks into the water or lays face
down in the water, ultimately drowning as a result of the poisoning. As such, COHb
concentrations are often lower than those typically associated with CO-related fatalit-
ies. As is true with all CO poisonings, there is no consistency regarding timing of the
COHb measurement. Reported COHb concentrations must be viewed as indicators
of exposure, because of the many factors impacting that measurement (i.e., oxygen
therapy duration, number of half-lives that have passed since the CO exposure ended,
activity level of the victim between exposure cessation and COHb measurement, etc.).
COHb analysis related to fatal boat-related poisoning may be complicated by
the fact that victims who are poisoned and subsequently drowned sometimes are
submerged for extended periods before their bodies are found. In such cases, body
decomposition, and related degradation of blood cells, may impact the ability to
accurately measure COHb.
7.4.2 BOAT-RELATED CARBON MONOXIDE POISONINGS

RESULTING IN DEATH
COHb concentrations were available for 50 of the 122 marine CO-related fatalities.
Six of those 50 people received oxygen therapy prior to blood analysis.
Reported COHb concentrations for people that were poisoned outside the cabin
ranged from 1.9% (after 440 min of oxygen treatment by intubation) to 100%. The
related estimated CO exposure duration for these victims was surprisingly short,
ranging from 30 s to 30 min.
Most (47) of the 56 people who died as a result of poisoning outside the boat’s
living quarters were in the water when they were exposed to CO, likely losing con-
sciousness and sinking. Reported duration of submersion for those that drowned as
a result of the poisoning ranged from 5 min to 13 days. Six people that died outside
the living quarters were on the boat, either sitting in the rear seat or rear section of
the boat, or leaning over the transom. Specific location of the victim was unknown
for three people.
COHb concentrations for deaths that occurred inside the cabin or a full canopy
ranged from 27% to 77%. With one exception, the people that died in the cabin of
boats had gone there to sleep or watch videos, and were found dead the following
morning when someone had noticed them missing, or when someone noticed the
boat adrift. The exception to this is an incident in which four children were found
8417: “8417_c007” — 2007/9/11 — 12:16 — page 164 — #8

Zippered
canopy
Swim opening
platform
FIGURE 7.4 Cabin cruiser with full canopy deployed.
unconscious (one of which was dead) in the rear area of a cabin cruiser boat that was
covered with a full canopy (see Figure 7.4). The children were using a shower device
that drew heated water from the operating propulsion engine’s cooling system. They
unzipped a panel in the canopy and stood on the swim platform to use the shower hose.
About 45 min after they had gone to the boat to shower, two of the boys were found
unconscious on the bed in the boat cabin, and two (one of which died) were found
unconscious in the area covered by the canopy. CO from the operating propulsion
engine had apparently entered the canopied area and cabin. The COHb of the boy
who died was 46.6%.
7.4.3 CARBON MONOXIDE EXPOSURE AS A DROWNING RISK

FACTOR
To assess the proportion of drownings and boat-related drownings for which CO
poisoning was a contributing or primary cause of death, drownings that occurred on
Lake Powell from 1994 to 2004 were identified.14 (The starting point was based on
the year of the first confirmed CO-related drowning at Lake Powell.)
Seventy-two people died of unintentional drowning within GCNRA during that
time period. Seventeen percent (12 of the 72 people) of all drownings were CO-related.
Twenty-five of the 72 people drowned in BARD-reportable boat-related accidents.
This meant that nearly half (12/25) of all the boat-related drownings at Lake Powell
were CO-related. These surprising numbers indicate that CO exposure may be a
factor in many more drownings on other open water bodies than are currently
recognized. This under-recognition is likely to continue unless COHb analyses are
routinely included in autopsy protocols and hospital emergency room patient care
procedures.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 165 — #9

To assess under-reporting of these cases, researchers examined the Coast Guard

surveillance system (BARD) data and found that one of three Lake Powell boat-
related drownings were absent from BARD, and that CO-related and nonCO-related
drownings were missing in similar proportions.
7.4.4 NON-FATAL BOAT-RELATED CARBON MONOXIDE

POISONINGS
As mentioned earlier, 485 people are known to have survived marine CO poisoning
that required at least emergency medical treatment. It is important to remember that
these patients did not always have easy access to transport (because they are out on
a lake, river, or ocean), and that transport to a local hospital where blood is drawn
(or equipment for breath analysis is available) often takes a while. These delays impact
the COHb result significantly, especially if oxygen is being delivered to the patient
during a lengthy transport.
COHb concentrations were available for 128 of these poisoning victims, of which
nearly half (i.e., 63) were treated with oxygen prior to COHb analysis (ranging from
10 to 645 min in duration). Measured COHb ranged from 1.9% to 47.8%. Information
about LOC was available for 250 of the 485 survivors of poisoning, and just over half
(i.e., 130) of the 250 were documented to have lost consciousness.
GCS ratings were available for 112 of the survivors, with assessment ratings
ranging from 3 (unconscious and unresponsive to deep pain) to 15 (indicating normal
mental status at the time of EMS response and assessment). It was striking to note
that 85 of the 112 GCS-rated patients were assessed as having normal mental status
(GCS of 15), despite the fact that nearly 40% (33) of those patients reported profound
LOC during their exposure to CO. Half of these patients (17/33) were evaluated by
EMS personnel within 10–35 min of experiencing LOC, indicating that CO poisoning
etiology can be easily missed if emergency response personnel do not know that CO
exposures outside of an enclosure can result in rapid, serious poisoning.
7.5 BOATS AND CO SOURCES RELATED TO MARINE

POISONINGS, WITH CASE STUDY
PRESENTATIONS
Types of boats involved in marine poisonings are categorized as follows: ski boats;
cabin cruisers; houseboats; other pleasurecraft (fishing boat, personal water craft,
etc.), or unspecified. The distribution of poisonings by boat is shown in Table 7.1,
and are discussed below.
7.5.1 SKI BOATS

Ski boats are typically a high-performance craft designed for speed and agility. They
vary in size, but are typically about 21 ft. long and about 8 ft. wide. Ski boats asso-
ciated with the outside poisonings (shown in Figures 7.4–7.6) are typically equipped
with a high horsepower (310–400 hp) gasoline-fueled inboard direct-drive propulsion
8417: “8417_c007” — 2007/9/11 — 12:16 — page 166 — #10

TABLE 7.1
Distribution of Marine Carbon Monoxide Poisonings by Boat Type
Fatal Nonfatal Total
Boat Inside Outside Unknown Inside Outside Unknown by Boat
Type Cabin Cabin Location Cabin Cabin Location Type
Ski 0 23 0 0 45 3 71
Cabin cruiser 31 3 4 102 14 4 158
Houseboat 2 23 0 177 54 1 257
Other or unknown 11 7 18 30 21 34 121
Total by outcome 44 56 22 309 134 42 607
Swim platform
Exhaust
Propeller
FIGURE 7.5 Ski boat transom, platform, exhaust, and propeller configuration.
engine, the propeller of which is under the belly of the boat and thus removed from
close proximity to the victim and thought by many to be out of harm’s way. Most of
these boats have a water level rear platform (see Figure 7.5) used to facilitate easy
access to the water and donning of ski and wakeboard gear. Engine exhaust is dual
piped, exiting the boat hull just below the rear platform.
All fatal and nonfatal poisonings associated with ski boats occurred outside of
any enclosure, and were caused by exposure to propulsion engine exhaust (this class
of boat is not typically equipped with a generator). The details of these incidents are
strikingly similar: 45 (63%) of 71 people poisoned on this type of boat were sitting on
or holding onto the swim platform; 18 of the 45 people died and 27 of them survived
poisoning (19 of whom experienced LOC). The ages of victims occupying the swim
platform ranged from 2 to 23 years. Five people were poisoned, four of whom lost
consciousness, while occupying the padded sunning deck at the rear of ski boats.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 167 — #11

(a) “Teak surfing” or “platform dragging”
Surfer
Displacement wave
Engine exhaust outlet (x2)

Vessel is operated at a speed that will create the largest
possible displacement wave, usually 10–12 miles per hour
(b)
Boat transom (rear wall)
FIGURE 7.6 Teak surfing. (a) Side-view showing position of surfer, (b) View toward transom
of boat showing same.
Much media, legislative, and educational attention has centered on a thrill seeking
activity that has been labeled as “teak surfing” or “dragging” (shown in Figures 7.6
and 7.7), in which the person holds onto the rear swim platform (which is often made
of teakwood, thus the term “teak surfing”) until the boat reaches a speed (typically
10–12 mph) and orientation that allows them to briefly release their grasp and be pulled
forward by the displacement wave directly behind the boat. Five of the poisonings
associated with platform occupancy (three fatal and two nonfatal) occurred during
teak surfing. The predominance of incidents associated with platform occupancy
actually occurred when the boat was moving at idle speed (5 mph) through the water,
while the victim was being pulled from here to there, or was sitting on the platform
dangling his/her feet in the water. In addition, one person who died on the platform
and eight survivors were sitting there while the ski-boat was not moving at all, but
the propulsion engine was idling. One of the later incidents involved the poisoning
of a 4-year-old girl exposed to propulsion engine exhaust while she sat on the swim
platform, playing with a shower device that is fed hot water by the operating engine.
(This is the same device previously mentioned in Section 7.4.2)
Case 1. In June 2001, an 18-year-old passenger of a ski-boat drowned in Lake
Powell as a result of CO poisoning. Three of the ten boat passengers were being
8417: “8417_c007” — 2007/9/11 — 12:16 — page 168 — #12

Exhaust
Propeller
FIGURE 7.7 Position of teak surfer relative to engine exhaust terminus and propeller.
pulled behind the boat, teak surfing. Approximately 2 min after they began, one of the
teenagers was unable to maintain her hold on the platform. She was reported as having
“jerky arm movements” and difficulty in communicating. She was pulled into the boat
by the passengers. Not recognizing the cause of her symptoms, another teen took her
position on the platform, and they began teak surfing again. Approximately 1–2 min
later, one of the three teens began to experience a severe headache and weakness.
This teen pulled herself up onto the swim platform while the boat continued to move
forward. The third surfer, still positioned for teak surfing, suddenly lost consciousness
and released his hold on the platform. He sank beneath the surface. His body was
retrieved from the water 3 days later. His COHb on autopsy was 57%.
Case 2. In July 2000, a 15-year-old girl survived CO poisoning while she was lying
on the rear padded sunning deck of a ski-boat on Lake Minnetonka in Minnesota. Boat
occupants were waiting for a fireworks display, with the propulsion engine operating
to power the onboard music system. Other occupants thought the girl was sleeping
until they tried unsuccessfully to awaken her. She had stopped breathing. Her COHb
measured upon transport to the local hospital was 30%.
Case 3. In July 2005, a 21-year-old woman drowned as a result of CO poisoning.
The woman was boating on the Gulf of Mexico with her husband and a group of
friends in a 21 ft. ski boat. She and her husband were floating in the water, holding
on to the swim platform of the stationary boat. The boat operator started the engine
and began moving at about 5 miles per h when the woman slipped from the platform
and sank. Her husband was unable to locate her. Several hours later, her body rose to
the water’s surface and was discovered. Her COHb upon autopsy was 67%.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 169 — #13

FIGURE 7.8 Cabin cruiser.
7.5.2 CABIN CRUISERS

A cabin cruiser is generally a large motorboat that has a cabin, and plumbing and
other conveniences necessary for living onboard. The hull of a cabin cruiser (see
Figure 7.8) is shaped like a conventional motorboat. Many modern cabin cruisers are
equipped with air-conditioning and other electrical appliances that necessitate the use
of an onboard generator for power.
Most (31) of the 44 deaths inside a boat of any kind occurred in a cabin cruiser.
Source of CO was known for 24 of the 31 deaths occurring in cabin cruisers:
18 of these deaths were associated with generator exhaust that infiltrated into the
cabin; 5 from propulsion engine exhaust; and 1 from emissions from both types of
engine. Four deaths occurred outside cabin cruisers, with the source being gener-
ator exhaust in three of the deaths and propulsion engine exhaust in the remaining
incident.
Case 4. In June 2005, a 36-year-old man and his 35-year-old wife died of CO
poisoning while inside a cabin cruiser. The couple’s boat was moored at Lake George
Inlet, Florida near two other boats; all three boats’ generators were operating to power
the air-conditioners. When the victims’ generator ran out of gas at about 10:00 a.m.,
friends opened the door and found the occupants’ bodies. When emergency workers
entered the cabin, they found high concentrations of CO. It was unclear if the CO was
from the boat on which the couple died, or from the combined exhaust of all three
generators. Autopsy results revealed that the husband’s COHb was 75.2%; the wife’s
COHb was 77.2%.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 170 — #14

Case 5. In September 2001, a 62-year-old man was swimming near his cabin
cruiser boat on Shasta Lake, California, talking with his wife who was on the boat. The
boat was not moving, but the propulsion engine was operating at the time, charging
the boat batteries. The man went to the swim platform at the rear of the boat to
rest. While resting for approximately 1–3 min, he started splashing water at his wife.
His eyes then became fixed, and he lost consciousness. His wife tried to pull him
from the water, but couldn’t. She moved his body to a nearby island. Although the
medical examiner initially listed his cause of death as a heart attack, this was changed
upon receipt of forensic toxicological test results two months later indicating that his
COHb concentration was 89.3%. The test was repeated, this time indicating a COHb
concentration of 80.8%.
Case 6. In August 2002, two 9-year-old girls were poisoned outside of a moored
cabin cruiser on Lake Powell. One girl died and one survived. They were observed
playing in shallow water (about 30 in. deep) near the rear of the boat very near the
exhaust terminus of the operating generator. One girl was called into the boat by her
parents, and when she climbed out of the water onto the swim platform, she stumbled
and fell onto the floor. She was thought to be suffering from dehydration. The other
girl was discovered to be missing about 15–30 min later. She was found lying on
the bottom of the lake. Attempts to resuscitate her were unsuccessful. The survivor’s
COHb was 15.1% after more than 70 min of oxygen therapy. The girl who died had
a COHb of 39% after more than 40 min of resuscitative efforts, including CPR and
intubation.
7.5.3 HOUSEBOATS
A houseboat typically looks a bit like a house trailer mounted on a large floating barge
(referred to as a monohull) or pontoons. Houseboats vary substantially in size, with
some reaching 90 ft. in length and 16 ft. in width, dependent upon the requirements of
the purchaser, restrictions of the water body on which the boat will be placed, transport
restrictions, and so forth. Many of the fatal poisonings associated with houseboats
(and all of the fatal houseboat poisonings on Lake Powell) were related to a specific
design, shown in Figures 7.9 through 7.11.
This design incorporates a monohull structure with an attached extended swim
deck at the rear of the boat. The rear deck structure creates a cavity or air space
between the hull and the water level swim platform. Exhaust of both propulsion
engines, and sometimes that of the gasoline-powered generator, is directed into this
air space, which came to be referred to by many as the “Death Zone” (a designation
that will be used throughout this chapter for ease of reference). When the engines or
the generator are operating, the build-up of CO in this cavity is so high that it creates
an imminent danger of death for anyone who enters the cavity. Exhaust lingers in this
cavity for on extended period following deactivation of either type of engine. CO
poisonings have also been associated with the very high CO concentrations measured
on or near the swim platform of houseboats. The common practice of continuous
generator operation to provide power for air-conditioning, entertainment centers, and
electronic suites while the houseboat is moored has exacerbated the problem.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 171 — #15

Living
quarters/
boat cabin
Swim
step Extended deck Engine compartment
FIGURE 7.9 Houseboat design associated with fatalities.
Propulsion
engine outdrive
Generator exhaust
terminus
FIGURE 7.10 Cavity beneath the extended deck (showing one of two propulsion engine
outdrives and rear-directed generator exhaust location) - commonly referred to as “the Death
Zone”.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 172 — #16

Exhaust from engine

and generator
DANGER! Air pocket

and water under step
Swim platform
Used by permission.
Copyrigh  2007 Boat Ed
www.boat-ed.com
FIGURE 7.11 Diagramatic depiction of exhaust accumulation within the cavity.
Poisonings outside of houseboats were typically attributed to exposure to gener-

ator exhaust (15 fatalities and 44 nonfatal poisonings). In other incidents, the source
of CO was a combination of generator and propulsion engine exhaust (two deaths
and one nonfatal). Five people were poisoned by propulsion engine exhaust only
(two deaths and three nonfatal).
All poisonings that occurred inside the living quarters of houseboats (2 fatalities
in 1 incident; 177 nonfatal poisonings in 27 incidents) were attributed to generator
exhaust infiltrating the cabin.
Case 7. In July 2004, a 34-year-old woman drowned and two others lost con-
sciousness as a result of CO poisoning on Perry Lake, Kansas. A group of women
were swimming behind several tethered (rafted) houseboats on which one or more
generators were operating. There was little wind when the incident occurred. Two
of the women were found unconscious. One of the women was not breathing, but
was revived, and the second was unconscious and unresponsive. A few minutes
later, someone noted that the third woman was missing. Her body was recovered
approximately 30 min later. On autopsy, her COHb was 45%.
Case 8. In September 2002, a 42-year-old man entered the airspace beneath the
extended rear deck of a houseboat on Lake Powell shortly after the propulsion engines
were deactivated. Just prior to the incident, the boat occupants were attempting to
moor the boat during windy weather. As they maneuvered the boat, the anchor ropes
became entangled in one of the engine propellers. Just after the engines were deac-
tivated (estimated to be more than 3, but less than 5 min), the victim entered the air
space beneath the stern deck to remove the lines from the propeller. He was wearing
8417: “8417_c007” — 2007/9/11 — 12:16 — page 173 — #17

a personal floatation device (PFD) at the time. After his first entry into the airspace
(estimated to have lasted about 2 min), he emerged and removed the PFD because he
was unable to access the propeller. After approximately 2 min, he entered the space
again, and stayed there about 2 min. He emerged for 2 min, and then re-entered the
space a third time. After about 2–3 min elapsed, he no longer responded to questions
from the boat occupants and failed to emerge from the space. His overall time of
exposure was thought to have been 6 min, with a total of approximately 15 min tran-
spiring before he was no longer heard from. Although divers made many attempts to
find him, they were unsuccessful. His body floated to the surface 3 days later. Autopsy
results indicated that his COHb was 51%.
Case 9. In June 1998, a 4-year-old girl was swimming at Lake of the Ozarks,
Missouri with a group of children behind the rear deck of a houseboat. She was
wearing a PFD, and was under the direct supervision of adult swimmers. She
swam to the swim platform, held on to the ladder while her mother applied sun-
screen to her face, and then swam away. Within moments she was observed floating
face up on the water, unconscious, and rigid. She was quickly brought into the
boat where her mother, a registered nurse, checked her for respirations and pulse.
She appeared pale and stiff at that time, was unresponsive with poor respiratory
effort. After 2–3 min of aggressive stimulation, the child began responding with
grunts but was described as disoriented and sleepy. Paramedics were called and
arrived 10–15 min later. They administered oxygen and transported the child to
the nearest hospital emergency department within 30–45 min. Her COHb level at
the hospital after approximately 1 h of oxygen therapy was 22.2%. Upon examin-
ing the houseboat during their next visit to the lake, the child’s parents discovered
that the exhaust terminus for the onboard generator that was operating at the time
of this poisoning was located at the edge of the swim platform, in the center of
the rungs of the ladder that the child was holding onto when the sun screen was
applied.
Case 10. In June 2000, 15 people, ages ranging from 16 to 47 years, were over-
come by CO on two rented houseboats on Lake Cumberland, Kentucky. The boats
were tied together and anchored in a cove. Both boats had gasoline-fueled generators;
the generator on one of the boats had a side-directed exhaust terminus. The exhaust of
one of the generators seeped into the adjacent boat through an open bathroom window.
CO was circulated through the full interior of the boat by the central air-conditioning
system. A few boat occupants awoke at about 5:00 a.m. with headaches and nausea.
Realizing they had a problem, the group radioed the marina and ambulances met the
boats at the shore. The water patrol officer that responded to the emergency witnessed
that two occupants were unconscious when he arrived, and others were drifting in
and out of consciousness. All 15 people were treated at the emergency department of
a nearby hospital; 3 were admitted as hospital inpatients for further treatment. There
were six CO detector/alarms on this boat, but none were properly connected when
the boat was inspected after the poisoning incident.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 174 — #18

7.6 AIRBORNE CARBON MONOXIDE

CONCENTRATIONS MEASURED ON AND NEAR
BOATS
7.6.1 METHODS FOR MEASURING, ANALYZING, AND EVALUATING
CO IN AIR
Four methods, each with different upper range of measurement capability, were used
for collecting CO air samples in the investigations discussed below.
Method 1: Direct-reading infrared instruments, more conventionally used to

measure CO directly in engine emissions, were used to analyze air samples of
very high CO concentrations (i.e., collected in the “Death Zone” and on swim
platforms) from houseboats and ski boats. Emissions analyzers are capable of
measuring CO in the ranges of percentages (1% CO equals 10,000 ppm) as high
as 100%, and also measure oxygen concentration.
Method 2: Direct-reading instruments with electrochemical sensor technology

were used to measure CO in concentrations between 0 and 1000 ppm. These
sensors can be damaged when exposed to excessive CO concentrations for
extended periods.
Method 3: Detector tubes (a glass tube filled with media that changes color
when CO in air is passed through it) capable of measuring 30,000–70,000 ppm
were used primarily to confirm ranges of measurement, and to indicate which
technology could be safely used in different locations of measurement.
Method 4: A laboratory analytical method was used to confirm very high CO

concentrations measured by direct-reading instruments. Grab samples were
collected using glass air-evacuated containers, which were then shipped to a
laboratory for analysis. The sample inside each vial was then analyzed for a
number of components, including CO and oxygen. This method was capable
of analyzing 0–100% CO, as well as oxygen concentration.
How much CO is too much? The answer to this question involves both duration
of exposure (minutes, hours, days, etc.) and air concentration. Units of measure for
CO concentration in air are parts of CO per million parts of air (ppm). Exposure
to CO concentrations (in ppm) results in a rise in CO in the blood, referred to as
COHb—expressed as percent saturation.
Various health and safety agencies recommend or require limits for CO in air.
Table 7.2 shows the limits for general populations and for workers. These limits are
presented as evaluation criteria to help the reader understand the hazard associated
with the air sampling data presented below.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 175 — #19

TABLE 7.2
Evaluation Criteria for Carbon Monoxide
Limit
(in parts
of CO per
million parts Time
Agency Intent of the Limit of air) Period
US Environmental Established to protect the most sensitive 35 1h
Protection Agency members of the general population by
maintaining increases in COHb to less than
2.1%77
9 8h
World Health Recommendations established to protect the 87 15 min
Organization general population by maintaining increases in
COHb to less than 2.5% when a normal subject
engages in light or moderate exercise78
52 30 min
26 1h
9 8h
National Institute The 8-h limit is recommended to protect workers 1200 IDLH
for Occupational from health effects associated with COHb levels value
Safety and Health in excess of 5%. The IDLH (Immediately
Dangerous to Life and Health) limit is
recommended as a concentration at which an
immediate or delayed threat to life exists or at
which an individual’s ability to escape unaided
from a space would be compromised. The
ceiling limit is recommended based on acute
effects of exposure79
200 Ceiling
limit—
never to be
exceeded
35 8h
American The 8 h limit is recommended to protect workers 125 Excursion
Conference of from increases in COHb levels in excess of limit (5 times
Governmental 3.5%. The agency similarly recommends a the 8-h
Industrial Biological Exposure Index (BEI) of 3.5% COHb standard)
Hygienists for end of shift exhaled breath analysis in
nonsmoking workers80
25 8h
Occupational Required limit to protect workers from health 50 8h
Safety and Health effects associated with a COHb of 8–10%81
Administration
8417: “8417_c007” — 2007/9/11 — 12:16 — page 176 — #20

7.6.2 CARBON MONOXIDE CONCENTRATIONS ASSOCIATED WITH

BOAT-RELATED CO POISONINGS
7.6.2.1 Ski Boats
In June and July 2001, CO air samples were taken as part of the investigation of the
Lake Powell death of an 18-year-old boy that succumbed to exposure to CO while
“teak surfing” (see Section 7.5.1, Case 1).20 CO concentrations as high as 23,800
ppm in the unobstructed airspace above the swim platform (where the teak surfer’s
face and upper body would be during the activity) were measured. When the airflow
above the platform was obstructed by a form simulating the shape of the upper torso
of a person with extended arms (like a “teak surfer”), CO concentrations above the
platform were consistently between 10,000 and 26,700 ppm, despite the fact that
engine maintenance had been conducted.
In an August 2001 Connecticut incident, a 15-year-old boy who was teak surfing
for an estimated 10–15 min, released the platform, floated briefly, lost consciousness,
and sank. His COHb measured during the autopsy was 56%.21 An investigation
involving air sampling for CO was later conducted on the boat. (The upper limit
for equipment used to measure CO in air for this investigation was 1000 ppm.) CO
concentrations exceeded 1000 ppm on the swim platform (and in what would be the
breathing zone of simulated swimmers/teak surfers) at varying engine speeds, and
whether the platform was obstructed or unobstructed. The operator of the boat was
exposed to as high as 410 ppm during idle, depending on the wind direction. Average
concentrations consistently exceeded 200 ppm during the entire sampling event when
the boat was operating.
7.6.2.2 Cabin Cruisers

In response to reported incidents that identified a potential safety problem involving
CO for recreational boater occupants inside cabins, the USCG contracted a study to
evaluate intrusion of CO into the passenger areas of gasoline-powered recreational
cabin cruiser boats.2 Tests results reported in 1991 showed that under certain condi-
tions, CO accumulated in the passenger spaces at a rate of about 1 ppm per minute,
posing a health and safety threat to occupants. In areas at the rear of the boat (on the
transom, in the rear deck passenger area, etc.) peak CO concentrations as high as 400
ppm, and 30 min average concentrations as high as 272 ppm, were measured. The
authors concluded: “It is apparent that the potential exists for a serious health and
safety problem for power boaters.”
Results of an interagency investigation of a fatal and nonfatal CO poisoning that
occurred near a cabin cruiser were reported in 2002.22 (See Section 7.5.2, Case 6
for details of the incident.) CO concentrations as high as 41,600 ppm, and oxygen
concentrations as low as 12%, were measured in the open air within inches of the
boat’s generator exhaust terminus (approximately where the poisoned children were
exposed). In addition, CO concentrations consistently exceeded the upper limit of
the air monitors placed on the swim platform and at locations one and five ft. from
the exhaust terminus, indicating that the CO concentration at these locations were
8417: “8417_c007” — 2007/9/11 — 12:16 — page 177 — #21

somewhere between 1000 ppm and the maximum value measured at the exhaust
terminus (41,600 ppm). Concentrations as high as 570 ppm were measured 10 ft.
away from the exhaust terminus.
7.6.2.3 Houseboats
Unless otherwise specified, all data discussed below pertain to houseboats designed
similarly to those in Figures 7.9 through 7.11.
In 1995, a 43-year-old man died in Florida as a result of a 5-min CO exposure
within the airspace under the extended rear deck of a houseboat. The boat’s electrical
power generator, which released exhaust into the airspace, was activated just before
he entered; propulsion engines were not operating. Approximately 5 min after he
entered the airspace, he was observed unresponsive floating face down in the shallow
water; he subsequently died. His COHb measured in the hospital 2 h after exposure,
and after more than an hour of oxygen therapy, was 29.7%. A forensic toxicologist
estimated that the man’s COHb was greater than 70% when he collapsed. CO con-
centrations measured as part of the investigation of this death were reported in legal
proceedings.17 Inspection and testing of the houseboat revealed that CO concentra-
tions in that airspace would reach 4,000–10,000 parts per million (ppm) within 2–5
min after the generator was activated.
As part of the initial investigation of the death of two young brothers exposed
to generator exhaust in the airspace under the rear deck of their family houseboat at
Lake Powell in August 2000, the NPS conducted air sampling to define the incid-
ent circumstances.23 CO concentrations at water level below the rear deck of three
similar houseboats exceeded 2000 ppm (which was the maximum CO concentration
the instrument could measure) approximately 10 min after generator activation. CO
measurements collected above the swim platforms of the three boats were 800, 100,
and 1156 ppm. Two of the people conducting the test began to experience headache,
mild nausea, and weakness while conducting this air sampling.
As other agencies joined in the investigation of past and ongoing CO poisonings at
Lake Powell, air sampling on houseboats continued. In September 2000, air sampling
in the airspace beneath the rear deck on similar houseboats documented the following
maximum CO concentrations: 13,000 ppm when one propulsion engine operated;
30,000 ppm when the generator operated (with concurrent oxygen deficiency, 13%
measured); and 20,000 when both the propulsion engines and generator operated. CO
concentrations exceeding 1200 ppm above and around the water level swim platform
were also measured. Repeat testing in November, 2000 confirmed these very high
concentrations of CO in the “Death Zone” of similar boats, as well as on the upper
rear deck and swim platform.24
Further testing was conducted on houseboats at Lake Cumberland in Kentucky,
with results reported in December, 2000.25 This study again documented that CO
concentrations near and under the rear deck of houseboats were very high (4,078 ppm
measured near water level off the back of the swim platform when the gasoline-
powered generator operated, and 10,224 ppm measured at the same location when
the generator and propulsion engines operated). These measurements confirmed that
there was significant potential for poisonings on houseboats on lakes other than Lake
8417: “8417_c007” — 2007/9/11 — 12:16 — page 178 — #22

Powell. This study was the first to characterize emissions of diesel generators—with a
maximum CO concentration of 10 ppm measured when the diesel generator operated
near the exhaust terminus and at the back of the boat. However, when the gasoline
motors were in operation, CO concentrations increased considerably, with 455 ppm
measured on the swim platform.
In March, 2003, an interagency group collected air sampling data as part of
a supplemental investigation of a fatal poisoning at Lake Powell.26 (See Section
7.5.3, Case 8 for a description of the death.) The purpose was to provide further
information about clearance of propulsion engine exhaust from the “Death Zone”
airspace. (Previously reported data documented that the CO concentration in that
airspace decayed to 0 ppm within 8 min following deactivation of the generator and
propulsion engines.27 ) In repeated tests conducted on two boats in the 2003 study,
rates of decay varied widely—with initial CO concentrations as high as 88,200 ppm
decaying to 2 ppm in widely varying time periods (from 10 to 30 min), indicating that
CO clearance times in that airspace are unpredictable, and could feasibly be much
longer than documented in these tests.
7.6.2.4 Carbon Monoxide Exposure in Areas of Congested

Boat Traffic (Lake Havasu Study)
During summer holiday weekends, as many as 700 boats collect in the Bridgewater
Channel of Lake Havasu at any given time, some moored to the shoreline of the chan-
nel and many cruising slowly through it (see Figure 7.12). The City of Lake Havasu
estimates that about one-third of the boats in the channel have engines operating,
representing exhaust roughly equivalent to that of 40,000 automobiles.
As a result of initial reports of poisonings occurring in an area congested with
boats at Lake Havasu, Arizona, CO exposures among municipal employees and vis-
itors at the Bridgewater Channel were evaluated through air sampling and expired
breath measurements. Air sampling was conducted during summer holidays in 2002
and 2003, when boat traffic within the channel was congested.28 Initial sampling
documented that CO concentrations in channel air exceeded all short-term exposure
criteria listed in Table 7.2; and 4 of 12 boating patients reporting to a local hos-
pital emergency department for any cause (injury, LOC, etc.) had COHb levels of
9.4–28.3%.
In response to these findings, an interagency investigation of CO concentrations
in the channel was conducted during Memorial Day weekend (May), and during
June–September, 2003. These surveys documented excessive CO exposure and con-
firmed the health risk among vacationers and employees working in the channel
near crowded motorboat gatherings, with a trend of higher exposures later in the
day. Sixty-nine percent of monitored workshift exposures of 36 municipal employ-
ees involved short-term CO exposures that exceeded the NIOSH ceiling limit of 200
ppm. Overexposures were confirmed by postshift measurements of exhaled breath:
67% of sampled workshifts resulted in estimated COHb levels equal to or greater
than the recommended limit of 3.5%,29 with the average COHb among nonsmoking
employees increasing from 1% in the morning to 6% in the afternoon (maximum
was 13%). Among 46 nonsmoking vacationers that volunteered for exhaled breath
8417: “8417_c007” — 2007/9/11 — 12:16 — page 179 — #23

FIGURE 7.12 Boats in Lake Havasu’s Bridgewater Channel.
measurements, the estimated COHb increased from a mean of 1% in the morning to

11% in the afternoon (maximum was 23%).19
7.7 PREVENTION EFFORTS

7.7.1 U.S. COAST GUARD REGULATION OF BOAT
MANUFACTURING/RECALL AUTHORITY
The USCG Recreational Boating Product Assurance Division of the Office of Boating
Safety is responsible for such things as developing and enforcing federal safety stand-
ards and investigating consumer complaints. Specific responsibilities of the division
includes inspecting and testing recreational boats for compliance, and issuing recalls
of recreational boats and associated equipment. For this reason, boats are excluded
from the jurisdiction of the US Consumer Product Safety Commission.
After the August 2000 deaths of two brothers at Lake Powell (and related
subsequent data collection), the Coast Guard determined that the houseboat design
shown in Figures 7.9–7.11 (specifically a houseboat with an extended rear deck,
water level swim platform, and rear-directed generator exhaust terminus that emp-
ties into the air space beneath the rear deck) was defective and fell within their
recall jurisdiction. The agency contacted 71 houseboat builders, informing them
about data documenting a deadly combination of generator exhaust and houseboat
swim platforms on boats of this design.30 As a result of this and following mailings,
8417: “8417_c007” — 2007/9/11 — 12:16 — page 180 — #24

six manufacturers agreed to a voluntary recall.31 Initially, it was estimated that more
than 2000 boats would have to be retrofitted with the new design suggested by
manufacturers (rerouting rear-directed generator exhaust terminus from within the
cavity to the side of the boat outside of the cavity), but that number was adjusted
downward to 1087 boats. Manufacturers self-certified that over 800 boats were retro-
fitted as a result of the voluntary recalls.32 Side exhaust was seen as an improvement
in design, but not a solution, as there had been poisonings associated with water
level exhaust outside of any enclosure (such as the one formed by the cavity beneath
the deck). In addition, the common practice of “rafting” several houseboats together
raised concerns about side exhaust that would now be directed towards the next boat
when the generator operated.
The Coast Guard also acted to prevent poisonings associated with the use of a
shower system at the back of the boat that draws heated water from the operating
propulsion engine (thus exposing the user to propulsion engine exhaust). This device
was associated with two poisoning incidents (one fatal and four nonfatal poisonings).
In December 2005, the Coast Guard notified the manufacturer of the system that there
were CO and other safety hazards for recreational boaters who use the shower system
directly connected to the operating propulsion engine. The manufacturer responded
with assurances that the system installation would be changed to eliminate connec-
tion to an operating propulsion engine, and that they would provide additional CO
warnings in printed materials.30
7.7.2 ENVIRONMENTAL PROTECTION AGENCY REGULATION OF

MARINE ENGINE EMISSIONS
The Clean Air Act directs Environmental Protection Agency (EPA) to regulate non-
road engines, a classification that includes marine propulsion engines and marine
generators. Initial EPA regulations for gasoline-powered marine engines, published
in 1996, pertained to outboard engines and personal watercraft only. These standards,
phased in from 1998 to 2006, achieved approximately a 75% reduction in hydrocar-
bons (HC) and oxides of nitrogen (NOx ) from new engines. Although a cap in CO
emissions was proposed in these regulations, none was finalized. However, some of
the methods used to reduce outboard engine HC emissions also result in a reduction
in CO emissions.33
In 2001, the California Air Resources Board adopted emission standards for new
gasoline-powered sterndrive and inboard marine engines that are expected to require
use of catalytic converters beginning in 2007.34
There are currently no federal standards for emissions from gasoline-fueled
sterndrive and inboard marine propulsion engines. Consequently, while these engines
remain uncontrolled, calculations conducted by Sonoma Technology, Inc. indicated
that an average marine engine emits the same amount of CO as 188 automobiles35
EPA gave notice of its intent to develop a proposal for these engines in 2002.34
EPA regulates new gasoline-powered marine generators under 2 sets of rules.
Rules for small generators (≤ 19 kW) vary depending on size of the engine, include
limits for HC, NOx , and CO, and were phased in from 1997 to 2005. Regulations
8417: “8417_c007” — 2007/9/11 — 12:16 — page 181 — #25

relevant to large gasoline-powered marine generators (>19 kW) were effective in

2004, with substantial reductions in HC, NOx , and CO required for 2007 models.34
7.7.3 STATE LEGISLATIVE ACTION

The National Association of State Boating Law Administrators (NASBLA) provides
model language for individual state legislation related to boating safety. In 2003, NAS-
BLA approved language related to marine CO poisoning, with revisions approved in
2005.36 The model act, referred to as the Safe Practices for Boat-Towed Watersports
Act, and aimed at the practice referred to as “teak surfing”, included the following
provisions:
[Requirements.]
a) No person shall operate a motorboat or have the engine of a motorboat run

idle while a person is teak surfing, platform dragging, or body surfing behind
the motorboat.
b) No person shall operate a motorboat or have the engine of a motorboat run

idle while a person is occupying or holding onto the swim platform, swim deck,
swim step, or swim ladder of the motorboat.
[Exemptions.]
The provisions of this act do not apply when a person is occupying the swim
platform, swim deck, swim step, or swim ladder while assisting with the docking
or departure of the motorboat, while exiting or entering the motorboat, or while
the motorboat is engaged in law enforcement activity.
To date, five states have promulgated legislation or regulations intended to pre-

vent marine CO poisonings, most using the model language proposed by NASBLA.
Pennsylvania was the first state to revise boating regulations related to unacceptable
boating practices in 2003.37 In 2004, the California State Legislature passed Assembly
Bill (AB) 2222, the Anthony Farr and Stacy Beckett Boating Safety Act of 2004.38
This bill added to the NASBLA language by:
1. Requiring that all state-sponsored or approved boating safety courses

include information about the dangers of CO poisoning at the stern of
a motorized vessel and how to prevent that poisoning;
2. Requiring that any new or used motorized vessel, when sold, bear warning
stickers as to the danger of CO poisoning on boats; and
3. Requiring that boat registration materials include similar informational
about the dangers of CO poisoning and boats.
Nevada followed in 200439 Oregon in 200540 and Washington state41 and Utah42
in 2006, each state adopting and embellishing the NASBLA model language.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 182 — #26

7.7.4 EFFECTIVENESS OF MARINE CARBON MONOXIDE

DETECTOR/ALARMS IN PREVENTION
7.7.4.1 Case-Based Data
Investigative records related to Lake Powell CO poisonings that occurred in the living
quarters of houseboats were evaluated with regard to CO detector/alarm presence and
function. From 1990 to 2004, 80 people survived poisonings inside houseboats in 15
incidents.14 Fifty of the eighty people were poisoned in boats that were known to have
had CO detectors in the living quarters. However, in only one incident (four people
poisoned) did the alarm sound alerting the occupants to the hazard. Twenty people
were poisoned in three separate instances involving disarmed or “broken” detectors.
Twenty-two people were poisoned in four instances involving functional detectors
that failed to sound during the poisoning incident. Ten people were poisoned in a
houseboat that was not equipped with CO detectors. Twenty people were poisoned
in houseboats in which the record failed to document the absence or presence of
detectors.
In May 2005, an analysis of boat-related CO poisonings conducted by the National
Marine Manufacturers Association (NMMA) pointed out that documentation was
available regarding CO detectors in 12% (66) of 506 cases they analyzed, and for half
of these cases (33), the exposure occurred with the CO detector disconnected.43 In 11
of the 66 cases (17%), the CO detector malfunctioned, and in only 1 of 66 poisonings
did a CO detector sound.
The failure of onboard CO detectors in the living quarters of houseboats on Lake
Powell, as well as data related to detectors elsewhere, raises concern about the impact
of such devices in CO-poisoning prevention. Improving the effectiveness of these
devices is complex, as there were four types of problems identified, each indicat-
ing a different corrective strategy. (Identified problems included: failure to alarm;
alarming when they shouldn’t; disarmed or dysfunctional detector; and absence of
detectors.) Failure of functional detectors to warn occupants of high CO concen-
trations and the sounding of alarms for no discernable cause are related to detector
sensor technology. A likely explanation of disabling of detectors by boaters is that
the detectors are sounding frequently and the boater either cannot identify a cause
for the alarm (also a detector technology issue) or cannot resolve the issue that is
causing CO to enter the cabin (an issue related to boat design, technology, and boater
education).
7.7.4.2 Marine Carbon Monoxide Detector Evaluation

In 2004, the USCG Office of Boating Safety Recreational Boating Product
Assurance Division released a contracted study of CO detector performance in the
marine environment.44 The purpose of the study was to evaluate the reliability of CO
detectors that, in 2002, were advertised as being suitable for marine environments.
About 90 detectors (five different products) were evaluated; 54 had metal-oxide
sensors, 18 had electrochemical sensors, and 18 had biomimetic sensors. The goal
was to determine the impact of one or more factors (humidity, salinity, temperature
8417: “8417_c007” — 2007/9/11 — 12:16 — page 183 — #27

variants, or out-gassing of new boat construction materials) on detector perform-

ance, using the Underwriter Laboratories, UL 2034 standard for CO alarms on
recreational boats, as evaluation criteria. The study also included data on perform-
ance of detectors following nonpowered storage, as is common with boats that have
seasonal use.
Results and related recommendations included:
1. Future studies should be performed over two season extremes. This was
based on results of the study that indicated the detectors may be affected
by seasonal changes.
2. The effect of long-term exposure of sensors to volatile compounds resulting
from out-gassing of vessel construction materials and related impact on
sensor functionality should be investigated. This was based on the data
characterizing new and increased existing volatile organic compounds over
time when ambient temperatures rise, and when the boat is in a “closed
condition” as would be the case during storage in nonuse seasons.
3. The recovery time of the sensors used and the time required for the sensor
to reach a stable operating condition should also be investigated. This was
based on results of nonpowered detector testing of the metal-oxide sensors,
documenting that prolonged storage in a nonpowered condition (as is the
case during storage of the boat or periods of nonuse) can affect sensor
performance. Instructions from the sensor manufacturer stated that these
conditions may require a longer preheating period to stabilize the detector
before use, and recommends storing the sensor in a sealed bag containing
clean air and nothing else (as in, no silica gel). No other manufacturer
provided any information regarding the proper or recommended storage
practices when the detector is not in use and not under power.
4. Detectors should be installed on various types of vessels and a portable
test chamber testing capability should be developed to allow for testing
detectors in a real world setting.
7.7.5 ENGINEERING CONTROL RESEARCH AND DEVELOPMENT

By the time data from the investigation of boat-related CO poisonings at Lake Powell
were published in December, 2000,13 work was already underway to develop controls
to reduce the CO hazard. Initial efforts were directed towards houseboat generators,
primarily based on Lake Powell data. (Of the 111 poisonings identified at Lake Powell
at that time, 74 were on houseboats, and 64 of these were associated with exposure
to generator exhaust. In addition, 7 of 11 deaths there had occurred near houseboats
of the design shown in Figures 7.9 through 7.11, and 5 of the deaths were associated
with generator exhaust only.)
In May 2001, the USCG Office of Boating Safety cosponsored the first of what
became a series of workshops to focus on ways to reduce marine CO hazards, bringing
together boat manufacturers, control innovators, and government agencies.45 The
initial meeting focused primarily on controls for houseboat generators, and spawned
a number of research and development projects.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 184 — #28

FIGURE 7.13 Generator exhaust terminus—side exhaust (marked by arrow).
Houseboat generator control alternatives developed and evaluated since then

included
1. Rerouting generator exhaust to the side (as shown in Figure 7.13)46,47

2. Rerouting generator exhaust through a hybrid wet/dry vertical stack
exhaust system with an exhaust terminus several feet above the upper deck
(Figure 7.14)27,47−57
3. Retrofitting generators with an emission control device (Figure 7.15)45,46,49,57
4. An electrical interlock device to deactivate the generator when conditions
indicated that boat occupants were swimming (lowered swim ladder, open
gates, etc.),48 and finally,
5. New manufacture of emission-controlled generators55,58
Control directly at the source is the preferred and most effective intervention.
In February, 2004, Westerbeke was the first manufacturer to introduce a series of
reduced-emission gasoline-powered marine generators, reporting a 99% reduction
in CO emissions. This reduction was confirmed through Coast Guard-sponsored
field testing of two generators (14 and 20 kW) installed and used on houseboats.55
Final results for the second round of tests, designed to assess longevity of effect-
iveness of the emission controls after 2300 and 1300 h of use (calculated to be
equivalent to 92,000 and 52,000 miles of use for an average automobile) indic-
ated little or no deterioration in the effectiveness of the controls.58 In 2005, Kohler
8417: “8417_c007” — 2007/9/11 — 12:16 — page 185 — #29

FIGURE 7.14 Stacks directing generator exhaust upward (marked by arrows).
FIGURE 7.15 Emission control device for retrofit of generators.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 186 — #30

also announced a new line of marine generators reported to reduce CO emissions

by 99%.59
By 2002, focus shifted towards control of propulsion engine emissions. The Coast
Guard and ABYC began semiannual meetings to discuss the progress of research and
development of all marine engineering controls. These are held in conjunction with
the International Boatbuilders Exposition meetings, and minutes of the meetings are
available through ABYC.
These periodic CO workshops have greatly facilitated the exchange of both
technical information among manufacturers and the public, and also the sharing of
accident information, and a general rising of awareness in the industry, government,
and the public. The workshops have also served to accelerate other mitigating tech-
nologies, including diesel engines as viable alternatives for recreational boats.60 The
result has been to highlight a much broader scope of research by adding express cruis-
ers, towed activities, etc. The workshops have also served as a catalyst for revision
of ABYC TH-23 (Design, Construction, and Testing of Boats in Consideration of
CO),61 and addition of dry stack exhaust considerations to ABYC P-1.62
Concurrent with the workshops, the Coast Guard sponsored evaluations of CO
emissions on recreational boats.63,64 Safe distances from the boat were evaluated to
address the issue of potential CO exposures for those being towed on water “toys”
behind the boat (tubers, wake surfers, etc.).65,66 Evaluations of the effectiveness of
devices designed to reroute the exhaust of ski-boat engines were also conducted.67,68
Prototype catalyzed propulsion engines were introduced at the International Boat
Builders Exposition in February 2005, and evaluations of the catalyzed engines
continued.69−71 In March 2006, Indmar introduced the first production inboard
propulsion engine with controlled emissions to be available on 2007 model year
boats.72
Most recently (January 2006), results of a Coast Guard sponsored study of CO
emissions and exposures on a class of cabin cruiser boat known as express cruisers
were reported.73 The study was performed to better understand how CO poisonings
may occur on express cruisers, identify the most hazardous conditions, and begin
the process of identifying controls to prevent/reduce CO exposures. Many of the
evaluated boats generated hazardous CO concentrations: peak CO concentrations
often exceeded 1100 ppm, while average CO concentrations were well over 100 ppm
at the stern (rear). Two boats with a combined exhaust system (exhausting at the sides
and underwater) had dramatically lower CO concentrations than any of the other
evaluated boats (about 40% lower).
7.7.6 INNOVATIONS IN EMS MEDICAL MANAGEMENT

With the support of the USCG, in 2003 NPS EMS personnel at Lake Powell
began using equipment that allowed noninvasive estimation of COHb concentra-
tion from analysis of exhaled breath samples for more effective triaging of patients.
The decision logic shown in Figure 7.16 is used to guide medics in acting upon
results of the analysis. A decision logic for use of a new noninvasive blood pulse-
oximeter has also recently been published74 and is discussed in another chapter of
this book.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 187 — #31

CARBON MONOXIDE POISONING

The micro CO meter is to be used on anyone who has been or you suspect has been exposed to or poisoned by carbon monoxide
(CO)
Consider scene safety for rescuers and patients.

Rapid extrication from hazardous environment PRN.(1)
Address airway issues
YES NO
Obtain exhaled CO Micro CO meter immediately Administer high flow oxygen
measurement available? via nonrebreather mask
Administer high
flow oxygen via Obtain exhaled CO
non-rebreather mask measurement as soon as
Micro CO meter available
Patient is pregnant, has abnormal YES

cerebellar testing, COHb level of 25% PATCH
or greater, exhibits cardiovascular
dysfunction, continued abnormal
mental status or had any of loss of
consciousness or seizure? Consider rapid transport to
hyperbaric facility(3)
NO
NO YES
Patient exhaled COHb level > 20%
Is patient Transport to nearest medical facility

YES
symtomatic? for continued evaluation, treatment,
and monitoring (2)
NO
PATCH
Consider field treatment with no transport.

Continue treatment with oxygen until COHb
is <5% (or 10% in smoker) (2)
(1) Rescuers will be equipped with SCBA for extricating patients from hazardous environmentt.
(2) Continue measuring exhaled CO with Micro CO meter Q 15–30 minute intervals.
(3) Initiate preplanning with receiving facility. Refer to transfer guideline: Hyperbaric oxygen therapy for acute carbon
monoxide poisoning.
FIGURE 7.16 Glen Canyon NPS decision logic for patient triage.
7.8 SUMMARY AND FUTURE DIRECTIONS

On the positive side, in a span of just over 5 years, many agencies, companies,
and individuals have joined to improve identification of marine CO poisoning, char-
acterize risks, develop and evaluate control technologies, and launch a plethora of
prevention programs. The resulting developments are impressive: (1) two manufactur-
ers successfully developed low-emission marine generators that are on the market; (2)
two major manufacturers developed and participated in testing of catalyzed inboard
propulsion engines, with one introducing it in production engines; (3) devices for
rerouting generator and propulsion engine exhaust were developed and tested; and
(4) several states passed safe towing legislation or regulations.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 188 — #32

So what is left to do? We must remember that there are more than 12.8 million
boats registered for use on US waters, more than 12 million of which are propelled
by outboard, inboard, or sterndrive engines.75 The only engines that have emission
controls are newly manufactured outboard engines and one inboard engine model.
Thus, a large proportion of the gasoline-powered marine engines in use have no
emission controls. Thus, even if EPA increases the scope of emission standards by
regulating inboard and sterndrive engines, a lot of boats will continue to emit a lot of
CO for a long time.
EPA regulation of CO emissions for all marine engines is vital for prevention of
marine CO poisonings. A reduction in automobile-related CO poisonings followed
the EPA Clean Air Act regulations requiring emission control on automobiles.76 The
same would likely follow similar requirements for marine engines.
The key to improved prevention of boat-related CO poisonings lies in improved
recognition and reporting. Until there is more comprehensive testing for COHb by
physicians and those involved in death investigation, CO poisoning cases will con-
tinue to be missed. This results in inadequate or incorrect treatment of cases, related
increased morbidity, inaccurate assignation of cause of death, and related failures
in prevention. Development of standardized autopsy protocols, combined with more
extensive training of the responding medical and law enforcement personnel to facil-
itate collection and transfer of adequate incident detail are needed for improved
recognition.
Adequately recognized poisonings must then be reported so that the full scope
of the problem can be defined for prevention programs. Reporting of marine CO
poisonings is low. This can be easily examined by comparing Lake Powell data for
fatal and nonfatal boat-related CO poisonings, identified, and documented through
extensive searches of EMS records, against Coast Guard data for reported CO pois-
onings. From 1990 through 2002, there were 13 fatal boat-related CO poisonings and
151 nonfatal boat-related CO poisonings requiring medical attention documented at
Lake Powell (thus a total of 164 boat-related CO poisonings at that one lake). In that
same period, Coast Guard data documented 170 injuries and 62 deaths related to CO
reported from incidents across the entire country. Clearly, marine CO poisonings are
not extensively reported to the Coast Guard.
State Boating Law Administrators (the recognized reporting authority) should
be encouraged to develop liaisons with other state-based surveillance efforts (i.e.,
state-based child fatality review boards, vital records departments, injury surveil-
lance efforts), as well as Federal surveillance systems, to enhance identification and
reporting of boat-related CO poisonings reporting. In addition, more extensive train-
ing of municipal law enforcement officers and first responders is needed, so that they
can better understand the need to report CO poisonings.
While developments in control have been impressive, implementation of new
controls and retrofitting existing boats remains problematic, as evidenced by recent
poisonings (not included in the 607 cases analyzed here). Recent Coast Guard data
identified two new poisonings (one fatal + one nonfatal) related to entry into the
“Death Zone”. Newspaper reports and related autopsy results document a June, 2006
death associated with ski-boat platform occupancy (COHb = 51%). The list of marine
CO poisonings continues to grow.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 189 — #33

Boat manufactures must assess and address design features that attract occupancy
at the back of the boat, redesigning to eliminate hazards. Consumer education, vital
to the success of retrofit programs, must continue and increase. If boat owners don’t
understand the problem, they won’t be motivated to go to the trouble of complying
with recalls and consumer advisories.
Failure of onboard CO detector/alarms raises concern about the impact of such
devices in CO poisoning prevention. Improving the effectiveness of these devices is
complex, as there were four types of problems identified, each indicating a different
needed corrective strategy. Failure of functional detectors to warn occupants of high
CO concentrations and the sounding of alarms for no discernable cause are related
to much needed improvements in detector sensor technology. A likely explanation of
disabling of detectors by boaters is that the detectors are sounding frequently and the
boater either cannot identify a cause for the alarm (also a detector technology issue)
or cannot resolve the issue that is causing CO to enter the cabin (an issue related to
boat design, technology, and boater education).
As has been documented here, quick death can and does occur from exposure to
CO in the marine environment, even in the open air. Acceptance of this possibility
remains difficult for technical audiences as well as consumers. Anecdotal accounts of
medical examiners and hospital emergency departments failing or refusing to order
COHb analyses related to incidents on boats continue. The need for increased aware-
ness of the problem was recently underscored in the following quote by a family
dramatically impacted by a teak surfing death:
Unfortunately, widespread understanding of the issue among the boating public

remains elusive. We still see kids teak surfing between wake boarding runs. Most
people we talk to are surprised at the severity of the CO risk behind a boat. In
spite of all our efforts, most seem not to realize the potential peril. I recently
talked to two families from Canada who said they teak surfed all the time and
were amazed when I told them it could be deadly. This risk is just not self-evident
and consequently more kids die unnecessarily every year.
Our family, the Dixey family, and people like us that have lost loved ones
certainly know the costs of not knowing about or under-estimating this hazard.
We who have now been warned have the responsibility to sound the dangers for
others. By lifting our voice in warning, we can do our part to make sure that
great kids don’t continue to die.
This quote certainly applies to all types of gasoline-powered marine engines, all
boaters, and to all deaths from this preventable cause of CO poisoning.
7.9 ACKNOWLEDGMENTS
Much of the work described in this chapter, including initial recognition of the problem
and call for engineering changes, would not have happened without my colleague
and good friend, Dr. Robert Baron. I am grateful for his guidance, wisdom, and
partnership. Thanks to Claire Babik, who has worked so hard for prevention in honor
of the brother she lost. In addition, gratitude is owed to Tim Radtke; to National Park
8417: “8417_c007” — 2007/9/11 — 12:16 — page 190 — #34

Service Rangers, Divers, Emergency Medical Technicians (EMTs), and Medics at

GCNRA; and to the researchers at the National Institute for Occupational Safety and
Health. All of these fine public servants have devoted many hours to the cause of
identifying and preventing marine CO poisonings. I pay respect to the families that
have lost so much to this unnecessary cause. And to my own family, I want to say
thanks for everything.
References
1. USCG, Boating Safety Circular 58, A good way to die, US Department of
Transportation, United States Coast Guard, June, 1984.
2. Simeone L.F. The intrusion of engine exhaust into the passenger areas of recreational
power boats, United States Coast Guard Office of Navigation Safety and Waterway
Services Report DOT-VNTSC-CG-91-1, 1991.
3. Simeone L.F. A simple carburetor model for predicting engine air-fuel ratios and
carbon monoxide emissions as a function of inlet conditions, United States Coast
Guard Report CG-079, 1990.
4. Silvers S.M. and Hampson N.B. Carbon monoxide poisoning among recreational
boaters. JAMA, 274, 1614–1616, 1995.
5. American Boat and Yacht Council, A-24, Carbon monoxide detection systems,
American Boat and Yacht Council, Inc., 1997.
6. Struttman T. et al. Unintentional carbon monoxide poisoning from an unlikely source,
JABFP, 11, 481–484, 1998.
7. Huff J.S. and Kardon E. Carbon monoxide toxicity in a man working outdoors with
a gasoline-powered hydraulic machine, NEJM, 320(23), 1564, 1989.
8. First M.W. and Murphy R.L.H. Carbon monoxide exposures from snow melting
machines. Am Ind Hyg Assoc J., 31, 754–757, 1970.
9. Hampson N.B. and Norkool D.M. Carbon monoxide poisoning in children riding in
the back of pickup trucks, JAMA, 267, 538–540, 1992.
10. Jumbelic M.I. Open air carbon monoxide poisoning, J Forensic Sci, 43, 228–230,
1998.
11. Easley R.B. Open air carbon monoxide poisoning in a child swimming behind a boat,
South Med J, 93, 430–432, 2000.
12. GCNRA NPS, Investigative Report, Case 94-4012, Glen Canyon National Recreation
Area National Park Service, 1994.
13. CDC, Houseboat-associated carbon monoxide poisonings on Lake Powell—Arizona
and Utah, 2000. Centers for Disease Control and Prevention, Morbidity and Mortality
Weekly Report, 49(49), 1105–1108, Dec 15, 2000.
14. NIOSH, Hazard evaluation and technical assistance report: 2000-0400 and
20020325, Glen Canyon National Recreation Area, Arizona and Utah, US
Department of Health and Human Services, Public Health Service, Cen-
ters for Disease Control and Prevention, National Institute for Occupational
Safety and Health, 2005. Available http://www.cdc.gov/niosh/hhe/reports/pdfs/2000-
0400-2956.pdf
15. US Coast Guard. Boating statistics—2004. US Department of Homeland Secur-
ity, COMDTPUB P16754.17, 2005. Available http://www.uscgboating.org/statistics/
accident_stats.htm
16. 33 CFR 173.55, Vessel Numbering and Casualty and Accident Reporting, Code of
Federal Regulations, Washington, DC, US Government Printing Office, Office of the
Federal Register, 2002.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 191 — #35

17. DOI, Boat-related CO poisonings on US waters national case listing, Safety Man-
agement Information System (SMIS), US Department of the Interior (DOI), 2006.
Available http://safetynet.smis.doi.gov/cohouseboats.htm
18. CDC, Carbon-monoxide poisoning resulting from exposure to ski-boat exhaust, June
2002 - Georgia, Centers for Disease Control and Prevention, Morbidity and Mor-
tality Weekly Report (MMWR) 51(37);829–830, September 20, 2002. Available
http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5137a3.htm
19. CDC, Carbon monoxide poisonings resulting from open air exposures to oper-
ating motorboats—Lake Havasu City, Arizona, 2003. Centers for Disease Con-
trol and Prevention, Morbidity and Mortality Weekly Report, 53(15), 314–
318, April 23, 2004. Available http://www.cdc.gov/mmwr/preview/mmwrhtml/
mm5315a3.htm
20. McCammon, J.B. Letter to Joe Alston, Park Superintendent, Glen Canyon National
Recreation Area from J.B. McCammon, NIOSH, CDC, PHS, DHHS, Denver,
Colorado, July 31, 2001. Available http://safetynet.smis.doi.gov/teakfinal.pdf
21. Ritter G. Air monitoring for Connecticut Department of Environmental Protection
Division of Conservation Law Enforcement boating accident reconstruction, Prepared
by TRC Environmental Corporation, September 13, 2001.
22. McCammon J.B. et al. Letter to Joe Alston, Park Superintendant, Glen Canyon
National Recreation Area, Page, Arizona, from J. McCammon (National Institute
for Occupational Safety and Health, Centers for Disease Control), T Radtke (US
Department of Interior), and DP Bleicher (National Park Service), December 3, 2002.
Available http://safetynet.smis.doi.gov/searayfinal.pdf
23. GCNRA NPS Investigative Report, Case 2000-4627: Carbon monoxide test-
ing, August 25, 2000, Glen Canyon National Recreation Area National Park
Service, 2000.
24. Hall, R.M. and McCammon, J.B. Letter to Joe Alston, Park Superintendent,
Glen Canyon National Recreation Area from R.M. Hall and J.B. McCammon,
NIOSH, CDC, PHS, DHHS, Cincinnati, Ohio, November 21, 2000. Available
http://safetynet.smis.doi.gov/HHE2powell.pdf
25. Hall, R.M. Letter to Dr. Rice Leach, Commissioner, Cabinet for Health Service,
Commonwealth of Kentucky from R.M. Hall, NIOSH, CDC, PHS, DHHS, Cin-
cinnati, Ohio, December 18, 2000. National Institute for Occupational Safety and
Health Hazard evaluation and technical assistance report: 2001-0026. Available
http://www.cdc.gov/niosh/hhe/reports/pdfs/2001-0026-letter.pdf
26. McCammon J.B. et al. Letter to Joe Alston, Park Superintendant, Glen Canyon
National Recreation Area, Page, Arizona from J. McCammon (National Institute
for Occupational Safety and Health, Centers for Disease Control), T. Radtke (US
Department of Interior), and D.P. Bleicher (National Park Service), March 17, 2003.
Available http://safetynet.smis.doi.gov/notoriousfinal.pdf
27. Dunn K. et al. An evaluation of an engineering control to prevent carbon monox-
ide poisonings of individuals on houseboats at Sumerset Custom Houseboats,
Somerset KY, National Institute for Occupational Safety and Health/Division of
Applied Research and Technology/EPHB Report 171-26a, May, 2001 Available
http://www.cdc.gov/niosh/surveyreports/pdfs/ectb-171-26a.pdf
28. NIOSH, Hazard evaluation and technical assistance report: 2002-0393, Lake Havasu
Municipal Employees, Lake Havasu City, Arizona, February 2004, US Department of
Health and Human Services, Public Health Service, Centers for Disease Control and
Prevention, National Institute for Occupational Safety and Health, 2004. Available
http://www.cdc.gov/niosh/hhe/reports/pdfs/2002-0393-2928.pdf
8417: “8417_c007” — 2007/9/11 — 12:16 — page 192 — #36

29. ACGIH, TLVs® and BEIs®. American Conference of Governmental Industrial

Hygienists, 2006.
30. McCormick D. Personal communication, 2006.
31. NBSAC, Minutes of the National Boating Safety Advisory Council 67th meet-
ing, Cleveland OH, 2001. Available http://www.uscgboating.org/nbsac/MtgMin/pdf/
67thNBSAC_Minutes.pdf
32. USCG, US Coast Guard Recall website, accessed Jan 30, 2006. Searched for
“generator exhaust” and “exhaust system”. Available http://www.uscgboating.org/
recalls/recalls_database.htm
33. EPA, Regulatory update: Overview of EPA’s emission standards for marine engines,
Publication EPA420-F-04-031, August 2004. Available http://www.epa.gov/otaq/regs/
nonroad/marine/420f04031.pdf
34. EPA, Program Announcement: Marine engine manufacturer develops low emis-
sion inboard marine engines. Publication EPA420-F-06-057, July 2006. Available
http://www.epa.gov/oms/regs/nonroad/marine/si/420f06057.htm
35. Roberts P. Unpublished data. Sonoma Technology Inc., Petaluma, California.
36. NASBLA, Safe practices for boat-towed watersports act. NationalAssociation of State
Boating Law Administrators, 2005. Available http://www.nasbla.org/pdf/Model%
20Acts/new/Boat-Towed%20Watersports%2092105.pdf
37. The Pennsylvania Code, 105.3 Unacceptable boating practices, 2003. Available
http://www.pacode.com/secure/data/058/chapter105/s105.3.html
38. California Code, Harbors and Navigation Code, Section 680-685, Anthony Farr and
Stacy Beckett Boating Safety Act of 2004, 2004. Available http://www.leginfo.ca.gov/
cgi-bin/displaycode?section=hnc&group=00001-01000&file=680-685
39. Nevada Code, NAC 488.435 Prima facie evidence of reckless or negligent operation,
2004. Available http://www.leg.state.nv.us/NAC/NAC-488.html#NAC488Sec435
40. Oregon Legislative Assembly, Senate bill 56 amending ORS 830.990, 2005. Available
http://www.leg.state.or.us/05reg/measpdf/sb0001.dir/sb0056.en.pdf
41. Washington Senate Bill 6364, An act relating to the regulation of recreational
vessels; amending 2 RCW 79A.60.610; adding a new section to chapter
79A.60 RCW; adding new sections to chapter 88.02 RCW; creating a new section;
prescribing penalties; and providing an effective date, 2005. Available http://
www.leg.wa.gov/pub/billinfo/2005-06/Pdf/Bills/Senate%20Passed%20Legislature/
6364.PL.pdf
42. Utah Administrative Code R651-224. Towed Devices. Available at http://www.rules.
utah.gov/publicat/code/r651/r651-224.htm R651-224-2. Unlawful Methods of
Towing.
43. Taylor R.K. et al. National Marine Manufacturers Association carbon monoxide label
study. Design Research Engineering, Novi, Michigan, 2005.
44. US Coast Guard Contract Number DTCG39-00-D-R0009, Task Order 01-F-00016,
September 2002.
45. US Coast Guard, Houseboat industry workshop on carbon monoxide, Summary
record, Lexington, Kentucky, May 3, 2001. USCoast Guard Office of Boating Safety,
Recreational Boating Product Assurance Division, Report prepared by Potomac
Management Group.
46. Earnest G.S., Beamer B., and Dunn K.H. Evaluation of side exhaust and prototype
and production emission control devices to prevent carbon monoxide poisonings
on houseboats. National Institute for Occupational Safety and Health/Division of
Applied Research and Technology/EPHB Report 171-29a, May 2002. Available
8417: “8417_c007” — 2007/9/11 — 12:16 — page 193 — #37

47. Dunn K.H. et al. Comparison of a dry stack with existing generator exhaust
systems for prevention of carbon monoxide poisonings on houseboats. National
Institute for Occupational Safety and Health/Division of Applied Research and
Technology/EPHB Report 171-28a, August 2001. Available http://www.cdc.gov/
niosh/surveyreports/pdfs/ectb-171-28a.pdf
48. Earnest G.S. et al. An evaluation of an engineering control to prevent carbon monox-
ide poisonings of individuals on houseboats. National Institute for Occupational
Safety and Health/Division of Applied Research and Technology/EPHB Report
171-25a, March 2001. Available http://www.cdc.gov/niosh/surveyreports/pdfs/ectb-
171-25a.pdf
49. Earnest G.S. et al. An evaluation of an emission control device, exhaust
stack, and interlock to prevent carbon monoxide poisoning of individuals on
houseboats. National Institute for Occupational Safety and Health/Division of
Applied Research and Technology/EPHB Report 171-27a, August 2001. Available
50. NIOSH, Evaluation of two exhaust stack configurations on two houseboats at Table
Rock Lake, Missouri. HETA Report Number 2003-0318-2936. Cincinnati, OH: US
Department of Health and Human Services, Public Health Service, Centers for Disease
Control and Prevention, National Institute for Occupational Safety and Health, June
2004. Available http://www.cdc.gov/niosh/hhe/reports/pdfs/2003-0318-2936.pdf
51. Earnest G.S. et al. An evaluation of vertical exhaust stacks and aged production
emission control devices to prevent carbon monoxide poisonings from houseboat
generator exhaust. National Institute for Occupational Safety and Health/Division of
Applied Research and Technology/EPHB Report 171-32a, October 2003. Available
52. Hammond D.R. and Marlow D.A. Followup evaluation of design changes to a house-
boat generator exhaust stack system. National Institute for Occupational Safety and
Health/Division of Applied Research and Technology/EPHB Report 171-34a2, July
2004. Available http://www.cdc.gov/niosh/surveyreports/pdfs/ectb-171-34a2.pdf
53. Hammond D.R., Earnest G.S., and Hall R.M. An evaluation of factors that
might influence exhaust stack performance to prevent carbon monoxide poisonings
from houseboat generator exhaust. National Institute for Occupational Safety and
Health/Division of Applied Research and Technology/EPHB Report 171-34a1, Janu-
ary 2004. Available http://www.cdc.gov/niosh/surveyreports/pdfs/ectb-171-34a1.pdf
54. Hammond D.R. et al. An evaluation of conditions that may affect the perform-
ance of houseboat exhaust stacks in prevention of carbon monoxide poisonings from
generators, JOEH, 3, 308–316, June, 2006.
55. ZimmerA.T., Earnest G.S., and Kurimo R.An evaluation of catalytic emission controls
and vertical exhaust stacks to prevent carbon monoxide poisonings from houseboat
generator exhaust. National Institute for Occupational Safety and Health/Division of
Applied Research and Technology/EPHB Report 171-36a, September 2005. Available
56. Dunn K.H. et al. Carbon monoxide and houseboats, Prof Safety, 48, 47–57, Nov.,
2003.
57. Earnest G.S. et al. An evaluation of an engineering control to prevent carbon monoxide
poisonings of individuals on and around houseboats, AIHAJ, 63, 361–369, May/Jun,
2002.
58. Earnest G.S. et al. An evaluation of catalytic emission controls to prevent
carbon monoxide poisonings from houseboat generator exhaust. National Insti-
tute for Occupational Safety and Health/Division of Applied Research and
8417: “8417_c007” — 2007/9/11 — 12:16 — page 194 — #38

Technology/EPHB Report 17/-38a, October 2006. Available http://www.cdc.gov/

niosh/surveyreports/pdfs/ectb-171-38a.pdf
59. Kohler, Press Release: New generation gasoline generators from Kohler reduce carbon
monoxide emissions by 99 Percent. Kohler Power Systems, Kohler WI. August, 2005.
Available http://www.kohlerpowersystems.com/pr/GasolineEmission_2005.html
60. Carlan K, Mastry Engine Center Presentation at the 2004 USCoast Guard / ABYC Car-
bon Monoxide Workshop, IBEX 2004, Miami Beach, FL, October 27, 2004. Minutes
of the meeting available at http://safetynet.smis.doi.gov/miami10-04_cominutes.pdf;
presentation slides available at http://safetynet.smis.doi.gov/miami10-04_slides.pdf
61. ABYC, TH-23: Design, Construction, and Testing of Boats in Consideration of Carbon
Monoxide, American Boat and Yacht Council, Edgewater, Maryland, 2004.
62. ABYC, P-1: Installation of Exhaust Systems for Propulsion and Auxiliary Engines,
American Boat and Yacht Council, Edgewater, Maryland, 2002.
63. Earnest, G.S. et al. Carbon monoxide emissions and exposures on recreational boats
under various operating conditions. Lake Mead, Nevada and Lake Powell, Ari-
zona. National Institute for Occupational Safety and Health/Division of Applied
Research and Technology/EPHB Report 171-05ee2, February 2003. Available
http://www.cdc.gov/niosh/surveyreports/pdfs/ectb-171-05ee2.pdf
64. Echt, A. et al. Carbon monoxide emissions and exposures on recreational boats under
various operating conditions, Lake Norman, N.C. National Institute for Occupa-
tional Safety and Health/Division of Applied Research and Technology/EPHB Report
171-31a, April 2003. Available http://www.cdc.gov/niosh/surveyreports/pdfs/ectb-
171-31a.pdf
65. USCG and ABYC, Carbon monoxide safe distance study. US Coast Guard, American
Boat and Yacht Council, September 2003.
66. Mann, L.W. Carbon monoxide exposure while operating an inboard boat and
related water sports activities. [Online] Available at http://www.FreshAirExhaust.com
(Accessed July 2004).
67. Marlow D.A. et al. Evaluation of the “Fresh Air ExhaustTM ” system to
reduce carbon monoxide exposure during motor boating and wake surfing.
National Institute for Occupational Safety and Health/Division of Applied
Research and Technology/EPHB Report 171-35a, August 2004. Available
68. Marlow D.A., Hammond D., and Earnest G.S. Evaluation of the SideswipeTM
exhaust system to reduce carbon monoxide poisoning during motor boating and
wake surfing. National Institute for Occupational Safety and Health/Division of
Applied Research and Technology/EPHB Report 171-37a, December 2005. Available
69. Hall R.M. et al. Case study: Evaluation of carbon monoxide emissions from engines on
recreational boats equipped with prototype catalysts, JOEH 3(2), D4-D7, Feb., 2006.
70. White J.J. Cleaner pleasurecraft: Lower-emission boats may be coming
soon to a lake near you, Southwest Research Institute, 2004. Available
http://www.swri.org/3pubs/ttoday/summer04/craft.htm
71. Carroll J.N. Catalyzed marine engine saltwater durability, Presentation by Southwest
Research Institute at the USCG Carbon Monoxide Seminar, Miami, Florida, October
2005, Minutes and presentation available from American Boat and Yacht Council,
3069 Solomon’s Island Road, Edgewater, Maryland, 21037.
72. Indmar, Press release: Indmar introduces first production catalyst for inboard
engines. Indmar Products Co., Millington TN. March 7, 2006. Available
http://www.indmar.com/news/newsroom_article.cfm?id=31
8417: “8417_c007” — 2007/9/11 — 12:16 — page 195 — #39

73. Garcia A., Beamer B., and Earnest G.S. In-depth survey report of carbon
monoxide emissions and exposures on express cruisers under various operat-
ing conditions, National Institute for Occupational Safety and Health/Division of
Applied Research and Technology/EPHB Report 289-11a, January 2006. Available
http://www.cdc.gov/niosh/surveyreports/pdfs/ECTB-289-11a.pdf
74. Hampson N.B. and Weaver L.K. Noninvasive CO measurement by first respon-
ders: A suggested management algorithm, Elsevier Public Safety, Lethal Exposure,
Supplement to J. Emerg. Med. Services, 10–12, Spring, 2006.
75. NMMA, 2005 recreational boating abstract. National Marine Manufacturers
Association, 2005. Available http://www.nmma.org/facts/boatingstats/2005/files/
populationstats3.asp
76. Mott J.A. et al. National vehicle emission policies and practices and declining US
carbon monoxide-related mortality. JAMA, 288, 988–995, 2002.
77. EPA, Air quality criteria for carbon monoxide, Publication No. EPA-600/8-90/045F,
US Environmental Protection Agency, Washington, D.C., 1991.
78. WHO, Environmental health criteria 213 - carbon monoxide (Second Edition), World
Health Organization, Geneva, ISBN 92 4 157213 2 (NLM classification: QV 662)
ISSN 0250-863X, 1999.
79. NIOSH, Recommendations for occupational safety and health: Compendium of policy
documents and statements, US Department of Health and Human Services, Public
Health Service, Centers for Disease Control and Prevention, National Institute for
Occupational Safety and Health, DHHS (NIOSH) Publication No. 92–100, 1992.
80. ACGIH, 2005 TLVs® and BEIs®: threshold limit values for chemical substances and
physical agents. American Conference of Governmental Industrial Hygienists, 2005.
81. CFR 29 CFR 1910.1000, Code of Federal Regulations, Washington, D.C., US
Government Printing Office, Office of the Federal Register, 1997.
8417: “8417_c007” — 2007/9/11 — 12:16 — page 196 — #40

8 Application of Warnings
and Labels for Carbon
Monoxide Protection
Gary Hutter
CONTENTS
8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198

8.2 Characteristics of Carbon Monoxide Formation and Movement as
Related to Warnings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
8.2.1 Physical Characteristics of Carbon Monoxide . . . . . . . . . . . . . . . . . . . . 198
8.2.2 Combustion and Formation of Carbon Monoxide . . . . . . . . . . . . . . . . 199
8.2.3 Buoyancy, Migration, and Dilution of Carbon Monoxide . . . . . . . . 200
8.2.4 Exit Velocity Considerations for Different Carbon Monoxide
Producing Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
8.2.5 Absorption and Dissipation of Carbon Monoxide . . . . . . . . . . . . . . . . 204
8.3 Common Sources of Carbon Monoxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
8.3.1 Consumer Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
8.3.2 Industrial Products and Processes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
8.4 What are Warnings, Labels, Instructions, and Other Forms of
Communications? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
8.5 The Content, Configuration, and Design of Labels, Warnings, and
Instructional Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
8.5.1 Design of Labeling and Warnings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
8.5.2 Written Instructional Information on Safety Materials . . . . . . . . . . . 209
8.5.4 Workplace Area Warnings and Labeling . . . . . . . . . . . . . . . . . . . . . . . . . . 211
8.5.4.1 Examples of General OSHA Warning Signage . . . . . . . . . 211
8.5.4.2 Examples of Specific OSHA Exposure Warning
Signage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
8.5.4.3 Other OSHA and Industry Warning Criteria . . . . . . . . . . . . 213
8.6 The Reasons, Rationale, and Scientific Bases for Providing Safety
Information for Carbon Monoxide with Selected Examples . . . . . . . . . . . . . 214
8.6.1 Ethical and Philosophical Reasons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
8.6.2 Research Basis: the Need for Safety Information . . . . . . . . . . . . . . . . . 216
8.6.3 Voluntary Standards for CO Warnings (Examples in
Chronological Order) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
8.6.4 Mandatory Safety Information Regulations . . . . . . . . . . . . . . . . . . . . . . . 219
197
8417: “8417_c008” — 2007/9/11 — 12:16 — page 197 — #1

8.6.4.1 State-Based Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223

8.6.4.2 Federal Government Examples. . . . . . . . . . . . . . . . . . . . . . . . . . 223
8.6.5 The Custom and Practice of Carbon Monoxide Safety
Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
8.6.6 Litigation-Driven Safety Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
8.1 INTRODUCTION
Other chapters in this book discuss about the chemical, physical, toxicological, and
health concerns of exposure to carbon monoxide (CO). This chapter addresses warn-
ings, labeling, instructions, and other communication modes (i.e., safety information)
about the hazards of exposure to CO. This chapter addresses (1) the conditions that
may lead to the static and variable formation of CO, (2) characteristics that may
impact the need to provide user-safety information, (3) common sources of CO, and
(4) the characteristic movement of CO through the air that may prompt a need for
user-safety information.
The mid-portion of the chapter discusses the functions, criteria, research, and
basis for CO safety information.
Finally, details about the current practices in communicating this information to
those who are potentially exposed to CO is provided. Examples of codes, standards,
and recommended practices are offered, and real-life examples of warnings, labeling,
and instructions are provided from products currently in the stream of commerce.
Under some subject headings, the impact of the subject on the need for warnings
are provided.
8.2 CHARACTERISTICS OF CARBON MONOXIDE

FORMATION AND MOVEMENT AS RELATED TO
WARNINGS
To understand better why CO exposure necessitates warning for the at-risk population,
it is helpful to know the characteristics of CO that influence its hazardous nature.
8.2.1 PHYSICAL CHARACTERISTICS OF CARBON MONOXIDE

CO is extremely toxic. It is often called a poison. It causes a broad array of symptoms
that precede possible death. Kandarjian1 list 22 different symptoms that arise from
CO exposure. (There are actually many more symptoms, signs, and conditions that
can result from CO exposure.)
CO is an invisible, odorless, nonirritating, and tasteless gas that in and of itself
has no (or extremely poor) sensory warning characteristics for those exposed. This
fact contributes to the insidious nature of CO exposure. However, some combustion
processes that release CO also produce aldehydes and/or particles that have distinct
odors, which may alert someone to the possible presence of CO. This is more likely
8417: “8417_c008” — 2007/9/11 — 12:16 — page 198 — #2

Application of Warnings and Labels for Carbon Monoxide Protection 199
to occur from the combustion of solid (wood, charcoal) or liquid fuels (fuel oil), and
is much less likely to occur from the refined gaseous fuels (e.g., methane, propane,
butane).
Natural gas and propane often have odorants added to the fuel mixture for leak
detection purposes. Unfortunately, they are typically consumed in the combustion
process, which eliminates the benefit of possible olfactory warnings. Under unusu-
ally poor combustion processes containing these odorants, some of the odor may
be detectable in the exhaust stream. (It has been suggested that a substance be used
that only gives odor when there is incomplete combustion, but none with complete
combustion.)
Labeling Implications: Carbon monoxide has poor sensory warning character-

istics.
8.2.2 COMBUSTION AND FORMATION OF CARBON MONOXIDE

Carbon-containing fuels that generate CO may originate as solids (e.g., charcoal,
wood), liquids (e.g., kerosene, gasoline), or gases (e.g., methane, propane, butane).
The combustion process actually takes place with the hydrocarbon components in the
gaseous phase even though they may originate from a solid or a liquid. Therefore, the
following discussion is written in terms of the gas phase of the fuel component.
Given the correct initial conditions, carbon (typically from a hydrocarbon
molecule) and oxygen will react to initially form CO as an intermediate com-
pound, with carbon dioxide as the final product of combustion. CO generation as
the final product of combustion is most often the result of interruption of the com-
bustion process at the intermediate step (quenching) and incomplete combustion of
carbon-containing materials.
Stoichiometry is the evaluation of the fuel and air ratios associated with com-
bustion efficiency of the gas phase. In theory, at an ideal stoichiometric mixture of
carbon-based fuels, there would be sufficient amounts of oxygen, on a mass basis, to
react with all of the carbon. This would allow the oxidation reaction to go to comple-
tion. In this case, there would be no CO remaining. An obvious corollary is that under
insufficient oxygen conditions, CO is produced because there are not enough oxygen
molecules to react completely with the carbon molecules. This incomplete combus-
tion can be done intentionally, as is the case with iron ore refining blast furnaces
or coke ovens. Both these processes intentionally operate under nonideal (oxygen-
deficient) stoichiometric conditions to refine (chemically reduce) the iron oxide to
iron, and to thermally distill coal into coke. In both these processes, large amounts of
CO are intentionally produced (1%), are an unavoidable part of the process, and
are recovered as a fuel for other operations.
Oxygen-deficient combustion that leads to the formation of CO can also be unin-
tentional. This occurs because of restricted air supplies, inadequate exhaust gas
management, or an overabundance of fuel. Examples of these scenarios include
(1) comfort heating furnaces that are installed in closet-like locations with insuffi-
cient fresh (combustion) air supply vents, (2) chimneys that have developed internal
blockages, and (3) burners that incorporate jets and/or orifices for the passage of
8417: “8417_c008” — 2007/9/11 — 12:16 — page 199 — #3

gaseous carbon fuels of different molecular weights (such as natural gas furnaces
fueled with propane). All of these conditions can result in combustion problems and
the subsequent generation and release of CO into people’s houses or other living
spaces.
A less obvious condition that leads to the formation of CO occurs when stoi-
chiometrically correct ratios of carbon and oxygen do not have sufficient time,
molecular intimacy, or energy (in the form of heat) in a combustion process to
completely react, resulting in an interrupted or incomplete reaction. To avoid these
conditions, gaseous fuels are often premixed with the combustion air in a turbulent
fashion to provide the time and contact for a complete reaction. For example, in gas
burners there are various fuel jets and mixing chambers that swirl the air and fuel
together. This facilitates the dispersion of oxygen within the combustible fuel, and
thus enhances combustion.
Combustion temperatures influence the creation of CO, and reflect the heat
available for a carbon–oxygen reaction to go to completion. These temperatures
are typically controlled so that there is sufficient energy available for completing
the reactions. Quenching the temperatures can shift the reaction kinetics and allow
some of the carbon to react only to the CO state, resulting in the presence of CO in the
exhaust. In low-temperature combustion processes, a catalyst can be used to lower
the threshold combustion temperatures. This method better assures complete combus-
tion and is used in catalytic heaters and reactors. Combustion processes depending on
catalytic reactions are prone to the elevated formation of CO if the catalyst becomes
damaged or poisoned.
Finally, even well-managed combustion mixtures of carbon-based fuels and air
can produce some finite quantity of CO simply because the processes are not ideal.
Incomplete combustion occurs, and if not vented properly can result in the formation
of even greater amounts of CO. For example, natural gas in a stove-top burner or
oven may produce several parts per million (ppm) of CO,2 especially when at low
settings. The CO is then diluted in the relatively large space of a typical home kitchen.
The same stove-top burner or oven, operating under the same conditions in a small
camper or motor home, results in higher concentrations of CO because of reduced
dilution and the decline of oxygen content in the surrounding feed-air supply.
These various combustion scenarios all produce CO concentrations, mass
amounts, and temporal exposure variations that may repeatedly alternate from harm-
less, to chronic exposure levels, to acute exposure levels. Because of this, exposures
may go unnoticed until a dangerous or even deadly situation occurs.
Labeling Implications: The combustion processes most often associated with

the production of CO can and do vary significantly due to small changes in
ambient or process conditions resulting in situations where little CO is produced
to situations where significant amounts of CO are produced.
8.2.3 BUOYANCY, MIGRATION, AND DILUTION OF CARBON

MONOXIDE
CO has a molecular weight close to that of air (CO = 28, Air = 28.9) and fol-
lowing the ideal gas laws would have approximately the same density as air across
8417: “8417_c008” — 2007/9/11 — 12:16 — page 200 — #4

the normal ambient temperature range. This would tend to make the gas neutrally
buoyant; neither wanting to migrate upward or downward due to density differences
when released into the environment. Therefore, if room temperature gas with CO
escaped from a calibration tank, it would not pool on the floor or raise to the ceiling.
Rather, its movement would be affected by the forces of the pressure drop at the leak,
concentration-driven diffusion, Brownian movement, and local air disturbances.3 [Of
course once mixing of the CO and air (oxygen, nitrogen, etc.) molecules occurs, sep-
aration of the molecules cannot take place once again without the addition of large
amounts of energy, lest it violate the Second Law of Thermodynamics.]
CO does not have any unusual adiabatic heat loss or gain characteristics. There-
fore, it does not have a vertical lapse rate that is different from air, and there are no
additional internal physical driving forces on it. Under local atmospheric “inversion”
conditions, ambient temperature CO would move similar to other neutrally buoyant
gases.
On the other hand, certain conditions can cause CO to move upward. CO recently
produced by a combustion process (e.g., in a residential furnace or boiler) is apt
to be warmer than the surrounding ambient air, and therefore will be thermally
buoyant and migrate upward, at least initially. For example, a few inches from the
combustion zone of propane heaters, gases are typically several hundred degrees
Fahrenheit above ambient temperatures. The extent of this temperature-driven dens-
ity difference and accompanying upward migration is controlled by the ability
of the exhausting plume to maintain the boundary of its thermal envelope. Fur-
ther, mechanisms like cross air movement and heat sinks cause a loss of heat and
buoyancy. These air movements and heat sinks can quickly inhibit further vertical
movement.
Certain conditions can also cause CO to migrate large distances. For example,
the continual release of hot exhaust gases that contain CO (e.g., a leak in a base-
ment furnace) produce their own convection currents that can effectively fumigate
distant areas. A furnace located in a cooler portion of a building can release products
of combustion—including CO—that move along the colder building cavities and
migrate several floors above the original leak. This phenomenon is enabled by
nonthermal volumetric expansions of simple gaseous carbon fuels due to changes
in pressure (e.g., from a pressurized cylinder), and by the increased number of
moles of gas in the by-products. The basic combustion equations for methane, eth-
ane, propane, and butane are given in Table 8.1. Comparing the volume (or moles
of gas) in the products of combustion (VP) to the volume of the reactants (VR),
reveals a positive volumetric change for all but methane during the combustion
process.
Dilution of airborne gaseous contaminants typically follows an exponential decay
profile. Ten thousand (10,000) cu. ft. of gas containing 500 ppm would require another
10,000 cu. ft. of clean dilution air to reduce the concentration to 250 ppm. It would
take another 20,000 cu. ft. of dilution air (10, 000 + 20, 000 = 30, 000 cu. ft.) to
reduce the concentration to 125 ppm; and a total dilution of 90,000 cu. ft. of dilution
air (100,000 − initial 10,000 cu. ft. = 90,000 cu. ft.) would be required to reduce the
level to the Occupational Health and Safety Administration (OSHA) 50 ppm PEL.
Hence, a short-term release of 10,000 cu. ft. of exhaust from a furnace containing
500 ppm CO would need to be diluted into 90,000 cu. ft. along its path before it is
8417: “8417_c008” — 2007/9/11 — 12:16 — page 201 — #5

TABLE 8.1
Volumetric Expansion Due to Reaction and Released Heat
from Reaction
Ideal combustion Heat released
Gas equation VP/VR (BTU/ft3 average)
Methane CH2 + 2O2 = CO2 + 2H2 O 3/3 1100
Ethane CH2 H6 + 3.5O2 = 2CO2 + 3H2 O 5/4.5
Propane C3 H8 + 5O2 = 3CO2 + 4H2 O 7/6 2500
Butane C4 H10 + 6.5O2 = 4CO2 + 5H2 O 9/7.5 3200
500
400
CO
(ppm) 300
200
100
20,000 40,000 60,000 80,000 100,000

Total Volume of Air to Dilute Concentration
FIGURE 8.1 Effect of dilution on concentration.
reduced to 50 ppm. This exponential dilution decay mechanism allows for significant
migration before the gas is rendered harmless (Figure 8.1).
Labeling Implications: On the basis of molecular weight, CO is neutrally buoy-

ant. It is thus more likely to move from its originating combustion process based
on its thermal buoyancy. This buoyancy is more profound under conditions of
greater temperature differential and with prolonged and continued releases com-
pared to smaller “puff” releases. There are additional forces driving volumetric
changes resulting from the combustion process. Owing to these characteristics,
the contaminant moves in significantly variable ways depending on seasonal
and ambient conditions.
8.2.4 EXIT VELOCITY CONSIDERATIONS FOR DIFFERENT CARBON

MONOXIDE PRODUCING EQUIPMENT
Some mechanisms that release CO have attached exhaust systems, so the exiting gases
may have considerable range in their volumetric flow rates and initial exit velocities.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 202 — #6

The role that this range in exit velocity plays in the dispersion and distribution of
the exhaust gas can be deceptive. For example, consider an engine’s exhaust system
(seen in cars, construction vehicles, small gasoline-powered engines). These exhaust
systems often accommodate at least an order of magnitude change in volumetric
exhaust flow across their normal operation range. Under low speed operation, the
volume of gases, and their ability to be propelled to a great distance from their exit
point, may be extremely low. At high speeds, the velocity at the point of exit may be
significantly greater. The general rule of thumb is that at approximately 20 diameters
away from the exit point, the horizontal velocity has dropped to 10% of its initial
value.4 Hence, it is a myth that the exhaust plume from most conventional engines
can be propelled great distances backwards from their initial release point.
In many combustion processes that have dedicated exhaust systems, there are rel-
atively small changes in volumetric exhaust flow during normal operation. Furnaces,
hot water heaters, and boilers are examples of this category of equipment, and they
are typically operated at a constant firing rate, with burners either on or off. Thermal
efficiency may be a consideration in the purchase of these devices, and therefore they
are designed to achieve low exhaust gas temperatures, and incorporate large diameter
exhaust systems to lower the resistance to exhaust flow, and sometimes utilize flue
dampers5 to conserve heat during the “off” portion of the operation. Owing to the rel-
atively small, thermally induced pressure differential in these systems, it is common
to see some back flow and/or spillage of combustion products into the area where
this equipment is located, at least under initial start-up conditions.6 The duration,
cause, and consistency of this is highly dependent on the maintenance of the system,
ambient conditions, and weather conditions. The footprint of this gas “spillage” is a
sooty residue on air supply vents in residential furnaces and hot water heaters. While
a small amount of “spillage” may be acceptable, consistent spillages can result in
chronic exposure conditions.
Some combustion sources do not have exhaust systems with well-defined exit
velocities (e.g., stoves, welding torches, portable heaters). These devices exhaust
directly into the area they operate in. They depend on their small size, clean exhaust,
or area ventilation for the removal and dilution of the CO that they may produce. This
is problematic when combustion is high in CO content, or when proper ventilation
is not present. There have been numerous situations in which industrial welders or
torch operators have been overexposed because they were working in a location with
a poor ventilation system. Further, campers have been injured or killed because they
used a portable heater or other combustion process in a tent that they thought was
naturally ventilated.
Labeling Implications: Automobile or other engine-powered equipment do not

project their exhausts a significant distance horizontally due to high exit velo-
cities; thermal effects are probably a greater driving force for movement of
the CO. Installations with low forced exit velocities and low thermal content
depend on their mechanical state of repair and ambient conditions for removal
of the products of combustion. For CO removal, systems without dedicated
exhaust systems rely heavily on the user, and on factors associated with their
location.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 203 — #7

8.2.5 ABSORPTION AND DISSIPATION OF CARBON MONOXIDE

CO is preferentially absorbed in comparison to oxygen. CO absorption and pur-
ging times (body dissipation time) from the blood system are a function of rate of
respiration, and the oxygen content in the subsequent breathing air.
The respiratory system is the only significant pathway for excretion of CO. The
removal of the CO-exposed person and the introduction of oxygen-enriched breathing
air are the most common means of removing CO from the body. The intravenous
addition of blood can dilute the total amount of CO held within the body, but there
are limits to the quantity of blood that can be added.
Absorbed mass or body burden is a function of the rate of uptake and the size of
the person. This means that large people generally take up a larger amount of CO
than people with smaller body mass. Individuals who have a higher respiratory (i.e.,
ventilatory) rate will also take up CO faster. People with compromised respiratory
systems, decreased levels of hemoglobin, or circulatory system problems are more
at risk.
Accelerated rate of absorption due to higher CO concentrations during exposure
may compress the normal display of symptoms, whereas slower or chronic exposure
may produce progressive steps in the symptomology. Symptoms may include flu-
like conditions, nausea, headache, sleepiness, and so forth followed by confusion,
decreased cognitive capability, and unconsciousness. Variants in the symptoms of
CO and variations among individuals can result in (1) CO exposures that go undia-
gnosed, (2) chronic exposures that cause long-term health problems, or (3) loss of
consciousness and death.
Safety and accident history information indicate that a significant percentage
of total CO nonworkplace deaths occur through the use of camping and heating
equipment that is used under poorly vented conditions, and from malfunctioning
furnaces and water heaters. Both groups could benefit from warnings.
8.3 COMMON SOURCES OF CARBON MONOXIDE

CO is associated with the following sources and locations:
8.3.1 CONSUMER PRODUCTS

• Automobile and truck internal combustion engines (typically gasoline-
fueled types)
• Cigarettes and side stream smoke
• Combustion of charcoal/gas grills (typically propane)
• Camp lanterns/stoves/gas or liquid fuel
• Fire places, chimneys, flues, flue damper
• Gasoline-powered engine implements:
Lawn mowers
Snow blowers
High-pressure washers
8417: “8417_c008” — 2007/9/11 — 12:16 — page 204 — #8

Power generators
Air compressors
Chain saws
• Hot water heaters (gas or oil)
• Kitchen stove burners and ovens (gas and propane fired)
• Motor boats, house boats
• Portable and stationary gas, liquid, and solid-fueled comfort and camp
heaters
• Residential furnaces (gas or oil)
• Wood burning furnaces and heaters
8.3.2 INDUSTRIAL PRODUCTS AND PROCESSES

• Blast furnaces
• Boiler operations (typically gas or oil)
• Calibration gases
• Cement kilns
• Chemical manufacturing (e.g., methanol, formaldehyde, ethylene)
• Coke ovens
• Construction site combustion space heaters
• Fuel gas generators (e.g., water gas, coal gas, producer gas)
• Gasoline engine powered tools and equipment
• Heat treating ovens (typically gas or oil)
• Industrial furnaces and space heating furnaces (gas, liquid, solid fuel)
• Industrial trucks and vehicles (typically gasoline or propane)
• Inert atmosphere generators
• Inhaling methylene chloride vapors (metabolizes to CO)
• Material handling/receiving docks (industrial trucks)
• Meat packaging additive
• Metallic ore refining
• Sewers and other underground structures
• Welding and torching operations
• Production of reducing oxides
This list demonstrates the ubiquitous consumer and industrial sources of CO. His-
tory has documented the numerous deaths and injuries associated with unexpected
exposures. This situation has prompted the codification of methods to provide warn-
ings and to investigate such hazards. Subsequent sections detail warning information;
and an example of this investigation codification in contained in ASTM E2292-03
Standard Practice for Investigating Carbon Monoxide Poisoning Incidents. The scope
of that standard states
1. Scope
1.1. This practice covers guidelines for collecting and preserving informa-
tion and physical evidence related to incidents involving the poisoning
of individuals by carbon monoxide.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 205 — #9

8.4 WHAT ARE WARNINGS, LABELS, INSTRUCTIONS,

AND OTHER FORMS OF COMMUNICATIONS?
In the context of this chapter, the phrase “warnings, labeling, instructions, and other
forms of communications” (e.g., safety information) refers to those textual messages,
words, pictures, alarms, and figures that are intended to eliminate or mitigate possible
exposure to CO by alerting the person at risk, informing the person at risk, or remind-
ing the person at risk of possible dangers. This last function of warnings to remind is
especially important in situations where use of the equipment is highly intermittent,
as may occur with seasonal camping equipment; or under circumstances where the
manifestation of the hazard of CO may be transient.
These warning items may appear as a label or tag directly on the product. They may
be a part of the products’ packaging that includes depictions of the use of the product,
or part of an instructional insert in the packaging. They may be a workplace sign or
alarm that indicates when elevated levels of CO are detected. Warnings can be found
in manuals that accompany products, or in locations where the products are used.
There are also brochures compiled by CO hazard awareness advocate organizations.
Many of these warning materials are now also available on the Internet or as inserts
in promotional literature.
This safety information is based on philosophical issues, safety research, vol-
untary, or regulatory requirements, the custom and practice of a given industry, or
is litigation driven. Regardless of the forces at play, safety efforts have manifested
themselves in a broad variety of warning schemes and warning materials.
8.5 THE CONTENT, CONFIGURATION, AND DESIGN

OF LABELS, WARNINGS, AND INSTRUCTIONAL
MATERIALS
The term “label” is generally associated with writing or graphics that are attached to
a product. Warnings may be a part of that “attached” label. They can also be included
in other locations such as on the product packaging or on inserts provided within
the packaging. The word warning itself is used as a specific “signal word” within a
properly designed warning scheme. Warnings or safety information may also be found
in certain locations rather than “on” a product. It is normal to see various warnings
in the instructions, on packaging, and on doors and in areas of industrial facilities.
Complicated safety instructions may be difficult to apply to an individual machine
or use in an industrial setting, whereas some consumer products will have fairly
detailed safety-related instructions. Instructions may also be a stand-alone item that
contains both operational instructions and safety information. Usually when there are
warnings that are difficult to actually apply to the product (e.g., or will not survive
the life or usage of the product), there will be an instruction on the product to consult
or “read the instructions” before operating or performing certain functions. On some
products, there may even be a pouch or a compartment so that one can store the safety
instructions along with the product.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 206 — #10

Arecent development on some chemical containers for lawn, garden, and pesticide
products is a “fold out” label that provides additional space for written information.
8.5.1 DESIGN OF LABELING AND WARNINGS

Although there is some latitude about the configuration and content of labeling, certain
designs have become generic. Typically, the signal words of “Notice,” “Caution,”
“Warning,” and “Danger” and the associated colors of blue, yellow, orange, and red
are progressive with increasing degrees of danger (Figure 8.2).
Often a triangular shape with an exclamation point or lightening bolt in its center
is used to denote that the sign is related to safety concerns (Figure 8.3). These are
called alert symbols.
Typically, product warnings have well-defined borders or textual boxes contain-
ing the different elements of the warning. Figure 8.4 identifies 8 of the common
elements of warnings based on a review of 11 publications addressing the design and
configuration of warnings as tabulated in Table 8.2.
Several American National Standards Institute (ANSI) standards provide
guidelines on the minimum size of font height for warning signs, with 0.08 in. height
being the minimum at a viewing distance of 1 ft. or less, 0.16 in. at 5 ft., and 0.22
in. at 8 ft. Konz,4 Woodson,7 McCormick and Sanders,8 and other authors on human
factors provide considerations on the display of information with respect to letter size
and readability, reading speeds, configurations, and conditions that may lead to errors
in reading content. Table 8.2 compares the contents of 11 publications9−19 on 12 items
related to the contents of warnings and shows a consistency across these items.
Your Your Your Your

Pictorial information Pictorial information Pictorial information Pictorial information
here here here here
FIGURE 8.2 ANSI 535.2 type labels.
FIGURE 8.3 Typical alert symbols ANSI, ASAE, and SAE.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 207 — #11

Signal word
Alert symbol
DANGER Symbols &

pictograph
Standard
border
Hazard
identification
Result of
ignoring
warning How to avoid
hazard
Additional
information
FIGURE 8.4 General vertical layout for warming label.
TABLE 8.2
Documents (A–K) Addressing Labeling/Warning Content Items (1–12)
Content items A9 B10 C11 D12 E13 F14 G15 H16 I17 J18 K19
1. Hazard alert/symbols x x x x x x x
2. Signal word x x x x x x x x
3. Color(s) x x x x x x x x
4. Symbols and pictures x x x x x x x x x x x
5. Hazard identification x x x x x x x x
6. Result of ignoring hazard x x x x x x x
7. How to avoid hazard x x x x x x
8. Location of label x x x x x x
9. Font size x x x x x x x x
10. Contents disposal x x
11. Material/durability x x x x
12. Fire/handling/Ref. x
One other labeling approach included in the ANSI standards is ANSI A13.120 for
piping systems. It includes a color scheme, arrows to indicate the direction of flow, a
signal word to identify the chemical content, and font size requirements.
In a section entitled “Legal/Litigation Driven” there is an additional discussion
of the contents of warnings/labels based on court decisions and legal interpretation.
There are other places and products where specific warning designs are required
by regulation and may not follow the criteria referred to above. The most ubiquitous
of these is one of the warnings on cigarette packages (Figure 8.5).
This particular warning is required by specific legislation,21 and does not conform
to the criteria often found in the standards and guidelines for CO warnings.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 208 — #12

“SURGEONS GENERAL’S WARNING: Cigarette Smoke

Contains Carbon Monoxide”
FIGURE 8.5 U.S. Government required CO warning on cigarette packages.
8.5.2 WRITTEN INSTRUCTIONAL INFORMATION ON SAFETY

MATERIALS
Safety instructions may require more space than is available on a label. In these
cases instructions are described in a separate document. It is common to see warnings
within manuals and other instructional materials that have the general appearance of
an “on-product” warning, but are instead found in the accompanying materials. An
example of a CO warning that is found with a product manual is shown in Figure 8.6.
These messages are either bundled together into one section of the product liter-
ature or distributed in various sections of the literature. The Handbook of Technical
Writing22 contains guidelines for the writing style that may be used in compos-
ing instructional materials. Writing and Designing Operator Manuals15 provides
strategies for the composition and detail of safety manuals. The content found in
these publications is the basis for instructional warnings about CO.
Various gasoline engine-powered equipment have warnings in their instructional
manuals about CO. Exemplar of this is the American Honda Motor Company Owner’s
Manual23 for a line of small displacement engines that are used on lawn mowers,
small portable generators, power washers, air compressors, and snow throwers. Their
manual starts out with a brief discussion of what “Safety Messages” mean and how
to identify them by a safety symbol, a signal word, and the box-like border around
them. The manual uses the word “Warning” and the alert symbol and addresses the
carbon monoxide hazard by stating
Carbon monoxide gas is toxic. Breathing it can cause

unconsciousness and even kill you. Avoid any areas or
actions that expose you to Carbon monoxide.
This manual also contains a text under the title “Safety Information” that states
Your engine’s exhaust contains poisonous carbon monoxide. Do

not run the engine without adequate ventilation, and never run the
engine indoors.
Under “Maintenance Safety,” there are additional “Safety Precautions” that state
Carbon monoxide poisoning from engine exhaust.

Be sure there is adequate ventilation whenever you operate the engine
(It should be noted that the exhaust emissions are met in part by the use of a catalytic
converter system.)
8417: “8417_c008” — 2007/9/11 — 12:16 — page 209 — #13

_____________________________________________________________________________________________
Instruction Manual: COSTAR® P-1, 9V Personal CO Detector QGI P/N 099-0062-01 REV 07/25/01 Page 1 of 4
Personal Carbon Monoxide Detector Model P-1
OWNER’S MANUAL
PLEASE READ AND SAVE!
Personal & Automobile Use
Dear New COSTAR® Model P-1 Owner,
Congratulations as you have taken steps to help insure the health and life safety of you and your family. We
are proud to offer you our unique, patented CO Sensor technology that detects CO in a manner similar to the
human body's response. The COSTAR® P-1 is an ideal and low-cost way of warning you of both the
acute and chronic effects of CO poisoning.
Please read this owner's manual carefully so you will have a better understanding of the effects of CO
poisoning and the COSTAR® P-1 Detector, as we work together pursuing a safer, healthier air quality for us
all.
To your good health and safety,

Mark Goldstein, Ph.D.
President
Quantum Group Inc.
1.0 WHAT YOU SHOULD KNOW ABOUT CO
Carbon monoxide (CO) is an insidious poison. It is a colorless, odorless and tasteless gas. It is a
cumulative poison. Even low levels of CO have been shown to cause brain and other vital organ damage
in unborn infants with no effect on the mother. The following symptoms are related to CARBON
MONOXIDE POISONING and should be discussed with all members of the household:
• MILD EXPOSURE: Slight headache, nausea, vomiting, fatigue (often described as “flu-like”
symptoms), giddiness
• MEDIUM EXPOSURE: Severe throbbing headache, drowsiness, confusion, fast heart rate
• EXTREME EXPOSURE: Unconsciousness, convulsions, cardio-respiratory failure, death
Many reported cases of CARBON MONOXIDE POISONING indicate that while victims are aware
they are not well, they become so disoriented they are unable to save themselves by either exiting the
building/automobile or calling for assistance. Also, young children and household pets may be the first
affected.
Your CO detector is designed to detect the toxic CO fumes that result from incomplete combustion,
such as those emitted from appliances, furnaces, fireplaces and auto exhaust. (DO NOT put your
detector next to the vehicle’s tailpipe. It may damage the detector permanently.)
A CO detector is NOT A SUBSTITUTE for fire alarms, smoke alarms, or other combustible gas
alarms. This carbon monoxide detector is designed to detect carbon monoxide gas from ANY source of
combustion.
CAUTION: This detector will only indicate the presence of carbon monoxide gas at the sensor. Carbon
monoxide gas may be present in other areas at a higher concentration than at the detector’s location;
therefore, immediately get fresh air.
WARNING: This product is not designed to comply with the Occupational Safety and Health
Administration (OSHA) commercial or industrial standards. Individuals with medical problems may
consider using detection devices that provide audible and visual signals for CO concentrations under 30
PPM.
2.0 WHAT YOU SHOULD DO IF THE ALARM SOUNDS

WARNING:
Actuation of this device indicates the presence of carbon monoxide (CO)
which can KILL YOU.
_____________________________________________________________________
FIGURE 8.6 COSTAR® P-1, 9 V, personal CO detector.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 210 — #14

All of these statements are part of an effort to alert users of the potential release of
CO, its poisonous nature, the need to vacate exposed areas, and means of mitigating
exposure.
More common are CO warnings found on internal combustion engine-powered
cars and trucks. In the Hyundai 2003 Santa Fe Model Owner’s Manual,24 there is an
entire page dedicated to engine exhaust emissions, predominantly associated with CO.
That warning uses the alert symbol, the signal word “Warning,” and black lettering
on a yellow background to garner attention and states that
Exhaust fumes contain carbon monoxide, a colorless gas that

can cause unconsciousness and death by asphyxiation.
Additional information is provided about how to avoid this hazard, what to look
for in detecting this hazard, and other precautions.
8.5.4 WORKPLACE AREA WARNINGS AND LABELING

Several ANSI and American Society for Testing and Materials (ASTM) standards
provide warning and labeling criteria for workplaces or work-related areas. They
include
Environmental and Facility Safety Signs, ANSI Z 535.2, 1991;

Accident Prevention Signs, ANSI Z 535.5, 1998;
Criteria for Safety Signs, ANSI Z 535.3, 1991.
Standard Practice for Labeling Art Materials for Chronic Health Hazards,
ASTM D 4236
On November 18, 1990 the Labeling of Hazardous Art Materials Act (Public
Law 100-695) went into effect which embraces ASTM D 4236 (above) and requires
labeling in that industry. OSHA has both general and activity-specific information
about warning signs. Below are several examples.
8.5.4.1 Examples of General OSHA Warning Signage

• For construction,
29CFR 1915.16(b) Posting of large work areas. A warning sign or label
required by paragraph (a) of this section need not be posted at an individual
tank, compartment or work space within a work area if the entire work area
has been tested and certified: not safe for workers, not safe for hot work,
and if the sign or label to this effect is posted conspicuously at each means
of access to the work area.
• For maritime,
1926.200(a) General. Signs and symbols required by this subpart shall be
visible at all times when work is being performed, and shall be removed
or covered promptly when the hazards no longer exist.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 211 — #15

• For general industry,

1910.145(a) Scope.
1910.145(a)(1) These specifications apply to the design, application, and
use of signs or symbols (as included in paragraphs (c) through (e) of this
section) intended to indicate and, insofar as possible, to define specific haz-
ards of a nature such that failure to designate them may lead to accidental
injury to workers or the public, or both, or to property damage. These spe-
cifications are intended to cover all safety signs except those designed for
streets, highways, railroads, and marine regulations. These specifications
do not apply to plant bulletin boards or to safety posters.
8.5.4.2 Examples of Specific OSHA Exposure Warning Signage

• For asbestos activities,
1910.1001(j)(3) Warning signs.
1910.1001(j)(3)(i) Posting. Warning signs shall be provided and displayed
at each regulated area. In addition, warning signs shall be posted at all
approaches to regulated areas so that an employee may read the signs and
take necessary protective steps before entering the area.
1910.1001(j)(3)(ii) Sign specifications.
1910.1001(j)(3)(ii)(A) The warning signs required by paragraph (j)(3) of
this section shall bear the following information:
DANGER
ASBESTOS
CANCER AND LUNG DISEASE HAZARD
AUTHORIZED PERSONNEL ONLY
• For cadmium and cadmium containing compounds, in all forms,

1910.1027(m)(2) “Warning signs.”
1910.1027(m)(2)(i) Warning signs shall be provided and displayed in reg-
ulated areas. In addition, warning signs shall be posted at all approaches to
regulated areas so that an employee may read the signs and take necessary
protective steps before entering the area.
1910.1027(m)(2)(ii) Warning signs required by paragraph (m)(2)(i) of this
section shall bear the following information:
DANGER
CADMIUM
CANCER HAZARD
CAN CAUSE LUNG AND KIDNEY DISEASE
AUTHORIZED PERSONNEL ONLY
RESPIRATORS REQUIRED IN THIS AREA
8417: “8417_c008” — 2007/9/11 — 12:16 — page 212 — #16

1910.1027(m)(2)(iii) The employer shall assure that signs required by this

paragraph are illuminated, cleaned, and maintained as necessary so that
the legend is readily visible.
1910.1027(m)(3) “Warning labels.”
1910.1027(m)(3)(i) Shipping and storage containers containing cadmium,
cadmium compounds, or cadmium contaminated clothing, equipment,
waste, scrap, or debris shall bear appropriate warning labels, as specified
in paragraph (m)(3)(ii) of this section.
1910.1027(m)(3)(ii) The warning labels shall include at least the following
information:
DANGER
CONTAINS CADMIUM
CANCER HAZARD
AVOID CREATING DUST
CAN CAUSE LUNG AND KIDNEY DISEASE
1910.1027(m)(3)(iii) Where feasible, installed cadmium products shall

have a visible label or other indication that cadmium is present.
8.5.4.3 Other OSHA and Industry Warning Criteria

There is no OSHA standard specific to CO signage. However, by following the criteria
for general signage, hazard mitigation, and OSHA’s criteria for other chemicals,
OSHA can utilize the “General Duty Clause” (a.k.a. “5a Clause”) to issue citations
about inadequate CO warning signs.
Other locations and industries have their own guidelines. Within the iron produ-
cing and steel manufacturing industry, CO is knowingly produced in blast furnace
operations and in coke ovens. It is common in these industries to see warnings about
the possibility of a CO outbreak (from a blast furnace) and the location of safe areas
and evacuation routes. Since CO is produced in large quantities and at such elevated
concentrations, blast furnace work areas are often equipped with monitors that con-
tinuously check CO concentrations. Safe rooms, rescue respirators, and signs about
CO hazards are typical in these facilities.
The Department of Defense Hazardous Chemical Warning Labeling System25
provides detailed labeling requirements including a mechanism for “rating” chem-
icals. The format for these Department of Defense (DOD) labels is unique, in part
because of their use in wartime or under the stress of military conditions. They do not
use the alert symbol, but do use the same hierarchy of signal words. They use a small
selection of pictograms, including a skull and cross bones, which under their criteria
is appropriate for CO. The DOD Hazardous Chemical Warning Label follows criteria
set out in the 29 CFR 1910.1200 Hazard Communication Standard. It is often 8.5 by
11 in. in size, and contains information about specific categories of health, contact,
fire, and reactivity. Following this format, a label for CO is rated as moderately toxic,
severely flammable, and slightly reactive.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 213 — #17

8.6 THE REASONS, RATIONALE, AND SCIENTIFIC

BASES FOR PROVIDING SAFETY INFORMATION
FOR CARBON MONOXIDE WITH SELECTED
EXAMPLES
Many books and treatises have been written about the issues surrounding informing
consumers of risk through warnings and informational means. A brief discussion of
some of the bases for warnings, labeling, instructions, and safety communications is
given below.
8.6.1 ETHICAL AND PHILOSOPHICAL REASONS

The concept of “do no harm”26 is widely accepted as a moral and ethical obligation
and may be an underling element (although not a part of) the Hippocratic Oath. All
products, processes, and workplaces are capable of harm, and at a minimum people
should be informed of hidden or unexpected risks that cannot be eliminated. Much of
“being informed” comes from education, life experiences, and simple observation.
With the added complications of modern technology, keeping oneself informed of
product and workplace risks is difficult and in some cases impossible. From an ethical
perspective, when hazards cannot be eliminated or reasonably safe guarded, it is
necessary to provide information in the form of warnings, labeling, instructions, or
other communications, so as to “do no harm.”
One of the basic safety philosophies originating in part from the engineering and
safety professional codes of ethics includes the notion that designers of equipment,
products, and process have the three basic goals of (1) making a product/process that is
functional, (2) making a product/process that helps human welfare by being available
(and therefore is cost efficient), and (3) making a product/process that is reasonably
safe. In other words, products have to work, be inexpensive enough for availability to
the general public, and be reasonably safe.27 User information (warnings, labeling,
and instructional information) affects all three of the items above.
For most products and processes to function, there is a need for some directions and
information. For most combustion processes, instructions about how the equipment
is to function includes information that reduces the formation, accumulation, and
concentration of CO.
Instructional manuals, labels, and warnings all increase the cost of products. They
can also delay the introduction of products into the market place. However, they can
significantly increase consumer awareness about safety concerns. On complicated
pieces of equipment and processes, informational materials often require testing and
evaluations via “task analysis” to determine the effectiveness of the material before
a new product enters the market place. Gordon and Hall28 and others suggest that a
task analysis of a product and its accompanying documents should be a part of the
normal product development sequence.
In essence, warnings, labeling, instructional materials, and other forms of safety
information should not be thought of as pieces of paper or decals, but as part of a vital
scheme to do no harm.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 214 — #18

Within the safety community there is a general consensus that has developed
over the years concerning steps that should be taken to ensure safe products and safe
workplaces.29 The order to these incremental steps are
1. Eliminate the hazard

2. Apply safeguarding technologies to control the hazard
3. Warn potential users about the hazard and its features
4. Provide training and/or personal protective equipment to mitigate exposure
to the hazard
With CO, there is no real opportunity to “eliminate” the hazard as long as carbon-
based fuels are in use. In some circumstances it is possible to provide safeguarding
technologies. These include CO alarms, oxygen deficiency sensors (ODS),30 CO
protective respirators, self-checking flue dampers, temperature sensors positioned
to detect hot gas leaks (indicative of a leak of combustion gases), timers on space
heaters, and the use of catalysts to enhance certain types of combustion processes.
Providing information manifests itself in items (3) and (4). Unfortunately, this
approach does not take into account the importance of safe product design in ensuring
user safety. This factor is especially important in our technological world.
Figures 8.7 through 8.12 show several excerpts of CO warning criteria that are not
based on general codes or standards (though they may follow code criteria for their
specific content and form), but are based on a range of ethical and philosophical safety
criteria. Some of these warnings originate from an association or industry collective,
while others are for products that do not produce CO, but are associated with other
equipment that may produce CO. A diverse product mix is provided for instructional
purposes.
“Carbon Monoxide Hazards from

Small Gasoline Powered Engines Fact Sheet
HS05-023B (8-05)
“Tool Rental Agencies Should:

…• Put warning labels on gasoline-powered tools.
For example:
WARNING - CARBON MONOXIDE PRODUCED
DURING USE CAN KILL - DO NOT
USE INDOORS OR IN OTHER SHELTERED
AREAS.”
FIGURE 8.7 Warning information from Texas Department of Insurance Division of Workers’
Compensation (TDI/DWC), E-mail: resource.center@twcc.state.tx.u
8417: “8417_c008” — 2007/9/11 — 12:16 — page 215 — #19

“Important Alert for Gasfitters

The Office of Gas Safety (OGS) alerts gasfitters that all open-flued
instantaneous gas water heaters (IGWH) in toilets and bathrooms
should be replaced as a matter of urgency….
Open-flued IGWH’s, in a confined living space, can expose gas users

to the risk of carbon monoxide poisoning….
• Provided the Important Safety Warning including the dangers and
symptoms of carbon monoxide poisoning .”
FIGURE 8.8 Warning information from Office of Gas Safety Level 1, Wool House, 369
Royal Parade, Parkville, Victoria 3052.
Informational Brochure On CO Poisoning From Boating
“Apply the enclosed carbon monoxide warning

decals at the helm and by the swim deck as a
reminder. Additional carbon monoxide warning
decals may be obtained from any boating
enforcement officer.”
FIGURE 8.9 Warning from National Marine Manufacturers Association, 200 East Randolph
Drive, Suite 5100 Chicago, IL 60601-6528, Available at: www.nmma.org.
8.6.2 RESEARCH BASIS: THE NEED FOR SAFETY INFORMATION

Much research into the efficacy of safety information has been conducted over the
years. The findings are curiously polarized: safety information either helps or is not
universal in effectiveness. However, there is not any substantial body of research
finding that safety information hurts or causes a new harm.
The literature repeatedly comes down to a discussion of to what extent safety
measures help. Somewhere on this continuum, between safety measures being mar-
ginal to them being extremely useful, there is a middle ground wherein safety efforts
help sufficiently and are reasonable in their design and application. A brief review
of the safety literature over the past 25 years indicates that safety efforts are direc-
ted at the performance and efficacy of warnings, and that whether or not warnings
should or should not be used is a less debated consideration. For example, in the
publication entitled “Human Factors Perspectives on Warnings, 1980 to 1993,”31 the
terms “effectiveness” or “effect(s)” appears in the title of listed papers and abstracts
at least 27% of the time. These researchers are attempting to establish the threshold
of effectiveness in warnings.
In volume 2 of the same publication,32 covering the period 1994–2000, the
research continues in that vein, with the added inquiries as to what aspects of a warn-
ing improves the effectiveness of the warning. Research into the colors by Braun et
al.,33,34 symbols by Caird et al.,35 risk perception and warning dilution from multiple
warnings by Chen et al.,36 legibility by Ringseis and Caird,37 and effects of age by
Hancock et al.38 on word content are all examples of the direction the most recent
8417: “8417_c008” — 2007/9/11 — 12:16 — page 216 — #20

1.1.SAFETY ALERT SYMBOLS AND SIGNAL WORDS
1.2.14.1. Generator
If your trailer is equipped with a gasoline or diesel generator, you must
have and follow the generator manufacturer's instructions. You must also
have one or more carbon monoxide detectors in the trailer's
accommodation spaces.
Carbon Monoxide is an odorless gas that can cause death. Be certain exhaust from a running
generator does not accumulate in or around your trailer, by situations such as :
Being drawn in by fans or ventilators operated in a trailer

;
Prevailing wind;
Being trapped between your trailer and other trailers, vehicles or
buildings; or
Being trapped between your trailer and, or in a snow bank, or other
nearby objects.
Operating gasoline and diesel generators can lead to

death or serious injury by:
Carbon Monoxide
Fire and Explosion
Electrocution
Have a working carbon monoxide detector in th e

accommodation spaces before operating a
generator.
Do not refuel a running generator or refuel near
ignition sources.
1.2.14.3. LP Gas Fuel System
You can die or be brain damaged by Carbon

Monoxide.
Make certain the exhaust from LP appliances is
directed to the outdoors.
Have a working carbon monoxide detector in th e
accommodation spaces of your trailer before
operating any LP gas appliance.
Do not operate portable grills or stoves inside th e
trailer.
FIGURE 8.10 Brockmann Trailers (a trailer manufacturer) (http://www.bockmann.co.uk/

supporta.asp), Lastrup, Northern Germany. Available at: http://www.bockmann.co.uk/
abouta.asp.
research has taken. None of these authors reports findings that product or process
warnings are useless.
Other research includes the role of “on-product” warnings, the interplay between
warnings and training, and the relative effectiveness of verbal and nonverbal warn-
ings. Examples of nonverbal warnings are included in Analysis of Workplace Factors
on Auditory Warning Alarm Location by Nanthavanij.39 CO alarms and oxygen
8417: “8417_c008” — 2007/9/11 — 12:16 — page 217 — #21

!WARNING
Failure to properly vent the water heater to the outdoors

as outlined above and in the following section can result in unsafe
operation of the water heater causing bodily injury, explosion, fire, or
death. To avoid the risk of fire, explosion, or asphyxiation from
carbon monoxide, NEVER operate this water heater unless it is
properly vented and has an adequate air supply for proper operation.
FIGURE 8.11 Rheem, water heater manufacturer power vent residential gas water heater,
use and care, AP10960-11 (04/02) manual (www.rheem.com/documents/resourcelibrary/use/
and care/rheempowervent/ap10960-11.pdf).
“Carbon Monoxide is a product of the reaction of H2 and CO2.
Carbon monoxide (CO) is a colorless, odourless, tasteless and flammable gas

which is acutely toxic. CO is introduced into the blood stream through the lungs
and binds with the hemoglobin preventing it from carrying oxygen around the
body. This can result in rapid damage to body tissues due to oxygen starvation.
Since CO is an accumulating toxin it can be dangerous even when present in quite
low concentrations over long periods of time.” ( note: triangular alert symbol is
used)
FIGURE 8.12 Carbolite “coal ash fusion furnace,” installation, operation, maintenance
instructions (industrial furnace), (industrial furnace) Hope Valley, S33 6RB, England CAF
16/38, MF45-1.0, Copyright 2002, 11/14/02.
depletion systems are prime examples of nonverbal warnings/alerting systems for

exposure to CO.
Some of the older research concerning the warning requirements for CO are
contained in the Encyclopedia of Chemical Labeling.14 This book classifies CO
as a chemical requiring identification as “extremely hazardous,” ”vapor extremely
hazardous,” “poisonous if inhaled,” and a “poison.”
The National Institute for Occupational Safety and Health (NIOSH) suggested in
1972 that CO cylinders and other containers be labeled “fatal if inhaled” and “do not
breath gas.”40 For areas where CO gas may be present, they recommended signage
stating that “high concentrations may be fatal.”
In 1974, the American Insurance Association referenced the research work of the
Consumer Product Safety Commission (CPSC) and concluded that,
The Consumer Product Safety Commission (CPSC) has repeatedly warned manu-
facturers of equipment which may produce carbon monoxide to label the equipment
so as to inform the general public of the toxic hazards if improperly vented or used.41
Further, in 1971 through 1985, the Environmental Protection Agency (EPA) had
a variety of research-based rules42 concerning ambient air quality. These included
rules concerning CO content and how to alert or warn the public if they are at risk
8417: “8417_c008” — 2007/9/11 — 12:16 — page 218 — #22

of overexposure. The 1986 40 CFR Ch. 1, Par. 51.16 “Prevention of air pollution
emergency episodes” required
“Communication procedures for transmitting status reports [on carbon monoxide] for
contact with the public officials, major emissions sources, public health, safety, and
emergency agencies and news media”43
to warn the public of air pollution episodes.
8.6.3 VOLUNTARY STANDARDS FOR CO WARNINGS (EXAMPLES IN

CHRONOLOGICAL ORDER)
Numerous nongovernmental codes, standards, and guidelines recognize the need to
inform workers and consumers of the hazards of CO. Figures 8.13 through 8.20 show
some examples for different time periods, locations, and products.
8.6.4 MANDATORY SAFETY INFORMATION REGULATIONS

Regulations regarding product warnings and labeling are based in part on our social
value system. Our society has allowed for the formation of various regulatory bodies,
and it is generally through these regulatory agencies that mandatory requirements
“Carbon Monoxide
DANGER! EXTREMELY FLAMMABLE
GAS UNDER PRESSURE
MAY BE FATAL IF INHALED
Avoid breathing gas.
Keep container closed.
Use with adequate ventilation.
Keep away from heat, sparks and open flames.
Never drop cylinder.
Keep cylinder out of sun and away from heat.”
FIGURE 8.13 Manufacturing Chemist Association (MCA) “Guideline for Chemical

Labeling.”
“NOTE: The 0.01% level of CO in indicated to protect the test personnel…”
FIGURE 8.14 Society of Automotive Engineers (SAE) “Test Procedure for Measuring Car-
bon Monoxide Concentrations in Vehicle Passenger Compartments,” SAE J989, Approved
1968, 1971 Handbook.
“Para:8.4.(b) WARNING: during an extended hot idle period of up to 60

minutes, idle exhaust CO from the test or traffic simulation vehicle will
probably remain constant.”
FIGURE 8.15 Society of Automotive Engineers (SAE) “Test Procedure for Measuring Car-
bon Monoxide Concentrations in Vehicle Passenger Compartments,” SAE J989a, Revised
1978, 1985 Handbook.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 219 — #23

“1.34.7 INSTALLATION MARKINGS…

(c)… This appliance must be properly connected to a venting system.
This appliance is equipped with a vest safety shutoff system.
WARNING: Operation of this wall furnace when connected to a properly
installed and maintained venting system or tampering with the vent safety
shutoff system can resultin carbon monoxide (CO) poisoning and possible
death.”
FIGURE 8.16 ANSI Z 21.49—1989 “Gas Fired Gravity and Fan Type Vented Wall Furnaces.”
“ Operation and Installation Instructions:

21.1J
The word “WARNING”, and the following or equivalent text:
“TO REDUCE THE RISK OF CARBON MONOXIDE POISONING,
TEST DETECTOR OPERATION IF NOT IN USE FOR 10 DAYS OR
MORE.
(K) The word “WARNING”, and the following equivalent text:
“This device may not alarm at low carbon monoxide levels. The
Occupational Safety and Health Administration (OSHA) has established that
continuous exposure to levels of 35 ppm should not be exceeded in an eight
hour period”
FIGURE 8.17 UL 1524 “Standard for Safety—Carbon Monoxide Gas Detectors for Marine
Use,” 1989.
WARNING!
Do not use equipment and tools powered by gasoline
engines inside buildings or other partially enclosed
spaces unless the gasoline engine can be placed
outdoors and away from air intakes
Tool rental Agencies Should:

• Place warning labels on gasoline-engine-powered tools. For
example:
WARNING – CARBON MONOXIDE PRODUCED DURING USE

CAN KILL – DO NOT USE INDOORS OR IN OTHER
SHELTERED AREAS.
FIGURE 8.18 National Institute for Occupational Safety and Health, NIOSH ALERT:1996,
Pub. No. 96-118 “Preventing Carbon Monoxide Poisoning from Small Gasoline-Powered
Engines and Tools.”
for product safety information are created. These efforts are a compromise between
no regulation at all, or banning products that cannot be made safe without such
information.44 Government regulatory agencies have to balance public demand,
business interests, and political intrigue, with a well-established scientific basis for
implementing safety standards.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 220 — #24

All unvented Servel Refrigerators shall have a warning label, protected from the
environment, that is either affixed to the outside of the front door or inside the
Un-Vented Servel Refrigerator in a location that is readily visible. The warning
label shall have the following wording: Warning This refrigerator is prone to the
production of Carbon Monoxide in levels that may be lethal. This refrigerator
may only be operated in an area that is isolated from the living space such as: a
remote shed, garage or open porch. The refrigerator shall be located a minimum
distance of 12 inches from any opening to the living space
FIGURE 8.19 Technical Standards and Safety Authority, 14th Floor, Certre Tower, Toronto,
Ontario, CAN, Ref. No FS-076-06, March, 2006, Director’s Order, Fuels Safety Program,
Roland Hadaller, Director.
Using a generator indoors WILL KILL LYOU IN

MINUTES.
Exhaust contains carbon monoxide, a poison gas
you cannot see or smell.
NEVER use in the home ONLY use outdoors and

or in partly enclosed far from open windows,
areas such as garages. doors, and vents.
FIGURE 8.20 UL offered CO warning label (www.ul.com/media/newsrel/generatorCO-

Marking.pdf).
Scientific research into chemical hazards often has implications for the develop-
ment of mandatory safety regulations. For example, Viscusi44 reports that “the Food
and Drug Administration’s (FDA’s) predominant risk standard for carcinogens is a
lifetime risk of 1/1,000,000. This is ten times as conservative as the 1/100,000 lifetime
risk embodied in California’s Proposition 65 (which mandated hazard warnings for
risks of cancer and reproductive toxicity from products, jobs, and the environment).”
Determining the accuracy of this risk assessment is difficult since there are no public
death registries for CO, whereas there are well-established registries and protocols
for cancer deaths (e.g., Centers for Disease Control, National Program for Cancer
Registries). Typically, manufacturers have an idea concerning the number of deaths
from CO related to their products. However, this information is generally kept at high
levels of confidentiality.
Burd (2002)45 reports in his research of a single county (Bexar County, Texas,
population 1.4 million, 2000 census) over a six decade time period (1935–1995) that
for the past 5 years (1991–1995) the death rate from CO was 40 deaths per 1,000,000
8417: “8417_c008” — 2007/9/11 — 12:16 — page 221 — #25

residents. Although a direct comparison to the FDA or the California numerical risk
criteria noted above is not perfectly correct, there is a close correspondence.
The CPSC has also investigated deaths from CO. In Non-fire Carbon Monox-
ide Deaths Associated with the use of Consumer Products 2001 Annual Estimates
(www.cpsc.gov/LIBRARY/coed04.pdf) the CPSC reports that:
In 2001 there were an estimated 130 unintentional non-fire CO poisoning deaths asso-
ciated with consumer products under the jurisdiction of the U. S. Consumer Product
Safety Commission. Fifty-eight percent of the estimated deaths in 2001 were associated
with the use of heating systems. Natural gas heating accounted for 37 percent and lique-
fied petroleum (LP) gas heating accounted for 35% of the estimated heating deaths. An
estimated 18% of the 2001 annual CO poisoning deaths were associated with engine-
powered tools, nine percent were associated with charcoal grills, eight percent were
associated with gas ranges and ovens, one percent were associated with camp stoves
and lanterns, and seven percent were associated with other or multiple appliances.46
Meanwhile, in the CPSC’S “Incidents, Deaths, and in-Depth Investiga-

tions Associated with Carbon Monoxide from Engine-Driven Generators and
Other Engine-Driven Tools, 1990–2004” (December 1, 2005, Natalie E. Marcy,
www.cpsc.gov/LIBRARY/coed05.pdf) there are reportedly 317 incidents of pois-
oning and 318 deaths from exposure. Both these reports indicate a significant death
rate from CO, and are the basis for some of the mandatory regulations concerning
warnings and labeling.
In 1993, the OSHA published their “Final Rule” on confined spaces:
Entry Permits: The entry permit.. shall identify: (7) The hazards
of the permit space to be entered.
Applying this criteria to a space that may contain CO requires postings about the
hazards of carbon monoxide.
In 1994, OSHA implemented mandatory rules about chemical hazards, primarily
directed at workers but that have labeling implications for both worker and consumer
populations. The “Hazard Communication” provision of 29 CFR 1910.1200 requires
specific information be provided on “all chemicals produced or imported” into the
country. This section specifically states that:
label means any written, printed, or graphic material displayed on or

affixed to containers of hazardous chemicals.
Section (f) contains detailed labeling requirements addressing the use of:
appropriate hazard warnings, or alternatives, words, pictures, symbols,

or combination thereof …regarding the hazards of the chemical …
Applying this criteria to containers of CO, there would be information about the
highly toxic nature of exposure, the need to take appropriate protection measures,
and the necessity of ceasing contact with the gas.
Figures 8.21 through 8.26 show examples of the current federal, state, and local
regulations concerning safety information for the hazard of CO.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 222 — #26

Tennessee OSHA has instituted a Special Emphasis Program for Carbon Monoxide (CO).
This is established to focus state-wide attention on carbon monoxide and to reduce
employee exposure to, and eliminate deaths from, carbon monoxide.
Carbon Monoxide Warning Sign
WARNING
CARBON MONOXIDE
A colorless, odorless, toxic gas is produced from incomplete
combustion of gas, oil, kerosene, and wood.
May cause dizziness, nausea, or headache
Excessive exposure may cause unconsciousness and death
May aggravate heart and artery diseases
FIGURE 8.21 State of Tennessee OSHA F:\TN Department of Labor and Workforce
Development.htm “Special Emphasis Program—Carbon Monoxide, The Hidden Killer.”
Mandatory Carbon monoxide Warning Decals Now Available
Contact: June lljana (916) 263-0788 e-mail: pubinfo.dbw.ca.gov
April 29, 2005
Sacramento – As of May 1, California boaters will be required to place stickers

on their newly purchased boats warning against the threat of carbon
monoxide poisoning
The Farr and Stacet Beckett Boating safety Act of 2004 requires that a
set of carbon monoxide warning stickers be placed on the transom and
helm of all new and used motorized boats sold in California. The bill,
AB 2222 (Koretz), was signed by Governor Schwarzenegger in
September.
FIGURE 8.22 State of California, Department of Boating and Waterways, F:\Mandatory

Carbon Monoxide Warning Decals Now Available.htm.
8.6.4.1 State-Based Examples

Figures 8.21 through 8.23 show state-based examples.
8.6.4.2 Federal Government Examples

Figures 8.24 through 8.26 show Federal Government examples.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 223 — #27

NECESSITY, FUNCTION, AND CONFORMITY: KRS 217.690 authorizes the Cabinet for
Human Resources to adopt administrative regulations to regulate the control of hazardous substances in
Kentucky. The purpose of this administrative regulation is to provide uniform standards relating to the
"conspicuousness" of labeling requirements; to specify requirements to identify hazardous substances
that present special hazards and require specialized labeling to protect the public health; and to prevent
the deceptive use of disclaimers on labelsof hazardous substances.Section 1.
Conspicuousness of Labeling Requirements.

(1) The signal word, the statement of the principal hazard or hazards, and instructions to read carefully
any cautionary information that may be placed elsewhere on the label shall appear together on the main
panel of the label. The information shall be placed together and distinctively apart from other wording or
designs. The necessary prominence shall be achieved by placement within the borders of a square or
rectangle with or without a borderline, and by use of suitable contrasts with the background achieved by
distinctive typography or color, and by both color and typography if needed.
Section 2. Special Labeling Requirements. In addition to the requirements of KRS 217.670 the following
hazardous substances are deemed to be misbranded unless the label includes the requirements stated
below: (1) Charcoal briquettes and other forms of charcoal for cooking or heating. Because inhalation
of the carbon monoxide produced by burning charcoal indoors or in confined areas may cause serious
injury or death, containers of the products shall bear the following borderlined statements: "WARNING;
Do Not Use for Indoor Heating or Cooking Unless Ventilation is Provided for Exhausting Fumes to
Outside. Toxic Fumes May Accumulate and Cause Death".
FIGURE 8.23 Labeling and identification standards. Relates to: KRS 217.650-217.710.
Statutory authority: KRS 194.050, 211.090, 211.180, 217.690, 902 KAR 47:020.
“(2) It shall be unlawful for any manufacturer or importer of cigarettes

to advertise or cause to be advertised (other than through the use of
outdoor billboards) within the United States any cigarette unless the
advertising bears, in accordance with the requirements of this section,
one of the following labels:… SURGEON GENERAL'S WARNING:
Cigarette Smoke Contains Carbon Monoxide.”
FIGURE 8.24 Title 15—Commerce and Trade Chapter 36—Cigarette Labeling and Advert-
ising §1331. Congressional declaration of policy and purpose.
8.6.5 THE CUSTOM AND PRACTICE OF CARBON MONOXIDE

SAFETY INFORMATION
“Custom and Practice” in the safety and engineering profession is a term referring to
common customs and practices in the design of equipment, products, and processes.
“Custom” having the meaning “a usual practice or habitual way of behaving; habit,”
and “practice” having the meaning “to do or be engaged in frequently or usually.”47
Custom and practice is all around us. It is what we are accustomed to and what
multiple designers have coalesced to in their designs. For example, the ubiquitous
use of right-handed threads on nuts and bolts is the custom and practice for most
threaded fastening applications. The introduction of metric threads was a change to a
preexisting custom and practice in the United States. For some things the custom and
practice is regional. In certain dryer portions of the southwest, residential air cooling
is performed by evaporative coolers, while in more humid areas, compression cycle
cooling units are used.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 224 — #28

To direct the Consumer Product Safety Commission to issue

regulations concerning the safety and labeling of portable
generators.
(1) WARNING: portable generator sold to the public for

purposes other than resale shall have a large, prominently
displayed warning label on the exterior packaging, if any,
of the portable generator and permanently affixed on the
portable generator regarding the carbon monoxide hazard
posed by incorrect use of the portable generator.
The warning label shall include the word

“DANGER” printed in a large font, and shall
include the following information, at a minimum, presented
in a clear manner:
(A) Indoor use of a portable generator can
kill quickly.
(B) Portable generators should be used
outdoors only and away from garages and open
windows.
C) Portable generators produce carbon monoxide,
a poisonous gas that people cannot see or smell.
(2) PICTOGRAM.--Each portable generator sold
to the public for purposes other than resale shall have a
large pictogram, affixed to the portable generator, which
clearly states “POISONOUS GAS” and visually depicts the
harmful effects of breathing carbon
monoxide.”
FIGURE 8.25 109th U.S. Congress (2005–2006) S. 2084: Portable Generator Safety Act,
Introduced December 13, 2005.
Requirements for Labeling of Retail Containers of Charcoal; proposed

Amendments AGENCY: Consumer Product Safety Commission.
ACTION: Proposed rule.\1\ SUMMARY: Under the Federal Hazardous
Substances Act, the Commission is proposing a rule to change the required
labeling for retail containers of charcoal intended for cooking or heating.
The labeling addresses the carbon monoxide hazard associated with
burning charcoal in confined spaces. The proposed amendments, which
include a pictogram, are intended tomake the label more noticeable and
more easily read and understood and to increase the label's ability to
motivate consumers to avoid burning charcoal in homes, tents, or vehicles.
DATES: Comments on the proposal should be submitted no later than
October 24, 1995.”
FIGURE 8.26 Consumer Product Safety Commission, 6 CFR Part 1500. Available at:
http://www.cpsc.gov/businfo/frnotices/fr95/95-40785.html.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 225 — #29

CAUTION
CARBON MONOXIDE
MAY BE PRESENT
FIGURE 8.27 Seton Style No. M9699.
In some cases, the custom and practice is dictated by codes, standards, or

guidelines. OSHA’s Standard Interpretations: Fluor Constructors, Inc. v. OSHRC,
No. 87- 4029 1/22/1988 states that,
a Commission judge agreed with the Secretary that both OSHA’s standards and industry
custom and practice plainly distinguish between lifelines and lanyards.
This concept, or pattern of practice, is generally accepted as a means to estab-

lish what safety measures are reasonable. Where codes do not require user-safety
information, custom and practice can be the basis for requiring notice of a hazard.
In the United States Occupational Safety and Health Review Commission’s decision
in “SECRETARY OF LABOR, Complainant, v. CITY OIL WELL SERVICE CO.,
Respondent”48 the Commission stated that,
The evidence indicates that the custom and practice of the industry is for the well servicer
to rely on the well operator or owner to advise it if H2 S hazards are present.
The phrase “of the industry” in the decision above associates custom and practice
with the multiple providers of goods or services that comprise an industry. When
there are few providers of these items, or if there is a single provider, it is difficult to
say this “practice” is the “usual” in that it may be the “only” practice. In this context,
one provider’s design does not establish enough of a pattern for there to be a custom
and practice to evaluate safe design.
In some cases, custom and practice criteria has been “institutionalized” into stand-
alone business enterprises. Several firms sell standard and customized warning signs
and labels, including ones directed at CO exposure (Emedco, Seton, VMC, Inc.,
Safetysign Inc.) (Figure 8.27).
There does appear to be a general custom and practice with respect to safety
warnings within the United States both with regard to their content and to a lesser
extent concerning on what products or places warnings should be applied. The con-
tent aspect is well documented. The content of warnings and other safety information
generally include (1) a signal word, (2) a pictorial, (3) a signal symbol, (4) a descrip-
tion of the hazard, (5) a statement of the consequences, and (6) a statement of how to
avoid the hazard.9,10,14,28 With regards to which products and places require safety
information, there is some variation among different products, product groups, and
processes.
With respect to CO, there are several established customs and practices for dif-
ferent products and places. In Carbon Monoxide: Its Hazards and the Mechanisms
8417: “8417_c008” — 2007/9/11 — 12:16 — page 226 — #30

INSTRUCTIONS “! WARNING! …
Actuation of your CO alarm indicates the presence of carbon
monoxide (CO) which can KILL YOU. …
CAUTION: This alarm will only indicate the presence
of carbon monoxide (CO) gas at the sensor. CO gas may be present in
other areas of the RV.”
FIGURE 8.28 Model 3400 Carbon Monoxide Gas Alarm 5052 (CO Alarm for RVs) CCI
Controls Co., Cecelia Street, South Gate, CA 90280. Phone: (323) 560-6060; Fax: (323)
560-1136; available at: www.ccicontrols.com.
Assembly Instructions
“WARNING
CARBON MONOXIDE HAZARD
Burning charcoal inside can kill you. It gives off carbon

monoxide, which has no odor. NEVER burn charcoal
inside homes, vehicles or tents….”
FIGURE 8.29 Tabletop Grill™ (Drum-type Barbeque Grill), New Braunfels, Model
03407610.
of Its Actions (1944), published by the United States Public Health Service,49 there
is reference to the early warning needs of CO exposure:
In spite of these efforts, a large number of accidents from CO poisoning occur which are
due to acts of carelessness of uninformed persons, as pointed out by Brumbaugh (1926)
and it seems to be imperative that the public be informed about the proper handling
of such appliances (bolding added for emphasis) [gas appliances and space heaters]… it
appears that the public is not aware of the dangers resulting from the improper handling
of gas appliances capable of forming CO due to incomplete combustion.
To the present time, the custom and practice is to continually inform the public of
the hazards of CO exposure. Automobile manuals, for example, consistently and for
many years have warned the public about the possibility of CO overexposure from
operation of a vehicle’s engine in an enclosed space.
Figures 8.28 through 8.31 show examples of warnings, labels, or instructions
about CO that are the result of an existing or emerging custom and practice, and are
not the result of voluntary standards or mandatory regulations.
8.6.6 LITIGATION-DRIVEN SAFETY INFORMATION

It is not unusual for an injured party to complain about insufficient warnings or
instructions in product defect or workplace injury cases. Since warnings and instruc-
tions can influence the safe use of a product, it is reasonable to consider them in
8417: “8417_c008” — 2007/9/11 — 12:16 — page 227 — #31

OPERATING INSTRUCTIONS AND OWNER’S MANUAL
“WARNING:
Early signs of carbon monoxide poisoning resemble the
flu, with headache, dizziness and/or nausea. If you have
these signs, heater may not be working properly. Get
fresh air at once! Have heater serviced. …
WARNING:
Combustion by-products producedwhen using this product
contain carbon monoxide, a chemical known to the State
of California to cause cancer and birth defects (or other
reproductive harm).”
FIGURE 8.30 Mr. Heater (Portable Propane Space Heater). Model # MH18B Enerco
Group, Inc., 4560 W. 160th Street, Cleveland, OH 44135 · 216-881-5500 05/04 Revision L1
# 78435.
Operating and Maintenance Instructions“

WARNING_Engine exhaust contains carbon monoxide, an odorless,
colorless, poison gas. …Breathing carbon monoxide can cause nausea,
fainting and death.”
FIGURE 8.31 Briggs & Stratton, Power Products Group, Model AA0101 (four-stroke
outboard engine), Milwaukee, WI, Form number MS-5682-9/03.
evaluating their effects on the safe use of the product or workplace. Safety cri-
teria can be based on the precedence of prior cases or regional law. Individual
states and the federal bench have their own juristically developed labeling, warn-
ing, and instructional criteria for product safety. Below is just one example of these
criteria:
Adequate warning: 1) it must be of such form that it could reasonably be expected to

catch the attention of the reasonably prudent man in the circumstances of its use; and
2) the content of the warning must be of such a nature as to be comprehendible to the
average user and to convey a fair indication of the nature and extent of the danger to the
mind of a reasonably prudent person … the question of whether or not a given warning is
legally sufficient depends upon the language used and the impression that such language
is calculated to make upon the mind of the average user of the product.
Implicit in the duty to warn with the degree of intensity that would cause a reasonable man
to exercise.… the caution commensurate with the potential danger… A clear cautionary
statement setting forth the exact nature of dangers involved would be necessary to fully
protect the seller… .50
Weinstein and Twerski51 in 1978 reports (p. 41)
8417: “8417_c008” — 2007/9/11 — 12:16 — page 228 — #32

The court in Muncy vs Magnolia Chemical Company set forth its requisite for
adequate warnings:
‘1) it must be of such form that it could reasonably be expected to catch the attention of
the reasonably prudent man in the circumstances of use; and 2) content of the warning
must be of such a nature as to be comprehensible to the average user… Implicit in the
duty to warn is the duty to warn with a degree of intensity that would cause a reasonable
man to exercise… the caution commensurate with the potential danger….’
In The Restatement of Torts: Products Liability,51 the American Bar Institute

addresses the relationship between dangerous product design and product warnings
and instructions. Section 402A states that,
One who sells any product in a defective condition unreasonably dangerous to the user
or to his property is subject to liability for physical harm thereby caused to the ultimate
user or consumer.
In The Restatement of Torts: Products Liability—the tension between product

design and product warning, the author explains that,
courts have interpreted [section] 402A to mean that a product may be unreasonably
dangerous because of a defect in manufacturing, design, or warnings and instructions.
Silvergate expands this by including the third restatements of torts language in

conjunction with proper design:
[A Product] is defective because of inadequate instructions or warnings when the fore-

seeable risks of harm posed by the product could have been reduced or avoided by the
provision of reasonable instructions or warnings by the seller or other distributor, or a
predecessor in the commercial chain of distribution and the omission of the instructions
or warnings renders the product not reasonably safe.
But that,
In general, when a safer design can reasonably be implemented and risks can reasonably
be designed out of a product, adoption of the safer design is required over a warning…
Hence, warnings and instructions have a second place behind proper design in
the eyes of the judicial system, but they are a vital element in deciding if a product
(or place or process) is reasonably safe.
Many of these and other legal admonitions are passed on to the engineering com-
munity. In a how-to book for engineers called What every Engineer Should Know
About Products Liability,52 the authors detail the four components of a warning
communication (p. 58):
1. Intelligibility
2. Adequacy
8417: “8417_c008” — 2007/9/11 — 12:16 — page 229 — #33

3. Completeness
4. Placement
and add that:
If a product has obvious dangers, there is no duty to warn, since the danger is a matter
of common knowledge. For example, the fact that a sharp knife is capable of cutting
a careless user is a matter of common knowledge and is obvious. Similar remarks can
be made for guns, blowtorches, and so forth. The hazards that must be warned against
are the less obvious ones - inherent, latent, or concealed dangers the manufacturer has
knowledge of but the user cannot foresee.
In the publication, Injury and Litigation Prevention, by Freeman, 1991,53 the

author presents eight considerations that may be used in a judicial evaluation of
warnings. Many of these are consistent with those detailed in previous sections, but
also included is the concept of “motivate[ing] behavioral change.” Warnings should
not just provide weak warnings, but the “consequences of continuing the forbidden
activity must be clear, not remote or hidden.”
Finally, while the courts and legal systems around the world have very different
traditions and bases for their laws and judicial processes, Huber54 discusses the pos-
itions of several judicial systems around the world in regards to safety, liability, and
warnings. Comparisons are made among English, Japanese, German, French, and
Canadian case law, and reports that product warnings may be a basis for concluding
a product is defective.
References
1. Kandarjian, L. Federal Policy Options for Indoor Air Pollution from Combustion
Appliances, Transactions of Combustion Processes and the Quality of the Indoor
Environment, Harper, J., ed., International Specialty Conference, Air and Waste
Management Association, September 1988.
2. Flickinger, J. Characterization of Emissions and Indoor Air Quality Impacts Resulting
from Vented and Unvented Gas Appliances in Ten Homes, Dames & Moore, Trans-
actions of Combustion Processes and the Quality of the Indoor Environment, Harper,
J., ed., International Specialty Conference, Air and Waste Management Association,
September 1988.
3. Carbon monoxide. Material Safety Data Sheet, Genium Publishing Company,
Schenectady, NY, January 1986, MSDS # 35.
4. Konz, S. Work Design, Industrial Ergonomics, 4th ed., Publishing Horizons, Inc.,
Scottsdale, AZ, p. 388, 1995.
5. Heating System Check Recommended for Carbon Monoxide—CPSC Release 88–92,
November 1989.
6. Fugler, D. Indoor Pollution Measured in Combustion Spillage Testing, Research Divi-
sion, Canada Mortgage and Housing Corp., Ottawa, ON, Transactions of Combustion
Processes and the Quality of the Indoor Environment, Harper, J., ed., International
Specialty Conference, Air and Waste Management Association, September 1988.
7. Woodson, W.E. Human Factors Design Handbook, 2nd ed., McGraw Hill, NY, 1992.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 230 — #34

8. McCormick, E.J. and Sanders, M.S. Human Factors in Engineering and Design,
5th ed., McGraw Hill, New York, NY, 1982.
9. Product Safety Label Handbook, Westinghouse Electric Company, Traftord, PA, 1981.
10. Product Safety Sign and Label System, FMC Corporation, Santa Clara, CA, Copyright
1990.
11. Holkham, T. Label Writing and Planning, Blackie Academic, London; Melbourne,
1995.
12. Holmes, N. Wordless Diagrams, Bloomsbury, USA, 2005.
13. Recchia, N. Warnings, Mark Batty, West New York, NJ, 2005.
14. Mellan, I. Encyclopedia of Chemical Labeling, Chem. Publishing Co., Newyork, NY,
1961.
15. Schoff, G. Writing & Designing Operator Manuals, Lifelong Learning Publications,
Division of Wadsworth, Inc., Belmont, CA, 1984.
16. Product Safety Signs and Labeling, ANSI Z 535.4, 1998; “Safety Color Code”,
ANSI Z 535.1, 1991; Environmental and Facility Safety Signs, ABSI Z 535.2, 1991;
Accident Prevention Signs, ANSI Z 535.5, 1998; Criteria for Safety Signs, ANSI Z
535.3, 1991.
17. For Hazardous Industrial Chemicals—Precautionary Labeling, ANSI Z 129.1, 1994.
18. Ryan, J. Design of Warning Labels and Instructions, Van Nostrand Reinhold,
1991.
19. Specifications for Accident Prevention Signs, USA Standard, USAS Z38.1, 1959.
20. Scheme for the Identification of Piping Systems, ANSI A13.1, 1981.
21. Title 15—Commerce and Trade, Chapt. 36—Cigarette Labeling and Advertising.
1999.
22. Brusaw, C. Handbook of Technical Writing, 5th ed., St. Martins Press, NY, 1997.
23. Honda Owner’s Manual, GC160–GC190, American Honda Motor Co., 31Zl8A08
00X31-ZL8-A080, 2003–2005 copyright.
24. 2003 Owner’s Manual—Santa Fe, Hyundai Motor Co., A030A020-AAT.
25. Department of Defense Hazardous Chemical Warning Labeling System, Office of the
Assistant Secretary of Defense, 93-17976, DOD 6950.5 H, June 1989.
26. Hippocrates, The Epidemics, Bk. I, Sect. XI.
27. Professional Engineering Codes of Ethics, 1999.
28. Gordon, S. and Hall, P. Systematic Training Program Design, University of Idaho,
Englewood, Cliffs, NJ, p. 70, 1994.
29. Barnett, R.L. and Brickman, D. Safety Hierarchy, Triodyne Inc., Safety Brief, Vol. 3,
No. 2 Revised, Niles, IL, June 1985.
30. A.G.A. Requirements for Carbon monoxide Safety Shutoff Systems and Carbon
Monoxide Warning Devices, Supplement to American National Standard for Auto-
matic Gas Ignition Systems and Components, ANSI Z 21.20, as contained in report
No. 2-86, Nov 10, 1986.
31. Human Factors Perspectives on Warnings, 1980–1993, Human Factors and Ergo-
nomics Society, 1994.
32. Human Factors Perspectives on Warnings, Vol. 2, 1994–2000, Human Factors and
Ergonomics Society, 2001.
33. Braun, C.C., Greeno, B., and Silver, N.C. Differences in Behavior Compliance as
a Function of Warning Color, Proceedings of the HFES 38th Annual Meeting, USA
pp. 379–383, 1998.
34. Braun, C.C., Holt, R.S., and Silver, N.C. Signal Word and Color Specifications
for Product Warnings, Proceedings of the HFES 38th Annual Meeting, USA,
pp. 1104–1108, 1994.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 231 — #35

35. Caird, J.K., Wheat, B., and McIntosh, K.R. The Comprehensibility of Airline Safety
Card Pictorials, Proceedings of the HFEWS 41st Annual Meeting, USA, pp. 801–805,
1996.
36. Chen, J.C., Gilson, W.D., and Mouloua, M. The Comprehensibility of Airline Safety
Card Pictorials, Proceedings of the HFES 41st Annual Meeting, USA, pp. 801–805,
1997.
37. Ringseis, E.L. and Caird, J.K. The Comprehensibility and Legibility of Twenty Phar-
maceutical Warning Pictorials, Proceedings of the HFES 39th Annual Meeting, USA,
pp. 974–978, 1995.
38. Hancock, H.E., Rogers, W.A., and Fisk, A.D. Understanding Age-related Differences
in the Perception and Comprehension of Symbolic Warning Information, Proceedings
of the HFES 43rd Annual Meeting, USA, pp. 617–621, 1999.
39. Nanthavanji, S. Analysis of Workplace Factors on Auditory Warning Alarm Location,
In Advances in Industrial Ergonomics and Safety VII, Taylor & Francis, Bristol, PA,
1995.
40. Criteria for a recommended standard …Occupational Exposure to Carbon monox-
ide, U.S. Department of Health, Education and Welfare, National Institute for
Occupational Safety and Health (NIOSH), 1972.
41. Chemical Hazards Bulletin, Engineering and Safety Services, American Insurance
Association, 85 John Street, NY 10038, August, 1974.
42. In 1971 Primary and Secondary National Ambient Air Quality Standards (36 FR
8186), in 1985 Review of the National Ambient Air Quality Standards for Carbon
Monoxide, 40 CFR Part 50, FR Vol. 50, No. 176, September 13, 1985.
43. 40 CFR Ch. 1, 7-1-86 Edition, Para 51.16(e) (3).
44. Viscusi, W.K. Product-Risk Labeling—A Federal Responsibility, Am. Enterprise
Institute, p. 1, Washington, DC, 1993.
45. Burd, L.T. CO Morbidity and Mortality Six Decades of Carbon Monoxide Poison-
ing, Master of Public Health Thesis, University of Texas, Health Science Center at
Houston, School of Public Health, August, 2002.
46. Non-fire Carbon Monoxide Deaths Associated with the use of Consumer Products
2001 Annual Estimates, CPSC, Susan Carlson, Directorate of Epidemiology, May
13, 2004.
47. Webster’s New World Dictionary of the American Language, 2nd College ed., World
Publishing Co., 1970.
48. Decision in Secretary of Labor, Complainant, vs. CITY OIL WELL SERVICE
CO., Respondent. OSHRC Docket number 81-1797, Buckley, Chairman; Cleary,
Commissioner, September 30, 1986.
49. von Oettingen, W.F. et al. Carbon Monoxide: Its Hazards and the Mechanisms of Its
Actions, Federal Security Agency, United States Public Health Service, Public Health
Bulletin No. 290, 1944.
50. Weinstein, A.S. and Twerski, A.D. Products Liability and the Reasonably Safe
Product, Wiley, p. 41, 1978.
51. Silvergate, S.H. The Restatement (Third) of Torts: Products Liability—The Tension
Between Product Design and Product Warnings, Florida Bar J., Vol. 75, pp. 10–17,
December 1, 2001.
52. Thorp, J.F. and Middendorf, W. What Every Engineer Should Know About Products
Liability, Marcel Dekker, Basel and NY, 1978.
53. Freeman, S.H. Injury and Litigation Prevention, Van Nostrand Reinhold, NY, 1991.
54. Huber, P.W. The Liability Maze, The Brookings Institute, Washington, DC, 1991.
8417: “8417_c008” — 2007/9/11 — 12:16 — page 232 — #36

9 Public Health
Surveillance for Carbon
Monoxide in the United
States: A Review of
National Data
Michael E. King and Joshua A. Mott
CONTENTS
9.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233

9.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
9.2.1 Public Health Surveillance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
9.2.2 Carbon Monoxide-Caused Mortality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
9.2.2.1 Recent Studies of National CO Mortality. . . . . . . . . . . . . . . 237
9.2.3 Carbon Monoxide-Caused Morbidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
9.2.3.1 Recent National Studies of Nonfatal
Carbon Monoxide Poisoning . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
9.3 Toward a National Surveillance System for Carbon Monoxide . . . . . . . . . . 243
9.3.1 Case Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
9.3.2 Mortality Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
9.3.3 Morbidity Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
9.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
9.1 OVERVIEW
This chapter provides a review of epidemiologic studies describing national surveil-
lance related to carbon monoxide (CO) poisoning in the United States. We discuss
sources of surveillance data and recent studies of CO-caused mortality and morbid-
ity. We conclude with a summary of national estimates and provide recommendations
for improving public health surveillance efforts and establishing a national system
specifically designed to track CO poisoning.
233
8417: “8417_c009” — 2007/9/11 — 12:15 — page 233 — #1

9.2 INTRODUCTION
CO poisoning is an important public health problem in the United States, resulting in
over 400 fatal unintentional poisonings annually.1 The burden of CO poisoning has
been shown to vary from state to state,2 and risk factors for poisoning may vary across
regions and populations, particularly in the wake of natural disasters.3,4 Ironically,
CO poisoning in indoor environments “is almost entirely preventable by the correct
installation, maintenance, and operation of devices that may emit CO indoors, com-
bined with the appropriate use of CO detectors” (i.e., alarms).5 Surveillance activities
are a critical component of any prevention effort and public health professionals rely
on timely accurate surveillance to describe the magnitude and etiology of poisonings
to better target interventions. Yet, despite increasing awareness of the dangers of CO
poisoning, a timely national surveillance system is not in place to monitor CO pois-
onings or deaths in the USA. The purpose of this chapter is twofold: (1) to provide a
review of epidemiologic studies describing national surveillance for CO poisoning in
the United States, and (2) to discuss issues relevant to the development of an ongoing,
national surveillance system for CO.
9.2.1 PUBLIC HEALTH SURVEILLANCE

Environmental public health surveillance is broadly defined as: “the ongoing system-
atic collection, analysis, and interpretation of data, closely integrated with the timely
dissemination of these data to those responsible for preventing and controlling dis-
ease and injury.”6 This definition emphasizes the importance of surveillance for CO
poisoning, since information about the epidemiological aspects of CO poisoning is
essential for targeting, monitoring, and evaluating risk factors and public health inter-
ventions. Surveillance activities are conducted routinely for diseases or conditions that
are highly prevalent, result in excess emergency department (ED) or hospital use, and
for which primary or secondary prevention is possible. However, with the exception
of isolated reports from states responding to public health emergencies7,8 surveil-
lance has not neen conducted for CO poisoning in the United States. The absence of
a national system has resulted in “basic gaps in our knowledge of the epidemiology
and outcomes of CO poisoning, including the long -erm sequelae, prevention, and
risk behaviors.”5
The primary challenge for environmental surveillance is relating health outcomes,
such as CO intoxication, to specific environmental exposures, and hazards.9 To facil-
itate this linkage, surveillance systems must collect, characterize, and disseminate
data on health hazards, exposures, and interventions, in addition to health outcomes.9
Environmental health hazards are defined as chemical, physical, and biologic agents
or biomechanical stressors in the air, food, and water.6 Exposure surveillance involves
monitoring individuals for “the presence of an environmental agent, its metabolites,
or its clinically inapparent (e.g., subclinical or preclinical) effects”.6 Tracking inter-
ventions includes monitoring programs and policies intended to prevent agents from
becoming hazards or minimize exposures and health outcomes. Given the variety of
data necessary to achieve these diverse surveillance objectives, no single dataset cur-
rently available is sufficient to serve as the sole source of surveillance data. Currently,
8417: “8417_c009” — 2007/9/11 — 12:15 — page 234 — #2

Public Health Surveillance for Carbon Monoxide in the United States 235
TABLE 9.1
Components of an Environmental Public Health Surveillance System and
Potential CO Poisoning Data Sources, United States
Surveillance Components Data Sources
Hazard surveillance • US Environmental Protection Agency National
Examples: Emission Inventory database
• CO emissions data • Utility companies/fire departments
• Gas leak public call-data
Exposure surveillance • Electronic lab reporting databases (e.g., the
Example: National Electronic Disease Surveillance System)
• Carboxyhemoglobin level data
Health outcomes • American Association of Poison Control Centers
Examples: Toxic Exposure Surveillance System (TESS)
• Nonfatal poisoning occurrences • Consumer Product Safety Commission CO
• Occupational poisoning data Poisonings database
• Outpatient visits • National Electronic Injury Surveillance System
• Emergency department visits (NEISS)
• Hospitalization data • National Ambulatory Medical Care Survey
• Mortality data • National Hospital Discharge Survey
• National Vital Statistics System
Interventions • Environmental Protection Agency
Examples: • Behavioral Risk Factor Surveillance System
• Data on the presence of working catalytic • Private national surveys (e.g., Porter-Novelli
converters in automobiles annual Healthstyles survey)
• CO-detector ownership data • National Conference of State Legislatures
• Gasoline-powered generator safety electronic legislation databases
knowledge/attitude data
• Emission and CO-detector legislation data
Source: Adapted from Thacker et al. Am. J. Publ. Health, 86, 633–638, 1996. The National Workgroup
on Carbon Monoxide Surveillance Carbon monoxide: A model environmental public health indicator.
ONLINE. 2006. Available: http://www.maine.gov/dhhs/eohp/epht/documents/CO_White.pdf. (Oct. 1,
2006).
“there are 31 federally funded national data systems that collect data on injury mortal-
ity, morbidity, and risk factors.”10 However, a limited number of them are potentially
useful for tracking CO poisoning. Table 9.1 provides examples of different types of
data and data sources for surveillance of environmental hazards, exposures, health
outcomes, and interventions related to CO in the United States.
To date, the majority of national surveillance studies for CO have focused on
health outcomes, and, in particular, estimates of mortality. This focus is perhaps due
to the absence of a single system that captures national CO-related morbidity reliably.
Likewise, CO is often only recognized as the cause-of-death upon postmortem exam-
ination, because the symptoms of CO poisoning are nonspecific and are frequently
attributed to other causes or overlooked altogether by clinicians and the public.11
8417: “8417_c009” — 2007/9/11 — 12:15 — page 235 — #3

Most of the studies reviewed in this chapter concern poisonings resulting from unin-
tentional and nonfire-related (NFR) events because both intentional and fire-related
CO poisonings have distinctly different public health prevention strategies and are
outside of the scope of this discussion.
9.2.2 CARBON MONOXIDE-CAUSED MORTALITY

All US death certificates are compiled by the National Center for Health Statistics
into the Multiple Cause-of-Death Mortality Database as part of the National Vital
Statistics System (NVSS).12 To date, this system has served as the primary source of
data for estimating national rates of poisoning death related to CO in the United States.
This database provides mortality data by multiple-cause of death for all fatalities
occurring in the United States. Each record is on the basis of information abstracted
from death certificates filed in vital statistics offices of each state and the district of
Columbia. Death certificates contain demographic and geographic information of the
decedent, as well as up to 20 conditions that contributed to the death, coded according
to the World Health Organization format International Classification of Diseases
(ICD).13,14 The causes of death codes are compiled in two forms: the entity axis,
which contains conditions as stated on the death certificate, and the record axis, which
is edited by a computer algorithm to remove duplicate records and revise ICD codes
to best describe the overall medical certification portion of the death certificate.15
It is important for researchers using Multiple-Cause data to be aware of the format
and structure of the data file prior to use; documentation accompanying public use
mortality data files specifies that the “record axis is designed for the generation of
person-based multiple-cause statistics” and “by definition, the entity data cannot meet
this requirement.”16 In other words, codes on the record axis may be used for CO
surveillance to count people with conditions, whereas the entity axis data may only
be useful for counting the number of times a particular condition was coded. The
8th Revision of ICD coding (ICD-8) was used for 11 years, from 1968 to 1978, the
9th Revision (ICD-9) from 1979 to 1998, and the 10th Revision (ICD-10) from 1999
to the present year. At the time of this writing, the most recent year of mortality data
available was 2003 as there is a delay of 2–3 years before national mortality data
are released to the public. As a reference, the ICD codes that have been used for the
national surveillance of CO are provided in Table 9.2.
One alternative source of national mortality data for the surveillance of CO is
the Centers for Disease Control and Prevention (CDC), Wide-ranging Online Data
for Epidemiologic Research (WONDER) data system. This system provides access
to an array of health-related datasets, including the NVSS underlying cause-of-
death files from National Center for Health Statistics (NCHS) and injury mortality
reports generated by the Web-based Injury Statistics Query and Reporting System
(WISQARS) maintained by the National Center for Injury Prevention and Control
(NCIPC). Although WONDER may be used to estimate national CO mortality, it
is important to note that the system can only identify cases of CO poisoning using
a single cause-of-death. Hence these data are likely to underestimate the burden
of unintentional poisoning because of the inability to select cases using codes in
multiple-cause-of-death codes fields.
8417: “8417_c009” — 2007/9/11 — 12:15 — page 236 — #4

TABLE 9.2
International Classification of Diseases Codes Specific to Carbon
Monoxide Poisoning and Additional Codes Useful for Case Identification
Underlying or contributing cause of death
Toxic effect of carbon monoxide
ICD-10: T58
ICD-8 and ICD-9: N986
Intent of death
Unintentional injury
ICD-10: X47
ICD-8 and ICD-9: E800-E949
Intent undetermined
ICD-10: Y17
ICD-8 and ICD-9: E980-E989
Mechanism of death
Fire-related (exclude from analyses)
ICD-10: X00-X09
ICD-8 and ICD-9: E837, E890-E899, E923, or N940-N949
Motor vehicle-related
ICD-10: No X-code specific to CO
ICD-9: E800-E807, E810-E825, E830-E836, E838, E840, E841, E843-E847, E868.2, E929.0,
E929.1, E952.0, E958.5, E958.6, E982.0, E988.5, and E988.6
ICD-8: E800-E807, E810-E825, E830-E836, E838, E840, E841, E843-E845, E873, E927,
E940-E941, and E952.0
Nonmotor vehicle-related
ICD-10: No X-code specific to CO
ICD-9: E867, E868.0, E868.1, E868.3, E919, E920, E951, E981
ICD-8: E870-E872, E874, E920, E927, E928, E951, E981
Source: Adapted from Mott et al., JAMA, 288, 988–995, 2006.
9.2.2.1 Recent Studies of National CO Mortality

The most recent analysis of unintentional, nonfire-related (UNFR), CO-related mor-
tality data in the United States was produced by the National Center for Environmental
Health’s Air Pollution and Respiratory Health Branch at the CDC.17 This study estim-
ated national and state-specific CO mortality rates and described the demographic,
seasonal, and geographic patterns in CO-related deaths for 1999–2002.17 Death cer-
tificate data were obtained from the NVSS using the underlying cause-of-death field
and record axis fields from the multiple-cause-of-death files. A case of unintentional
CO-related death was defined as a US resident death coded using the ICD-10 code
T58 (toxic effect of CO) as the underlying cause-of-death and for which poisoning
by accidental exposure (code X47) or exposure of undetermined intent (code Y17)
to gases or vapors was listed as a contributing cause in one of the 20 record axis
fields. All records of deaths caused by intentional or fire-related exposure to CO,
coded as X00–X09, were excluded from the analysis. This study found that during
8417: “8417_c009” — 2007/9/11 — 12:15 — page 237 — #5

1999–2002, there were 11,132 CO-related deaths among U.S. residents, of which
1982 (18%) were classified as both unintentional and NFR. From 1999 to 2002, the
total average age-adjusted annual death rate for UNFR CO was 1.8 deaths per million
persons in the United States. Death rates were highest for adults (2.2 per million for
adults aged >18 years, versus 0.4 per million for children aged 17 years or younger)
and men (2.8 per million men versus 0.9 per million women). The average daily num-
ber of CO-related deaths was greatest during the months of January and December
(277 per month) and lowest during the months of July and August (91 per month). For
the period from 1999 to 2002, only 32 states had a sufficiently large number of UNFR
CO-related deaths to calculate stable mortality rates (Figure 9.1). Although numer-
ous states had rates above the national four-year average annual rate, the state with
the highest statistically reliable UNFR CO mortality rate was Nebraska (5.6 deaths
per million person-years) and the state with the lowest reliable rate was California
(0.6 average deaths per million) (Figure 9.1). Reporting of acute CO poisoning by
healthcare providers was mandated for 13 states, although there was no clear pattern
of differences in CO-related mortality between states with mandated reporting and
those without.17
In 2005, the National Center for Injury Prevention, in conjunction with the
National Center for Environmental Health, used multiple-cause-of-death mortality
data to estimate the annual incidence of fatal and nonfatal unintentional, NFR CO
poisoning.18 This study provided national crude annual UNFR CO death rates by
demographic characteristics and season for 2 years: 2001 and 2002. In addition,
case-fatality rates were estimated by dividing the number of deaths by the sum
UR (NH)
UR UR UR (VT) UR (ME)
MR UR/MR (MA)
UR UR
UR MR UR (RI)
MR UR (CT)
UR MR (NJ)
UR MR UR/MR (DE)
MR UR (DC)
MR
UR
MR
UR
UR(HI) CO Death Rate*

> U.S. Average
= U.S. Average
*Rate per 1,000,000 population per year, age-adjusted to 2000 US Census population
. < U.S. Average
UR = Rate unreliable
MR = CO death is a mandated reportable condition
FIGURE 9.1 Average rates of unintentional, nonfire, carbon monoxide-related death, by

state, United States, 1999–2002.
8417: “8417_c009” — 2007/9/11 — 12:15 — page 238 — #6

of UNFR CO deaths and nonfatal exposures multiplied by 100. Fatal UNFR CO

poisoning deaths were identified as those with ICD-10 code T58 as a leading or
contributing cause of death and an underlying cause of death code indicating either
accidental poisoning (X47) or poisoning of undetermined intent (Y17). This study
found that the crude annual UNFR CO death rate for the United States was 0.17 per
100,000 with an average of 480 deaths per year. The annualized incidence of fatal
UNFR CO exposures was greatest during the fall and winter months, with more
cases occurring during December (56 deaths) and January (69 deaths). The crude
death rate from CO was highest for adults over 65 years (0.32 per 100,000), males
(0.24 per 100,000), and nonHispanic whites and blacks (0.17 per 100,000). The
case-fatality rate increased with age, ranging from 0.6% for children under 4 years
to 5.5% for adults aged 55–64 years. Overall, males had a 2.3 times higher case-
fatality rate than women and 23.5% of deaths occurred among adults aged 65 years
or more.18
In 2002, Mott et al.19 conducted an ecological analysis of national CO death data
from the NVSS and annual CO emissions estimates for light-duty vehicles obtained
from the US Environmental Protection Agency. This study estimated the percent
change in CO emissions and CO mortality rates by intent and mechanism for the
years 1968–1998. The main outcome from this study was US resident deaths from
1968 to 1998 coded with ICD-8 or ICD-9 code N986 (toxic effect of CO) as a
contributing cause -of -death or those records with an ICD external cause of injury
code exclusive to CO poisoning (Table 9.2). From 1968 through 1998, this study
identified 116,703 NFR deaths due to CO in the United States. During this same time
period, crude mortality rates associated with NFR CO declined from 20.2 deaths to
8.8 deaths per million person-years in the United States. There were 2.2 intentional
deaths for every one unintentional NFR CO-related death. Between 1968–1998, motor
vehicles were identified as the mechanism for 70.6% of deaths and, following the
introduction of the catalytic converter in 1975, annual estimates of NFR CO emissions
decreased by 76.3%. In addition, unintentional motor vehicle-related NFR CO death
rates declined by 81.3% and rates of motor vehicle-related NFR CO suicides declined
by 43.3%.19
In 1996, the CDC published a summary of findings from an investigation of deaths
associated with multiple motor-vehicle related CO poisonings in Colorado and New
Mexico, together with national estimates of CO deaths from 1979–1992.20 This report
presented the geographic pattern of national unintentional CO mortality associated
specifically with stationary motor vehicles (UMVR). Deaths were identified using
ICD-9 code E868.2, a code specific to deaths due to accidental poisoning by CO or
another utility gas from the following sources: farm tractor not in transit, gas engine,
motor vehicle not in transit, or any type of combustion engine not in a watercraft.
This analysis found that crude death rates for UMVR CO were highest in states in
the northern regions of the United States from 1979 to 1992, although no specific
national or state rates were provided.20
In the early 1990s, Cobb and Etzel2 published a descriptive analysis of uninten-
tional NFR CO-related deaths (UNFR) in the United States. This study calculated
crude and age-adjusted death rates by demographic characteristic, season, and state
for the years 1979–1988. All US resident deaths coded using ICD-9 code N986 were
8417: “8417_c009” — 2007/9/11 — 12:15 — page 239 — #7

identified between 1979–1988, excluding those coded as intentional (codes E950–

E959 and E960–E969), intent undetermined (E980–E989), and those owing to fire
(E837, N940–N949, E923). The 10-year average crude UNFR CO death rate for males
(0.78 per 100,000 persons) was found to be 3 times higher than that for females (0.26
per 100,000 persons). Overall, 83% of UNFR CO-related deaths occurred among
whites, yet “race-specific death rates were more than 20% higher for blacks (0.63
per 100,000) than for whites (0.51 per 100,000).”2 Annual incidence of unintentional
CO-related death was found to be highest in the month of January, with an average of
181 deaths, and lowest in the month of July, with an average of 44 deaths. The state
with the highest age-adjusted UNFR CO death rate for the 10-year period was Alaska
(2.72 per 100,000) and the lowest rate was found in Hawaii (0.05 per 100,000).2
In addition to the above-mentioned surveillance reports produced by the CDC,
national estimates of CO-related mortality produced by the Consumer Products Safety
Commission (CPSC) may provide useful national surveillance for CO-related mor-
tality. These reports combine data from multiple sources, such as the NVSS and
National Electronic Injury Surveillance System (NEISS), with proprietary CPSC
datasets21 on the basis of death certificate data purchased directly from states and
in-depth follow-up investigations of select deaths. These data are used to produce
annual national estimates of unintentional NFR CO exposures and fatalities by
product, victim age, and incident location. However, the reports currently available
are limited to estimates associated specifically with the use of consumer products
under the jurisdiction of the CPSC (e.g., engine-powered tools, charcoal grills, gas
ovens, and other appliances). CPSC reports do not capture fatalities associated with
motor vehicles, despite the fact that CO in motor-vehicle exhaust has been found
to account for the majority of poisoning deaths in the United States.22 The most
recent year for which a complete CPSC report of CO-related mortality is avail-
able is 2002.23 In 2002, CPSC identified 188 unintentional, nonfire CO-related
deaths and the average annual estimate from 1999 to 2002 was 141 deaths. Heat-
ing systems were associated with 55% of deaths and engine-powered tools were
associated with 28% of deaths in 2002. An estimated 71% of CO deaths occurred in
the home. Overall, 81% of fatal CO incidents involved a single death, with adults
over 45 years of age accounting for 55% of all unintentional, nonfire CO deaths
in 2002.23
Although CPSC memoranda have been published covering both non-fatal CO
exposure incidents and fatalities for the periods 1990–2004 and 2002–2005, respect-
ively, CPSC notes that the counts for recent years contained in these reports may
not be complete.24 Briefly, from 1990 to 2004, there were 318 nonfire CO-related
deaths identified by CPSC, 274 of which were associated with the use of gen-
erators. Most of these deaths (39%) occurred during winter and 77% took place
in the home. From 1990 to 2004, an estimated 33% of deaths occurred among
adults aged 45–64 years and 75% of all “engine-driven tool” CO-related deaths
were males.24 From 2002 to 2005, preliminary counts from CPSC indicate that 253
nonfire CO-related deaths were associated with engine-driven tools in the United
States. An estimated 218 (86%) of CO deaths were associated with gasoline-powered
generators.20
8417: “8417_c009” — 2007/9/11 — 12:15 — page 240 — #8

9.2.3 CARBON MONOXIDE-CAUSED MORBIDITY

National surveillance for nonfatal CO exposures is challenging, given the limited
number of data sources currently available (Table 9.1). One potential source for
national nonfatal CO poisoning data is the NEISS, operated by the CPSC since 1971.25
The NEISS has been used to monitor consumer product-related injuries resulting
from consumer products under the regulatory jurisdiction of the CPSC. In 2000, the
surveillance system was expanded to collect data about all types and causes of injuries
and poisonings treated in hospital EDs, whether or not they were associated with
consumer products. This expanded system is called the NEISS All Injury Program
(NEISS-AIP). NEISS data are currently collected from a nationally representative
sample of 100 EDs, selected from a stratified probability sample of all US hospitals
that have at least 6 beds and provide 24-h emergency services, while NEISS-AIP data
are collected from a 66 hospital subsample. Data from approximately 500,000 injury-
related ED cases are collected annually by NEISS and coded for cause and intent of
injury using guidelines consistent with coding guidelines in the ICD-9-CM.25 Use
of these data for CO surveillance is limited, however, they exclude incidents owing
to occupational exposure, those associated with motor vehicles, and cannot provide
state or local estimates.25
Another source of national nonfatal CO poisoning surveillance data is the National
Hospital Ambulatory Medical Care Survey (NHAMCS) conducted by the National
Center for Health Statistics of the CDC. The NHAMCS collects data on the utiliza-
tion and provision of ambulatory care services in hospital emergency and outpatient
departments. Findings are on the basis of a national sample of visits to the emer-
gency and outpatient departments of noninstitutional general and short-stay hospitals,
exclusive of federal, military, and veterans administration hospitals, located in the 50
States and the District of Columbia. Annual data collection began in 1992.
Similarly, the national Toxic Exposure Surveillance System (TESS) operated by
the American Association of Poison Control Centers has demonstrated capability to
serve as a source of CO data for both state-based26 and national studies.27 Implemen-
ted in 1983, TESS is a comprehensive database that contains detailed toxicological
information about over 24 million poisoning incidents reported to 61 poison control
centers in the United States.28 The calculation of rates of calls to TESS for a given
exposure is facilitated because participating poison control centers also report the size
of the population they serve. However, the use of TESS data for CO surveillance may
be limited by the fact that most reports pertain to the treatment of nonfatal poisonings.
Therefore, TESS data may underestimate the incidence of fatal CO exposures. As of
October, 2006, annual reports summarizing TESS data are available online for the
years 1983–2004 (http://www.aapcc.org/annual.htm).
Finally, recent studies have suggested that data obtained from hyperbaric oxygen
(HBO) therapy centers may serve as an indicator of CO burden.27,30 HBO is typ-
ically used to treat those patients with the most severe CO poisoning,27 comprising
approximately 6% of all patients seen for CO in EDs.30 According to Hampson, “as
long as the spectrum of severity of the condition and treatment practices has not been
recognized to have changed significantly during the period surveyed, the national rate
of HBO therapy should serve as a qualitative marker of total disease incidence”.27
8417: “8417_c009” — 2007/9/11 — 12:15 — page 241 — #9

Using data from HBO therapy facilities for national CO surveillance is limited by
the lack of general availability and the fact that these data do not include less severe
cases, those not referred for HBO, or those treated at EDs using HBO. Likewise,
a patient’s state of residence may differ from that in which they received the HBO
therapy, complicating the calculation of valid state rates.
9.2.3.1 Recent National Studies of Nonfatal

A 2005 study by Hampson27 used TESS data and information from a recent survey of
HBOT facilities29 to compare trends in the annual incidence of nonfatal CO poisoning
in the United States. Hampson27 searched TESS records from 1985 to 2002 for all calls
received by poison control centers regarding cases of CO exposure, and calculated
annual call rates per million person-years. This study found that annual call rates for
CO poisoning increased from 31.1 per million in 1985 to 95.4 per million in 1996.
Although the rate decreased to 54.5 per million persons from 1996 to 2002, there was
a significant increase in calls for the entire 18 year period (p = .0022).27 Rates of
HBOT for CO poisoning and rates of calls to poison control centers were also found
to correlate strongly (r = 0.82, p = .0002).27
Also in 2005, the CDC published national estimates of unintentional, NFR CO
exposures for 2001–2003 using NEISS-AIP data.18 Nonfatal cases of CO poison-
ing were considered to be those coded as “CO exposure” or “CO poisoning” in
NEISS-AIP hospital data. Additional criteria used to identify nonfatal CO poisonings
included: (1) intent of injury unintentional or undetermined, (2) principle diagnosis of
“poisoning” or “anoxia”, and (3) additional narrative information indicative of CO.
This study estimated that 15,200 patients were treated in EDs annually for nonfatal,
unintentional, NFR CO poisoning during 2001–2003. Overall, the rate of nonfatal
exposure was similar for males and females, although the rate was highest for children
under 4 years of age (8.2 per 100,000 person). While most nonfatal CO exposures
(64.3%) occurred in the home, only 9.3% of patients reported owning a CO detector.18
This study also provided national estimates of CO-related mortality discussed earlier
in this chapter.
Although not pertaining to CO specifically, a CDC report published in 1999 used
data from the NHAMCS to describe poisoning-related ED visits from 1993 to 1996
in the United States.31 This study used a variety of ICD-9-CM codes to identify all
injury-related ED visits, but did not provide sufficient detail to describe the etiology,
intent, or mechanism of CO exposures. From 1993 to 1996, toxic effect of CO
(ICD-9-CM code N986) was the sixth leading principle diagnosis identified among
poisoning-related ED visits in the United States An annual average of 34,000 ED
visits were attributed to CO during the 4 year study period.31
In an earlier study, Hampson30 reported a synthetic national estimate of nonfatal
CO poisoning on the basis of a survey of EDs located in Washington, Idaho, and
Montana. This survey collected data about the total number of patients seen for acute
CO poisoning in 1994, although details regarding the etiology, source, or intent of
the poisoning were not collected. This analysis extrapolated data from the three states
to the United States as a whole and used state-specific age-adjusted CO death rates
8417: “8417_c009” — 2007/9/11 — 12:15 — page 242 — #10

to adjust for the fact that CO poisoning is more common in the Pacific Northwest
region. The study estimated 42,890 ED visits for CO poisoning occurred in 1994 in
the United States, resulting in a national ED visit rate of 16.5 per 100,000 persons for
CO poisoning.30
The public health literature describing surveillance of CO poisoning included
in this chapter is characterized by its variability. Although most studies focus on
fire-related CO poisoning, different data sources and definitions of poisoning are
used for case ascertainment, thus limiting the comparability of national estimates.
In addition, other national estimates of nonfatal CO poisoning are available in the
literature, although many of these focus only on poisonings resulting from a specific
mechanism (i.e., consumer products) or are very old. For example, recent reports
from the CPSC have used data from the NEISS to estimate that approximately 5000
nonfatal CO poisonings associated with motor-driven appliances occur annually in the
United States.32 Likewise, a 1974 study estimated that 10,000 nonfatal CO poisonings
occurred annually in the United States.33 The variability among national studies and
estimates of CO poisoning underscores the need for a national surveillance system
for CO.
9.3 TOWARD A NATIONAL SURVEILLANCE SYSTEM

FOR CARBON MONOXIDE
The lack of information linking environmental hazards to health outcomes has con-
tributed to what has been labeled the “environmental health gap” by The Pew
Environmental Health Commission.34 To address this gap and coordinate the devel-
opment of CO surveillance in the United States., the CDC established The National
Workgroup on Carbon Monoxide Surveillance, a partnership of public health profes-
sionals and agencies spanning private and federal jurisdictions from environmental
health and injury prevention to emergency response.5 This workgroup has noted that
the need for nationwide CO surveillance “is recognized in the Healthy People 2010
goal for the United States of ‘increasing the number of Territories, Tribes, and States,
and the District of Columbia that monitor carbon monoxide poisoning from 7 to
51.’”5 One indicator of progress toward the Healthy People 2010 goal for CO is
the number of states that mandate reporting of CO poisoning as part of the National
Notifiable Disease Surveillance System. The National Council of State and Territorial
Epidemiologist (CSTE) serves as a partner in the National Workgroup for CO Sur-
veillance and monitors patterns of nonnotifiable diseases and conditions; currently,
15 states mandate reporting of CO poisoning (http://www.cste.org/NNDSSSurvey/
2004NNDSS/NNDSSstatechrreporcondnona2005.asp). Similarly, although a num-
ber of states have mandated the installation of CO detectors, surveillance data are
needed to evaluate the effectiveness of ongoing legislative interventions for CO.
9.3.1 CASE DEFINITIONS

Although there is a growing recognition of the need to develop a national surveil-
lance system for CO, methods of case identification and ascertainment remain a
8417: “8417_c009” — 2007/9/11 — 12:15 — page 243 — #11

serious challenge for injury and death due to CO. The ability to accurately describe
CO-related morbidity and mortality depends heavily on the methods used to identify
and classify cases, the rules for coding data, and the type of data available. Effective
public health surveillance relies on the establishment of a widely adopted, clear, and
reliable case definition that includes criteria describing person, place, and time.35 To
date, no such consensus has been reached in the case of CO poisoning. According to
the National Workgroup on Carbon Monoxide Surveillance, 2 national public health
organizations have published their own version of a case definition for public health
surveillance of CO poisoning; The CSTE and the State and Territorial Injury Preven-
tion Directors Association (STIPDA). The STIPDA definition is more conservative
than the CSTE definition, that is, if both definitions are applied to the same dataset,
the STIPDA definition will identify fewer records as having CO poisoning than the
CSTE definition. The CSTE definition has been modified by at least one state to make
it a more conservative definition.36
The case definitions suggested by CSTE and STIPDA are listed in Table 9.3.
A formal evaluation of these case definitions is the focus of a current study by the
National Workgroup for CO Surveillance.36
9.3.2 MORTALITY LIMITATIONS

In addition to achieving a uniform case definition for tracking CO poisoning, under-
standing the limitations of available data is vital to improve national CO surveillance
efforts. For mortality specifically, the use of death certificate data for national CO
TABLE 9.3
Two Case Definitions for the National Surveillance of Carbon Monoxide in the
United States
Carbon Monoxide Case Definitions
CSTE
Confirmed case:
ICD-9 Coded Data: (1) a record in which the Nature of Injury code N-986 “Toxic effect of CO” is listed
or (2) a record in which an External Cause of Injury (E-Code) indicating exposure to carbon monoxide
(exclusively) is listed such as E868.3, E868.8, E868.9, E952.1, or E982.1.
Probable case:
ICD-9 Coded Data: A record in which an E-code indicating acute carbon monoxide poisoning inferred
from motor-vehicle exhaust gas exposure is listed, ie. E868.2, E952.0, or E982.0.
STIPDA
Records must have the N-code for CO poisoning (986) in the principal diagnosis field. Because this case
definition relies only upon the presence or absence of the N-code, it does not define classification of cases,
such as confirmed and probable.
Source: National Workgroup on Carbon Monoxide Surveillance. Project to evaluate carbon monoxide
surveillance CSTE and STIPDA case definitions with hospital data. 2006 Unpublished report. Obtained
October, 2006 from the Air Pollution and Respiratory Health Branch, U.S. Centers for Disease Control
and Prevention.
8417: “8417_c009” — 2007/9/11 — 12:15 — page 244 — #12

surveillance is limited by recent changes in the way deaths are coded and classified.
As noted earlier, the Tenth Revision of the ICD (ICD-10) was implemented in 1999
and is currently used to categorize causes of death in the NVSS. Earlier versions
of ICD coding provided external cause of injury codes exclusive to CO poisoning
(ICD 8 = E874, E875, and E952.1; ICD 9 = E868.3, E868.8, E868.9, E952.1, and
E982.1) or poisoning from motor vehicle exhaust (ICD8 = E873 or E952.0; ICD9 =
E868.2, E952.0, and E982.). Since CO is the only acutely poisonous gas in motor
vehicle exhaust, in the past these latter codes could also be used to indicate CO pois-
oning. Although numerous ICD-10 codes mention CO, the Tenth Revision of ICD
has only one code specific to CO:T58. While the latest revision contains codes that
indicate the manner or intent of injury or can identify fire-related exposures, much
detail has been lost in the ability to describe certain etiologic mechanisms (such
as motor vehicle-related injuries) that could be described previously using E-codes
related to CO. Hence, estimates of CO deaths classified using ICD-9 are not directly
comparable to estimates derived from ICD-10 coded data. In addition to limitations
regarding case ascertainment, data from NVSS do not become available to the public
in a timely manner and lack other etiologic detail necessary for CO surveillance. For
example, NVSS data do not contain sufficient detail to identify multiple victims of
the same incident of CO poisoning. Estimating CO-related death from the NVSS
requires a thorough understanding of the ICD coding rules and the limitations for
case identification imposed by selection criteria for specific years of data.
To summarize, the most important implication of the implementation of ICD-10 is
that the already limited etiologic detail in the NVSS has been further reduced, thereby
substantially decreasing its usefulness as national CO surveillance system. The use
of the NVSS for national mortality surveillance indicates that there are roughly 500
unintentional, NFR CO deaths per year in the United States. Follow-up epidemiolo-
gic investigations of these deaths, however are needed to provide us with enough
etiologic information to suggest meaningful public health interventions. Such invest-
igations remain a high priority until alternate sources of surveillance data can be made
available.
9.3.3 MORBIDITY LIMITATIONS

Data currently available for the surveillance of nonfatal CO poisoning are limited
by difficulties with case ascertainment. CO poisoning is characterized by nonspecific
signs and symptoms that are often ignored or attributed to another condition by the
public and medical professionals alike.11 In addition, those who experience nonfatal
CO poisoning but do not present for medical treatment are not counted in morbidity
estimates from medical records. Hence, all estimates of nonfatal poisoning are likely
to underestimate the actual incidence of CO poisoning. The variation of populations
and methods evident among datasets currently available suggest that no single source
can serve as a comprehensive surveillance source for CO poisoning morbidity in the
United States at this time.
Despite these limitations, the continuing morbidity and mortality associated
with acute CO poisoning in the United States necessitates the establishment
of a national surveillance system. The goals of a national surveillance system
8417: “8417_c009” — 2007/9/11 — 12:15 — page 245 — #13

for CO range from planning for rapid public health responses following dis-
asters and tracking the burden of CO poisoning over time, to improving our
understanding of exposure sources, related hazards, and facilitating research.5,36
To promote and facilitate surveillance at the local, state, and federal levels,
the National Workgroup for CO Surveillance has produced a summary report
describing the attributes of CO as a model environmental public health indicator
(http://www.maine.gov/dhhs/eohp/epht/documents/CO_White.pdf).5
9.4 CONCLUSIONS
Although the overall national rate of death from NFR CO declined steadily from 1968
to 1998 in the United States,19 the average annual rate of unintentional nonfire-related
(UNFR) CO-related death has remained relatively stable over the past 4 decades.17–19
From 1968 to 1998, the crude death rate for UNFR CO was 7.06 deaths per 1 million
person-years.19 The most recent analysis of mortality data from 2002 reported a crude
UNFR CO death rate of 1.8 deaths per 1 million person in the U.S.17 There are an
average of 494 accidental, NFR deaths and approximately 1,747 intentional deaths
due to CO-poisoning each year in the United States.19 Evidence from a recent eco-
logical study suggests the decrease in all CO-related deaths in the United States was
driven primarily by a reduction in deaths from exposure to motor-vehicle exhaust.9
This reduction has been attributed to the national implementation of the 1970 Clean
Air Act. In contrast to trends in national CO mortality, the incidence of nonfatal CO
poisonings rose, while rates of death fell from 1985 to 1996 in the United States27 The
annual number of nonfatal poisonings then decreased from 1996 to 2002.27 Published
estimates of ED visits due to UNFR CO poisoning suggest there are approximately
15,200 ED visits per year, with 1,676 resulting in subsequent hospitalization.18 Simil-
arly, unpublished estimates from the 2002 National Hospital Discharge Survey using
the CSTE-proposed case definition identified 1496 hospitalizations due to UNFR CO
poisoning in the U.S.37 Across studies, men and older adults (ranging from over 45
to over 65 years) are most at risk for unintentional death or injury from CO.17–19
Despite the lack of uniform national surveillance data, the burden of CO poisoning in
the United States may be summarized using estimates from a variety of related data
sources (Figure 9.2). The substantial health burden of unintentional CO poisoning
illustrated in Figure 9.3 suggests the need to put CO morbidity and mortality under
ongoing public health surveillance.
The de facto national surveillance system for CO poisoning in the United States has
been the NVSS. However, this system has several limitations for the surveillance of
CO poisoning. Although the data have taken two years to process following collection,
the recent implimentation of ICD-10 codes has further limited the utility of the NVSS
for CO surveillance due to the removal of several important ICD codes that denote
etiologic mechanism of poisoning. Public health professionals must consider these
recent changes in the ICD coding system when assessing CO-related mortality derived
from the NVSS. When estimating nonfatal CO poisoning, data from multiple sources
should be used to cross-validate estimates of morbidity, to reduce the likelihood of
under-estimation from any single source of data. To obtain a complete picture of
8417: “8417_c009” — 2007/9/11 — 12:15 — page 246 — #14

Suicides
12 120
CDR per1,000,000 person-
CO Emissions control timeline
10 100
Grams of CO emited per VMT

1970:Congress enacts Clean Air Act. CO
8 80 emissions standard at 34.0 g/mile.
years
1975: Catalytic converter introduced on new

6 60
passenger cars to meet new CO emissions
standard of 15 g/mile.
4 40
1978:1975 and newer model year cars
2 1975– catalytic converters required 20 make up 34% of the U.S.passenger
on all new passenger cars vehicle fleet.
0 0
1980: All new passenger cars required to
Unintentional deaths meet new CO emissions standard of 7.0
5 120 g/mile. 1975 and newer model year cars
make up 50% of US.passenger vehicle
Gramsof CO emited per VMT

CDR per1,000,000 person-
fleet.
4 100
80 1981:All newcars required to meet

3 new CO emissions standard of 3.4 g/mile.
years
60
2 1990:1975 and newer model year cars
40 make up 91% of the U.S.passenger vehicle
fleet.
1 20
1992: Standards setting emission limits for
0 0
carbon monoxide at temperatures <20°F
are established. Oxygenated gasoline is
68
70
72
74
76
78
80
82
84
86
88
90
92
94
96
98
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
introducedin cities with high CO levels.
CDR: Motor vehicle CDR: Nonmotor vehicle COEmissions (g/mile)

Source: EPA CO emissions inventory
data,and Fact Sheet OMS-12.
FIGURE 9.2 Annual crude death rates (CDR) from carbon monoxide poisoning, and annual
estimated grams of CO emitted per vehicle mile traveled (VMT), United States, 1968–1998
(From Mott et al., JAMA, 288, 988–995, 2006.)
494 deaths17
1,496 hospitalizations37
13,201 emergency department visits18
15,633 calls to poison control centers27
FIGURE 9.3 The nonfire-related, carbon monoxide poisoning pyramid, 2002, U.S.
CO poisoning, surveillance systems must be established in a variety of settings and

must be comprehensive, capturing the incidence, causes, and circumstances of both
nonfatal and fatal poisonings in a timely manner. Likewise, innovative approaches
may be necessary to gather surveillance data on nonresident populations, such as
8417: “8417_c009” — 2007/9/11 — 12:15 — page 247 — #15

undocumented immigrants who may be at increased risk for CO poisoning, but for
whom the calculation of valid rates is challenging. Ultimately, the standardization of
an ICD-coded case definition for CO-related morbidity and mortality is a necessary 1st
step to describe the national burden of CO poisoning, to better target interventions, and
to evaluate the impact of public health prevention efforts.5 Until additional resources
can be identified and allocated to surveillance at the federal, state, and local levels,
epidemiologic investigations of recent CO-related deaths will need to remain the
foundation for obtaining the etiologic information necessary to mount effective public
health campaigns.
References
1. U.S. Centers for Disease Control and Prevention. Poisonings fact sheet.
ONLINE. 2006. National Center for Injury Prevention and Control. Available:
http://www.cdc.gov/ncipc/factsheets/poisoning.htm. (Oct. 1, 2006).
2. Cobb, N. and R. A. Etzel. Unintentional carbon monoxide-related deaths in the United
States, 1979–1988. JAMA 266: 659–663, 1991.
3. U.S. Centers for Disease Control and Prevention. Carbon monoxide poisoning after
Hurricane Katrina—Alabama, Louisiana, and Mississippi, August–September, 2005.
MMWR, 54: 996–998, 2005.
4. Daley, W. R., A. Smith, E. Paz-Argandona, J. Malilay, and M. McGeehin. An outbreak
of carbon monoxide poisoning after a major ice storm in Maine. J. Emerg. Med. 18:
87–93, 2000.
5. National Workgroup on Carbon Monoxide Surveillance. Carbon monoxide: A
model environmental public health indicator. ONLINE. 2006. Available: http://
mainegov-images.informe.org/dhhs/eohp/epht/CO_WHITE.pdf. (Oct. 1, 2006).
6. Nsubuga, P., M. White, and S. Thacker et al. Public health surveillance: A tool for
targeting and monitoring interventions. In Disease Control Priorities in Developing
Countries, 2006, D. Jamison, J. Breman, A. Measham, G. Alleyne, M. Claeson,
D. Evans, P. Jha, A. Mils, and P. Musgrove, eds., pp. 997–1015, New York, NY:
Oxford University Press.
7. U.S. Centers for Disease Control and Prevention. Unintentional carbon monox-
ide poisonings in residential settings—Connecticut, November 1993–March 1994.
MMWR 44: 765–767, 1995.
8. Girman, J., Y. Chang, S. Hayward, and K. Liu. Causes of unintentional deaths from
carbon monoxide poisonings in California. W. J. Med. 168: 158–165, 1998.
9. Thacker, S. B., D. Stroup, R. Parrish, and H. Anderson. Surveillance in environ-
mental public health: Issues, systems, and sources. Am. J. Publ. Health 86: 633–638,
1996.
10. Institute of Medicine, Surveillance and data. In: Reducing the Burden of Injury:
Advancing Prevention and Treatment, Bonnie, R., C. Fulco, and C. Liverman, eds.,
1999, 60–81. National Academy Press, Washington, DC.
11. Raub, J., M. Mathieu, N. Hampson, and S. Thom. Carbon monoxide poisoning—A
public health perspective. Toxicology 145: 1–14, 2000.
12. U.S. Centers for Disease Control and Prevention. U.S. census populations with
bridged-race categories. ONLINE. 2004. National Center for Health Statist-
ics. Available: http://www.cdc.gov/nchs/about/major/dvs/popbridge/popbridge.htm.
(Oct. 1, 2006).
8417: “8417_c009” — 2007/9/11 — 12:15 — page 248 — #16

13. World Health Organization. Manual of the International Statistical Classification of

Diseases, Injuries, and Causes of Death, 1967: 8th revision. Geneva.
14. World Health Organization. Manual of the International Statistical Classification of
Diseases, Injuries, and Causes of Death, 1977: 9th revision. Geneva.
15. U.S. Centers for Disease Control and Prevention. Public use data file documentation:
multiple cause of death for ICD-9, 1994. 1996, Hyattsville, MD: National Center for
Health Statistics.
16. U.S. Centers for Disease Control and Prevention. Public use data file document-
ation: mortality for ICD-10, 1999. 2002, Hyattsville, MD: National Center for
Health Statistics. Available: http://0-www.cdc.gov.mill1.sjlibrary.org/nchs/data/dvs/
Mort99doc.pdf. (Oct. 1, 2006).
17. U.S. Centers for Disease Control and Prevention. Update of unintentional, non-fire,
carbon monoxide-related death rates, United States, 1999–2002. 2006, Unpublished
data, Air Pollution and Respiratory Health Branch.
18. U.S. Centers for Disease Control and Prevention. Unintentional non-fire-related
carbon monoxide exposures in the United States, 2001–2003. MMWR 54: 36-39,
2005.
19. Mott, J., M. Wolfe, C. Alverson et al. 2002. National vehicle emissions policies and
practices and declining U.S. carbon monoxide-related mortality. JAMA 288: 988–995,
2006.
20. U.S. Centers for Disease Control and Prevention. Deaths from motor-vehicle related
unintentional carbon monoxide poisoning, Colorado, 1996, New Mexico, 1980, 1995,
and United States, 1979–1992. MMWR 45: 1029–1032, 2006.
21. Consumer Product Safety Commission. Non-fire carbon monoxide fatalities asso-
ciated with engine driven generators and other engine driven tools in 2002
through 2005. ONLINE. 2006, August. Available: http://www.cpsc.gov/library.
(Oct. 1, 2006).
22. Baker S., B. O’Neill, M. Ginsburg, and G. Li. The Injury Fact Book, 1992, 2nd ed.,
New York: Oxford University Press.
23. Consumer Product Safety Commission. Non-fire carbon monoxide deaths associated
with the use of consumer products: 2002 annual estimates. ONLINE. 2006, Available:
http://www.cpsc.gov/library.
24. Consumer Product Safety Commission. Incidents, deaths, and in-depth invest-
igations associated with carbon monoxide from engine-driven generators and
other engine-driven tools, 1990–2004. ONLINE. 2005, December. Available:
http://www.cpsc.gov/library.
25. Kessler, E. and T. Schroeder. The NEISS sample: Design and implementation. 2000.
Washington, DC: U.S. Consumer Product Safety Commission.
26. U.S. Centers for Disease Control and Prevention. Monitoring poison control center
data to detect health hazards during hurricane season—Florida, 2003–2005. MMWR
55: 426–428, 2006.
27. Hampson, N. Trends in the incidence of carbon monoxide poisoning in the United
States. Am. J. Emerg. Med. 23: 838–841, 2005.
28. Watson, W., T. Litovitz, C. Rubin et al. Toxic Exposure Surveillance System [abstract].
In: Syndromic Surveillance: Reports from a National Conference, 2003. MMWR 53:
262, 2004.
29. Hampson, N. and C. Little. Hyperbaric treatment of patients with carbon monoxide
poisoning in the United States. Undersea Hyperb. Med. 32: 21–26, 2005.
30. Hampson, N. Emergency department visits for carbon monoxide poisoning in the
pacific northwest. The J. Emerg. Med. 16: 695–698, 1998.
8417: “8417_c009” — 2007/9/11 — 12:15 — page 249 — #17

31. McCaig, L. and C. Burt. Poisoning-related visits to emergency departments in the

United States, 1993–1996. J. Toxicol. 37: 817–826, 1999.
32. Mah, J. C. Non-fire carbon-monoxide deaths and injuries associated with the
use of consumer products: annual estimates - October 2000. ONLINE. 2001,
August. U.S. Consumer Product Safety Commission. Available: http://www.cpsc.gov/
library/co00.pdf.
33. Schaplowsky, A., F. Oglesbay, and J. Morrison et al. Carbon monoxide contamina-
tion of the living environment: A national survey of home air and children’s blood.
J. Environ. Health 36: 569–573, 1974.
34. Pew Environmental Health Commission. America’s Environmental Health Gap: Why
the Country Needs a Nationwide Health Tracking Network, 2000. Technical Report.
Baltimore, MD: Pew Environmental Health Commission.
35. Teutsch, S. Considerations in planning a surveillance system. In Principles and Prac-
tice of Public Health Surveillance, 2000, S. Teutsch and R. Churchill, eds., pp. 17–29.
New York: Oxford University Press.
36. National Workgroup on Carbon Monoxide Surveillance. Project to evaluate carbon
monoxide surveillance CSTE and STIPDA case definitions with hospital data. 2006
Unpublished report. Obtained October, 2006 from the Air Pollution and Respiratory
Health Branch, U.S. Centers for Disease Control and Prevention.
37. U.S. Centers for Disease Control and Prevention. Estimates of unintentional, non-fire,
carbon monoxide-related hospital discharges, National Hospital Discharge Survey,
United States, 2002. 2006 Unpublished data, Air Pollution and Respiratory Health
Branch.
8417: “8417_c009” — 2007/9/11 — 12:15 — page 250 — #18

10 Carbon Monoxide
Sensors and Systems
Kosmas Galatsis and Wojtek Wlodarski
CONTENTS
10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251

10.2 Sensor Technologies for Carbon Monoxide Gas Detection . . . . . . . . . . . . . . . 253
10.2.1 Semiconducting Metal Oxide (SMO) Gas Sensors . . . . . . . . . . . . . . . 254
10.2.2 Optical Gas Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
10.2.3 Electrochemical Gas Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
10.3 Sensor Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
10.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
10.6 Appendix 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
10.1 INTRODUCTION
Carbon Monoxide (CO) is one of the most important gases in the field of sensor
technology. This is because its toxicity combined with its properties of being odorless,
colorless, tasteless, and nonirritating to the respiratory tract. Attempts at detection of
CO date back to the famous French physiologist, Claude Bernard, circa 1846,1 who
performed experiments with CO poisoning dogs. Small birds and mammals were
used for decades in mines as living CO detectors. CO has been called the “silent
killer,” the “stealthy-poison,” and even the “smart poison” because it enters the body
without notice and leaves so quickly with little trace. See an earlier discussion of CO
detectors by Kwor in Carbon Monoxide Toxicity, 2000.
Today’s CO detectors/alarms are small electronic devices. Such devices are
installed in homes near heating devices or in garages where sources of CO such as
combustion burners and/or motor vehicles may potentially pollute the breathing space.
If a sufficient level of CO is detected, the device audibly alarms, giving occupants a
chance to ventilate the area or safely vacate. Unlike smoke, CO is undetectable by
the unaided human senses, and hence, people often find themselves in environments
polluted with CO without knowing it.
251
8417: “8417_c010” — 2007/9/11 — 12:15 — page 251 — #1

During the mid 1990s, the installation of residential, wall-mounted CO alarms

grew rapidly in the United States. This was a result of the availability of low cost
CO alarms, marketing campaigns, education campaigns by the American Lung Asso-
ciation and advocacy by the US Consumer Products Safety Commission (CPSC).
In 1995, Chicago was the first municipality in the United States to mandate CO
alarms in all single and multiple family housing, and in some special use buildings
such as schools, churches, theaters, museums. By 1998, approximately 20 million
units had been shipped and an estimated 8–15% of homes had at least one CO alarm
installed.2
CO alarms today retail for $20–$60 (U.S.), are widely available, and can either be
battery-operated or AC (mains) powered. Other locales such as Massachusetts have
followed Chicago in acknowledging CO poisoning as a priority and have enacted
state legislation which requires CO detectors in homes. Such domestic detectors are
required to meet performance standards such as BS EN 50291 (UK and Europe), UL
2034 (USA), and CSA 6-19-01 (Canada). The Canadian standard has been argued to
be superior as it requires “time of manufacture” and test for “lifetime reliability.” As
an example, the British–European standard requires domestic CO alarms to behave
in the following way:
• No alarm within 60 min at 45 ppm CO

• Alarm within 30 min, but not less than 10 min, at 150 ppm CO
• Alarm within 6 min at 350 ppm CO
• Recovery from the alarm state within 6 min in clean air
Although these design standards help set a benchmark for performance, they do
not specify the degree to which alarms must maintain their performance, and hence
has contributed to CO detectors known to becoming notoriously unreliable with age.
Poor sensitivity at low humidity was a major problem in one field experiment, showing
as much as 79% of alarms failed when tested at 5% RH and that 3 of the 10 brands
tested worked well.3
Other than the domestic need for CO alarms, another important CO sensing
application that has recently gained considerable interest is vehicle cabins. Exhaust
pollutants find their way into the cabin through the ventilation system, also known as
the heating, ventilation, and air conditioning (HVAC) system. Independent studies4−7
have shown that vehicle cabins commonly show concentrations of toxic gases such as
CO, hydrocarbons (HC), volatile organic compounds (VOC), and oxides of nitrogen
(NOx ) higher than safety limits set by Occupational Safety and Health Administration
(OSHA) and World Health Organization (WHO). Among the array of toxins found
in vehicle exhaust gases, CO is the most deadly poison. It is a major subject of
overlooked issues concerning motor-vehicle cabin air quality and suicides involving
CO.8,9 Of the 2320 suicides registered for the year 2002 in Australia, 416 persons
(18%) died from use of motor-vehicle exhaust gases.10 By understanding the prob-
lem of CO pollution within vehicle cabins, CO sensor technology can be employed
to circumvent the danger. See Chapter 9 of Carbon Monoxide Toxicity, 2000, and
Appendix 1 for additional discussion of this issue.
8417: “8417_c010” — 2007/9/11 — 12:15 — page 252 — #2

Carbon Monoxide Sensors and Systems 253
10.2 SENSOR TECHNOLOGIES FOR CARBON

MONOXIDE GAS DETECTION
At the heart of any gas alarm and detecting system are the sensors. The sensor detects
target gases, and then converts the information into an electrical signal for processing.
There are numerous ways to sense gas. However, as cost, size, and simplicity are
critical sensor attributes, three main sensing technologies have dominated domestic
CO alarms. These are
1. Semiconducting metal oxide (SMO) technology

2. Electrochemical (EC) technology
3. Infra-red/optical technology (IR)
SMO gas sensors are currently the smallest CO sensors available. These sensors
have a small heated element, causing reducing/oxidizing gases to react with the
surface of a metal oxide film, changing the semiconductor’s conductivity propor-
tionally to the gas concentration. Electrochemical gas sensors have electrodes placed
in contact with a liquid electrolyte to form an Electrochemical sensor. As the gas
diffuses, it reacts with the working electrode, changing its electrical potential propor-
tional to the gas concentration. And third are IR sensors, where the optical sensing
element undergoes light transmission changes when exposed to the target gas.
Table 10.1 compares the technologies against seven key sensor device
criteria. Domestic CO alarms predominantly employ either the semiconductor or
Electrochemical sensor. Semiconductor based CO sensors have penetrated the market
with companies such as Figaro (Japan), Microchemical (MiCS) (Switzerland), and
FiS (Japan). Some Electrochemical sensor manufactures include City Technology
(U.K.), Monox (UK), and Kidde (USA). Optical CO sensors have been pioneered by
Quantum (USA). For vehicle cabin air quality monitoring installed within the HVAC
systems of vehicles, metal oxide sensors have dominated as they are small, have a
long lifetime and the technology allows for the sensor element to be conveniently
TABLE 10.1
Comparison of Three Gas-Sensing Technologies with Respect to
Desirable Carbon Monoxide Domestic and Vehicle Air Quality
Monitoring Criteria
Criteria Infra Red—Optical Electrochemical Metal Oxide
Cost <US$15 <US$10 <US$5
Life time >6 years 2–5 years >6years
Sensitivity Very good Very good Very good
Selectivity Excellent Very good Poor
Response time Seconds Seconds Seconds
Size Medium Medium Small
Ease of use Good Excellent Excellent
8417: “8417_c010” — 2007/9/11 — 12:15 — page 253 — #3

optimized for various toxic gases. However, for aftermarket vehicle CO detectors,
optical sensors from Quantum have dominated.
10.2.1 SEMICONDUCTING METAL OXIDE (SMO) GAS SENSORS

SMO gas sensors are relatively small, reliable, durable, and have low cost. Traditio-
nally, the major disadvantages of SMO as gas sensitive devices has been their poor
gas selectivity, and the influences of humidity and temperature.11,12 The introduction
of noble metal catalysts (such as platinum and palladium), filters (activated carbon),
and modification to the SMO microstructure and nanostructure has enabled SMO
sensor manufacturers to improve selectivity and stability performance.13 For redu-
cing gases such as CO, molecules react with adsorbed oxygen ions (from ambient
air) on the surface of the oxide. The adsorbed oxygen loses its electron by reacting
with the reducing gas molecule, thereby increasing the films conductivity. A simple
model consisting of three possible reactions is shown below14,15 :
2CO + O2− ←→ 2CO2 + e−

CO + O− ←→ CO2 + e−
CO + O2 ←→ CO2 + 2e−
From this reaction it is obvious that a change in ambient oxygen concentration

will also change the rate of this redox process and influence the output signal of
the sensor. The relationship between film conductivity (σ ) and gas concentration (c)
follows a power law that can be described by16 :
σ = kcn
where k is a measured proportionality constant unique to the film/sensor and the

exponent n can range from 0.3 to 0.8. Owing to this intrinsic nonlinear semicon-
ducting nature, linearization circuitry within hardware/software is usually required.
In addition, for the SMO material to react with a gas, the material is elevated to
temperatures between 90◦ C and 250◦ C enabling the reduction/oxidization process to
occur. Elevating the sensor to high temperature requires an integrated heater circuit
to be fabricated below or adjacent to the sensing element. Owing to this high tem-
perature requirement, SMO gas sensors require relatively high power consumption.
Traditional SMO sensors fabricated on alumina substrates typically consume above
350 mW. One way of reducing power consumption is by fabricating gas sensors using
a thin Si membrane as done by MiCS in Switzerland. Power consumption of the MiCS
sensor is about 30–50 mW. Typically for detecting CO gas, the sensitive film material
used in SMO sensors is tin oxide (SnO2 ). Other transitional metal oxides such as
tungsten oxide (WO3 ), indium oxide (In2 O3 ), chromium titanium oxide (Cr2 TiO3 )
fabricated at a thickness between 200 nm and 10 µm have also been shown to be
effective CO sensing materials.17−24
For automobile air quality monitor (AQM) applications, MiCS manufactures dual
element sensors for detecting both reducing gases such as CO and HC’s and oxidizing
8417: “8417_c010” — 2007/9/11 — 12:15 — page 254 — #4

gases such as NO2 and O3 . This allows for detection of gasoline pollution from cars
and motorbikes, and diesel pollution from diesel-powered cars as well as trucks and
buses. The Si based sensor chips (Figure 10.1a) are bonded to either transistor outline
(TO) packages or SMD (surface mount device) packages. Figure 10.1b shows two
sensor chips on the same housing developed for the automobile industry, capable
of detecting both CO and NOx . Typically, the sensor is integrated on a printed cir-
cuit board with peripheral electronics and packaged as a CO alarm for domestic or
automobile use.
(a) (b)
(c) (d)
FIGURE 10.1 Photographs of commercial CO gas sensors. (a) The MiCS Micro-Electro-
Mechanical Systems (MEMS) chip 2.1 × 2.3 mm. This chip is made up of a thin Si membrane,
which is a micromachined silicon platform that includes an integrated heater and interdigital
electrodes. The sensitive layer is a thin (about 200 nm in. thickness) metal oxide polycrystalline
film; (b) Dual MiCS MEMS chips mounted and bonded onto a TO5 package. These sensors
are used in automobile applications to control the HVAC system, (c) The FIS SB series sensing
element made up of a platinum coil heater (as shown in the inset) with a sensing platinum
electrode in the middle of the coil. The structure is coated and covered with sensitive semi-
conducting metal oxide, tin dioxide (SnO2 ). The SnO2 material is made up of many small
particles in the size range of submicron to several tens of microns, (d) The FIS SB series sensor
package for integration within CO detectors. These sensor structures are encapsulated in nickel
plated brass (as shown in the inset) with an attached active charcoal filter and then enclosed in
an outer plastic housing. These sensors are typically employed within domestic CO detectors.
(Courtesy of MicroChemical of Switzerland and FIS of Japan).
8417: “8417_c010” — 2007/9/11 — 12:15 — page 255 — #5

Figure 10.1c and d shows the sensor design of the FiS (Japan) sensor employed
in domestic CO detectors. Compared to MiCS, the FiS design is based on a tin
dioxide coated over a platinum coil and a sensing electrode. The coil heats the struc-
ture at an elevated temperature while the working electrode senses the conductance
changes. Figure 10.2a shows the FiS sensor responding to CO gas. Some important
characteristics to note are: (1) The sensor signal returns to its original baseline. This
characteristic ensures that the sensor remains calibrated, free of drift that will result
in signal errors. (2) The response time and decay time of the sensor is a few minutes
which is adequate for either domestic or automobile applications. (3) The magnitude
of sensor signal changes as a function of concentration (i.e., exponent n, based on
Equation 10.1). This permits a greater signal to noise ratio and signal dynamic range.
Figure 10.2b is a typical selectivity test performed to determine cross-sensitivity
to other gases. As shown, the SMO sensor does not offer absolute CO selectivity.
Improving selectivity requires optimizing calatyst concentration (such as platinum),
optimizing material annealing temperature and optimizing crystal and grain proper-
ties. Figure 10.2c shows a stability test of the FiS CO sensor. The slight change in
baseline is a result of the complicated nature of the crystallization process due to oper-
ation at elevated temperatures. The sensor is extremely stable over 1000 days—even
after 1000 days a baseline change of only 10% occurs, which is usually mitigated
by intelligent microprocessor algorithm programming. Improving this characteristic
of SMO sensors is a great challenge that drives the active research disciplines of
semiconductor metal oxide gas sensors.
10.2.2 OPTICAL GAS SENSORS

IR-based sensors are relatively physically small, consume low power, are select-
ive, and are rapidly decreasing in cost. These sensors are considered as solid state
and have a lifetime of over 6 years (depends on IR source degradation/failure) with
good resolution, relatively high selectivity, and broad dynamic range. These sensors
identify gases by taking advantage of a gas’s unique IR absorption spectra. Most
gases (more than one type of atom) can be detected by measuring their absorption
at a particular IR wavelength, which corresponds to the resonance of the molecular
bonding between dissimilar atoms. Figure 10.3a shows an IR absorption spectrum of
some common gases. For example, to detect CO, the wavelength at which one carbon
atom and one oxygen atom resonate in a carbon monoxide molecule is 4.7 µm.
Therefore, the IR system will be filtered to detect radiation at a bandwidth centered
at 4.7 µm.
There are certain basic components common to all IR gas sensors: an IR source
(e.g., incandescent lamp), an IR detector (e.g., thermopiles, pyroelectric detectors,
photodiode), a means to select appropriate wavelengths (e.g., band pass interference
filter) and a sample cell. The simple sensing setup is shown in Figure 10.3b. The
IR source is at one end and the IR sensor at the other. The band pass optical filter
must correspond with the absorption wavelength of the gas being measured. As the
concentration of the gas being measured increases, the output signal from the sensor
reduces as the IR is absorbed by the target gas molecules. The relationship between
8417: “8417_c010” — 2007/9/11 — 12:15 — page 256 — #6

(a)
5.0
CO 100 ppm clean air CO 300 ppm clean air CO 100 ppm clean air
4.0
VRL (V )
3.0
2.0
1.0
0
0 5 10 15 20 25 30
(b) Time (min)
100000
10000
1000
Rs (Kohm)
100
10
1
Air
CO 30 ppm
CO 60 ppm
CO 100 ppm
CO 200 ppm
CO 400 ppm
Methane 500 ppm 2hr

Butane 300 ppm 2hr
Heptane 500 ppm 2hr
Ethyl acetate 200 ppm
Isopropyl alcohol 200
Carbon dioxide 5000p
Air
CO 30 ppm
CO 60 ppm
CO 100 ppm
CO 200 ppm
CO 400 ppm
FIGURE 10.2 FIS SB series CO sensor data showing, (a) exposure to 100, 300, and 100 ppm
CO with clean air cycles. The x-axis represents time (minutes) and the y-axis represents the
proportional voltage response from the sensing element. The sensor is stable when responding
to CO gas, returning to baseline within a few minutes. In addition, the repeatable response
characteristic at 100 ppm also is another desirable sensing attribute, (b) selectivity to other
gases following the UL2034 specification. The largest cross-sensitivity is with heptane (error
bars overlap with an equivalent CO error signal of about 30 ppm), (c) long-term stability. The
sensor’s stability over 1000 days shows extremely stable characteristics, even after 1000 days
a baseline change of only 10% occurs, which is usually mitigated by intelligent microprocessor
algorithm programming. (Courtesy of FIS, Japan).
8417: “8417_c010” — 2007/9/11 — 12:15 — page 257 — #7

(c)
10
R /R0(CO100)
R /Ro R /R0(CO300)
0.1
0 100 200 300 400 500 600 700 800 900 1000 1100
Time (days)
FIGURE 10.2 Continued.
IR transmission, I, and gas concentration, c, can be explained by Beers Law of

Absorption:
I = Io e(−kc l)
where I is the intensity of IR radiation at the IR detector, Io is the IR radiation

emitted from the IR source, k is the absorption coefficient, and l is the optical path
length.
To improve the signal to noise ratio, an improved IR detection setup as shown in
Figure 10.3c was developed. The difference is that a reference detector was added to
compensate for humidity, vibration, source intensity deterioration, detector contam-
ination, vibration, and aging. As a result, a dual beam topology is typically employed
with most IR gas sensors. A reference detector senses IR at a neutral wavelength
where almost no absorption takes place (i.e., 4 µm). By taking the ratio of both
detector voltage U1 and reference signal U2 , the common Io coefficient is cancelled,
and the target gas signal component remains which corresponds to the target gas
concentration.
The Quantum Group (US) is the leading manufacturer of optical gas sensors
for domestic CO alarms. They have developed a unique type of solid state IR gas
sensor based on the “biomimetic” phenomena. The company has been successful in
developing a broad range of domestic CO alarms and it is the first company to offer an
aftermarket CO detector for vehicle safety applications. The IR-based “biomimetic”
sensors are designed to replicate the CO uptake by hemoglobin in the blood, hence
the name “biomimetic.” In doing so, the sensing element can be set to alarm based
on the replicated blood level of carboxyhemoglobin (COHb).
Figure 10.4a shows the elements of the biomimetic CO sensor. The patented
sensing elements are made from a porous transparent disk coated with a monolayer
8417: “8417_c010” — 2007/9/11 — 12:15 — page 258 — #8

(a)
1014
Absorption (arbitrary value)

NO
(5.3)
101 H2O
(1.4) SO2
(4) CO
(4.7)
HC
(3.4)
H2O CO2
0 (1.9) (4.3)
1 2 3 4 5 6
Wavelength (µm)
(b) IR bandpass filter
IR detector
IR source
(c)
U1∝I0 exp (kcl)
U2∝I0
Gas filter
Reference filter
IR light
FIGURE 10.3 IR sensor design that is based on, (a) the absorption bands of various gases in
the IR region, (b) A simplified diagram of a single beam IR absorption gas detector and, (c) an
improved version of the dual beam arrangement makes the gas detector insensitive to source
performance deterioration. (Courtesy of PerkinElmer Optoelectronics, Germany).
of supramolecular organometallic complex. This complex is formed through a self-

assembly process to generate the sensing elements that mimic hemoglobin. Upon
exposure to CO, one or both of the sensing elements changes its spectral character
and absorbs photons of light at a rate dependent on the concentration of CO in the
surrounding environment. The sensing elements reverse their spectral shift by a self-
generation process whose rate depends upon the decrease of CO in the environment.
8417: “8417_c010” — 2007/9/11 — 12:15 — page 259 — #9

(a) Retainer
clip
Filter
element
Red
sensing Yellow
element sensing
Sensor element
housing
Window Photodiode
IR lED
Sensor
holder
(b)
FIGURE 10.4 A commercialized optical CO gas sensor, (a) components and structure of
Quantum’s biomimetic CO gas sensor, and (b) a photograph of the CO sensor cell incorporated
within Quantum’s line of domestic and automobile CO detectors. (Courtesy of the Quantum
Group, USA).
8417: “8417_c010” — 2007/9/11 — 12:15 — page 260 — #10

CO
Preconditioning Exposure Recovery period
120
110
100
90
80
70
Temperature = 0°C/15% RH
I%
60
50 Temperature = 25°C/30% RH
40
30 Temperature = 50°C/30% RH
20
10
0 RH = Relative
0
60
120
180
240
300
360
420
480
540
600
humidity
Time (min)
FIGURE 10.5 Response curves of a Quantum biomimetic sensor towards 100 ppm CO for
90 min with varying ambient temperatures from 0◦ C to 50◦ C and relative humidity from 15%
to 30% RH. It can be seen that varying these parameters influences the response time and
baseline of the optical sensor. (Courtesy of the Quantum Group, USA).
This mechanism acts as a variable IR bandpass filter. By monitoring the rate of change
in the amount of light transmitted through the sensing elements, the concentration
of the CO in the surrounding environment can be determined accurately. The sens-
ing elements are held in optical alignment by the sensor-housing placed between
an IR light emitting diode (LED) and a photodiode. Pulses of light emitted by the
LED pass through the first sensor-housing window and are attenuated by the sens-
ing elements. The attenuated light exits through the second sensor-housing window
and is then detected by the photodiode. The light transmittance follows Beer’s Law
(Equation 10.2).
Figure 10.5 shows the response of the Quantum Group’s biomimetic sensor over
90 min to a 100 ppm concentration of CO at varying temperatures and relative
humidity. After the 90 min, the CO is removed and the sensor begins regenerat-
ing. When exposed to CO, the sensor rapidly absorbs photons at 940 nm. In a fixed
alarm point detector, a value of 30% is typically set as the alarm point. Thus in such a
detector, the alarm would be triggered within 40 min of exposure, which is well within
the UL safety guidelines. When the sensor is again exposed to clean air (marked as
recovery time), the biomimetic component begins a self-regenerating process. As the
sensor reverses its spectral shift, the signal increases and within hours the sensor has
fully recovered. Although the response time is slower than the SMO sensor, incor-
porating sensing algorithms based on the rate of change of signal, improves response
time.
10.2.3 ELECTROCHEMICAL GAS SENSORS

Electrochemical gas sensors are also small electronic devices. A City Technology CO
sensor commonly found in CO domestic alarms is shown in the inset of Figure 10.6.
In their simplest form they are comprised of two electrodes: sensing and counter,
8417: “8417_c010” — 2007/9/11 — 12:15 — page 261 — #11

Capillary diffusion
barrier
Sensing
electrode
Separator
Counter
electrode
Current
collectors
Electrolyte
Sensor pins
FIGURE 10.6 A simplified schematic of a two electrode electrochemical cell manufactured

by City Technology. The inset show photographs of the complete sensor package. These sensors
are employed within domestic CO detectors. (Courtesy of City Technology, U.K.).
separated by a thin layer of electrolyte. The structure is enclosed in a plastic housing

that has a small capillary tube to allow gas entry to the sensing electrode and
includes pins which are electrically attached to both electrodes and allow easy external
interface. These pins may be connected to a simple resistor circuit that allows the
voltage drop resulting from any current flow to be measured. Gas diffusing into
the sensor is either oxidized or reduced at the sensing electrode and, coupled with
a corresponding (but converse) counter reaction at the other electrode, a current is
generated through the external circuit. Since the rate of gas entry into the sensor is
controlled by the capillary diffusion barrier, the current generated is proportional to
the concentration of gas present outside the sensor. Of great importance to any elec-
trochemical gas sensor is the design of the diffusion barrier, which limits the flow of
gas to the sensing electrode. The electrode is therefore able to react with all target gas
as it reaches its surface, and still has electrochemical activity in reserve. The reactions
that take place at the electrodes in a CO sensor are:
Sensing: CO + H2 O− CO2 + 2H+ 2e−

Counter: 12 O2 + 2H+ + 2e− > H2 O
And the overall reaction is: CO + 12 O−
2 > CO2
Similar reactions take place for all other toxic gases that are capable of being
electrochemically oxidized or reduced. From the reaction at the counter electrode, it
is evident that oxygen is required for the current generation process to take place. This
is usually provided in the sample stream by air diffusing to the front of the sensor, or by
diffusion through the sides of the sensor (a few thousand ppm is normally sufficient).
8417: “8417_c010” — 2007/9/11 — 12:15 — page 262 — #12

However, continuous exposure to an anaerobic sample of gas may result in signal

drift, despite the oxygen access paths which may cause the sensor to be poisoned.
Similar to SMO gas sensors, electrochemical gas sensors are also affected by
temperature variations. The baseline signal of most electrochemical sensors tends to
increase exponentially with temperature, approximately doubling for every 10◦ C rise
in temperature which proves problematic for domestic applications as the baseline
shift with temperature could seriously affect the ability to measure these gases accur-
ately and result in false alarming. Nevertheless, by compensating for this drift either
in the hardware or software, such temperature influences can be reduced.
10.3 SENSOR SYSTEMS

The CO sensor is the main component within all domestic CO detectors. Support
electronics are also required to provide the sensor with intelligence so that it will
actuate alarms according to compliant standards. Most detectors incorporate at least
one microprocessor that allows them to be quickly reprogrammed and the behavior
of the alarm to be altered to suit various applications or standards. For domestic
applications, CO alarm design and alarm requirements are well defined by associated
performance specifications. However, in emerging CO and air quality monitoring
applications such as monitoring vehicle cabin air quality, specifications, and stand-
ards have yet to evolve. Vehicle cabin air quality concerns are usually generated
by the following four scenarios: (1) Pollutant gases entering the vehicle through
the ventilation system, (2) A lack of fresh airflow resulting in low oxygen and high
carbon dioxide concentrations due to occupant respiration, (3) Pollutant gases enter-
ing from the external environment through window openings, imperfect seals, and
other holes, and (4) Toxic gases entering the vehicle cabin by redirected exhaust
fumes for self-harm (i.e., suicide) purposes.
Currently, no system or aftermarket product addresses all four vehicle AQM
concerns. Only two commercial AQM solutions currently exist for vehicles: (1) The
most common are AQM systems controlling HVAC ventilation flaps, and (2) Less
common are aftermarket toxic gas alarms for vehicle cabin applications, such as
that commercialized by the Quantum Group (U.S.). Currently, the demand for AQM
systems is driven by the increasing concern for passenger safety, health, comfort, and
by automakers aiming for features and attributes that differentiate their vehicles. In
turn, this growth has increased demand for reliable automotive air quality sensors.
Figure 10.7 shows a simplified view of an AQM system controlling the HVAC vent-
ilation flap. External gases enter the vehicle cabins through the ventilation system.
Mounted in the air intake of the HVAC system, the AQM sensor sends a signal to
the fresh air inlet flap to close when pollutant gases are detected and automatically
reopen when the external air quality returns to an acceptable level. Although a driver
could close the air inlet manually, forgetting to reopen it could cause the oxygen con-
centration in the cabin to decrease and carbon dioxide levels to increase. Therefore,
a compromise must be reached. One way of tackling the problem to implement with
the system an air quality factor. For instance, the absolute concentration of particular
gas (Cx ) in the vehicle cabin is dependent on the exhaust flow rate (F), time (T ), cabin
8417: “8417_c010” — 2007/9/11 — 12:15 — page 263 — #13

O2
CO2
3. Ventilation system/flap
OPEN CLOSED 2. Electronics/algorithms
1. Sensor response
CO
NOx
HC
Particulates
FIGURE 10.7 An overview of a typical automobile air quality monitor (AQM) employed
within a heating, ventilation and air conditioning (HVAC) system. When the ventilation flap
is open, dangerous pollutants such as CO and NOx may enter the cabin. To mitigate this,
electronics automatically close the ventilation flap. High carbon dioxide and low oxygen con-
centrations may result through occupant respiration. High carbon dioxide and low oxygen
concentrations are dangerous because they induce fatigue and drowsiness, reducing driver
attention and response times.
volume (V ), and cabin seal (S). Therefore,
Cx = f (F, T , V , S)
Concentrations of carbon monoxide (CCO ), and oxygen (CO2 ) have been identified
as important gas species contributing to poor cabin air quality. The summation of
each absolute gas species concentration gives rise to an air quality factor (AQcabin )
such as:
AQcabin = αCO + δ(CO2 )−1
Where α, and δ are proportionality coefficients. It should be noted that other gas
species such as hydrocarbons and nitrogen oxides have been ignored. Absolute
threshold limits could then be set for scenarios such as suicide (AQsuicide ) and driver
fatigue (AQfatigue ). For increased reliability and effective suicide attempt identifica-
tion, the change of air quality with time (dAQcabin /dt) should also be incorporated
into the driver fatigue and suicide detecting algorithms:
dAQcabin ∂CCO ∂(CO2 )−1

=α +δ
dt ∂t ∂t
An alarm threshold, dAQsuicide /dt, could also be incorporated as done so by

Quantum Group. Therefore, the cabin gas-sensing system should include both
absolute and changing air quality factors, to determine if alarms need to be activated.
Software solutions to improve CO detectors are commonplace. In addition,
rate of change, humidity compensation (through humidity sensors) and temperature
8417: “8417_c010” — 2007/9/11 — 12:15 — page 264 — #14

compensation (through temperature sensors) are also common within sensor systems.
However, compensation for environmental variables increases the cost of the CO
detector. Data referenced above showing that 79% of alarms failed when tested at 5%
RH3 is compelling evidence of performance standards not meeting real-life long-term
requirements for adequate domestic CO monitoring.
10.4 CONCLUSIONS
Detection of CO has gone far beyond the primitive approach of Claude Bernard
and others. CO sensors and detectors employ advanced materials, electronics, and
software to ensure reliable and selective performance while maintaining economic
sensitivity and feasible for the domestic market. This chapter has discussed the three
major sensing techniques employed in mainstream CO detector/alarms. SMO sensors
depend on chemi-absorption between the oxide and CO molecules for CO detection.
Various methods are employed by industry to increase selectivity through the intro-
duction of catalysts such as Pt and Pd, and filters using activated carbon. Optical
sensors depend on CO energy absorption by incident photons. Humidity, temper-
ature, and pressure are environmental factors that may affect sensor components.
Electrochemical sensors are also vulnerable to cross sensitivity, temperature, and
humidity variations.
The intrinsic deficiencies of materials and electronic components that make up
commercial CO sensors and systems have been documented and are well known.
These issues have plagued manufacturers and the research community for many
years and continue to be areas of active scientific interest. Nevertheless, economic
forces, government legislation, competition, and customer demand drive CO detector
products to be sold at the lowest possible prices, while high customer standards,
product superiority, competitive advantage and market reputation drive product
quality and innovation. Hence, these forces lead to the classic economic balance
between price and performance.
10.5 ACKNOWLEDGMENTS
The authors kindly thank all contributors including Dr Mark Goldstein from Quantum
Group, Inc. (USA), Dr Herve Borrel from MiCS (Switzerland), Dr Nobuaki
Murakami from FiS (Japan), Dr Jürgen Schilz from PerkinElmer Optoelectronics
(Germany), and City Technology Sensors (UK).
10.6 APPENDIX 1 (D.G. PENNEY)

Suicide in Australia, especially that of young men, had attained an alarming rate in
recent years, higher than that in the USA and most other countries. The use of motor-
vehicle exhaust gas for this purpose was the most popular method. For this reason, in
1998 the Australian Medical Association in cooperation with other governmental and
industrial groups as well as various individuals in Australia, invited me to provide
conceptual solutions for reducing this tragic loss of young life.
8417: “8417_c010” — 2007/9/11 — 12:15 — page 265 — #15

Almost immediately, it was felt that by limiting the availability of the lethal
component of motor-vehicle exhaust gas, carbon monoxide (CO), the use of this
method of commiting suicide would decline, possibly saving several hundred lives
per year. The use of CO was often chosen because of its availability, ease of use,
and supposed painless induction of unconsciousness. Making CO unavailable would
defeat this approach.
Several solutions were considered: (1) Accelerate the rate of installation of effect-
ive catalytic converters on Australian motor vehicles, possibly by instigating a retro-fit
program. (2) Require the sale or retroinstallation of CO detectors on all motor
vehicles in Australia that would warn drivers/occupants of the danger, and/or, imme-
diately shut-down the engine and prevent restarting. (3) Place a distinctive odorant
in petrol/gasoline that would give motor-vehicle exhaust an unpleasant odor and
thus discourage/warn potential suicide attempters. (4) Design ignition systems that
would prevent motor vehicles from remaining in an “idle” mode for more than a
short time.
Catalytic converters are expensive, eventually wear-out, are slow to come on-line
in Australia owing to a long mean vehicle life, and retro-fitting would be difficult and
place financial burdens on people least able to pay. Also, current catalytic converters
still permit exhaust gases to contain lethal CO concentrations. Finally, the fact that
catalytic converters only become effective in reducing CO at elevated temperatures
means that exhaust gases would continue to contain supra-lethal concentrations of CO
during the “warm-up” period. CO detectors that produce engine “shut-down” would
have to be carefully designed so as not to exacerbate traffic problems due to elevated
ambient CO concentration. This approach appeared to be the best overall solution, and
could have provided some additional health benefits separate from the suicide issue.
Solutions involving odorants in motor-vehicle fuel might cause public discomfort and
complaints and undesirable environmental pollution. Most motor vehicles need the
capability to idle, for example, waiting for traffic or stop lights, taxis, and vehicles
being repaired.
Australia represented just 1% of the world motor-vehicle market. The average age
of Australian motor vehicles (8 × 106 ) in 1997 was 12–14 years. It was my charge
in visiting Australia in early April, 1998, to recommend to the Australian Medical
Association (AMA) and the Working Group on Motor Vehicle Exhaust Suicide, a CO
concentration that might be set as the threshold at which engine shut-down would
occur. Mathematical modeling of motor vehicle exhaust gas revealed a “unique sig-
nature” that might be used to quickly and unequivocally identify a suicide attempt,
distinct from simple leakage of outside gases into the vehicle.
Motor vehicles would be equipped with a sensor array in the passenger compart-
ment that was sensitive to: carbon monoxide, carbon dioxide, and oxygen. Sensor
output would be directed to a microchip with an embedded program such that: (1)
measured CO concentration was integrated over time in a manner modeling human
CO uptake, and thus provides a Low warning alarm at 35 ppm (7% COHb), and a
High warning alarm at 100 ppm (14% COHb), and (2) A CO concentration at 100
ppm and above, as well as rapidly rising CO2 concentration and rapidly falling O2
concentration would immediately trigger engine shut-down.
8417: “8417_c010” — 2007/9/11 — 12:15 — page 266 — #16

There were several advantages to this scheme: Low and high alarms give
warning of CO presence at levels shown to impair psychometric performance
(“drowse alarm”) and known to produce health damage in at-risk groups (congestive
heart failure (CHF), coronary artery disease (CAD), fetus). These would give warn-
ing of elevated ambient CO and/or abnormal exhaust gas leaks into the motor-vehicle
driver/passenger compartment.
With concentration × time computer integration, neither heavy cigarette smoking,
auto tunnels, nor congested roads would be likely to trigger even the low CO alarm.
Use of CO2 concentration and O2 concentration changes along with CO concentration
would prevent “false positives,” that is, inappropriate engine shut-down. Changes in
the concentrations of these three gases would provide a unique “signature” of the
suicide attempt.
A second sensor array might be placed outside the motor-vehicle, preferrably near
the rear tailpipe. This would cause engine shut-down in those instances where people
attempt to commit suicide outside of the car, behind the tailpipe (in a garage, outside,
etc.). If the cost of the three-sensor array proved too great, only one sensor responding
to CO might instead be used. In this event, threshold CO concentration might be set
somewhat higher in order to avoid inappropriate engine shut-downs.
CO detectors are standard equipment in households in the USA, warning of
furnace malfunction, etc. They are also required in motor homes, recreational power-
boats, and other devices where people are fully or partially enclosed and in proximity
to an internal combustion engine. Why shouldn’t such devices now become standard
equipment in motor vehicles, considering that cars are such prodigious generators of
CO and in such close proximity to the driver and passengers, and the fact that cars
already incorporate minimally several microcomputers in their normal operation. For
further details of the proposed scheme, see www.coheadquarters.com/CO1.htm.
References
1. C. Bernard, Introduction a l’etude de la medecine experimentale. Paris: J.B. Bailliere
et Fils, 1865, pp. 85–92, 101–104, 107–112, 265–301.
2. Revised Standards to Improve Carbon Monoxide Alarm Performance, Gas Research
Institute Digest (GRID), Chicago, vol. 21, pp. 24–25, 1998.
3. P. K. Clifford, Evaluating the Performance of Residential CO Alarms, Mosaic
Industries, Inc., Newark GRI–02/0112, 2002.
4. K. Galatsis, W. Wlodarski, Y. X. Li, and K. Kalantar-zadeh, Vehicle cabin air
quality monitor using gas sensors for improved safety, presented at COMMAD
2000 Proceedings. Conference on Optoelectronic and Microelectronic Materials and
Devices (Cat. No. 00EX466). IEEE. 2000, pp. 65–68. Piscataway, NJ.
5. K. Galatsis, W. Wlodarski, L. Yongxiang, and K. Kalantar-zadeh, Ventilation con-
trol for improved cabin air quality and vehicle safety, presented at IEEE VTS 53rd
Vehicular Technology Conference, Spring 2001. Proceedings (Cat. No. 01CH37202).
IEEE. Part vol. 4, 2001, pp. 3018–3021, Piscataway, NJ.
6. K. Galatsis, B. Wells, and S. McDonald, Vehicle cabin air quality monitor for fatigue
and suicide prevention, SAE Transactions, vol. 2000-01-0084, 2000.
7. S. Sato, R & D Review of Toyota CRDL 39, vol. 1, 36, 2004.
8417: “8417_c010” — 2007/9/11 — 12:15 — page 267 — #17

8. V. H. Routley and J. Ozanne-Smith, The impact of catalytic converters on motor

vehicle exhaust gas suicides, Med. J. Australia, vol. 168, pp. 65–67, 1998.
9. M. A. Skopek and R. Perkins, Deliberate exposure to motor vehicle exhaust gas: the
psychosocial profile of attempted suicide. Australian and New Zealand J. Psychiatry,
vol. 32, pp. 830–838, 1998.
10. Motor Vehicle Census, Australian Bureau of Statistics, Canberra Cat. 0309.0, 2001.
11. W. Gopel, New materials and transducers for chemical sensors, presented at Sensors
and Actuators B (Chemical), vol. B18, no. 1–3, March, 1994, pp. 1–21. Switzerland.
12. N. Yamazoe, New approaches for improving semiconductor gas sensors, presented at
Sensors and Actuators B (Chemical), vol. B5, no. 1–4, Aug.–Dec., 1991, pp. 7–19.
Switzerland.
13. G. Sberveglieri, Recent developments in semiconducting thin-film gas sensors,
presented at Sensors and Actuators B (Chemical), vol. B23, no. 2–3, Feb. 1995,
pp. 103–109. Switzerland.
14. M. J. Madou and S. R. Morrison, Chemical Sensing with Solid State Devices:
Academic Press, San Diego, 1989.
15. S. R. Morrison, In Semiconductor Sensors, S. M. Sze, ed., New York: J. Wiley, 1994.
16. H. Meixner, J. Gerblinger, and M. Fleischer, Sensors for monitoring environmental
pollution, presented at Sensors and Actuators B (Chemical), vol. B15, no. 1–3, Aug.,
1993, pp. 45–54. Switzerland.
17. G. Kiriakidis, N. Katsarakis, M. Katharakis, M. Suchea, K. Galatsis, W. Wlodarski,
and D. Kotzias, Ultra sensitive low temperature metal oxide gas sensors, presented
at 2004 International Semiconductor Conference. CAS 2004 Proceedings (IEEE Cat.
No. 04TH8748). IEEE. Part vol. 2, 2004, pp. 325–331, vol. 2. Piscataway, NJ, USA.
18. K. Galatsis, Y. Li, W. Wlodarski, C. Cantalini, M. Passacantando, and S. Santucci,
MoO3 , WO3 single and binary oxide prepared by sol-gel method for gas sensing
applications, J. Sol-Gel Sci. Tech., vol. 26, pp. 1097–1101, 2003.
19. S. Kaciulis, L. Pandolfi, S. Viticoli, G. Sberveglieri, E. Zampiceni, W. Wlodarski,
K. Galatsis, and Y.X. Li, Investigation of thin films of mixed oxides for gas-sensing
applications, Surface and Interface Analysis, vol. 34, pp. 672–676, 2002.
20. L. M. Cukrov, P. G. McCormick, K. Galatsis, and W. Wlodarski, Microcharacterisation
and gas sensing properties of mechanochemically processed nanosized iron-doped
SnO2 , presented at Proceedings of IEEE Sensors 2002. First IEEE International Con-
ference on Sensors (Cat. No. 02CH37394). IEEE. Part vol. 1, 2002, pp. 443–447,
vol. 1, Piscataway, NJ.
21. Y. X. Li, D. Wang, Q. R. Yin, K. Galatsis, and W. Wlodarski, Microstructural char-
acterization of sol-gel derived Ga2 O3 -TiO2 thin films for gas sensing,” presented
at COMMAD 2000 Proceedings. Conference on Optoelectronic and Microelec-
tronic Materials and Devices (Cat. No. 00EX466). IEEE. 2000, pp. 363–366.
Piscataway, NJ.
22. Y. X. Li, K. Galatsis, W. Wlodarski, J. Cole, S. Russo, J. Gorman, N. Rockelmann,
and C. Cantalini, Polycrystalline and amorphous sol-gel derived WO3 thin films and
their gas sensing properties, presented at COMMAD 2000 Proceedings. Conference
on Optoelectronic and Microelectronic Materials and Devices (Cat. No.00EX466).
IEEE. 2000, pp. 206–209. Piscataway, NJ.
23. L. Yongxiang, W. Wlodarski, K. Galatsis, S. H. Moshli, J. Cole, S. Russo, and N.
Rockelmann, Gas sensing properties of p-type semiconducting Cr-doped TiO2 thin
films, presented at Transducers ’01. Eurosensors XV. 11th International Conference
on Solid-State Sensors and Actuators. Digest of Technical Papers. Springer-Verlag.
Part vol. 1, 2001, pp. 840–843, vol. 1. Berlin, Germany.
8417: “8417_c010” — 2007/9/11 — 12:15 — page 268 — #18

24. K. Galatsis, L. Yongxiang, W. Wlodarski, E. Comini, G. Sberveglieri, C. Cantalini, S.

Santucci, and M. Passacantando, Comparison of single and binary oxide MoO3 , TiO2
and WO3 sol-gel gas sensors, presented at Transducers ’01. Eurosensors XV. 11th
International Conference on Solid State Sensors and Actuators. Digest of Technical
Papers. Springer-Verlag. Part vol. 1, 2001, pp. 836–839, vol. 1, Berlin, Germany.
25. For availability: www.coheadquarters.com/CO1.htm
8417: “8417_c010” — 2007/9/11 — 12:15 — page 269 — #19

8417: “8417_c010” — 2007/9/11 — 12:15 — page 270 — #20
11 Marketing of Carbon
Monoxide Information
and Alarms in Europe
and Beyond: Use of the
World Wide Web in
Saving Lives
Rob Aiers
CONTENTS
11.1 My Introduction to CO Alarms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271

11.2 Use of the Internet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
11.3 Department of Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
11.4 The Department of the Environment, Transport, and the Regions . . . . . . . 281
11.5 The Department of Trade and Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
11.6 The Health and Safety Executive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
11.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
11.8 Appendix 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
11.1 MY INTRODUCTION TO CO ALARMS

I have been involved with carbon monoxide (CO)-related issues for 8 years. My
work and indeed my passion for getting the message out has resulted in our website
becoming the number one CO-related issues website based on Google searches. From
time to time we swap places with David Penney’s website.
My background is in marketing, although I am an aircraft engineer by trade,
serving 13 years in the Royal Air Force (RAF) and culminating in the first Gulf War.
On leaving the RAF, I completed a diploma in sales and marketing management.
I have worked for a number of large organizations and have been consulting in my
own marketing business for 5 years.
271
8417: “8417_c011” — 2007/9/21 — 22:04 — page 271 — #1

My entre’ to the problem of CO exposure came when working for the world’s
largest water utility company, as a consultant, with a remit for new and innovative
products. My goal was to find a product that could be known outside the water com-
pany’s geographical area (water utilities in the United Kingdom cover a geographic
area for their core business, but can own a number of such businesses in various areas)
and thus make nonregulated profits (core business profits are regulated in the United
Kingdom).
The first product that I was asked to market was a product called Water Fuse. It
was a water-related product that the water company took on before I joined. Although
Water Fuse was a good product, it had a number of problems. It measured water usage
and if a leak was detected, would shut off the water. We did a number of commercial
trials for the product and found it to be very reliable, however that was not the whole
story. We found that the cost of the product combined with the installation costs made
it prohibitively expensive, around $550. Most people in the market for this type of
product (i.e., the domestic user) were happy if they had a problem their insurance
would cover. Indeed most people’s insurance premiums were less than the cost of
Water Fuse and covered all elements of liability including fire, theft, and so forth.
I set out to find a replacement for Water Fuse that would be cheaper and easier to
install.
It was at this time that we came across many wacky products. One that springs to
mind is Eco-Balls. This was a product that one used in place of detergent. Its marketing
blurb read, “unleash the ionic power of your washing machine.” The premise of this
product was that when you put the balls in with your washer, the power of ions would
clean your clothes. When tested in the lab, it was found to be only marginally more
effective than water alone.
On my travels I met with a company that had done a trial of a product that was
similar to Water Fuse, but was much less expensive and easier to install. This product
could also be expanded to cover small leaks in all parts of a property by utilizing the
electrical ring main of the house to send a signal back to the master unit, turning off
the water.
At the end of the meeting I asked what other products they were working on. “We
have a product that we think could be a major lifesaver, but it does not fit the water
company profile.” The product it turns out was a CO detector. We spent more time
talking about this product than about the Water Fuse replacement. I was amazed how
little I knew about CO.
I spent the next few weeks researching the topic and discovered I was by no
means alone in my ignorance. I was further astounded at the conflict in the numbers
of deaths and injuries from CO exposures/poisonings. The figures ranged from an
official government statistic of around 50 deaths a year to as many as 1500 deaths
per year from nonofficial sources. I was appalled that there wasn’t more information
available and that the government did not and still doesn’t, recognize or promote CO
safety. Every now and then, late at night on TV, there is a puny campaign about CO
exposure consisting of a public relations commercial.
I am quite cynical about why the government does not publicize the CO problem.
I believe that if they made people more aware of CO, the government would be
compelled to follow up with a plan to resolve the problem. The government in the
8417: “8417_c011” — 2007/9/21 — 22:04 — page 272 — #2

Marketing of Carbon Monoxide Information and Alarms in Europe and Beyond 273
UK owns about 2.5 million homes, mostly “social housing.” If it publicized the
problem, it would have to protect the residents from danger—the costs would be
great. A CO alarm costs around $20.00. Multiply that by 2.5 million homes and the
total is roughly $50 million.
I saw the CO alarm to be a perfect public relations vehicle for a water company.
It could be a true win-win situation. The water company had around 11 million
customers and had access to the rest of the 60 million UK population through other
water utilities. They could publicize the dangers of CO and also sell a product that
could protect its users from future exposure. This might drastically reduce deaths.
The company who first had the CO alarm was small, but it was innovative. It
also had all the problems inherent to small companies, namely cash on hand. The
product had been prototyped, with a short manufacturing run in Taiwan. However,
the company had no money to move into full production. We concluded a deal where
we had exclusive access to the product. The CO alarm was inserted directly into the
wall electrical power receptacle. It incorporated some clever technology that allowed
it to predict carboxyhemoglobin (COHb) levels, that is, blood CO content. We agreed
to pay for an increase in product run once we had seen the product being built in a
factory in Taiwan and were convinced that all was okay.
Some colleagues and myself flew out to Taipei, Taiwan to verify the situation.
This took a total of 5 days, either on the road or at manufacturing facilities. At every
location we were dressed in full medical whites (“clean garb”), as the facilities also
produced computer and automotive industry components. My view at that time was
that Taiwan was just a producer of cheap toys for Christmas, crackers, and so forth.
Nothing was further from the truth. All the facilities we saw were high-tech—at the
top end of the computer industry.
The costs of the CO alarms were a problem. If the selling price was too high,
we’d be unable to sell many units. Not only would we not make projected profits,
but more importantly, people would not be able to protect themselves from CO pois-
oning. Clearly the profit element was important and getting unit costs down was
vitally important. We had to have the best product at the best price, so that when we
advertised, we had an affordable solution for all.
Ideally I would have liked to get the product down to a price of around £10 ($17).
That is about the same price as a smoke alarm. We soon realized that the CO alarm
was a much more complicated product and it may never be possible to make it as
cheap as a smoke alarm. The smoke alarm industry had become a mature market,
unlike that for CO alarms. It is worth drawing some parallels with what had allowed
the smoke alarm industry to gain so much ground.
In the 1980s, UK market penetration for smoke alarms was around 9% of
households. At that time the government mounted a big campaign through the fire
service and TV advertising. The “Smoke kills” campaign was responsible within a
9-year period of raising market penetration to around 65%. Recently, government
legislation has caused that penetration to be near 100%. By law now, all new houses
must have a smoke detector hard-wired into the construction, and it is now considered
the norm to have such a device in place. The smoke kill’s campaign still goes on, but
the focus is now more on maintaining existing devices. I think that the world would
be a safer place if the same regulations existed for CO alarms.
8417: “8417_c011” — 2007/9/21 — 22:04 — page 273 — #3

We had to look for manufacturing facilities that would get CO alarms to market at
a price that would ensure that most people would buy them. Many of the Taiwanese
factories also had Chinese manufacturing facilities. These involve much reduced cost,
since labor costs in the Republic of China are less than that of Taiwan. This would
have been the next move had not something happened.
It became clear that the price that I was trying to achieve was not realistic and that
CO alarms would never be as cheap as smoke alarms. As stated above, this is because
smoke alarms detect danger as the danger is occurring (by either optical sensors or
ion detection), but do not require the complex circuitry required in a CO alarm. It
is possible to have a large release of CO that causes an immediate audible alarm,
or to have much lower levels of CO released over prolonged periods of time. The
best type of CO alarm is one that can make the correct calculations regarding human
CO uptake. As stated above, I also wanted to sell the best alarm, so it had to be a
unit that measured CO concentration over time. Inexpensive card-type detectors are
available that change color in the presence of CO, but these can be contaminated by
chemicals, don’t give satisfactory low-level indications, and have no audible alarm.
I am not totally opposed to this type of CO detector, as they play a role in raising CO
awareness. Nonetheless, they are not sufficiently reliable for constant usage.
We had by this stage fully evaluated the CO alarm market and were excited to get
on with marketing the product and making the public aware that there was a problem.
We hit on a major problem at this time—the water company was likely to be sold.
This was happening at a high level in the company so none of us was aware of the
impending sale. Orders came down from on high that we were to stop all unnecessary
marketing and to get back to the core business.
My vice-president at that time was a very talented woman who was not held in
high regard by the company board. There followed a free for all with junior managers
stepping on each others heads to try to gain political high ground. In particular, while I
was on leave, an alliance was formed between two managers who were determined to
steal our thunder, or better still, to totally kill our work. In the end they won and all of
the people who had been working for me were let go or moved to other departments.
With that, all of our efforts were dashed.
I then left the company to move on to other things, but the company had no
idea what contracts had been signed. Consequently a legal dispute ensued, because
contracts had been signed for the water company (by me) that tied them to marketing
and supply agreements. The company with whom we had been dealing needed the
agreements in order to move forward, but foolishly had signed an agreement with
a third party to sell the product before the legal wrangle was resolved and they
ultimately lost everything.
11.2 USE OF THE INTERNET

I resumed my marketing business at this time, after working initially on a con-
sulting basis for the water utility. I worked for a number of large clients, but
also felt the government could and should publicize the CO issue. It was around
this time that I spawned the idea of a CO information website, that is,
www.carbonmonoxidekills.com. I immediately began getting it set up.
8417: “8417_c011” — 2007/9/21 — 22:04 — page 274 — #4

Unfortunately I knew nothing about the internet, so I needed to team up with

someone who was familiar with its complexity. It was at this time I met Frank Tiarks,
a Danish chap who had been educated in the UK. Frank had a very good understand-
ing of what was required to make website content user friendly, and of the content
necessary to cause the major web search engines to list it first. I soon learned that it is
no use just having a website—it has to be seen. I would liken it to having a shop that
sells great products that people really need, but the shop is located down a dead-end
road or in the depths of the countryside where no one will find it. One of the most
important factors for success here is having a “Frank” who knows how to do it.
There are many reasons why the world wide web (WWW) is an excellent way
of getting your message out. First, the WWW is democratic. You have the same
chance of getting your message across to the public if you are a small corner toy
shop, as Toys-R-Us does. From that perspective it didn’t matter where you were
geographically located as long as you could speak English. Anyone who has visited
our website and has acted on the information there owes a debt to Frank Tiarks who
has worked tirelessly to get CO-related information to the public.
We needed to have a website with immediate impact, so we came up with the
name, www.carbonmonoxidekills.com. As a capitalist, I had to generate a profit from
the website simply to be able to continue running it. Over the past 5–6 years, some
$80,000 has been spent on it, of which only a small percentage has been recouped. I
am proud of the fact that we have become one of the most highly utilized, web-based
CO information sites in the world. Approximately 5500 people per day look at the
information on www.carbonmonoxidekills.com (see Figure 11.1). If we have saved
one life or helped one sufferer improve his life, it has been money well spent. Some
of the respondents have been referred to Dr. Penney for evaluation. More recently we
have added some law firms as consultants, so some of our viewers have been helped
with their legal problems.
As you might guess, the website is quite comprehensive. Where questions are not
answered, we pass the visitor on to an expert to answer the question (see Figure 11.2).
It should also be noted that what you can see here is only the front page of the website.
There are many additional layers, so almost any subject relating to CO poisoning can
be found.
It took some time to get us where we are now in terms of a web presence. We
constantly re-examine the content of the website to insure that it has the best inform-
ation for our visitors. We now get around 2 million visitors a year, many of whom
e-mail us to thank us for the website.
It is a sad indictment of those official government sources that should be doing
this job and not be relying on individuals like me to do the job for them. Unfortunately
we only get visitors who already are aware that there is a problem involving CO. In
the words of Donald Rumsfeld, “we don’t know what we don’t know.” A clumsy way
of saying that is, if you have never heard of CO, why would you look for information
about it. That is the main issue here. If governments would just get the word out,
people would be better able to decide for themselves whether there is a problem.
The United States is much better at this than the United Kingdom. It may have to
do with the fact that America is a more litigious society. We do not have a punitive
damages system in the United Kingdom, so where someone may get an award of
8417: “8417_c011” — 2007/9/21 — 22:04 — page 275 — #5

FIGURE 11.1 The front page of www.carbonmonoxidekills.com
several hundred-thousand dollars in the United States, in the United Kingdom the
award is more likely to be £10,000 at best.
The questions we are asked through the website are diverse to say the least. In
some cases they are downright stupid. In one case an American youth asked me if it
was dangerous to travel in the trunk of his friend’s car. The great bulk of the questions
from an uninformed public are completely justified (see Figure 11.3).
About 75% of the website’s visitors are American—15% are from the United
Kingdom (we have a lot of UK links from other websites, especially government
sites), with the remaining visitors coming from the rest of the world. The reasons for
this are several-fold. First, the United States has a larger population than the United
Kingdom. Second, America is more web-savvy than most of the world. However the
most important reason is that Americans are more aware of the dangers of CO, and
are thus better able to seek out information. It should be noted that our website is in
English alone, so other language speakers may not be able to use it.
While there are charities and experts who work hard to get the message out, the
same cannot be said of everyone. People with vested interests in CO issues have not,
8417: “8417_c011” — 2007/9/21 — 22:04 — page 276 — #6

FIGURE 11.2 The law connection to the Web.
in my view, been ready to sing from the same song sheet. There are three or four
government departments in the United Kingdom who have some responsibility with
regard to CO. From a public health perspective there is the Department of Health,
from an environmental point of view (i.e., housing, etc) there is the Department of the
Environment, and from a business perspective there is the Department of Trade and
Industry. There is also the Health and Safety Executive who looks at the industrial
side of safety.
11.3 DEPARTMENT OF HEALTH

This department’s charge is the state of the nation’s health, with its remit covering
doctors, nurses, and the general health infrastructure. Health concerns cover every
part of people’s lives and some areas are covered better than others. Let me draw a
parallel if I may. Meningitis affects about the same number of people in the United
Kingdom as those who are affected by CO, based on government statistics. It is a
8417: “8417_c011” — 2007/9/21 — 22:04 — page 277 — #7

FIGURE 11.3 The forum associated with www.carbonmonoxidekills.com.
very serious disease that can kill or maim. Unchecked, CO poisoning can also kill, or
damage people’s lives either through an acute poisoning episode or through chronic
exposure.
You would be hard pressed to find a parent in the United Kingdom who was not
aware of the symptoms of meningitis and what to do if he/she suspected his/her child
was suffering from it. This is just as it should be and is so because the government and
the media have spent a good deal of time and effort publicizing its dangers. There was
scarcely a time 3 years ago when you turned on the TV that this issue was not being
aired. Please excuse me if I shout, “CO EXPOSURE IS AS DANGEROUS AS THE
MANY STRAINS OF BACTERIA AND VIRUS THAT CAUSE MENINGITIS.” It
is about time that politicians and health experts in the United Kingdom took a lead
from some of the US states and began taking the CO exposure/poisoning issue more
seriously.
It is my opinion that most medical practitioners here and in the United States
are only slightly better than useless when it comes to diagnosing and treating their
8417: “8417_c011” — 2007/9/21 — 22:04 — page 278 — #8

patients exposed to CO. If you visit your physician with flu-related symptoms, CO
poisoning is usually the last cause he/she will include in the differential diagnosis, if it
is included at all. Physicians are poorly trained about the dangers of CO poisoning—
after all, the concern is primarily toxicological, not internal medicine. Physicians
generally don’t know the correct questions to ask. For example, they usually fail to
ask how your house is heated? Whether several people and even animals are sick at the
residence? I very much doubt it. However, this is not an excuse for complacency. We
must train our medical personnel better to recognize the symptoms of CO exposure.
There are few ways for a physician to investigate the living circumstances of the
person who has presented for treatment. That patient could be residing in a property
where someone previously suffered similar symptoms. It is possible that over an
extended period of time many people suffered CO poisoning in that building because
no medical professional was alert. This scenario could go on for years.
Early in 2006 I had a meeting at the Department of Health with a number of
people. This included a lawyer who had represented victim’s interests in CO cases
that had gone to court in the United Kingdom. There were a couple of representatives
from a CO charity and the head of the Health Protection Agency, whose responsibility
it is to deal with preventive health measures.
I found the way the meeting was conducted to be astounding. We were informed
that a leaflet had been sent to all UK physicians regarding CO. The leaflet was fairly
brief. We asked if the department had followed up the leaflet to make sure that
physicians had received it, and whether they had read and understood the information.
The answer was no. We were also told that if there was a high demand for the leaflet
there would be a reprinting, and that it was also available on the Department of
Health’s website. If the physicians had not read the information in the first place how
was there going to be a high demand?
I told department personnel that I would help them in whatever way I could to get
the message out, but that they needed to help me. One of the ways that this could be
done quite simply and at very little cost, was for them to call TV stations to alert them
to the problem. In the United Kingdom we have a morning program called GMTV.
They spearheaded a brilliant campaign for meningitis and could do the same for CO.
I have been unsuccessful in getting through to them. However one call from the chief
medical officer would spur them into action at the cost of a phone call.
At the meeting, a vice-president of CO Awareness said that she had tested the
department’s system by calling “NHS Direct” (a phone hot line that gives health
advice). She was told that the best people to speak with about CO was CO Awareness,
her own charity. When questioned about the computer menu on their help line which
advises what should be done when certain syptoms are mentioned, the official said
that unexplained headaches and drowsiness would NOT be related to CO exposure
by the operator. Of course this is incorrect.
The problem goes much deeper. I asked the chief medical officer what would be
the recommendation from physicians should they suspect CO poisoning? He replied
that “the patient should be tested.” Do they have the appropriate equipment I asked?
“I’ll have to get back to you,” he said. “So let’s accept the fact that they have the
relevant testing equipment,” I said. “When do you recommend that the patient be
tested?” “Well, during their consultation” was his reply. I had to point out that if the
8417: “8417_c011” — 2007/9/21 — 22:04 — page 279 — #9

FIGURE 11.4 www.codetection.com, which is lined to form carbonmonoxidekills.com
patient had been away from the source of the CO poisoning for more than a few hours,
then they would likely give a negative test.
In this event, someone who was suffering an episode of CO poisoning, but who
happened to get an appointment at the end of the day might have CO dismissed as a
potential cause of his/her illness, and then might go on being exposing needlessly to a
life threatening situation. In some ways this is much worse than not being diagnosed
at all. This problem was dismissed almost out of hand, likely because I was not a
medical professional.
The Department of Health should in my view be recommending that all homes be
fitted with CO alarms (see Figure 11.4). There are a number of different views about
CO alarms and their efficacy, but I think they are a good weapon in the arsenal for
combating CO poisoning. They should also be recommending that all combustion
appliances be checked annually. Only landlords have to perform an annual safety
check. I am sorry to be bashing the Department of Health, because on the whole they
do a great job. They are often hog-tied by a lack of government funding.
8417: “8417_c011” — 2007/9/21 — 22:04 — page 280 — #10

11.4 THE DEPARTMENT OF THE ENVIRONMENT,

TRANSPORT, AND THE REGIONS
This department is responsible for a numbers of areas. It is concerned with amongst
other things, the air we breathe. Many people think that CO is only caused by vehicu-
lar emissions. There are major crossovers with other departments and I am loathe
to list the full responsibility of this department, since by the time this book goes to
print, it may have changed or there may be a new administration. This Department
is responsible for dangers that lurk in our homes. It is responsible for planning reg-
ulations and for new building laws, for example, requiring newly built homes to
have hard-wired smoke detectors. In my view, it is remiss in this respect in not also
requiring CO detectors.
The Department of the Environment requires landlords to have a safety check each
year of accommodations where fossil or wood burning appliances are used. Because
many people in the United Kingdom own there own homes, it means that a large
fraction of residences are not covered by any regulation. It is unfortunately the way
of the world that we buy “stuff” for which we can see a direct benefit. A CO detector
costs about the same as a take-away meal, and at the end of that meal we are happy
to part with our cash. Because it costs a bit more to get our furnace/boiler serviced
(i.e., $100/200), many of us might shy away from the cost. It usually doesn’t make
us feel good—in fact quite the reverse. We will only pay this charge if compelled to
do so. We do not like paying our house or car insurance, but do it because we are
compelled to. New York has the right idea in that landlords must by law install a CO
alarm, and in turn can collect $25 from each tenant to help pay for it.
Do yourself a favor. Purchase a recommended CO alarm to protect yourself and
your family. Better yet, buy two CO alarms. By its nature CO is colorless, odorless,
and has no taste. It does not occur to us that there will ever be a problem. You would
not leave small children near a staircase lacking a gate to prevent them from falling.
11.5 THE DEPARTMENT OF TRADE AND INDUSTRY

This department deals with the interests of trade, but also regulates some of the
peripheral issues related to CO. One of the main areas of authority regarding CO
are the gas supply companies, which until recently was reduced to just one supplier,
“British Gas.” British Gas lost its monopoly and now there are many suppliers in the
market place. These companies came into being as a result of denationalization, a
trend in the United Kingdom. The majority of gas users still use British Gas. Some
credit has to be given to British Gas for recent initiatives, but they are not on the
whole completely altruistic in their nature.
I will paraphrase a conversation that I had with one senior British Gas executive.
He said that “if we publicize the fact that gas could be dangerous and that in certain
circumstances gas appliances will give off CO, then we denigrate our product.” This
is a bit like saying that we should not install seat belts in automobiles because it
makes them appear unsafe. From a trade point of view they are missing the point.
Their product is not the problem, it is what is done with that product when it leaves
the pipeline that is key.
8417: “8417_c011” — 2007/9/21 — 22:04 — page 281 — #11

British Gas now sells CO detectors. In fact, they are the biggest UK distributor
for the SF-350 CO detector, the best selling unit in Europe. Most of these have been
given away with service, which is sold on a commercial basis. British Gas has a
large service division that must make a profit. In my opinion, if they played the
white knight and told people that there was a major problem, they would be seen
as a responsible company with their customer’s interest at heart. They would benefit
from their customers coming to them and asking for service. Also, people would in
turn buy the CO detectors, and not expect them to be free. The company would not
be held up as irresponsible as they are now by some groups for not publicizing the
CO danger. In this way, they could benefit financially from making CO a safety and
health issue.
11.6 THE HEALTH AND SAFETY EXECUTIVE

This department, HSE, is responsible primarily for safety. I recently attended a meet-
ing of their Gas Safety Committee. I had been invited by the head of a leading charity
to present my views to the committee. A number of things came up in that meeting
that I found astounding and frustrating. The Department had a £100,000 surplus at
the end of the financial year which they had to spend quickly, otherwise it would dis-
appear from the following year’s budget. They didn’t ask the charities who endeavor
to get the message about the best course of action. They didn’t call on our firm for
suggestions. Instead they called CORGI. CORGI is the regulatory body for heating
engineers. You cannot work on gas appliances in the United Kingdom unless you are
qualified and a member of CORGI. CORGI’s suggestion was to give it to them and
they would do a survey; so they did.
The HSE could not provide us with a qualitative or quantitative evaluation of what
they were going to get for their money. I guess they will get a survey back which will
form part of the departments future strategy that says, “CORGI are really good guys
and are in the opinion of the people who where surveyed, a top organization?”
The HSE Department has tried to develop a cohesive plan to alert people to CO
dangers. They have not called on me. I have on occasion called them, but they seem to
have no clue as to what is going on in the world outside government. An independent
enquiry stated in 2000 that it would be desirable to impose a fuel levy on gas, oil,
and coal of about 0.25%, which would be used to fund CO awareness publicity.
The HSE has now said that they will not implement the tax after being lobbied hard
by the gas companies. They can see this going the way of all taxes, that is, once
introduced, the taxes will only go up. Consequently, they don’t want even a small
tax that they could absorb. Instead they have said to the HSE, “We have spent £2
million this year on CO awareness.” In actual fact, they had a campaign about service
that was high profile, with many UK celebrities in ads. Unfortunately the ads never
mentioned CO. The ads were mainly about getting the British public to buy their
service—service which I might add, will go on making millions for years because it
is annual service. The net effect of this is that less people will be harmed by CO, but
don’t for a minute think that they have done it for purely altruistic reasons, they want
your buck.
8417: “8417_c011” — 2007/9/21 — 22:04 — page 282 — #12

All of these government departments have budgets which in and of themselves

are not really that much, maybe a couple hundred thousand pounds each. If however,
they combined their resources, they could do some real good. Also, if they had one
team looking at this rather than many disparate groups, they would not be duplicating
effort and could reduce spending. The effort could also be better coordinated so that
we have something tangible to show for it. People would readily be able to see the
dangers immediately, not months or years later. People would also be more aware
about protecting themselves in the first place. In the last few months I have been
to meetings at the House of Lords and subsequently with Lord McKenzie who is
tasked with getting all the information on CO together in one place with one agency
responsible. It seemed at last that there was some momentum, however that was some
months ago and I have heard nothing since.
We have created a website brand www.carbonmonoxidekills.com which is well
recognized throughout the English-speaking world. Since I am a capitalist, our web-
site is linked to a number of others. I submit that most problems in the world could be
resolved by entrepreneur’s looking at them as opportunities to earn revenue. The more
money that we make in this venture, the more we are able to plow it into CO aware-
ness. It’s a win-win situation. Also, the more we get out there, the more information
that we acquire, the more we can educate people about the issue.
Dr. Penney asked me to write this chapter partly because we have a relationship
going back a few years. He had kindly agreed to be our online “doctor” some years
ago. As such, he responds to enquiries regarding CO from the public. We feel that
he is an integral part of our online organization. We also have relationships with law
firms who are able to answer legal questions. Therefore, we are able to get the best
representation for people who have suffered CO poisoning episodes.
As mentioned above, we sell CO detectors through another link on our website,
that is, www.carbonmonoxidekills.com. This has been hard work and it has taken a
great deal of effort to get where we are. People often contact me, asking how they
can set up there own website. They are more than welcome to try, but we are the
experts and that is why over the years we have arrived at the position that we hold. To
achieve this, you have to get as much information about your subject on the website.
That information needs to be updated regularly so that the search engine spiders
can recognize new information. This keeps the rankings up. A spider is a computer
routine that runs on search engines like Google, Yahoo, and so forth. Working for
a high ranking is a bit of a catch 22 situation. Unless you have a good ranking,
you can’t get one. Another factor in terms of search engine rankings, is how many
other websites link to yours and the type of organizations they are. The best types of
organizations are government departments, the next are educational establishments
such as universities, and so forth. Media institutions are also good in that they attract
lots of visitors on a given subject such as CO that appears in the news. Thus, you will
get renewed interest at the time of newsworthy items.
One has to be very careful when approaching these kinds of websites. They will
not link to you if your information is commercial or inflammatory. Many charities
suffer for the latter reason, because they will often hold someone accountable for an
incident making them on occasion being seen as biased and having a narrow manifesto.
No government department will link to them. It is also sometimes difficult to get
8417: “8417_c011” — 2007/9/21 — 22:04 — page 283 — #13

a government body—and most educational institutions are government influenced—

to link to you unless another government institution has linked first. No one wants
to be first to link because they are scared that it will impact their impartiality and
position. The first one to come on board is the hardest to get. If I call an institution
now and ask them to link to us, they are only too willing because we have 100s of
other links. There is now a comfort factor for them.
There has not been a willingness in the United Kingdom to grasp the nettle and
do something about the issue of CO. Proper unbiased research need to be done, and
we need to have a better understanding of the issue. Statistics on CO need to be
consistent. Some of the country’s top toxicologists state numbers that differ greatly
from the government’s official figures.
I mentioned above that most physicians are ill-prepared to diagnose and treat CO
poisoning. Moreover, if a physician has not identified the connection to CO, then
people may never know that they had CO-exposure. They may have relocated from
their house, their job, left home, etc. They may exhibit effects of CO in later life
that will never be attributed to CO. Death certificates may not tell the whole story
either for very similar reasons. The physician may list just one of the symptoms as
the cause of death. The coroner’s office in the United Kingdom is not attached to the
Department of Health—it is part of the Home Office. I was amazed to learn that until
recently, the job of coroner was a part time job. I am glad to say that we are now
going to get a full-time chief coroner. This may clarify the position and he/she will
liaise closer with the Health Department. In short, we need to know who and how
many people are affected. Only then can the problem be properly addressed.
In Paris, France all cadavers are tested for poisoning of all types. As a result, more
people are recorded to have died from CO than in all of the United Kingdom. This
is probably not because Paris has a worse problem, but just that better monitoring is
in place.
Take for example the murderer Harold Shipman, who was a UK general practi-
tioner. No one is quite sure how many people he killed. His story has been recognized
as one shortcoming of the coroner’s office. As a result, certain remedial measures
were suggested. The one I am most interested to see implemented is the checking of
all bodies for poisons, including CO, as a matter of course. This could prove a real
breakthrough and give us proper statistics so that the problem will be highlighted and
solutions found.
11.7 CONCLUSIONS
There is much work to be done. Combining the budgets of disparate government
departments, along with some coordinated thinking would result in a better under-
standing. I have offered government departments use of our website facility. If they
wish, we are able to run studies and research at little cost through our valued visitor
stream. We are very well-placed to be able to assist. While recognized for the efforts
that we have made, we are never seen as part of the solution.
My ambition is that www.carbonmonoxidekills.com will be the world-wide mar-
keting tool for all CO-related issues. I want to make myself redundant, in as much as
8417: “8417_c011” — 2007/9/21 — 22:04 — page 284 — #14

we have done the job so well, there is no more need for publicity because the whole
world has got the message and deaths and injury from CO are a thing of the past. I
would like people to know as much about CO as about meningitis. If the government
picked up the ball and ran with it, the issue of CO may actually pay for itself. Get the
message out and people will be empowered to protect themselves. People would not
be taking up the valuable time of physicians with mysterious illnesses or have to be
hospitalized.
I make no apologies for the fact that this chapter lacks high aca-
demic content. Instead, I hope you will appreciate the journey that we at
www.carbonmonoxidekills.com have made over the last few years. We have tried
to be the most open information platform possible, and not an exclusive high brow
intellectual website that turns visitors off, indeed those that need us most. It has some-
times been tough, and I have thought about giving it up at times. It takes a lot of my
time. Then I get an e-mail from a distraught mother, father or other relative or friend,
that thanks us for our help and it all seems worthwhile.
It is sad to say that unless governments do more, we are the only outlet for some
people. I only wish that we could get through to more people, but language barriers
make that difficult. It would be difficult for us to translate our website into other
languages, such as Chinese. The paradox is that as the markets in the developing
world get bigger, it is they who will need it most.
11.8 APPENDIX 2
Carbon Monoxide Headquarters known as COHQ, or “coheadquarters.com,” is one
of the oldest carbon monoxide (CO) information sites on the web. It was started in
1996, so is now more than 11 years old. That is very old in terms of the history of the
web! Initially COHQ ran on an old Macintosh computer in my research laboratory
at the Medical School at Wayne State University in Detroit, MI. It had a long URL
because it did not have its own registered domain name. In the late 1990s the website
received its own domain name and I migrated it over to a commercial server in Texas.
The new URL and the one used today is “coheadquarters.com/CO1.htm.” From a
collection of a few dozen linked pages in the early days, COHQ has grown to a
site containing hundreds of pages dealing with many subtopics in the field of CO
toxicology: chronic CO poisoning, dangers to high risk groups, neuropsychological
effects, FAQs, CO alarms/detectors, and so forth.
COHQ was begun as a way to provide the public with straightforward, unbiased,
information about CO and its effects on humans. It has been my operation from the
beginning. With the exception of a few dozen pages written at first by my student
Amy Derusha, all of the content, architectural design, art (whether good or bad), and
maintenance was been done by me. The website was produced by writing HTML in
“text” by hand—it never involved using a web-editor. As crude and lacking in flash
as COHQ is, it still comes up in the top 15 sites in doing a Google search using the
terms “carbon monoxide poisoning.”
The goals of COHQ were: (1) To act as a platform for public information about CO,
(2) To act as an educational resource for all viewers, including medical professionals,
8417: “8417_c011” — 2007/9/21 — 22:04 — page 285 — #15

(3) To act as a data resource—“look-up” and retrieval of information, and (4) To

present new ideas and information at the forefront of research in the field. Like other
good expert sites, it (1) Provides reliable, in-depth information in a narrow, defined
span of knowledge, (2) Is accessible, understandable, and useful to the lay public
as well as professionals in field, (3) Remains current as the field progresses and as
viewers needs change, and (4) Is unbiased; that is not supported or influenced by a
commercial interest in the field.
The web was a very different place a decade ago. The number of websites was
miniscule as compared to today. Google was not yet operating. “My Space” was years
away. There were search engines running there, but they didn’t have the power and
speed of today’s. It is hard to believe that in the middle 1990s it was difficult for us to
even find sources of instruction for the HTML language. Much of the early writing
was hit or miss—whatever worked was used.
The objective of COHQ was never to sell new furnaces, to provide furnace main-
tenance, and so forth as so many thousands or tens of thousands of websites found by
a “carbon monoxide” search pulls up today. I have been accused of using the site as
an advertising vehicle. Any good information source, whether a book, storefront, or
website inadvertently advertises the author/owner when people go there. Nonetheless,
unlike most CO sites, COHQ has no axe to grind. It provides both simple and tech-
nical information free about CO to whoever wishes to look at it. The constant theme is
public health, for groups and for individuals—educating and protecting people from
this age-old poison [see other chapters in this book on misconceptions about CO
(Chapter14) and on public perceptions of CO (Chapter15)]. The moto on the home
page reads, “CARBON MONOXIDE HEADQUARTERS, (to) provide information,
public service, help people, maintain health, save lives.”
I was pleased when “carbonmonoxidekills.com” came along in the late 1990s as
another major CO information source, in this instance, emanating from the United
Kingdom. As you have now read, its major objective is to safe-guard everyone by
everyone having CO alarms. Carbonmonoxidekills.com is currently number 1 when
the same search noted above for COHQ is run on Google. So I was thrilled some
years ago, when Rob Aiers, the owner, asked me to answer questions on CO from
his website. I have continued in that role to this day. Thus, it was natural when this
book was being planned that I would ask Mr. Aiers to tell us about his experience in
developing a very successful website similar to COHQ.
8417: “8417_c011” — 2007/9/21 — 22:04 — page 286 — #16

12 Investigating Carbon
Monoxide Poisonings
Thomas M. Dydek
CONTENTS
12.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287

12.2 Signs and Symptoms of Carbon Monoxide Poisoning . . . . . . . . . . . . . . . . . . . . 288
12.3 Assessments of Carbon Monoxide Exposure Level and Duration. . . . . . . . 289
12.3.1 Carboxyhemoglobin Levels as a Measure of Carbon Monoxide
Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
12.3.2 Occupational Exposure Standards for Carbon Monoxide . . . . . . . . 291
12.3.3 Community Exposure Standards and Guidelines for Carbon
Monoxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
12.3.4 Carbon Monoxide Exposure Duration Assessments . . . . . . . . . . . . . . 292
12.4 Treatments for Carbon Monoxide Poisoning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
12.5 Other Factors to Consider in Investigations of Carbon Monoxide
Poisoning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
12.6 Case Study of Carbon Monoxide Poisoning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
12.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
12.1 INTRODUCTION
Carbon monoxide (CO) is a colorless, odorless, tasteless, and potentially toxic gas.
These properties have earned it the title of “the silent killer”.1 CO poisoning is
responsible for more than one half of the poisoning fatalities reported in this country
every year. It is the leading cause of death in industrial accidents as well. Fatalities
and CO-related injuries are also common throughout the world. Another factor that
makes CO an especially dangerous toxin is that the early symptoms of poisoning are
easily confused (and often misdiagnosed) as the onset of a cold or the flu, stomach
virus, or other common diseases.2
CO is produced by natural sources and by man-made sources. Natural sources
include forest fires, oxidation of nonmethane hydrocarbons, and oxidation of meth-
ane. Plants can also emit CO as a metabolic by-product.3 Anthropogenic sources of
CO are mostly associated with incomplete combustion of organic materials such as
287
8417: “8417_c012” — 2007/9/11 — 12:15 — page 287 — #1

gasoline, fuel oil, natural gas, wood, or plastics. Common man-made sources of
CO include improperly vented cooking or heating devices, tobacco smoke, agricul-
tural burning, and internal combustion engines. Exposure to CO is one of the chief
dangers associated with the fighting of fires in buildings and forest fires. Workers in
occupations in which routine exposures to CO occur include truck and bus drivers,
mechanics, highway toll takers, garage attendants, and police officers. Everyone who
drives a car or truck, especially in areas of congested traffic, has some exposure
to CO.4
This chapter covers topics related to the investigation of CO-poisoning events.
These investigations can occur in a number of different situations. Oftentimes these
investigations are undertaken as part of legal proceedings. Many people have com-
mented that we in the United States today are the most litigious society in the history
of the world. While this may be a somewhat dramatic statement, there is a large
element of truth in it. Many people in this country feel that pursuing legal remedies
for perceived injuries which they believe they have suffered should be a first, rather
than a last, resort. Expert witnesses in engineering, industrial hygiene, and toxicology
are often called to investigate CO-poisoning incidents and to render opinions about
whether or not the CO exposure reported was of at a sufficient level or duration to
have caused the health harm alleged.
The topics covered in this chapter include signs and symptoms of CO poisoning,
exposure level and duration assessments, treatments for CO poisoning, differences
in susceptibility between people, how the above factors affect the acute effects of
exposure, and the long-term prognosis for CO victims, and other toxic exposures
or conditions that mimic CO toxicity (differential analysis). All of these factors are
important in the investigation of CO poisonings.
12.2 SIGNS AND SYMPTOMS OF CARBON

MONOXIDE POISONING
CO poisoning results in a decreased level of oxygen in the body. The brain and
other parts of the central nervous system are the areas of the body which are among
the most sensitive to oxygen lack.5 When oxygen levels in tissues fall, aerobic
metabolism decreases and lactic acid accumulates. Neurons begin to break down,
leading to cell death and brain damage.6 The longer the brain is deprived of adequate
oxygen, the more widespread the damage will be. Similar effects occur in muscle
tissues deprived of oxygen. This is of special concern when the muscle involved is in
the heart.
For many years most toxicologists believed that COs toxicity was fully explained
by this hypoxic effect. More recent research has shown, however, that CO exerts
direct toxic effects by inhibiting the activity of cytochrome a3 oxidase and by
causing lipid peroxidation. These latter findings help to explain the clinical exper-
ience that carboxyhemoglobin (COHb) levels (an indicator of the risk of tissue
hypoxia) are a very poor predictor of a patient’s medical condition and his or her
prognosis.2,5,7
8417: “8417_c012” — 2007/9/11 — 12:15 — page 288 — #2

Investigating Carbon Monoxide Poisonings 289
The acute symptoms seen with CO-poisoning depend on the concentration of

CO and the duration of the exposure. At low levels of exposure there may be subtle
changes in time discrimination, visual vigilance, and choice response. Exposure to
higher levels will aggravate preexisting angina pectoris. Symptoms seen in people
with higher level CO exposure include severe headache, dizziness, nausea, vomiting,
mental confusion, visual disturbances, reddening of the skin (not always), com-
partment syndrome, loss of muscle tissue, fatigue, hypotension, and coma. Severe
exposure can of course be fatal.8
The COHb levels measured in the CO-poisoned patient are sometimes used to
classify CO exposures in terms of mild, moderate, or severe poisonings. In this
system, COHb levels of less than 30% are termed “mild” poisonings. “Moderate”
poisonings are those in which the victim has a COHb level of from 30% to 40% and
“severe” poisonings occur when COHb levels are greater than 40% . This construct,
while sometimes useful at a simplistic level,7 should be used with great caution since
other signs and symptoms are known to be far more important in determining how
and when to treat a patient poisoned by CO.
Depending on the level and duration of CO exposure, delayed or prolonged
symptoms can also occur in CO-exposed individuals. Some of these effects can
be severe and can last for years. These symptoms can include sleep disturbances,
vision problems, hearing loss, tinnitus, peripheral neuropath, mental deficits, memory
problems, and difficulty concentrating, to name but a few. While there is often some
recovery of mental function in CO poisoned individuals over time, many brain tissues
damaged by CO show little, if any capability for regeneration. Similarly, the neur-
ological or other damage done in cases of compartment syndrome is not generally
reversible. Therefore, many of the chronic medical conditions brought on by severe
CO poisonings are likely to be permanent.5,9,10
12.3 ASSESSMENTS OF CARBON MONOXIDE

EXPOSURE LEVEL AND DURATION
In any poisoning case the investigator must try to determine the amount of exposure
an individual has had. The cornerstone of toxicology is the “dose makes the poison.”
Knowledge of the “dose” as reflected by the exposure a person has experienced
helps to assess the potential for adverse health effects the person may exhibit, guides
treatment that the individual will require, assists in assessing the patient’s follow-up
care needs, and determines his/her prognosis.
In the legal arena, a large part of job of the expert witness as a poisoning incid-
ent investigator is to determine what the exposure level was. Exposure assessments
are also required in the workplace to ascertain whether or not occupational exposure
standards were exceeded. Community exposure standards and guidelines have also
been established by federal and state environmental agencies and by private organ-
izations. Monitoring ambient air levels of CO and keeping outdoor CO levels below
applicable standards and guidelines is one of the jobs of federal and state air pollution
control agencies.
8417: “8417_c012” — 2007/9/11 — 12:15 — page 289 — #3

12.3.1 CARBOXYHEMOGLOBIN LEVELS AS A MEASURE OF CARBON

MONOXIDE EXPOSURE
When measurements of the CO level in the air have not been made or if made are not
representative of the actual CO exposure levels a poisoning victim has experienced,
other methods have been used to estimate CO exposure levels. One such method is
to rely on COHb levels in the victim’s blood.
When CO is inhaled, it is readily absorbed into the blood. Once there, the vast
majority of the CO binds to hemoglobin in the erythrocytes to form COHb. CO binds
to hemoglobin with an affinity more than 200 times that of oxygen. Thus CO displaces
the oxygen from the hemoglobin, impairing oxygen delivery to critical body tissues
such as the central nervous system, heart, and other organs.5
COHb levels in people exposed to CO reach a peak or plateau if the exposure
lasts for 5–10 h. Further exposure to CO will not increase the COHb saturation of the
hemoglobin if the air CO concentration remains constant. This is referred to as the
“equilibrium” (or steady-state) COHb level. Estimates have been made as to what
COHb level will be reached on the basis of concentration of CO in the air to which
a person is exposed. For example, a person exposed to 30 ppm CO for an extended
period of time will eventually have about 5% of their hemoglobin as COHb. An
exposure to 100 ppm CO will yield an equilibrium COHb level of 20%. Exposures
to 600 ppm gives a COHb level of more than 50%.11 In the latter case 50% of the
hemoglobin is in the “carboxy” form. Whether more subtle toxic effects occur after
this “plateau” COHb level is attained is the object of ongoing research.
Various investigators have attempted to correlate health effects with COHb level.
Such data nearly always show a huge variability, presumably because it is not just
the effect of CO on hemoglobin that is important, but also the effects of CO on tissue
and cells (e.g., on the cytochromes) and the effects of lipid peroxidation.
Some studies have shown decreased vigilance in subjects with only 2–3%
COHb.12,13 Others have shown no effects on vigilance or other health endpoints
at COHb levels up to 12.6%.14−27 Conversely, there are studies that showed effects
in addition to decreased vigilance at relatively low (less than 13%) COHb levels.
These effects included small changes in the electrocardiogram, increased minute
volume, reduction in exercise stamina, driving skill deficit, increased heart rate,
visual sensitivity decrement, fatigue, and increased reaction time.28−41
Mild headaches are reported in people with COHb levels of 13–20%,16 but other
effects have not generally been reported until COHb levels exceed about 30%.31,42−49
Symptoms seen at COHb levels of from 30% to 40% have been associated with
severe headache, dizziness, difficulty concentrating, nausea, vomiting, polycythemia,
and loss of consciousness.50,51 Subjects with COHb levels from 40% to 45% were
unable to perform any tasks requiring even minimal physical exertion.52 Coma and
convulsions usually occur at COHb levels of 50–60%, or below.51 Acute COHb levels
near 70% are almost always lethal. Note that the median COHb saturation of people
dead from CO-poisoning is near 53%.
While such crude relationships between COHb levels and health effects may
serve as a general guide, the toxicologist must be cognizant of the fact that different
individuals are affected differently by CO. This issue is discussed more fully below.
8417: “8417_c012” — 2007/9/11 — 12:15 — page 290 — #4

12.3.2 OCCUPATIONAL EXPOSURE STANDARDS FOR CARBON

MONOXIDE
Many CO exposures (and poisonings) occur in occupational settings. There are
two main types of occupational exposures to CO: accidental exposures and routine
exposures. In the case of accidental poisonings, the challenge for the investigator (who
is usually not present at the accident itself) is to determine what the exposure levels
might have been. It is rare for there to be air quality measurements in such situations
and the actions of emergency response personnel may complicate the exposure level
assessment. Obviously, the first duty of safety personnel is to assist the CO-poisoning
victim. In this effort, first responders may open doors or windows or turn on fans
or institute other ventilation efforts before they take measurements of ambient CO
levels. Measurements made after ventilation has taken place will be lower that those
responsible for the poisoning.
CO levels during routine industrial operations are easier to obtain and interpret.
Airborne CO levels are routinely monitored in some industrial settings, but in some
cases where CO intoxication is suspected, monitoring may not be available. A typical
function of the occupational safety investigator is to go to the place of business in
question and to obtain CO levels in the air under normal plant operating conditions.
These measured CO levels can then be compared to the existing occupation exposure
standards.
Occupational exposure standards have been set to protect worker’s health. There
are three major types of occupational exposure standards and guidelines for CO in
this country. The current Occupational Safety and Health Administration (OSHA)
standard (which carries the force of law) is an 8-h average of 50 ppm. The National
Institute of Occupational Safety and Health (NIOSH) recommended standard is an
8-h average of 35 ppm.53 The American Conference of Governmental and Indus-
trial Hygienists threshold limit value (TLV) over 8 h is 25 ppm.54 OSHA originally
proposed lowering their standard to that recommended by NIOSH, but this rule was
remanded by the U.S. Circuit Court of Appeals.55
These occupational standards and guideline limits were supposedly set to protect
against adverse cardiovascular, respiratory, and neurobehavioral effects. These limits
were set to also be protective of pregnant workers and their unborn children, and
other workers at high risk, although whether this is actually true is open to question.
NIOSH has also established a “ceiling” exposure limit for CO of 200 ppm. This
level is not to be exceeded at any time during a working day. The “Immediately
Dangerous to Life and Health” (IDLH) for CO is 1200 ppm. Exposure to levels of
CO greater than the IDLH “is likely to cause death or immediate or delayed permanent
adverse health effects or prevent escape from such an environment.”53
12.3.3 COMMUNITY EXPOSURE STANDARDS AND GUIDELINES FOR

CARBON MONOXIDE
Air pollution investigators are often called upon to access the levels of CO in
community air. CO levels are typically highest near highways or major industrial
facilities having combustion sources. As in the case of the industrial environment,
8417: “8417_c012” — 2007/9/11 — 12:15 — page 291 — #5

there are methods available to measure the levels of CO in community air. These
measured levels can then be compared to air quality standards or guidelines set by
federal or state agencies or by other organizations.
A variety of community exposure guidelines have been set up to protect public
health from air pollutants, including CO. Table 12.1 summarizes the current com-
munity exposure standards and guidelines for this chemical. The agencies and
organizations that have set community exposure limits for CO include the Amer-
ican Industrial Hygiene Association (AIHA),56 the U.S. Environmental Protection
Agency (EPA),57 and the California Air Resources Board (CARB).58,59 Like the
occupational exposure standards mentioned in the previous section, these limits have
been set to protect members of the general population from the adverse effects of CO
exposure.60 Since these standards and guidelines are subject to change, the reader
is urged to consult with the agency or organization involved to get the most current
exposure levels of interest.
12.3.4 CARBON MONOXIDE EXPOSURE DURATION ASSESSMENTS

As mentioned briefly above, the CO exposure duration as well as the exposure level
has to be considered in investigations of CO poisonings. In occupational and in some
nonoccupational settings, the frequency of exposure is also important. One general
rule in toxicology that reflects the importance of both exposure level and duration is
referred to as Haber’s Law or Haber’s Rule.
This rule can be stated as follows:
C×t =k
where C is the toxicant concentration or exposure level, t is the time of expos-

ure, and k is a constant reflecting the severity of the toxic effect. For example, if a
toxic substance obeys Haber’s Law, a 30 min exposure to 100 ppm of the chemical
should give a similar level of toxic effect that a 60 min exposure to 50 ppm would
(30 min × 100 ppm = 3000 ppm-min = 60 min × 50 ppm).61
The formation of COHb in CO-poisoned people does not seem to follow Haber’s
Law. Data from research studies and from clinical experience is summarized in
Table 12.2.62 Three different ranges of COHb levels are shown below: 2.0–2.7%,
7.0–8.5%, and 11.0–12.6%. These data show that the product of CO exposure level
and exposure duration is in no way indicative of COHb levels. This may be partially
explained by the fact that COHb levels reach an equilibrium concentration at some
point in time and do not increase even when exposure duration does. If C remains
constant and t is increasing, k, the COHb level in this case will not increase. The
deviation from Haber’s Law may also explained by the fact that responses to CO
exposures are highly variable from one individual to another.
In any case, the duration of exposure is important to the CO investigator. Unless
CO levels are very high, an exposure of a few minutes will rarely if ever cause adverse
effects. On the other hand, long-term exposures to quite low levels of CO for extended
periods of time (months to years) can lead to serious health consequences.
8417: “8417_c012” — 2007/9/11 — 12:15 — page 292 — #6

TABLE 12.1
Summary of Existing Community Air Quality Standards and Guidelines
Agency or Organization Type of Standard Carbon Monoxide Level (ppm) Averaging Time
AIHA TEEL-0 50 15 min
AIHA ERPG-1 200 1h
AIHA ERPG-2 350 1h
AIHA ERPG-3 500 1h
EPA NAAQS-1 35 1h
EPA NAAQS-2 9 8h
CARB AQS-1 0.25 1h
CARB AQS-2 0.04 24 h
CARB REL 20 1h
The AIHA is the American Industrial Hygiene Association. The EPA is the U.S. Environmental Protection
Agency. The CARB is the California Air Resources Board.
TEEL-0, the threshold concentration below which most people will experience no appreciable
risk of health effects.
TEEL-1, the maximum concentration in air below which it is believed nearly all individuals
could be exposed without experiencing other than mild transient adverse health effects or
perceiving a clearly defined objectionable odor.
could be exposed without experiencing or developing irreversible or other serious health
effects or symptoms that could impair their abilities to take protective action.
could be exposed without experiencing or developing life-threatening health effects.
ERPG-1, the maximum concentration in air below which it is believed nearly all individuals
could be exposed for up to 1 h without experiencing other than mild transient adverse health
effects or perceiving a clearly defined objectionable odor.
could be exposed for up to 1 h without experiencing or developing irreversible or other serious
health effects or symptoms that could impair their abilities to take protective action.
could be exposed for up to 1 h without experiencing or developing life-threatening health
effects.
NAAQS-1, the 1-h average National Ambient Air Quality Standard as established by the US
EPA.
NAAQS-2, the 8-h average National Ambient Air Quality Standard as established by the US
EPA.
AQS-1, the 1-h average air quality standard as established by the CARB.
AQS-2, the 8-h average air quality standard as established by the CARB.
REL, the Reference Exposure Level as established by the CARB. RELs are levels at or below
which even the most sensitive members of the community would not suffer any adverse health
effects.
8417: “8417_c012” — 2007/9/11 — 12:15 — page 293 — #7

TABLE 12.2
Carboxyhemoglobin Levels for Different Carbon Monoxide Exposure Levels
and Durations
CO Exposure Level (C in ppm) Exposure Duration (t in h) COHb Level (%) C × t (ppm-h)
50 1.33 2.0 67
12 192 2.4 2304
50 0.42 2.7 21
100 2.5 7.0 250
50 192 7.1 9600
650 0.75 8.5 488
75 168 11.0 12600
100 8.0 12.0 800
200 2.67 12.6 534
Another factor that comes into play in evaluating a person’s risk of CO exposure
is the frequency of exposure as compared to the half-life of CO excretion in the
body. For example, in the occupational world, workers normally work 8 h shifts.
Unless a worker has significant CO exposures outside of the work place (a possibility,
especially if the worker has a long daily commute through heavy traffic) he or she
will have 16 h away from work to recover from CO exposure on the job. If a person
takes in CO more quickly than it can be eliminated from the body, elevated COHb
will tend to persist during off-work times.
The half-life (expressed in units of time) of an exogenous chemical in an organism
is the time it takes for a given level of that chemical to be reduced by one-half. In two
half-lives, the chemical level would be 25% of the original level, after three half-lives,
it would be 12.5%, and so on. In CO poisonings, it is sometimes useful to know what
initial COHb level in a particular individual. This can be done if the COHb level is
measured a short time after the CO-poisoning, and then back-calculating to get the
initial level known at to , based on how many COHb half-lives have transpired.
The half-life for COHb of a person breathing ambient or room air is roughly
4–5 h.63 Next consider the previously mentioned example of a worker exposed for
8 h on the job and then having little or no CO exposure for the following 16 h. Sixteen
hours is approximately four half-lives for COHb for people breathing ambient air.
After four half-lives have passed, COHb levels should be reduced by 94%. While
this is a large reduction, it should be pointed out that the COHb levels would not
return to the baseline levels before the worker went back to work the next day, again
to be exposed to CO. On the second day of exposure the COHb level would start at
a higher baseline and would then reach a higher level than reached on the first day.
The intervening 16 h “rest” periods would decrease the COHb levels by 94% each
day, but there would be some accumulation over the work week. Fortunately, the 63 h
between 5:00 p.m. Friday afternoon and 8:00 a.m. Monday morning affords almost
16 half-lives, during which the COHb level would be reduced to less than 0.002% of
the level on Friday afternoon.
8417: “8417_c012” — 2007/9/11 — 12:15 — page 294 — #8

12.4 TREATMENTS FOR CARBON MONOXIDE

POISONING
Medical treatments administered to CO-poisoned patients are designed to reverse
the effects that CO has on the body; namely, the effects on the blood oxygen car-
rying capacity, the binding to myoglobin in muscle, the interference CO exerts on
the cytochrome oxidase system, effects on lipid peroxidation in the brain and other
tissues, and so forth. The type of treatment provided can influence the investig-
ator’s conclusions concerning the nature and extent of the damage caused by CO
poisoning.
Emergency medical technicians (EMTs) and other “first responders” are trained
to administer 100% oxygen to people they suspect have been poisoned by CO. This
treatment has the effect of reducing the COHb levels in the patient’s blood at a faster
rate than would be expected without such treatment. As mentioned earlier, the half-
life of COHb without supplemental oxygen is 4–5 h. The half-life of COHb in patients
breathing 100% oxygen centers around 60–80 min.63,64 Oxygen treatment functions
to promote the dissociation of COHb65 and to reduce tissue hypoxia.
A more controversial treatment regime for the CO-poisoning victim is hyperbaric
oxygen treatment (HBOT). See in-depth discussions of this treatment modality else-
where in this book. HBOT is accomplished by placing the patient in a large sealed
chamber, giving the patient 100% oxygen, and gradually increasing the pressure
inside to levels several times greater than of atmospheric. It is clear is that HBOT
can greatly decrease the half-life of COHb—to roughly 20–30 min.63,65,66 HBOT has
also been shown to promote CO dissociation from cytochrome a3 oxidase in animal
studies and to reduce brain lipid peroxidation.65
The efficacy of HBOT in humans has not been conclusive. Some studies have
shown marked reductions in the number and severity of both acute symptoms
and the incidence of delayed neuropsychological sequelae.67,68 Other investigat-
ors have reported either no advantage to HBOT69 or have even found that HBOT
worsened the patient’s condition.70 One explanation for these conflicting findings,
besides the variability in susceptibility expected in those CO-poisoned, is that
the groups of patients studied came from a wide variety of CO poisoning situ-
ations. Some HBOT was done at less than the optimal pressures of from 2.5 to
3.0 atm., while in other studies there had been a delay in initiating the HBOT.
Some groups of CO-poisoned individuals studied, included individuals who had
lost consciousness, while others did not and some studies had flaws in design and
execution.5,65,71
The current consensus seems to be that HBOT should be applied selectively. There
are risks of side-effects to this treatment and the transport of critically injured patients
to the nearest HBOT center also poses risks. HBOT may not be the best approach for
all patients. HBOT is more likely to have benefits outweighing its risks for patients
having one or more of the following characteristics:
1. Severe intoxication as evidenced by coma, seizures, focal neurological

deficits or cardiac effects65
2. Those who can be given HBOT within 6 h of the poisoning event67
8417: “8417_c012” — 2007/9/11 — 12:15 — page 295 — #9

3. Those presenting with COHb levels in excess of 25%.5,71

4. Some studies suggest that patients with less severe CO poisoning may do
just as well after normobaric treatment (i.e., sea level) with 100% oxygen.66
12.5 OTHER FACTORS TO CONSIDER IN

INVESTIGATIONS OF CARBON MONOXIDE
POISONING
The investigator of CO-poisoning events needs to be aware of other causes of the
symptoms that are associated with overexposure to CO. The process by which other
causes are dealt with is called the “differential diagnosis.” Other conditions that can
have symptoms in common with CO poisoning include viral infections, food poison-
ing, depression, anxiety, transient ischemic attacks, coronary artery disease, cardiac
arrythmias, Parkinson’s disease, meningitis or encephalitis, epilepsy, migraine, drug
overdose, ethanol intoxication, pneumonia, and sinusitis.10,65
Clues exist by which to discern whether the case being investigated is actually a
CO-poisoning case. For example, if all of the occupants of a residence or business
establishment are affected, many of the above alternate explanations can be elimin-
ated. If symptoms improve when the victims depart the site, this usually points the
finger at suspected site as the likely culprit. Indications of incomplete combustion,
such as yellow flames on gas appliances or heaters provide an indication that CO
exposure is possible/likely.70 In the occupational environment, CO exposures can
occur wherever there is a source of combustion. Some examples include propane
or gasoline powered forklifts (i.e., hi-lows, lift trucks), generators, bobcats, scissor
lifts, cherry pickers, concrete saws and chain saws, floor strippers and polishers, and
so forth.
The toxicity of CO is enhanced at high altitudes, at elevated temperatures, and in
people having increased ventilation or metabolic rates. The human fetus is at particu-
larly high risk from CO poisoning because of its normal development in a somewhat
oxygen-deprived environment. Children are at higher risk of CO poisoning because
they generally have higher metabolic rates than adults. People with preexisting dis-
eases such as anemia, cardiovascular, or cerebral vascular disease, hypovolemia, or
those with increased endogenous CO production are also at higher risk of adverse
effects from CO exposure.10
Exposures to some other chemicals either in the home or in the workplace can
mimic CO exposures and are easy to confuse with those exposures. For example,
exposure to methylene chloride (i.e., dichloromethane) can cause the same types of
effects as exposure to CO. This is because methylene chloride is metabolized to CO
and carbon dioxide in the body. An 8-h exposure to 150 ppm of methylene chloride
produces the same elevated carboxyhemoglobin level that a 35 ppm exposure to CO
would produce over the same time period. While causing less of an effect per ppm of
exposure, it has been found that the half-life of methylene chloride-induced COHb
is greater than that of CO-induced COHb, so effects may last for longer periods of
time.7 Exposure to cyanide can mimic CO intoxication, mainly because cyanide also
disrupts the functioning of cytochrome a3 oxidase.
8417: “8417_c012” — 2007/9/11 — 12:15 — page 296 — #10

12.6 CASE STUDY OF CARBON MONOXIDE

POISONING
The following is an account of an actual case in which three individuals were severely
poisoned by CO. This particular situation involved a faulty swimming pool heater at
a hotel. Malfunctions of heaters are a common cause of CO-related poisonings and
fatalities. The names of the victims, the other individuals involved in this incident,
and the name of the hotel have been omitted to maintain privacy.
The three individuals involved were staying at a hotel near Denver, Colorado.
Victim #1 went jogging the evening of October 31 and returned to his room at about
7:00 p.m. He soon felt ill and probably passed out around 8:00 p.m. that evening.
Housekeeping staff entered his room the next morning at 8:20 a.m., but seeing him
on the bed, assumed he was just sleeping soundly and left the room. A colleague of
Victim #1 who was staying at another hotel called Victim #1’s hotel at 9:45 a.m. on
November 1 and asked the desk clerk on duty to check on his friend. The clerk went
to Victim #1’s room, knocked on the door, but did not enter. He stated that it was
against hotel policy to enter a room unless the guest gave them authorization to do so.
Later that morning after repeated calls from the colleague, the hotel staff did enter
Victim #1’s room. Hotel records show that Victim #1’s door was opened at 11:04 a.m.
and 11:15 a.m. by the desk clerk trying unsuccessfully to rouse him. The colleague
finally came to the hotel to check on Victim #1 at about 11:30 a.m. and went to his
room. Finding him unconscious, the colleague asked the desk clerk to call 911 and
to summon emergency medical personnel. Victim #1 was finally removed from the
room at about 11:55 a.m. Victim #1 therefore had an approximately 17-h exposure to
CO. In this case the duration of exposure could be obtained from what is known as
an “audit trail.” At some hotels each time the card key is used to open the door, the
time and whose card was used (guest, housekeeping, desk clerks, etc.) is recorded on
the hotel computer. This victim suffered severe brain damage.
Victim #2 has stated that she got into her room at the hotel at about 5:00 p.m. on
October 31. After 15–20 min she felt confused and nauseated. Maids working at the
hotel entered her room at about 8:30 a.m. the next day and found her passed out on the
floor and thought that she had had too much to drink (example of a “misdiagnosis”).
Upon reporting this to the hotel management, the maids apparently were told to leave
Victim #2 alone. Nothing was done to aid her until the emergency personnel (who
had been summoned to assist Victim #1) arrived and removed her from her room just
before noon. She was therefore exposed to the CO for a total of about 19 h. This victim
incurred cognitive deficits and had to have a leg amputated because of compartment
syndrome she suffered caused by the CO exposure.
Victim #3 entered her room at about 5:00 p.m. on October 31. She believed she
passed out by about 6:00 p.m. Victim #3 awoke at about 6:30 a.m. the next morning,
showered, got dressed, and passed out again. She called some friends to come and get
her at about 8:00 a.m. Sometime shortly thereafter, she also called the front desk to
let them know she had vomited and that the carpet needed to be cleaned. Her friends
arrived at about 9:00 a.m. and seeing her condition took her to the hospital. Therefore,
she was in her room and exposed to the CO for about 15.5 h. Her outcome involved
limited neuropsychological deficits.
8417: “8417_c012” — 2007/9/11 — 12:15 — page 297 — #11

Blood tests done after these three victims reached the hospital showed that each
had elevated COHb levels. This confirmed that they had been exposed to CO. All three
people were diagnosed by physicians at the hospitals as having CO-poisoning. Their
symptoms at the emergency room (ER) included severe headache, dizziness, nausea,
vomiting, reddening of the skin, compartment syndrome, loss of muscle tissue,
fatigue, hypotension, and coma. These are consistent with what would be expected
from CO exposures. In addition, all three victims suffered ongoing health dam-
age involving sleep disturbance, vision problems, hearing loss, tinnitus, peripheral
neuropathy, mental deficits, memory problems, and difficulty concentrating.
During their investigations of this incident, local fire department personnel
measured CO concentrations in the rooms occupied by the three individuals. They
were all in excess of 200 ppm. These readings were taken after the doors to the rooms
had been opened and some ventilation of these rooms had occurred. Because of this,
the CO levels to which the victims were exposed would certainly have been higher
than that recorded by the fire department. The source of the CO was traced to the
swimming pool heater in the hotel. CO levels in this mechanical room were found to
be “extremely high” according to fire department personnel.
Police investigating this incident looked for, but did not find any evidence of
the presence of drugs, drug paraphernalia, or alcohol in the Plaintiffs’ hotel rooms.
Furthermore, there was no evidence of any violence or trauma to the victims. Victim
#1 suffered occasional headaches and visual problems prior to the incident, but these
became much worse afterwards. Otherwise, the Plaintiffs’individual medical histories
prior to this incident were unremarkable. Thus, it was possible to rule out other
possible causes for the victims’ adverse health effects.
The conclusion reached in this CO-poisoning investigation was that the adverse
health effects suffered by the victims were caused by their exposures to CO at the
hotel. The bases for this conclusion are as follows:
1. Elevated levels of CO in excess of 200 ppm were found in the hotel rooms
occupied by the individuals even after some ventilation of these rooms had
occurred
2. The victims were exposed to high levels of CO for periods of time ranging
from 15.5 to 19 h
3. The presence of elevated levels of COHb in the blood of these three
individuals confirms that they did sustain an exposure to CO
4. Both the acute and the long-term symptoms exhibited by the three victims
were entirely consistent with those associated with an overexposure to CO
5. Finally, the three victims in this case had no significant medical or psy-
chiatric conditions or problems prior to the incident, and no other likely
explanations could be found for the health harm suffered by the individuals.
Another conclusion was that the victims’ injuries would have been less severe if
they had been removed from their hotel rooms earlier. The housekeeping staff went
into each of the three victims’ rooms at about 8:30 a.m. on November 1. Although the
first two victims were found unconscious at that time, they were not taken out of their
rooms until about noon. This resulted in each of them sustaining an extra three and
8417: “8417_c012” — 2007/9/11 — 12:15 — page 298 — #12

one-half hours of CO exposure. The odds of more serious injuries go up with each
additional hour of exposure. Victim #3 was fortunate to have been able to summon
aid on her own. Notably, she suffered fewer and less severe effects than the other two
individuals who had 2–3 h more exposure to CO at the hotel.
The basis for this finding is that neurological damage from CO poisoning is
progressive. It becomes more severe the longer tissues are without adequate oxygen as
explained above. If Victim #1 had been taken out of his room earlier and had received
medical treatment sooner, he would have probably suffered less severe brain damage.
If Victim #2 had been rescued earlier, she may not have suffered such a severe injury
to her leg. Victim #3’s continuing injuries (problems with memory loss and other
mental deficits), while not as great as those suffered by the other two victims, would
most likely not have been as severe or taken as long to overcome had she been able
to get to the hospital sooner.
12.7 SUMMARY
CO poisoning is a leading cause of accidental injuries and fatalities in this country
and throughout the world. Exposures to CO are possible wherever there is a source of
combustion; for example, heating systems, fires, petroleum product fueled vehicles,
and industrial equipment, to name a few.
This chapter has been an overview of how investigations of CO-poisoning incid-
ents are carried out and the types of information required in such investigations.
Investigations include evaluations of the signs and symptoms of the intoxication,
assessments of exposure level and duration, reviewing the medical treatments that
may have been administered, and the ruling out of other factors that may have caused
the poisoning.
The investigator first of all should be familiar with the signs and symptoms of
CO poisoning. Many of the symptoms are common to other conditions such as viral
infections, alcohol intoxication, coronary disease, and other disease states. This can
complicate a positive determination of CO’s involvement. Other chemicals can cause
similar symptoms and should be also investigated as possible causes.
It is crucial in an investigation of CO poisonings to determine how great a “dose”
of CO the victim received. This may be done using CO measurement instruments, or
by determining COHb levels subsequent to the poisoning incident. Measured airborne
CO levels can be compared to occupational and community exposure standards and
guidelines, to assess health risk in more routine human exposure scenarios. The
medical treatment that an individual received can influence the investigation of CO
poisonings. Whether a victim received no supplemental oxygen, was given 100%
oxygen to breathe, or had HBOT, will influence the immediate health condition and
the long-term prognosis for that individual.
CO poisonings have occurred for tens of thousands of years, and probably even
before humans inhabited the earth (from volcanic activity, forest fires, etc.). The
advent of man’s use of controlled fires for warmth and cooking no doubt signaled a
sharp rise in the risk and incidence of CO poisonings. Even with all of the technolo-
gical advances that have been made over the millennia, CO poisoning is still common
8417: “8417_c012” — 2007/9/11 — 12:15 — page 299 — #13

and continues to be a major threat to public health. One objective of this chapter is
to assist those who are involved in investigations of CO poisoning. This chapter and
the others in this book also serve to raise awareness of this threat and to hopefully
reduce the number of CO poisonings.
References
1. Shephard, R.J. Carbon Monoxide, The Silent Killer, Charles C. Thomas Publisher,
Springfield, Illinois, 1983.
2. Kindwall, E.P. Carbon monoxide, In: Occupational Medicine, 3rd ed., Zenz, C.,
Dickerson, O.B., and Horvath, E.P., eds., Mosby, St. Louis, 1994, Chapter 29.
3. U.S. Environmental Protection Agency, Air Quality Criteria for Carbon Monoxide,
USEPA Publication No. EPA 600/P-99/001F, U.S. Government Printing Office,
Washington, DC., 2000.
4. Lipsett, M.J., Shusterman, D.J., and Beard, R.R. Inorganic compounds of carbon,
nitrogen, and oxygen, In: Patty’s Industrial Hygiene and Toxicology, 4th ed., Vol. II,
Part F, Clayton, G.D., and Clayton, F.E., eds., John Wiley & Sons, Inc., New York,
1994, pp. 4523–4552.
5. Tomaszewski, C. Carbon monoxide, In: Goldfrank’s Toxicological Emergencies, 6th
ed., Goldfrank, L.R., et al., eds., Appleton & Lange, Stamford, Connecticut, 1998,
pp. 1551–1563.
6. Victor, M., and Adams, R.D. Metabolic Diseases of the Nervous System, In:
Harrison’s Principles of Internal Medicine, 9th ed., McGraw-Hill Book Company,
New York, 1980, pp. 1978–1979.
7. Hathaway, G.J., Proctor, N.H., and Hughes, J.P. Carbon monoxide, In: Proctor
and Hughes’ Chemical Hazards of the Workplace, 4th ed., Van Nostrand Reinhold,
New York, 1996, pp. 113–116.
8. Smith, R.P. Toxic responses of the blood, In: Casarett and Doull’s Toxicology, The
Basic Science of Poisons, 5th ed., Klaassen, C.D., Amdur, M.O., and Doull, J., eds.,
McGraw-Hill, New York, 1996, Chapter 11, pp. 343–344.
9. Smith, J.S., and Brandon, S. Morbidity from acute carbon monoxide poisoning at
three-year follow-up, Br. Med. J. 1, 318, 1973.
10. Respiratory Toxicology, In: Ellenhorn’s Medical Toxicology, 2nd ed., Ellenhorn, M.J.,
et al., eds., Williams & Wilkins, Baltimore, pp. 1465–1476, 1997.
11. Seinfeld, J.H. Air Pollution, Physical and Chemical Fundamentals, McGraw Hill
Book Company, New York, 1975, pp. 22.
12. Fodor, C.G., and Winneke, C. Effect of Low CO concentrations on resistance to
monotony and on psychomotor capacity, Staub-Reinhalt Luft, 32, 46, 1972.
13. Beard, R.R., and Grandstaff, N.W. Carbon monoxide and human functions, In:
Environmental Science Research, Vol. 5, Behavioral Toxicology, B. Weiss and V.G.
Laties, eds. New York, Plenum Press, 1975,pp. 1–27.
14. Sievers, R.F., Edwards, T.I., and Murray, A.L. A Medical Study of Men Exposed to
Measured Amounts of Carbon monoxide in the Holland Tunnel for 13 Years, Public
Health Bulletin No. 278, U.S. Government Printing Office, Washington, DC., 1942.
15. Mikulka, P.R., et al. The effect of carbon monoxide on human performance, Ann. N.Y.
Acad. Sci. 174, 409, 1970.
16. Stewart, R.D., et al. Experimental human exposure to carbon monoxide, Arch.
Environ. Health 21, 154, 1970.
8417: “8417_c012” — 2007/9/11 — 12:15 — page 300 — #14

17. Horvath, S.M., Dahms, T.E., and O’Hanlon, J.F. Carbon monoxide and human vigil-
ance, a deleterious effect of present urban concentrations, Arch. Environ. Health 23,
343, 1971.
18. O’Donnell, R.D., et al. Low level carbon monoxide exposure and human psychomotor
performance, Toxicol. Appl. Pharmacol. 18, 583, 1971.
19. Raven, P.B., et al. Effect of carbon monoxide and peroxyacetyl nitrate on man’s
maximal aerobic capacity, J. Appl. Physiol. 36, 288, 1974.
20. Ettema, J.H., et al. Effects of alcohol, carbon monoxide and trichloroethylene
exposure on mental capacity, Int. Arch. Occup. Environ. Health 35, 117, 1975.
21. O’Hanlon, J.F. Preliminary studies of the effects of carbon monoxide on vigilance
in man, In: Behavioral Toxicology, Weiss, B., and Laties, G., eds., Plenum Press,
New York, 1975, pp. 61–75.
22. Benignus, V.A., et al. Lack of effects of carbon monoxide on human vigilance,
Perception and Motor Skills. 45(3, Pt 1), pp. 1007–1014, 1977.
23. Luria, S.M., and McKay, C.L. Effects of low levels of carbon monoxide on vision of
smokers and nonsmokers, Arch. Environ. Health 34, 38, 1979.
24. Davies, D.M., et al. The effects of continuous exposure to carbon monoxide on
auditory vigilance in man, Int. Arch. Occup. Environ. Health 48, 25, 1981.
25. DeLucia, A.J., Whitaker, J.H., and Bryant, L.R. Effects of combined exposure to
ozone and carbon monoxide (CO) in humans, In: Advances in Modern Environmental
Toxicology, Vol. 5, Lee, S.D., Mustafa, C.G., and Mehlman, M.A., eds., Princeton
Scientific Publishers, Princeton, N.J., 1983, pp. 145–159.
26. Mihevic, P.M., Gliner, J.A., and Horvath, S.M. Carbon monoxide exposure and
information processing during perceptual-motor performance, Int. Arch. Occup.
Environ. Health 51, 355, 1983.
27. Benignus, V.A., et al. Effect of low level carbon monoxide on compensatory tracking
and event monitoring, Neurotoxicol. Teratol. 9, 227, 1987.
28. McFarland, R.A., et al. The effects of carbon monoxide and altitude on visual
thresholds, J. Aviat. Med. 15, 381, 1944.
29. Ray, A.M., and Rockwell, T.H. An exploratory study of automobile driving perform-
ance under the influence of low levels of carboxyhemoglobin, Ann. N.Y. Acad. Sci.
174, 396, 1970.
30. Bender, W., Goethert, M., and Malorny, G. Effect of low carbon monox-
ide concentrations on psychological functions, Staub-Reinhalt Luft, 32, 54,
1972.
31. Ekblom, B. and Huot, R. Response to submaximal and maximal exercise at different
levels of carboxyhemoglobin, Acta. Physiol. Scand. 86, 474, 1972.
32. McFarland, R. Low-level exposure to carbon monoxide and driving performance,
Arch. Env. Health 27, 355, 1973.
33. Ramsey, J.M. Effects of single exposures of carbon monoxide on sensory and
psychomotor response, Am. Ind. Hyg. Assoc. J. 34, 212, 1973.
34. Wright, G., Randell, P., and Shephard, R.J. Carbon monoxide and driving skills, Arch.
Env. Health 27, 349, 1973.
35. Drinkwater, B.L., et al. Air pollution, exercise, and heat stress, Arch. Env. Health. 28,
177, 1974.
36. Gliner, J.A., et al. Man’s physiologic response to long-term work during thermal and
pollutant stress, J. Appl. Physiol. 39, 628, 1975.
37. Horvath, S.M., et al. Maximal aerobic capacity at different levels of carboxyhemo-
globin, J. Appl. Physiol. 38, 300, 1975.
8417: “8417_c012” — 2007/9/11 — 12:15 — page 301 — #15

38. Putz, V.R., Johnson, B.L., and Setzer, J.V. Effects of CO on Vigilance Performance.
Effects of Low-Level Carbon monoxide on Divided Attention, Pitch Discrimina-
tion, and the Auditory Evoked Potential, Publication Number DHEW (NIOSH)
77–124, U.S. Department of Health, Education, and Welfare, National Institute of
Occupational Safety and Health, Cincinnati, Ohio, 1976.
39. Putz, V.R., Johnson, B.L., and Setzer, J.V. A comparative study of the effects of carbon
monoxide and methylene chloride on human performance, J. Env. Pathol. Toxicol. 2,
97, 1979.
40. Davies, D.M. and Smith, D.J. Electrocardiographic changes in healthy men during
continuous low-level carbon monoxide exposure, Env. Res. 21, 197, 1980.
41. Bunnell, D.E. and Horvath, S.M. Interactive effects of heat, physical work and CO
exposure on metabolism and cognitive task performance, Aviat. Space Env. Med. 60,
428, 1989.
42. Chevalier, R.B., Krumholz, R.A., and Ross, J.C. Reaction of non-smokers to carbon
monoxide inhalation, cardiopulmonary responses at rest and during exercise, JAMA.
198, 1061, 1966.
43. Pirnay, F., et al. Muscular exercise during intoxication by carbon monoxide, J. Appl.
Physiol. 31, 573, 1971.
44. Vogel, J.A. and Gleser, M.A. Effect of carbon monoxide on oxygen transport during
exercise, J. Appl. Physiol. 32, 234, 1972.
45. Vogel, J.A., et al. Carbon monoxide and physical work capacity. Arch. Environ. Health
24, 198, 1972.
46. Parving, H.H. The effect of hypoxia and carbon monoxide exposure on plasma volume
and capillary permeability to albumin, Scand. J. Clin. Lab. Invest. 30, 49, 1972.
47. Stewart, R.D., et al. Effect of carbon monoxide on time perception, Arch. Environ.
Health 27, 155, 1973.
48. Hudnell, H.K. and Benignus, V.A. Carbon monoxide exposure and human visual
detection thresholds, Neurotoxicol. Teratol. 11, 363, 1989.
49. Schulte, J.H. Effects of mild carbon monoxide intoxication, Arch. Environ. Health 7,
524, 1963.
50. DiMarco, A. Carbon monoxide poisoning presenting as polycythemia, N. Engl. J.
Med. 319, 874, 1988.
51. Chiodi, H., et al. Respiratory and circulatory responses to acute carbon monoxide
poisoning, Am. J. Physiol. 134, 683, 1941.
52. Stewart, R.D. The effect of carbon monoxide on humans, Ann. Rev. Pharmacol. 15,
409, 1975.
53. NIOSH Pocket Guide to Chemical Hazards, Department of Health and Human Ser-
vices, Centers for Disease Control and Prevention, National Institute for Occupational
Safety and Health, DHHS (NIOSH) Publication Number 97–140, U.S. Government
Printing Office, Washington, DC, 2003.
54. 2005 TLVs and BEIs, American Conference of Governmental Industrial Hygienists,
ACGIH Signature Publications, Cincinnati, Ohio, 2005.
55. Department of Labor, Occupational Safety and Health Administration, 29 CFR Part
1910, Air Contaminants, Federal Register, 54, 2651, 1989.
56. Current values for the American Industrial Hygiene Association can be found at the
Internet Web Site http://www.eh.doe.gov/chem_safety/teel.html.
57. Current National Ambient Air Quality Standards can be found at the USEPA Internet
Web Site http://www.epa.gov/air/criteria.html.
58. Current values for California Air Quality Standards can be found at the CARB Internet
Web Site http://www.baaqmd.gov//pln/air_quality/ambient_air_quality.htm.
8417: “8417_c012” — 2007/9/11 — 12:15 — page 302 — #16

59. Current values for the California Reference Exposure Limits can be found at the
CARB Internet Web Site http://www.oehha.ca.gov/air/pdf/acuterel.pdf.
60. Craig, D.K. Derivation of temporary emergency exposure limits (TEELs), J. Appl.
Toxicol. 20, 11, 2000.
61. Pastenbach, D.J. The History and Biological Basis of Occupational Exposure Limits
for Chemical Agents, In: Patty’s Industrial Hygiene and Toxicology, 5th ed., Volume
3, VI, Harris, R.L., ed., John Wiley & Sons, Inc., New York, 2000, pp. 1953–1954.
62. National Research Council, Carbon monoxide, In: Review of Submarine Escape
Action Levels for Selected Chemicals, National Academy Press, Washington, DC,
2002, pp. 69–96.
63. Varon, J. and Marik, P.E. Carbon monoxide poisoning, J. Emerg. Int. Care Med. 1, 1,
1997.
64. Weaver, L.K., et al. Carboxyhemoglobin half-life in carbon monoxide-poisoned
patients treated with 100% oxygen at atmospheric pressure, Chest. 117, 801, 2000.
65. Gill, A.L. and Bell, C.N.A. Hyperbaric oxygen: Its uses, mechanisms of action, and
outcomes, Q. J. Med. 97, 385, 2004.
66. Leach, R.M., Rees, P.J., and Wilshurst, P. ABC of oxygen: hyperbaric oxygen therapy,
Brit. Med. J. 317, 1140, 1998.
67. Thom, S.R., et al. Delayed neuropsychological sequelae after carbon monoxide pois-
oning: Prevention by treatment with hyperbaric oxygen, Ann. Emerg. Med. 25, 474,
1995.
68. Ducasse, J.L., Celis, P., and Marc-Vergnes, J.P. Non-comatose patients with acute car-
bon monoxide poisoning: Hyperbaric or normobaric oxygenation?, Undersea Hyperb.
Med. 22, 9, 1995.
69. Weaver, L.K., Hopkins, R.O., and Larson-Lohr, V. Neuropsychological and functional
recovery from severe carbon monoxide poisoning without hyperbaric oxygen therapy,
Ann. Emerg. Med. 27, 736, 1996.
70. Scheinkestel, C.D., et al. Hyperbaric or normobaric oxygen for acute carbon monoxide
poisoning: A randomized controlled clinical trial, Med. J. Aust. 170, 203, 1999.
71. Harper, A. and Croft-Baker, J. Carbon monoxide poisoning: undetected by both
patients and their doctors, Age Ageing 33, 105, 2004.
8417: “8417_c012” — 2007/9/11 — 12:15 — page 303 — #17

8417: “8417_c012” — 2007/9/11 — 12:15 — page 304 — #18
13 Carbon Monoxide
Detectors as Preventive
Medicine
James W. Rhee and Jerrold B. Leikin
CONTENTS
13.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305

13.2 Carbon Monoxide Detector Technology: A Brief Review . . . . . . . . . . . . . . . . 306
13.3 Setting an Initial Standard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
13.4 The Chicago Experience. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
13.5 Clinical Implications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
13.6 Some Data Regarding Effectiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
13.7 Current and Future Directions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
13.1 INTRODUCTION
Carbon monoxide (CO) is odorless, colorless, tasteless, and nonirritating. As such,
CO has no warning properties that can alert unwary individuals to its presence. When
people become ill owing to CO toxicity, the symptom complex can be nonspecific and
variable making the diagnosis of CO poisoning difficult. Given its lack of warning
signs and lack of distinct clinical features, CO has been labeled by some as a “silent
killer.”
Data analyzed by Cobb and Etzel from the National Center for Health Statistics,
attributed 53,133 deaths to CO from 1979 to 1988—making it the most common
cause of acute poisoning deaths.1 When Shepherd and Klein-Schwartz examined the
mortality data from 1979 to 1994 for persons aged 10–19 years, they found 3034 out
of 7936 poisoning deaths were attributable to carbon monoxide.2 As these numbers
were derived from databases, they are likely to be significant under representation
of the true number of deaths caused by CO. They also do not take into account the
significant morbidity that may occur from severe CO exposures.
Given the “silent” nature of CO and its significant impact on public health, a need
for residential-based CO detectors was identified. In 1989, the Consumer Product
Safety Commission (CPSC) urged Underwriters Laboratories (UL) (Northbrook,
Illinois) to develop a standard that would serve as a guideline for residential
305
8417: “8417_c013” — 2007/9/11 — 12:15 — page 305 — #1

CO detectors. This standard (UL 2034) applying to single and multiple station CO
detectors was published in April, 19923 —subsequently, the first two CO detectors
were listed later that year.4
13.2 CARBON MONOXIDE DETECTOR

TECHNOLOGY: A BRIEF REVIEW
Most of the residential CO detectors utilize one of the two methods of sensing the
presence of CO—either a biomimetic sensor or a metal oxide sensor.
The types of technology usually involved in residential sensors include the “Gel
cell” metal oxide and electrochemical sensor. The nondispersive infrared technology
(NDIR) is usually utilized for industrial purposes. The “Gel cell” based sensor sends a
light beam through a biomimetic sensor to a photosensitive component which alarms
at appropriate set points. This colorimetric sensor essentially mimicked the hemo-
globin uptake of CO, thus changing the spectral response. These types of sensors
exhibit proper sensitivity to CO and can achieve a unique accumulation of CO. There
is some interference with other gases and vapors along with humidity, and these types
of sensors take a longer time to recover in the setting of CO removal (slower response
reversibility). The “Gel cell” detectors are rarely equipped with digital displays.
The metal-oxide (usually tin-oxide) sensor detects CO by measuring the resistivity
of the metal component through oxidation of CO to carbon dioxide, which reduces
the resistance of the sensor. These sensors usually require main (i.e., AC) power and
cannot use batteries as a primary power source.
Upon exposure to CO, the electrochemical sensors will generate electricity
through an acid electrolyte (usually sulfuric acid or phosphoric acid). This technology
has been utilized as portable or fixed gas monitors in industry owing to its resolution
(down to 1 ppm) and stability, and is probably the primary type of sensor used in
residential detectors. These sensors can also be affected by cross-contamination of
gases.
The NDIR can measure CO concentrations with an accuracy of ±2 ppm, but are
expensive and therefore usually not utilized in residential detectors. Metallocorroles
utilizing cobalt (111) can selectively absorb CO gas on a molecular complex and may
provide a future matrix in sensor development.5
13.3 SETTING AN INITIAL STANDARD

UL is a well-known organization located in Northbrook, Illinois that develops
standards and test procedures for materials, components, assemblies, tools, equip-
ment, and procedures, chiefly dealing with product safety and utility. The CPSC
worked closely with UL to develop standards for CO detectors. Part of setting this
standard required UL to create thresholds at which the detectors would alarm.
The threshold at which the residential CO detectors would alarm presented a
unique challenge. Initially, set points were generated on the basis of extrapolating
what levels of CO over what period of time would generate a carboxyhemoglobin
(COHb) concentration of 10% in a nonsmoking individual. These values were based
8417: “8417_c013” — 2007/9/11 — 12:15 — page 306 — #2

Carbon Monoxide Detectors as Preventive Medicine 307
TABLE 13.1
UL 2034 (1992–1996). Carbon Monoxide Con-
centration Versus Time for Alarm Test Points
Air Concentration (ppm) Maximum Response Time (min)
100 90
200 35
400 15
on the Coburn–Forster–Kane equation. (In addition, the CO detector should alarm

at an exposure of 6000 ppm within 3 min.)6 This initial standard (UL 2034) was set
in place in 1992.3 Soon afterwards CO detectors were sold across the United States.
Table 13.1 illustrates the original set points for activating the audible alarm in the CO
detectors.
Other standards, such as loudness of the alarm and the response to other gases
needed to be clarified as well. Similar to smoke detectors, the CO detector was
set to emit an 85-dB alarm (at 10 ft.), which is loud enough to wake most people
when sounding outside a bedroom through a closed door. However, unlike smoke
detectors, CO detectors have computer processor-based software which allow it to
alarm at certain set points. Sensors were set to not alarm when exposed for 2 h to
methane (at 500 ppm), butane (at 300 ppm), heptane (at 500 ppm), ethyl acetate
(at 200 ppm), isopropyl alcohol (at 200 ppm), carbon dioxide (at 5000 ppm), toluene
(at 200 ppm), and acetone (at 200 ppm).7 Other major features include a red light-
emitting diode (LED) as a trouble signal, a green LED indicating normal operation
and a reset mechanism for testing and resetting purposes. Some detectors utilize a
digital display (over 30 ppm).
13.4 THE CHICAGO EXPERIENCE

The state of Illinois, in the past had one of the highest fatality rates from CO—
accounting for 8.7% of all unintentional CO deaths nation-wide, with a rate of about
0.9 deaths per 1000 people between 1979 and 1988.1 In light of this statistic, Chicago,
Illinois, passed a mandatory CO detector ordinance in March, 1994 requiring CO
detectors to be present in all homes, apartments, class “B” and “C” buildings with
heat sources that generate CO.8
During the first 3 months of the ordinance implementation, 68 individuals were
transported with suspected CO poisoning attributed to CO detector alarms.9 However,
during this time period (October 1–December 31, 1994), the Chicago Fire Department
received over 12,000 calls of CO detector alarming. In about 85% of these cases, less
than 9 ppm of CO was reportedly measured by the Chicago Fire Department.9–12 This
was climaxed on December 21–22, 1994, when the Chicago Fire Department was
involved in 3464 CO investigations.11 It was subsequently determined that a thermal
air inversion (upper atmosphere warmer than lower atmosphere) had caused a fivefold
8417: “8417_c013” — 2007/9/11 — 12:15 — page 307 — #3

elevation of ambient CO concentration, of 13–20 ppm, thus setting off multiple CO

alarms. CO detector sensitivity or resistance specifications were 15 ppm over an 8-h
period at that time.9,10
In response to the inordinate number of “nuisance alarms,” UL increased the
15 ppm resistance specification from 8 h to 30 days in the 1996 revised standard.7
The UL standard 2034 is under a continuous maintenance process and was last revised
in 2006.
13.5 CLINICAL IMPLICATIONS

Given the variable presentation of CO toxicity—it is apparent that healthcare pro-
viders could use assistance in diagnosing occult CO poisoning. Patients with CO
poisoning are often misdiagnosed as having a viral syndrome or food poisoning.13
Earlier studies have shown that among adult patients presenting to an emergency
department during the winter months with complaints of headache or dizziness, 3–5%
have COHb levels greater than 10%.14
Because CO is a colorless, odorless, tasteless, and nonirritating gas, the physician
and the patient have few clues as to the contribution of CO to the illness (see other
chapters in this book). The CO detector can act as a useful screening tool to identify CO
exposure. The CO detector should allow the clinician to obtain a history of exposure
and prompt the clinician to begin investigating the potential for CO poisoning.
13.6 SOME DATA REGARDING EFFECTIVENESS

CO detectors/alarms have made death from CO poisoning completely preventable.
CO detectors can save lives.
Despite the high number of false alarms in Chicago after the initial mandatory CO
detector ordinance was established in 1994, only one CO-related death was reported
in the Chicago media between September 1994 through February 1998. Compared to
other cities during this same time period, Chicago (the only city to have a CO detector
ordinance during the study period) had the lowest case fatality rate.15 And despite
the loudly voiced criticisms and objections to the ordinance from various sources, a
Chicago Sun-Times poll (though hardly scientific) reported that 77% of respondents
supported such a mandatory ordinance.
Krenzelok et al.16 found that in an investigation of emergency responses to
possible CO poisoning, residences with CO alarms had lower CO concentrations
(18.6 ppm with CO detectors versus 96.6 ppm without CO detectors) and fewer
symptomatic patients (11.7% with CO detectors vs 63.4% without CO detectors).16
They concluded that audible CO detectors were effective in alerting people to the
presence of abnormal levels of CO, thus resulting in less exposure to CO. This
subsequently lead to a lower incidence of CO-related symptoms.16
A study conducted by Bizovi et al.10 found similar results where only about
5% of individuals at a site where a CO detector had activated displayed signs of
CO poisoning, suggesting that the CO detector may have prevented more serious
exposure.10
8417: “8417_c013” — 2007/9/11 — 12:15 — page 308 — #4

Yoon et al.17 estimate that 78 out of 136 unintentional deaths due to CO poisoning
that occurred during a 15-year period in New Mexico, may have been prevented if
audible CO detectors were in use.17
In North Carolina, a couple of studies evaluated the impact of a local CO detector
ordinance (which exempted all-electric heated homes) was conducted.18,19 While, the
ordinance did not seem to decrease the amount of CO exposures, the relative incid-
ence of severe poisoning requiring hyperbaric oxygen treatment was diminished.18
Another study evaluating a CO poisoning outbreak during a winter storm in 2002
demonstrated that the North Carolina county CO detector ordinance did not eliminate
a CO poisoning outbreak, but it did mitigate its effects.19 Of note, the study found
that none of the patients who developed symptoms of severe CO poisoning had a
functioning CO detector.19
Despite the apparent effectiveness of CO detectors at saving lives, only 29% of
respondents to a survey conducted by Runyan et al.20 reported the presence of CO
detectors in their homes.20 The lack of widespread use of these detectors intuitively
limits the effectiveness of these devices to have a profound impact on mortality and
morbidity due to CO poisoning.
13.7 CURRENT AND FUTURE DIRECTIONS

Newer aspects of CO detector specifications include an alarm reset within 6 min if the
CO concentration exceeds 70 ppm, secondary power supply considerations, and alarm
tests point revision (revised November, 2001—see Table 13.2). Usage in recreational
vehicles/marine units and unconditioned areas was added in 1997.7 The revised UL
2034 standard is similar to that of the International Approved Services (IAS) and the
British Standards Institute (BSI).
While these new standards and new sepcifications for CO detectors facilitate the
design and development of newer and better CO detectors, the potential impact these
CO detectors can have on public health is limited by the prevalence of their use.
TABLE 13.2
UL 2034, revised November, 2001. Carbon Monoxide Con-
centration Versus Time for Alarm Test Points Based on
10% Carboxyhemoglobin (COHb)
A. Carbon Monoxide Concentration (ppm) and Response Time
Concentration, ppm Response time, min
70 ± 5 60–10
150 ± 5 10–50
400 ± 10 4–15
B. False Alarm-Carbon Monoxide Concentration Resistance Specifications
Concentration, ppm Exposure time, (no alarm)
30 ± 3 30 days
70 ± 5 60 min
8417: “8417_c013” — 2007/9/11 — 12:15 — page 309 — #5

This problem can potentially be remedied by CO detector ordinances set at the state
and local levels, since the ordinances already in place have had a demonstrable impact
on public health.
References
1. Cobb, N. and Etzel, R. A. Unintentional carbon monoxide-related deaths in the United
States, 1979 through 1988, JAMA, 266, 659, 1991.
2. Shepherd, G. and Klein-Schwartz, W. Accidental and suicidal adolescent poisoning
deaths in the United States, 1979–1994, Arch. Pediatr. Adolesc. Med., 152, 1181,
1998.
3. Standard for Safety UL 2034 Single and Multiple Station Carbon Monoxide Detectors,
Underwriters Laboratories, Northbrook, IL, 1st ed., 1992.
4. Hrones, T. L. and Patty, P. E. Carbon monoxide detectors: protection against the silent
killer, In Poisoning and Toxicology Compendium, Leikin, J. B., and Paloucek, F. B.,
eds., Lexicomp, Hudson, Ohio, 1998, 630.
5. Barbe, J. M., Canard, G., Brandes, S., Jerome, F., Dubois, G., and Guilard, R. Metal-
locorroles as sensing components for gas sensors: remarkable affinity and selectivity
of cobalt(III) corroles for CO vs. O2 and N2 , Dalton Trans., 1208, 2004.
6. Coburn, R. F., Forster, R. E., and Kane, P. B. Considerations of the physiolo-
gical variables that determine the blood carboxyhemoglobin concentration in man,
J. Clin. Invest., vol. 44(11), 1899–1910, 1965.
7. Standard for Safety UL 2034 Single and Multiple Station Carbon Monoxide Detectors,
Underwriters Laboratories, Northbrook, IL, 2nd ed., 1996.
8. City Council of the City of Chicago: Amendment of Title 13, Chapter 64 of the
Municipal Code of Chicago by addition of new sections 190 through 300 requiring
CO detectors in various buildings. Meeting of March 2, 1994.
9. Leikin, J. B. Carbon monoxide detectors and emergency physicians, Am. J. Emerg.
Med., 14, 90, 1996.
10. Bizovi, K. E., Leikin, J. B., Hryhorczuk, D. O., and Frateschi, L. J. Night of
the sirens: analysis of carbon monoxide-detector experience in suburban Chicago,
Ann. Emerg. Med., 31, 737, 1998.
11. Eversole, J. M. Carbon Monoxide Detector Ordinance: Review of the Chicago Fire
Department experience, In Poisoning and Toxicology Compendium, Leikin, J. B., and
Paloucek, F. B., eds., Lexicomp, Hudson, Ohio, 1998, 633.
12. Leikin, J. B., Krenzelok, E. P., and Greiner, T. H. Remarks to the Illinois House
of Representatives Executive Committee hearing regarding state carbon monoxide
detector act (HB 603), J. Toxicol. Clin. Toxicol., 37, 885, 1999.
13. Barret, L., Danel, V., and Faure, J. Carbon monoxide poisoning, a diagnosis frequently
overlooked, J. Toxicol. Clin. Toxicol., 23, 309, 1985.
14. Heckerling, P. S., Leikin, J. B., and Maturen, A. Occult carbon monoxide poisoning:
validation of a prediction model, Am. J. Med., 84, 251, 1988.
15. Clifton, J. C. N., Leikin, J. B., Hryhorczuk, D. O., and Krenzelok, E. P. Surveillance
for carbon monoxide poisoning using a national media clipping service, Am. J. Emerg.
Med., 19, 106, 2001.
16. Krenzelok, E. P., Roth, R., and Full, R. Carbon monoxide: the silent killer with an
audible solution, Am. J. Emerg. Med., 14, 484, 1996.
17. Yoon, S. S., Macdonald, S. C., and Parrish, R. G. Deaths from unintentional carbon
monoxide poisoning and potential for prevention with carbon monoxide detectors,
JAMA, 279, 685, 1998.
8417: “8417_c013” — 2007/9/11 — 12:15 — page 310 — #6

18. Tomaszewski, C., Lavonas, E., Kerns, R., and Rouse, A. Effect of a carbon monoxide
alarm regulation on CO poisoning, J. Toxicol. Clin. Toxicol.,41(5), 167–168, 2003.
19. Lavonas, E., Tomaszewski, C., Kerns, W., and Blackwell, T. Epidemic carbon
monoxide poisoning despite a CO alarm law, J. Toxicol. Clin. Toxicol.,41(5),
711–712, 2003.
20. Runyan, C. W., Johnson, R. M., Yang, J., Waller, A. E., Perkis, D., Marshall, S. W.,
Coyne-Beasley, T., and McGee, K. S. Risk and protective factors for fires, burns, and
carbon monoxide poisoning in U.S. households, Am. J. Prev. Med., 28, 102, 2005.
8417: “8417_c013” — 2007/9/11 — 12:15 — page 311 — #7

8417: “8417_c013” — 2007/9/11 — 12:15 — page 312 — #8
14 Misconceptions About
Carbon Monoxide
David G. Penney
CONTENTS
14.1 Properties, Presence, and Detection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313

14.2 Physiology of Carbon Monoxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
14.3 Symptoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 318
14.4 Treatment and Outcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
14.5 Miscellaneous . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
14.1 PROPERTIES, PRESENCE, AND DETECTION

There are many many misunderstanding about carbon monoxide (CO), even today,
by the general public, healthcare professionals, and others who ought to be better
informed. The first grouping of these misconceptions are shown in Table 14.1.
CO is known as the “Silent Killer” or the “Stealthy Poison” because it is impossible
to detect by humans—we cannot smell, taste, or see CO gas, no matter what the
concentration. For this reason gas supply companies often put smelly tracers in fuel
gases to assist us in detecting their presence. With school groups and nonscientists I
often refer to CO as a “smart poison,” as opposed to “dumber poisons” like hydrogen
sulfide and others that we have warning of by our chemoreceptors (i.e., taste, smell).
CO is also smart in that it enters our bodies simply by our breathing, and does not have
to be “picked up” by ingestion through the mouth as lead largely is. Another element
of “smartness” is that it leaves the body quickly, although damage may already have
been done, while dumb poisons remain (lead, mercury, etc.) to be detected days,
weeks, even years later.
CO is of course slightly lighter than air, being made up of one carbon atom and
one oxygen atom. It has about the dimensions of an oxygen molecule (i.e., diatomic
oxygen), contributing to several others of its properties (i.e., ability to diffuse easily,
attach to hemoglobin where oxygen sits). The molecular weight of CO is roughly 28
(see Reference Data, Chapter 35). Calculating the molecular weight of air, a mix-
ture of nitrogen (28), oxygen (32), argon (40), carbon dioxide (44), and water vapor
(18), gives a weighted average of approximately 29. Thus, CO is about 4% lighter than
313
8417: “8417_c014” — 2007/9/11 — 12:15 — page 313 — #1

TABLE 14.1
Misconceptions: Properties, Presence, and Detection
• CO is easy to detect.
• CO is lighter than air and therefore rises (to the ceiling) and stays there.
• CO is not combustible.
• CO adsorbs and absorbs to fabric, crockery, walls and thus remains long after it has left the air.
• CO and natural gas are the same thing.
• Natural gas contains considerable amounts of CO.
• You can always tell if CO is present because of a peculiar odor that is present.
• A brand new, well designed, perfectly “tuned” heating/cooking device cannot produce toxic/lethal
amounts of CO.
• Diesel engine exhaust never contains enough CO to cause harm.
• Heating, ventilation, and air conditioning (HVAC) and gas company service personnel always check
for CO when performing maintenance/service on home heating systems.
• CO will be detected immediately by service personnel if it is present in a home heating system.
• When your home CO detector shows low levels of CO, it is probably just an instrument malfunction.
• Cracks in heat exchangers are responsible for the production of CO.
• Home CO detectors/sensors are the best devices to ferret out CO because they react to very low levels
of the gas.
air. However, the assertion that CO rises to the ceiling when released into the air of
a structure is incorrect. The CO is always mixed with vast quantities of excess air,
mixes thoroughly and completely, and cannot unmix lest it violate the Second Law of
Thermodynamics. The key to observations of higher CO concentration near ceilings
stems from the fact that hot air rises, and exhaust containing CO is often warmer than
the surrounding air.
CO is of course combustible and has been used as a fuel gas. It will burn to
carbon dioxide, which it does with the release of additional heat. CO was used as a
fuel source in past years, for both lighting and heating. This property should be kept
in mind when working around high concentrations of CO, where ignition may occur.
We may forget about this property because usually the toxic effects are of greatest
significance at far lower concentrations.
CO does NOT attach in any way to fabric, glass, crockery, and so forth. I heard
that a physician when seeing a patient who had sustained acute CO poisoning, told
his patient’s mother to clean everything in the apartment after the incident because
CO “sticks.” She did this as well as giving away all the furniture and clothing that had
been in the apartment. This was of course entirely needless; the result of inadequate
education in toxicology.
Natural gas as it is used in the United States, contains little CO. Its main component
is methane, CH4 . I understand there are also small amounts of other gases too, of the
aliphatic series (e.g., ethane). CO may only be present around a combustion device
using natural gas after the gas is combusted, and when that combustion process is
incomplete, that is, not all of the gas is burned to form carbon dioxide and water
vapor. Other possible incomplete combustion products include soot (i.e., elemental
carbon).
8417: “8417_c014” — 2007/9/11 — 12:15 — page 314 — #2

Misconceptions About Carbon Monoxide 315
All common combustion devices involve some amount of incompelete combus-

tion, although under certain circumstances this may be very small. So, all such devices
produce some CO. Usually the concentration of CO generated is below levels set as
safe by regulatory standards. But even these devices may under certain conditions
produce extraordinarily high, even lethal concentrations of CO if something else in
the heating system is inadequate or fails (compromised combustion air source, broken
exhaust system), or if slow or catastrophic failure of the heating device itself takes
place through corrosion, lack of cleaning, physical impact by outside forces, and so
forth.
Diesel engines generally produce lower CO concentrations in exhaust gases than
gasoline-fueled engines. On the other hand, the Environmental Protection Agency
(EPA) over the past several decades has required gasoline-fueled vehicles in the
United States to have catalytic converters that work on the exhaust gases. This device
converts the CO to carbon dioxide using various catalysts. They only work effectively
when they are hot, that is, they are not effective at cold start. The key point here is that
diesel engine equipped vehicles lack catalytic converters, so ALL the CO generated
by the engine goes out the tailpipe. Thus, the CO emissions from many diesel powered
vehicles are above those from gasoline-fueled vehicles.
Heating and cooling technicians and gas company personnel should always
monitor for CO when checking a combustion device. Unfortunately, this is not always
done. “Sniffing” for natural gas or propane (LPG) is more common, to detect leaks
that could lead to fire or explosion. Checking for CO that can lead to injury and/or
death is sadly often not done. Unfortunately, even when CO is checked, a problem
may be missed. A high CO reading is proof-positive of a problem with a combustion
device or the system it is operating in. Failure to find CO on one attempt does not
entirely rule out the possibility of a problem during that time frame and under slightly
different environmental conditions. Of course measurements must be made when
the combustion device is operating, and for sufficient time to catch possible delayed
build-up. Often measurements will have to be made several times, and under slightly
different conditions, for example, when other combustion devices are also operating,
when exhaust fans are being run, when outside wind conditions change, and so forth.
One of the most common, possibly lethal mistakes made by people is to assume
that a residential CO alarm (i.e., detector) showing low, or even high levels of CO,
is malfunctioning. “It never did before,” “it must be the battery,” “The smoke alarm
is not sounding,” “take the battery out and go back to bed.” I recommend that people
have a minimum of two CO alarms in their residence, one located very near where
they sleep, and another near where they are when awake, reading, watching TV,
and so forth. Two alarms sounding simultaneously make it much less likely that a
person will come to the conclusion that the warning is due to malfunction. If there
are two or more floors to the residence, an additional alarm should be installed for
each additional floor.
Home CO alarms are not the best devices to discover where CO is coming from
in a breathing space. After all, they are built as “alarms,” not as instantaneous sensing
devices. They have mathematical algorithms built in to simulate human CO uptake,
and thus to provide warning before we have taken very much of the CO up into our
blood and tissues. Also, their cost is minimal, making them affordable to the general
8417: “8417_c014” — 2007/9/11 — 12:15 — page 315 — #3

public, but by no means professional grade in accuracy. The best devices to ferret
out CO sources and monitor it accurately are what we call “monitors.” They measure
CO almost instantaneously and produce a reading in parts per million (ppm). They
are usually portable and the readings are highly reproducible. As one might imagine,
these kind of devices are more costly than home CO detectors.
14.2 PHYSIOLOGY OF CARBON MONOXIDE

CO binds to hemoglobin reversibly (Table 14.2). While this binding is avid, as we
know it is some 230 times that of oxygen, it is nonetheless a competitive process
with oxygen. As the oxygen partial pressure rises, the CO is forced off the site
on the hemoglobin molecule and oxygen takes its place. This mechanism is one
of the major reasons for using hyperbaric (high pressure) oxygen in treating CO
poisoning.
CO-induced hypoxia is far more serious than hypoxic hypoxia. The presence of
CO in the blood occupying the oxygen sites on the hemoglobin decreases the oxygen
transport capability of the cardiovascular system, not unlike increasing high altitude
does by desaturating arterial blood as lung partial pressure of oxygen falls. In addition
to this, however, CO causes the remaining oxyhemoglobin to hold its oxygen more
strongly, thus shifting the oxygen dissociation curve (ODC) to the left and making it
harder to unload what oxygen remains in the blood to the tissues. People tend to forget
this second action of CO on the ODC. Hypoxic hypoxia does just the opposite, shifting
the ODC to the right and enhancing the unloading of oxygen to the tissues. Thus, CO
inflicts a “double-whammy” on the body in terms of oxygen delivery capability that
is much more serious than is that of hypoxic hypoxia.
CO poisoning is also far more serious than a comparable simple anemia, where it
may appear that there is the same amount of total hemoglobin able to carry oxygen.
Like hypoxic hypoxia, anemia causes a right-shift in the ODC, which enhances oxygen
TABLE 14.2
Misconceptions: Physiology
• CO binding to hemoglobin is irreversible.
• CO (caused) hypoxia is no more serious than any other type of hypoxia.
• CO poisoning is no more serious than an anemia in which there is a comparable amount of hemoglobin
able to carry oxygen.
• Small animals (birds, mice, etc.) die more quickly because their hemoglobin binds CO more avidly
than that of humans, thus they were used as alarms for CO in mines.
• The fetus is protected from CO by the maternal body.
• Good COHb measurements can be obtained 1 day to a week after a person leaves the site of the CO
poisoning.
• Breathing “clean” air for 2–3 h will eliminate all CO from the body.
• Breathing 100% oxygen for 20–30 min will eliminate all CO from the body.
• Breathing (filter) masks protect the wearer from the inhalation of CO.
8417: “8417_c014” — 2007/9/11 — 12:15 — page 316 — #4

unloading to the tissues, whereas CO does the opposite. A condition where the body
has only one-half of its normal hemoglobin concentration (i.e., anemia) is well-
tolerated, whereas a condition where half of the hemoglobin has CO attached [i.e.,
50% carboxyhemoglobin (COHb) saturation] is usually quickly lethal.
Small birds and mammals were at one time used as CO alarms in mines and
other closed places where CO occurred. Canaries and other like small birds were
especially popular in this capacity because they would fall off their perches when
incapacitated, making it obvious to the observer there was a problem. The reason
for their usefulness was not that the small birds or mammals were necessarily more
sensitive to CO, but because their warm-bloodedness and high surface to volume
ratio caused their metabolic rate to be necessarily much higher than that of humans,
making their ventilation rate greater than that of humans. Hence they take up the CO
much faster. This provided a living “early warning” system for CO.
The fetus residing inside the maternal body is subject to the CO breathed by the
mother. Being a nonirritating gas, unlike chlorine or nitrogen dioxide, it is taken up
silently and without notice through the maternal lungs. The CO reaches the placenta
through the circulating maternal blood, and again, it passes silently and readily across
the membranes into the fetal circulation where it attaches to the fetus’ hemoglobin.
While CO compromises oxygen delivery to tissues in the maternal circulation by
the mechanisms described above, the situation is much more serious for the fetus,
which normally is operating in an oxygen-depleted environment. The presence of
CO makes the situation worse by further degrading the oxygen carrying capacity of
the fetus.
The right time to draw blood for accurate measurement of COHb is one of the
biggest misunderstandings by lay people, and even members of the medical com-
munity. The half-life of COHb in humans breathing sea level ambient air centers
around 4–5 h. That is, one-half of the COHb will be gone in 4–5 h, and half of what
remains will be gone in another 4–5 h (i.e., two half-lives). Within one day, COHb
level in the blood will be at or near background (0.4–1.4%) no matter how high it
was to start with. For accurate measurement, blood sampling should be done within
2–4 h of leaving the site of the CO poisoning. Physicians must not say when contacted
by a patient on Friday or on the weekend, “just come to the office on Monday and
we’ll do the COHb test.” It will be too late and the results will be useless. On the
basis of what we know of COHb half-life, it is clear that breathing “clean” air for
2–3 h. will not eliminate all the CO from the body—approximately 24 h is required.
In the same way, breathing 100% oxygen for 20–30 min at sea-level pressure, known
as normobaric oxygen therapy, will not eliminate all the CO from the body, since
the half-life of COHb when breathing 100% oxygen under normal conditions centers
around 50–70 min. A number of hours will be required to do this.
Another misconception involves the use of masks in polluted environments. Dust
masks or more complicated filtering respirators will remove particulates of various
sizes. Virtually none of these devices will remove CO. Donning a mask that does
not remove CO when the wearer believes it will, can be a very dangerous, even
deadly mistake. In order to safely enter an atmosphere containing CO, self-contained
breathing apparatus (SCBA) is required, such as that used by firefighters and hazmat
personnel.
8417: “8417_c014” — 2007/9/11 — 12:15 — page 317 — #5

14.3 SYMPTOMS
While CO poisoning is classically described as causing the skin, nail beds, and
exposed mucous membranes to turn pink or bright red, this is rarely seen in actual
fact. Often dead CO victims appear gray or yellow. The COHb saturation must be
high enough for the COHb to show; people dying in the water from acute CO pois-
oning usually reach only 40–45% COHb, the level at which incapacitation occurs.
Breathing then ceases below the water’s surface. The pink color is of course difficult
to see in darkly pigmented people (i.e., negroids), those with severe sunburn, and so
forth. (see Table 14.3)
Fever is rarely associated with CO poisoning, especially chronic CO poisoning.
Nonetheless, thermoregulatory dysfunction is a common outcome of CO poisoning,
and is thought to be due to brain damage caused by the poison. Victims with this
condition usually feel cold in thermally neutral surroundings, and occasionally report
being hot, however, in the latter instance actual body temperature is rarely above
normal. Fever is reported following severe acute CO poisoning, probably associated
with cerebral tissue damage and edema, gastrointestinal damage, and so forth.
The lungs along with the nasal passages, throat, and trachea are generally unaf-
fected by CO. Most of the CO uptake by the body occurs through the former route,
and the CO goes directly into the blood. While congestion, cough, and hoarseness are
not caused by CO inhalation, other substances that often accompany CO, particulates,
nitrogen oxides, sulfur oxides, aldehydes, and so forth will do this.
A recent book on CO (Dwyer et al., 2003, p. 5)1 lists “wheezing” or bronchial
constriction” and “persistent cough” as “signs and symptoms of CO poisoning.” In
my experience this is incorrect. Such symptoms may be due to other components of
TABLE 14.3
Misconceptions: Symptoms
• The skin, nail beds, and so forth of people with CO poisoning are invariably red or pink in color.
• Fever is a common symptom of CO poisoning.
• Nasal congestion, cough, and hoarseness are symptoms of CO poisoning.
• The lungs are inflammed by low to moderate levels of CO and appear abnormal by x-ray.
• Hyperventilation is a response to low and moderate CO poisoning
• Symptom clusters involving prolonged headache, dizziness, nausea, and fatigue of the whole family
should be blamed on viruses, bad food, or group craziness.
• Everyone responds to CO in the same way, that is, all people show the same symptoms.
• Depression is not a residual effect of CO poisoning.
• Loss of short-term memory capability is not a residual effect of CO poisoning.
• Muscle and/or joint pain is not a residual effect of CO poisoning.
• Difficulty with attention and concentration is not a residual effect of CO poisoning.
• Blurry vision is not a residual effect of CO poisoning.
• Personality change is not a residual effect of CO poisoning.
• More people experience acute CO poisoning than the chronic type
• There is a good dose–response relationship between CO in the air and COHb and immediate symptoms
or long-term health damage.
8417: “8417_c014” — 2007/9/11 — 12:15 — page 318 — #6

exhaust gases (e.g., Particulates, aldehydes, sulfur, and/or nitrogen oxides), but not
to the CO specifically.
As noted above, the lungs are virtually transparent to CO. In life-threatening CO
poisonings, the lungs become edematous and begin to fill with fluid. This congestion
is coughed up as a light to more darkly tan fluid, often mistaken for vomitus. It is
usually observed at scenes of lethal CO poisonings, combined or not combined with
vomitus.
Tachypnea is not a symptom of low to moderately severe CO poisoning, since
the carotid chemoreceptors are insensitive to percent saturation of the arterial blood,
responding only to changes in arterial PO2 and pH, which generally change little until
the CO poisoning becomes very severe. That is why hyperventilation with simple CO
poisoning is so rare, and its presence suggests a more complicated poisoning.
As discussed at length earlier, CO poisoning usually presents with a number of
nonspecific symptoms; few if any of the symptoms are specific (i.e., pathognomonic)
to CO poisoning. We often speak of symptom clusters, since CO usually induces
symptoms involving so many organ systems. “Flu-like” is the way the symptoms are
often described. Yet the “flu” may continue for weeks or months, affect everyone
in the same breathing space (i.e., house, apartment) at the same time, subside when
an individual leaves the space, etc. characteristics which are very “un-flu-like.” His-
torically, chief among misdiagnoses, has been viral or bacterial flu, food poisoning,
psychosomatic behavior, and so forth. Sometimes medical providers will continue to
insist on traditional causes for weeks and months, when they are clearly impossible,
and fail to recognize the ear-marks of a site-specific poisoning. See my chapter on
misdiagnosis of CO poisoning (Chapter 19).
Although there is a commonality of responses to CO, not all people respond exactly
the same, or have the same overall sensitivity to CO. This is the basis for the problems
with tables of symptoms seen in articles and books on CO, as they occur at increasing
concentrations of CO or COHb. I’ve often said that “people are not lab rats.” We vary
in age, gender, race, height, body weight, and so forth and in ways that are difficult
to describe or at present unknown, that influence our sensitivity (or tolerance) to CO.
In any group of people exposed to what appears the same CO concentration for the
same period of time, a diversity of responses develop. Some individuals suffer few
immediate or even long-term health effects, while others can be severely affected, or
even die during the initial CO exposure. Then there is a third group who respond in an
intermediate manner, with mild initial symptoms and mild long-term effects. Thus,
we usually see a “normal distribution” of responses, approximating a bell-shaped
curve. This pattern appears to occur whether the CO poisoning is extremely mild, or
extremely severe.
Clinical depression is seen with a high frequency in people who have suffered
long-term health damage from CO exposure. This is documented by the Utah group,
principally Dr Ramona Hopkins (see chapter 22 in this book). The depression appears
to have at least two components, that caused by the realization of the loss of func-
tionality by the CO-injured individual, and by direct CO-induced damage to the
brain. The latter pathway is often overlooked by those health professionals unfamiliar
with CO poisoning. The depression is usually treatable with standard antidepressant
medications.
8417: “8417_c014” — 2007/9/11 — 12:15 — page 319 — #7

Along with chronic fatigue, decrement in short-term memory capability is prob-

ably the most common long-term damage caused by CO poisoning. It is described by
victims as involving problems in remembering appointments and what was already
said in a conversation, losing keys, wallet/purse, finding the right word, and so forth.
It is embarrassing, affects a person’s confidence and self-esteem, and often forces
that person to become less social, and even a recluse. Damage to longterm or remote
memory, is far less common. In fact, it is rarely seen by this author, although it is
reported in the literature.
In the CO Support Study (see Hay et al., Carbon Monoxide Toxicity, 2000),2
muscle and joint pain were the most common long-term outcome of CO poisoning. My
studies of chronic CO poisoning in particular, verify the frequency of these symptoms.
Because clinical testing of the complaining muscles and joints rarely show evidence
of local damage, it is my working hypothesis that the damage is wholey in the brain,
and that the pain is referred, ie. sited in the brain. I don’t suggest that it is imagined
or fabricated, that it is in fact real, but is simply another product of the CO-induced
brain damage, not unlike that that causes the well-recognized cognitive and memory
deficits.
Like short-term memory damage, problems with attention–concentration are
extremely common after CO poisoning. People become more distractible, with even
minor auditory or visual distractions. Combined with this, and closely related to
attention–concentration, is usually a vastly reduced ability to do more than one task
at a time (i.e., multitasking). Having to have the TV off when talking on the phone, the
radio off when entering a freeway, and having to study in a totally quiet environment,
is often reported.
While blurry vision is a common complaint during CO poisoning, it can continue
for some time after poisoning has ended. It is one of several dozen visual com-
plaints that may be persistent and residual. See Dr Helffenstein’s chapter in this book
(Chapter 23).
One of the most common statements by family and friends of a CO victim is that
he/she is no longer the same—there has been “a personality change.” This usually
occurs in the direction of increased irritability, anxiety, depression, apathy, and so
forth. CO poisoning does not cause personalities to improve!
I have written elsewhere (Chapter 1) that the most frequent kind of CO poisoning
is the chronic type, that it is the least likely to be diagnosed, and probably results
in the largest amount of total injuries. New studies make it clear that brain damage
from CO poisoning is only very poorly correlated with severity of poisoning, whether
that be gauged by air CO concentration, COHb saturation, loss of consciousness, and
so forth.
14.4 TREATMENT AND OUTCOME

I often see cases where nasal prongs were used following CO poisoning. Just because
the oxygen in the tank is 100%, that doesn’t mean the oxygen coming through the
nose to the lungs will be 100%. It could be just a bit over 21%, that is, room air,
owing to dilution with copious amounts of fresh air (see Table 14.4).
8417: “8417_c014” — 2007/9/11 — 12:15 — page 320 — #8

TABLE 14.4
Misconceptions: Treatment and Outcome
• Inhalation of 100% oxygen from a rebreathing mask or from nasal prongs are the best immediate
means of removing CO from the body.
• Victims of CO poisoning should be released from medical care immediately following 1–2 h of oxygen
treatment, whether or not their symptoms have disappeared.
• There is no need for repeat COHb measurements, psychometric tests, or other clinical tests following
medical treatment for CO poisoning.
• People who recover from CO poisoning are always completely normal.
• Use of 100% oxygen therapy at normal atmospheric pressure (NBO) is a proven approach to eliminate
the residual effects of CO poisoning.
• Use of 100% oxygen at increased atmospheric pressure (HBO) is a proven approach to eliminate the
residual effects of CO poisoning.
• CO exposure never produces brain damage unless there is a period of unconsciousness.
• Low/moderate CO exposure cannot produce brain damage or significant changes in functional
performance.
• Venous blood measurement of COHb is not as accurate as arterial.
CO poisoning victims should be kept in-hospital for observation, usually for a

minimum of one night. All too often patients are seen for 2–3 h, given oxygen,
then released and continue to be symptomatic—headache, nausea, dizziness, and so
forth. Patients should never be discharged until they are completely symptom-free.
An extra night in the hospital gives staff a chance to observe them. Delayed sequelae
are such well-known complications of CO poisoning, even in patients who appear
to be doing very well patient. Discharge documents should clearly warn of possible
delayed effects, that is, sequela and how to recognize them.
CO washout rates vary tremendously among people, even when body mass,
gender, age, and so forth are taken into consideration. Second COHb measurements
should be done to confirm that normal or near-normal COHb levels exist after treat-
ment. This is an area that needs great improvement. COHb should be measured
increasingly by the newer methods (breathlyzer, multiwavelength pulse-oximeters).
Their advantage is that they can be done as often as necessary until COHb is suffi-
ciently reduced. Psychometric testing should also be done on a more frequent basis in
the emergency room (ER), especially in those patients who presented with symptoms
of altered consciousness or gait/balance problems.
It is recognized that a fraction of people who recover from CO poisoning are
left with decrements in function. Mainline medicine has been slow to recognize the
physical symptoms that often persist, and also the sensory, motor, cognitive, and
psychological deficits that many patients are left with, instead only chosing to focus
on the gross neurologic problems.
The use of 100% normobaric oxygen (NBO) has been used for many years to treat
CO poisoning. On a mechanistic basis, it should be efficacious simply because it more
quickly removes CO from the body than breathing room air. In recent studies, it was
the therapy to which hyperbaric oxygen (HBO) treatment was compared. The use of
8417: “8417_c014” — 2007/9/11 — 12:15 — page 321 — #9

NBO is proven only in the sense that aside from HBO, it appears to be the best approach
available. See the chapters by Drs Tomaszewski (Chapter 17) and Scheinkestel and
Millar (Chapter 18) on the use of NBO and HBO therapies.
Studies during the past 20 years make it abundantly clear that loss of consciousness
is not required for development of neurologic sequelae. Even more recent studies
demonstrate the lack of correlation between brain damge and air CO concentration,
COHb saturation, and many other traditional markers of poisoning severity.
For the measurement of blood CO alone, a venous sample is just as good as
an arterial sample. The use of noninvasive methods (breathylizer, new generation
pulse-oximeters) for this purpose is encouraged.
14.5 MISCELLANEOUS
Medical students and residents receive virtually no training in the diagnosis and
treatment of CO poisoning, and usually have little or no contact with CO-poisoned
patients in their practice (see Table 14.5). This is a large part of the reason for the high
misdiagnosis rate of CO poisoning by physicians. As a result few physicians have
an adequate index of suspicion for CO poisoning when presented with appropriate
symptoms and when they take a situational history.
The gold standard for the cognitive-memory problems that occur with such a
frequency following CO poisoning is neuropsychological testing. This should be
done by a neuropsychologist who understands the pathology of CO poisoning and
who has evaluated many patients with this condition. Psychiatry is well-suited to
assist medically in the treatment of emotional and personality problems that often
arise after CO poisoning. Neurology can be brought to bear for patients with gross
neurologic problems (i.e., aphasia, gait disturbance) caused by CO poisoning. It must
TABLE 14.5
Misconceptions: Miscellaneous
• Physicians receive adequate training in the diagnosis and treatment of CO poisoning in medical school
and residency/fellowship.
• Physicians get adequate experience with CO poisoning in treating their patients.
• Physicians generally have a high enough index of suspicion relative to CO poisoning in order to
diagnose it reliably.
• Psychiatrists and neurologists are the medical professionals of choice to determine the extent of central
nervous system (CNS) damage caused by CO.
• High-tech imaging devices (CT, MR, SPECT) always show areas of brain damage from CO poisoning,
if they exist.
• Two wavelength pulse oximeters are excellent devices for determining oxygen saturation and also
whether a person is suffering from CO poisoning.
• In environments containing CO, the levels of CO2 , oxygen and other gases are unimportant in the
degree of poisoning.
• The presence of green plants in a closed space will remove CO from the atmosphere.
• Toxicology is central to medicine.
8417: “8417_c014” — 2007/9/11 — 12:15 — page 322 — #10

be kept in mind however, that professionals in neither of these disciplines of medicine

have training in toxicology.
Computed tomography [CT, Computerized Axial Tomography (CAT)] is a form
of computer-enhanced x-ray scanning. Magnetic resonance imaging (MRI) shows
internal anatomy without the use of ionizing radiation. Both of these techniques
allow us to look at structure only, not function. Single photon emission computed
tomography (SPECT) permits the imaging of function in living tissue (see Chapters
by Drs. Heuser [20] and Hipskind [21]).
Pulse oximeters currently in use are “blind” to COHb, thus giving incorrect values
for oxygen saturation in the presence of COHb. These pulse oximeter devices should
never be used when there is suspicion of CO poisoning. In that case arterial blood
should immediately be drawn for measurement of total blood gases, which includes
COHb and oxygen saturation. See the chapter by Dr. Hampson [33] and mine [19],
regarding the new generation of pulse CO-oximeters marketed by Masimo.
Elevated carbon dioxide concentrations causes hyperventilation and an inc-
reased rate of CO uptake. Depressed oxygen (i.e., depletion) does the same, and
also increases the final COHb saturation by shifting the Haldane equation.
A recent article on residential CO poisoning published in the January, 2007 issue
of “Woman’s Day” magazine, states that the presence of green plants will remove CO
from the atmosphere. This was reportedly told to a patient by a treating physician.
There is in fact no basis for this statement. Maybe he was confused by the fact that
plants take up carbon dioxide (not carbon monoxide), converting it to oxygen through
the process of photosynthesis.
Toxicology might be likened to a poor relation to medicine. For most physicians
the normal circles of training and practice experience in internal medicine rarely
intersect with the toxicology circle. Thus, the constant complaints I receive from
CO-victims, saying that few if any physicians they see are helpful in the diagnosis,
evaluation, or treatment of their conditions.
References
1. Dwyer, Leatherman, Manclark, Kimball, Rasmussen. Carbon Monoxide: A Clear
and Present Danger, ESCO Press, USA, 3rd ed., 2003 (www.escoinst.com).
2. Hay, A.W.H., Jaffer, S., Davis, D. Chronic carbon monoxide exposure: The CO
Support study. In: Carbon Monoxide Toxicity, D.G. Penney, ed., CRC Press, NY,
2000, pp. 419–438.
8417: “8417_c014” — 2007/9/11 — 12:15 — page 323 — #11

8417: “8417_c014” — 2007/9/11 — 12:15 — page 324 — #12
15 A Survey Study of Public
Perceptions About
Carbon Monoxide
David G. Penney and Linda M. Penney
CONTENTS
15.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325

15.2 Electric Generator Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
15.3 Propane Radiant Heater Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
15.4 Recreational Powerboat Safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
15.5 Conditions Affecting Exhaust Gas Danger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332
15.6 Safe Use Indoors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333
15.7 The Greatest Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
15.8 The Worst Poison . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
15.9 The Best Advice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336
15.10 Duration of Time in the Body . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
15.11 Properties of CO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
15.12 Experience with CO Poisoning? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
15.13 Use of CO Emitting Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
15.1 BACKGROUND
On the basis of the large numbers of accidents involving carbon monoxide (CO),
recent studies have asked questions about the public’s knowledge of the dangers of
this gas.1 Our survey was conducted in the latter half of 2006 by sampling the opinions
of people in Michigan (MI) and Florida (FL). The data derived from this survey are
shown in Tables 15.1–15.13. Adult respondents belonged to civic and athletic social
clubs. These data were analyzed by age groupings and by gender.
Youth data were obtained by surveying students in both junior high and senior high
school classes. The Michigan school was a large public school in Benzie County (first
set of four rows). In Florida, one school was a large public high school in St. John’s
County (first set of four rows), while data from two smaller parochial schools in the
county were combined (second set of four rows). The youth data were also analyzed
325
8417: “8417_c015” — 2007/9/11 — 12:15 — page 325 — #1

TABLE 15.1
Assume the Electric Power is Off, but Can be Restored by the Use of a
Gasoline-fueled, Portable Generator. If Started in a Room, How Much
Ventilation is Adequate for the Generator to be Run Safely?
Percent of Total
Male Female Range Mean Age A B C D E

Michigan
38 — 14–15 14.4 13.2 10.5 26.3 26.3 23.7
— 49 14–15 14.4 28.6 18.4 24.5 18.4 10.2
29 — 16–18 16.7 0 37.9 17.2 24.1 20.7
— 48 16–19 16.8 12.5 22.9 27.1 31.3 6.3
20 8 30–59 48.5 7.1 14.3 3.6 21.4 53.6
31 14 60–69 64.9 11.6 14.0 7.0 18.6 48.8
34 7 70–81 74.2 13.2 10.5 5.3 36.8 34.2
85 — 31–81 64.9 12.1 10.8 6.0 26.5 44.6
— 29 30–79 62.3 7.7 19.2 3.9 23.1 46.2
Florida (rows 1–4 public; 5–8 private)
38 — 14–16 15.6 15.8 13.2 23.7 23.7 23.7

— 43 15–16 15.5 21.4 19.1 14.3 28.6 16.7
15 — 17–18 17.1 8.3 0 0 41.7 50.0
— 14 17–18 17.2 7.1 28.6 21.4 35.8 7.1
9 — 13–15 13.9 22.2 22.2 0 33.3 22.2
— 14 11–15 13.5 7.7 15.4 30.8 15.4 30.8
44 — 16–18 16.5 4.7 4.7 9.3 46.5 34.9
— 40 16–18 16.6 2.6 15.4 20.5 33.3 28.2
56 18 18–59 44.8 1.4 2.8 4.2 15.7 75.0
16 2 60–83 71.1 5.6 11.1 16.7 22.2 44.4
72 — 19–83 50.8 2.9 5.7 7.1 14.3 70.0
— 20 26–69 46.9 0 0 5.3 31.6 63.2
*20 *5 26–80 57.5 4.4 8.7 8.7 21.7 56.5
A = One window open; B = Two windows open; C = Three or more windows open; D = All windows
and the door open; E = None of the above.
∗ “medical”—all adults, MI and FL, with self-reported medical, biological occupations.
by age group and gender. The questions put to respondents in the survey are printed
above each table, and the “correct” answer is indicated in bold. The survey asked only
for age, gender, and occupation to be written in. Before the survey was administered,
respondents were instructed to answer quickly, not to agonize over their responses,
and told the survey was not an intelligence test. When we could not be present for
administration of the survey, the survey takers (teachers) were provided instructions to
be read to the class prior to starting. There was an initial un-numbered fill-in question
that asked yes or no, whether respondents (or if students, their parents) had rented
or owned various small combustion devices (Table 15.12). A final question on the
survey asked respondents about their experience with CO poisoning incidents. This
8417: “8417_c015” — 2007/9/11 — 12:15 — page 326 — #2

A Survey Study of Public Perceptions About Carbon Monoxide 327
TABLE 15.2
Portable Propane Radiant Heaters are Available to Provide Heat, for
example, when Camping. Which One of the Situations Below is Safe?
Percent of Total

Michigan
38 — 14–15 14.4 0 7.9 34.2 18.4 39.5
— 49 14–15 14.4 2.0 14.3 49.0 4.1 30.6
29 — 16–18 16.7 0 3.5 41.4 13.8 41.4
— 48 16–19 16.8 2.1 4.2 66.7 8.3 18.8
20 8 30–59 48.5 0 3.6 10.7 17.9 67.9
31 14 60–69 64.9 2.3 2.3 11.6 11.6 72.1
34 7 70–81 74.2 2.7 10.8 13.5 24.3 48.7
85 — 31–81 64.9 2.5 6.2 12.4 21.0 58.0
— 29 30–79 62.3 0 3.7 11.1 7.4 77.8
38 — 14–16 15.6 5.3 21.1 44.7 5.3 21.7

— 43 15–16 15.5 0 18.6 44.2 28.6 16.7
15 — 17–18 17.1 7.7 0 30.8 15.4 46.2
— 14 17–18 17.2 0 15.4 53.9 7.7 23.1
9 — 13–15 13.9 0 0 55.6 11.1 33.3
— 14 11–15 13.5 0 7.1 64.3 0 28.6
44 — 16–18 16.5 2.3 9.1 36.4 4.6 47.7
— 40 16–18 16.6 0 15.0 37.5 5.0 42.5
56 18 18–59 44.8 1.3 0 22.7 5.3 70.7
16 2 60–83 71.1 0 0 55.6 0 44.4
72 — 19–83 50.8 1.4 0 25.0 4.2 69.4
— 20 26–69 46.9 0 0 45.0 5.0 50.0
*20 *5 26–80 57.5 0 4.2 16.7 8.3 70.8
A = Use in a closed tent while sleeping;

B = Use in a tent with the door flap loose/open; C = Use for a short time before going to sleep in order
to warm up the closed tent; D = Use in a tent with two windows open; E = None of the above.
question had no “correct” answer. Sources of CO involved in CO poisonings were

written in on the survey by respondents, and are compiled in Table 15.13.
15.2 ELECTRIC GENERATOR USE (QUEST. 1)

In most cases a majority of respondents believed that a gasoline-fueled electric gen-
erator could be operated inside a closed space (i.e., room) if some ventilation were
provided. This is of course false—no amount of ventilation will make this scenario
safe. The generator MUST be operated outside, preferably 20–25 ft. from the nearest
structure. In general, adults responded with the highest percent of correct answers,
8417: “8417_c015” — 2007/9/11 — 12:15 — page 327 — #3

TABLE 15.3
When Using a Recreational Powerboat, What is Safe (one only)?
Percent of Total

Michigan
38 — 14–15 14.4 7.9 5.3 5.3 79.0 2.6
— 49 14–15 14.4 6.3 2.1 8.3 66.7 16.7
29 — 16–18 16.7 6.9 3.5 0 89.7 0
— 48 16–19 16.8 4.4 4.4 4.4 82.6 4.4
20 8 30–59 48.5 3.7 0 0 81.5 14.8
31 14 60–69 64.9 0 0 0 97.6 2.4
34 7 70–81 74.2 2.8 0 0 91.7 5.6
85 — 31–81 64.9 1.3 0 0 91.1 7.6
— 29 30–79 62.3 3.9 0 0 92.3 3.9
38 — 14–16 15.6 13.2 2.6 7.9 68.4 7.9

— 43 15–16 15.5 7.1 0 9.5 73.8 9.5
15 — 17–18 17.1 6.7 6.7 0 73.3 13.3
— 14 17–18 17.2 14.3 0 0 71.4 14.3
9 — 13–15 13.9 33.3 0 0 55.6 11.1
— 14 11–15 13.5 28.6 0 0 71.4 0
44 — 16–18 16.5 13.6 0 2.3 68.2 15.9
— 40 16–18 16.6 7.7 7.7 0 82.1 2.6
56 18 18–59 44.8 2.8 0 0 93.1 4.2
16 2 60–83 71.1 0 0 0 93.8 6.3
72 — 19–83 50.8 1.5 0 0 94.1 4.4
— 20 26–69 46.9 5.3 0 0 89.5 5.3
*20 *5 26–80 57.5 0 0 0 96.2 3.9
A = Sitting on the transom with the boat engine idling; B = Teak surfing (hanging on to the swim
platform behind the boat); C = Wake surfing (body surfing in the wake immediately behind the boat);
D = Water skiing 100 ft. behind the boat; E = Sitting on the boat moored near other boats whose
engines are idling.
youths the least. People in healthcare-science fields answered correctly no more fre-
quently that other adults generally. On the basis of the responses of many people, and
especially so youths, who thought that opening one window will be enough, many of
them would not have survived that scenario if it had really happened. In fact, for many
of the cohorts of teenagers, the vast majority of the individuals would be at risk for
death or injury from CO poisoning from a generator based on their faulty knowledge
and perceptions of this gas. People die on a regular basis from CO poisoning because
of improper use of generators, and especially after severe storms and hurricanes when
main electrical power is not available.
The study by Hampson and Zmaeff1 cited a number of incidents of injury and
death from improper use of generators following storms and hurricanes. This included
8417: “8417_c015” — 2007/9/11 — 12:15 — page 328 — #4

TABLE 15.4
What Weather Condition Might Greatly Affect the Levels of Exhaust Fumes
on Board a Boat?
Percent of Total

Michigan
38 — 14–15 14.4 13.9 16.7 8.3 22.2 38.9
— 49 14–15 14.4 4.2 10.4 10.4 31.3 43.8
29 — 16–18 16.7 0 17.2 3.5 20.7 58.6
— 48 16–19 16.8 2.0 10.2 6.1 14.3 67.4
20 8 30–59 48.5 8.0 8.0 0 28.0 56.0
31 14 60–69 64.9 2.6 0 7.7 33.3 56.4
34 7 70–81 74.2 0 2.9 0 25.7 71.4
85 — 31–81 64.9 2.8 2.8 2.8 27.8 63.9
— 29 30–79 62.3 3.7 3.7 3.7 33.3 55.6
38 — 14–16 15.6 16.2 8.1 5.4 37.8 32.4

— 43 15–16 15.5 7.3 9.8 9.8 26.8 46.3
15 — 17–18 17.1 7.1 14.3 7.1 35.7 35.7
— 14 17–18 17.2 0 21.4 0 50.0 28.6
9 — 13–15 13.9 22.2 0 0 33.3 44.4
— 14 11–15 13.5 14.3 14.3 21.4 28.6 21.4
44 — 16–18 16.5 6.8 18.2 2.3 29.6 43.2
— 40 16–18 16.6 7.5 12.5 5.0 25.0 50.0
56 18 18–59 44.8 2.9 7.1 4.3 34.3 51.4
16 2 60–83 71.1 13.3 6.7 0 26.7 53.3
72 — 19–83 50.8 4.6 9.1 3.0 31.8 51.5
— 20 26–69 46.9 5.6 0 5.6 33.3 55.6
*20 *5 26–80 57.5 0 0 4.2 25.0 70.8
A = Sun out/overcast; B = Air temperature; C = Rain; D = Humidity; E = Wind direction.

26 people in January, 1993 (Washington), 74 in January, 1998 (Maine), 71 in Decem-

ber, 2002 (North Carolina), and 2 in September, 2003 (Hurricane Isabel). Lovanas,
at the Centers for Disease Control and Prevention (CDC), found that about one-third
of generator incidents involved non-English speaking patients.2
15.3 PROPANE RADIANT HEATER USE (QUEST. 2)

A small percentage of respondents believed that it is safe to sleep in a tent with an
operating propane radiant heater. While these heaters produce little CO when burning
in the open air, they can produce prodigious amounts of CO if oxygen in the ambient
air becomes depleted and limiting to complete combustion. A somewhat larger per-
centage, but usually still far less than a majority, realized that when using these much
smaller (i.e., than generators) sources of CO, the secure opening of two windows
8417: “8417_c015” — 2007/9/11 — 12:15 — page 329 — #5

TABLE 15.5
Which One of the Following May Safely be Used Indoors?
Percent of Total

Michigan
38 — 14–15 14.4 5.3 76.3 2.6 2.6 13.2
— 49 14–15 14.4 2.1 60.4 4.2 0 33.3
29 — 16–18 16.7 17.2 72.4 0 3.5 6.9
— 48 16–19 16.8 6.3 68.8 0 2.1 22.9
20 8 30–59 48.5 0 76.9 0 0 23.1
31 14 60–69 64.9 2.3 84.1 0 4.6 9.1
34 7 70–81 74.2 0 84.6 2.6 0 12.8
85 — 31–81 64.9 0 82.7 0 2.5 14.8
— 29 30–79 62.3 3.6 82.1 3.6 0 10.7
38 — 14–16 15.6 5.3 73.7 2.6 0 18.4

— 43 15–16 15.5 9.3 58.1 2.3 2.3 27.9
15 — 17–18 17.1 0 73.3 6.7 6.7 13.3
— 14 17–18 17.2 0 78.6 0 0 21.4
9 — 13–15 13.9 11.1 55.6 0 0 33.3
— 14 11–15 13.5 0 50.0 7.1 0 42.9
44 — 16–18 16.5 9.1 79.6 4.6 0 6.8
— 40 16–18 16.6 10.0 60.0 2.5 0 27.5
56 18 18–59 44.8 4.1 82.4 0 0 13.5
16 2 60–83 71.1 6.7 66.7 0 0 26.7
72 — 19–83 50.8 2.9 79.4 0 0 17.7
— 20 26–69 46.9 10.0 80.0 0 0 10.0
*20 *5 26–80 57.5 8.0 72.0 4.0 0 16.0
A = Hibachi grill; B = A dozen candles; C = Gasoline-powered generator; D = Gasoline-powered

pressure washer; E = I don’t know.
*“medical”—all adults, MI and FL, with self-reported medical, biological occupations.
of a tent, guarantees safe operation. This requirement is specified by manufacturers.

Many people believed that warming up the tent prior to going to bed was safe. This,
in fact, may be one of the most unsafe practices that has resulted in may deaths:
(1) CO can accumulate in the closed tent during warm-up, which then poisons the
person after he/she comes inside to sleep, or (2) the heater is operated while people
are inside the tent in order to heat up, the people then fall asleep, CO accumulates,
and death occurs silently. The warmed inside environment, combined with fatigue,
CO uptake, and possible alcohol use readily induces sleep. An opened door flap is not
safe, because unless it is secured it could inadvertently be brushed closed or be closed
by wind. In either case, the end result is predictable. One manufacturer’s propane
radiant heaters are known to have killed over 75 people during the past 16 years.
8417: “8417_c015” — 2007/9/11 — 12:15 — page 330 — #6

TABLE 15.6
Which One of Those Below is Known on Average to Represent the Greater
Risk of Immediate Injury or Death from Carbon Monoxide Poisoning?
Percent of Total

Michigan
38 — 14–15 14.4 50.0 15.8 7.9 2.6 23.7
— 49 14–15 14.4 38.6 6.8 20.5 6.8 27.3
29 — 16–18 16.7 79.3 0 3.5 0 17.2
— 48 16–19 16.8 76.6 8.5 4.3 2.1 8.5
20 8 30–59 48.5 76.0 8.0 16.0 0 0
31 14 60–69 64.9 88.6 0 9.1 0 2.3
34 7 70–81 74.2 91.9 0 8.1 0 0
85 — 31–81 64.9 88.8 2.5 8.8 0 0
— 29 30–79 62.3 80.8 0 15.4 0 3.9
38 — 14–16 15.6 68.4 5.3 15.8 2.6 7.9

— 43 15–16 15.5 73.2 4.9 9.8 0 12.2
15 — 17–18 17.1 53.9 15.4 15.4 7.7 7.7
— 14 17–18 17.2 64.3 14.3 7.1 0 14.3
9 — 13–15 13.9 55.6 0 22.2 11.1 11.1
— 14 11–15 13.5 64.3 0 28.6 0 7.1
44 — 16–18 16.5 65.1 7.0 11.6 9.3 7.0
— 40 16–18 16.6 59.0 2.6 15.4 10.3 12.8
56 18 18–59 44.8 77.5 2.8 14.1 0 5.6
16 2 60–83 71.1 62.5 12.5 18.8 6.3 0
72 — 19–83 50.8 75.0 2.9 14.7 1.5 5.9
— 20 26–69 46.9 72.2 11.1 16.7 0 0
*20 *5 26–80 57.5 96.0 0 4.0 0 0
A = An automobile; B = A gasoline-powered lawn mower; C = A ski-boat with inboard gasoline engines;

D = A snow thrower; E = An electric water heater
15.4 RECREATIONAL POWERBOAT SAFETY (QUEST. 3)

While the vast majority of respondents believed that the safe activity was water
skiing 100 ft. behind the powerboat, some respondents believed that sitting on the
boat transom (the side to side structure at the back of a ship or boat) with the engine
idling, teak surfing (holding onto the swim platform and being dragged through the
water), wake surfing (being pulled behind the boat on a board via a very short rope),
and sitting on a boat near other boats whose engines are idling was also safe. All four
of the latter behaviors of course expose individuals to the potential for injurious or
lethal CO poisoning. In general, younger people were more likely to endorse these
8417: “8417_c015” — 2007/9/11 — 12:15 — page 331 — #7

TABLE 15.7
Which Poison Debilitates or Accidentally Kills the Most People Each Year in
the USA?
Percent of Total

Michigan
38 — 14–15 14.4 0 0 97.4 2.6 0
— 49 14–15 14.4 6.3 2.1 83.3 2.1 6.3
29 — 16–18 16.7 0 0 92.9 3.6 3.6
— 48 16–19 16.8 2.1 0 79.2 8.3 10.4
20 8 30–59 48.5 0 0 100 0 0
31 14 60–69 64.9 2.3 0 97.7 0 0
34 7 70–81 74.2 0 0 97.4 0 2.6
85 — 31–81 64.9 1.2 0 97.5 0 1.2
— 29 30–79 62.3 0 0 100 0 0
38 — 14–16 15.6 2.6 5.3 68.4 5.3 18.4

— 43 15–16 15.5 4.9 4.9 78.1 7.3 4.9
15 — 17–18 17.1 7.1 0 85.7 0 7.1
— 14 17–18 17.2 0 7.1 85.7 0 7.1
9 — 13–15 13.9 0 0 100 0 0
— 14 11–15 13.5 0 0 78.6 0 21.4
44 — 16–18 16.5 4.6 4.6 81.8 4.6 4.6
— 40 16–18 16.6 2.5 5.0 87.5 5.0 0
56 18 18–59 44.8 1.4 0 90.0 0 8.6
16 2 60–83 71.1 6.7 0 93.3 0 0
72 — 19–83 50.8 3.0 0 91.0 0 6.0
— 20 26–69 46.9 0 0 88.2 0 11.8
*20 *5 26–80 57.5 0 0 96.0 0 4.0
A = Chlorine; B = Hydrogen sulfide; C = Carbon monoxide; D = Cyanide; E = Carbon dioxide

risky behaviors, especially, sitting on the transom with the engine running by the
Florida youth.
15.5 CONDITIONS AFFECTING EXHAUST GAS

DANGER (QUEST. 4)
There was considerable confusion among respondents about what conditions will
affect exhaust fume distribution on a boat. Greater fractions of adult respondents
thought wind direction would be the major factor, while fewer young people made
this choice. Approximately one-third of respondents thought humidity would be most
important. Some even thought sunshine versus overcast, air temperature, or rain might
be major influences. Studies show that wind velocity and wind direction, combined
8417: “8417_c015” — 2007/9/11 — 12:15 — page 332 — #8

TABLE 15.8
If Carbon Monoxide Exposure is Suspected, What is the Best Advice?
Percent of Total
Michigan
38 — 14–15 14.4 36.8 7.9 52.6 2.6 0
— 49 14–15 14.4 30.4 4.4 63.0 2.2 0
29 — 16–18 16.7 37.9 3.5 51.7 6.9 0
— 48 16–19 16.8 16.7 0 75.0 6.3 2.1
20 8 30–59 48.5 44.4 0 55.6 0 0
31 14 60–69 64.9 34.1 0 65.9 0 0
34 7 70–81 74.2 51.3 0 48.7 0 0
85 — 31–81 64.9 48.8 0 51.2 0 0
— 29 30–79 62.3 25.0 0 75.0 0 0
38 — 14–16 15.6 31.6 7.9 52.6 5.3 2.6

— 43 15–16 15.5 23.8 2.4 69.1 2.4 2.4
15 — 17–18 17.1 21.4 0 78.6 0 0
— 14 17–18 17.2 38.5 7.7 46.2 7.7 0
9 — 13–15 13.9 33.3 11.1 55.6 0 0
— 14 11–15 13.5 21.4 0 78.6 0 0
44 — 16–18 16.5 25.0 9.1 59.1 2.3 4.6
— 40 16–18 16.6 32.5 0 65.0 2.5 0
56 18 18–59 44.8 35.7 1.4 62.9 0 0
16 2 60–83 71.1 31.3 0 68.8 0 0
72 — 19–83 50.8 34.3 1.5 64.2 0 0
— 20 26–69 46.9 38.9 0 61.1 0 0
*20 *5 26–80 57.5 44.0 0 56.0 0 0
A = Go outside and breathe some fresh air; B = Ignore it and it will go away; C = Go immediately to
an emergency center and be evaluated by a physician; D = Call your doctor for an appointment in a
few days; E = Take an aspirin or Motrin for headache.
with boat speed and direction, greatly influence the concentrations of CO found on
boat decks, in cabins and immediately behind the boat.
15.6 SAFE USE INDOORS (QUEST. 5)

A high percentage of respondents recognized the burning of a dozen candles as being
the safe choice indoors. Candles are small combustion sources and produce very
little CO. Nevertheless, a significant number of people didn’t know the answer (in one
instance >40%), and some even endorsed the use of a hibachi or an internal combustion
engine indoors. Younger respondents appeared to be the less knowledgeable, and in
some cases significant percentages of the youth (>10%) thought very risky, even
lethal behaviors were okay (e.g., use of a hibachi indoors).
8417: “8417_c015” — 2007/9/11 — 12:15 — page 333 — #9

TABLE 15.9
How Long Does Carbon Monoxide Stay in the Human Body?
Percent of Total
Michigan
38 — 14–15 14.4 10.5 5.3 10.5 15.8 57.9
— 49 14–15 14.4 6.4 2.1 8.5 2.1 80.9
29 — 16–18 16.7 10.3 17.2 10.3 17.2 44.8
— 48 16–19 16.8 4.2 18.8 4.2 6.3 66.7
20 8 30–59 48.5 0 3.7 11.1 7.4 77.8
31 14 60–69 64.9 4.6 11.4 9.1 6.8 68.2
34 7 70–81 74.2 5.1 0 5.1 10.3 79.5
85 — 31–81 64.9 3.6 4.8 8.4 7.2 75.9
— 29 30–79 62.3 3.7 7.4 7.4 11.1 70.4
38 — 14–16 15.6 7.9 15.8 13.2 10.5 52.6

— 43 15–16 15.5 11.9 4.8 9.5 2.4 71.4
15 — 17–18 17.1 20.0 13.3 6.7 6.7 53.3
— 14 17–18 17.2 0 7.1 7.1 7.1 78.6
9 — 13–15 13.9 0 0 0 0 100
— 14 11–15 13.5 7.1 0 7.1 0 85.7
44 — 16–18 16.5 11.4 11.4 13.6 11.4 52.3
— 40 16–18 16.6 10.0 10.0 5.0 7.5 67.5
56 18 18–59 44.8 5.6 8.5 12.7 15.5 57.8
16 2 60–83 71.1 6.7 6.7 6.7 6.7 73.3
72 — 19–83 50.8 6.1 10.6 12.1 16.7 54.6
— 20 26–69 46.9 5.3 0 10.5 5.3 79.0
*20 *5 26–80 57.5 0 12.5 16.7 20.8 50.0
A = Once breathed, it stays always; B = For 65 days; C = For 1 week; D = For 1 day; E = I don’t
know.
* “medical”—all adults, MI and FL, with self-reported medical, biological occupations.
15.7 THE GREATEST RISK (QUEST. 6)

The vast majority of respondents viewed the automobile as the greater risk from
CO poisoning. This of course is out of line with the fact that automobiles produce
only extremely low levels of CO today when properly used and maintained, thanks
to the US Environmental Protection Agency (US EPA) and the catalytic converters
required at manufacture. Internal combustion engine-driven devices without catalytic
converters represent far greater hazards today for the individual user. The greatest
of these are those with the largest engines, therefore the greatest CO generating
capacity. That would be gasoline-powered powerboats—ski-boats, cabin cruisers,
fishing boats, and so forth, especially those with inboard engines. Generally, the
bigger the boat and engines, the more dangerous. Only 10–20% of respondents were
aware of this fact. A few thought that lawn mowers or snow throwers were most
8417: “8417_c015” — 2007/9/11 — 12:15 — page 334 — #10

TABLE 15.10
Which One of the Following is a Property of Carbon Monoxide?
Percent of Total
Michigan
38 — 14–15 14.4 23.7 5.3 34.2 21.1 15.8
— 49 14–15 14.4 6.4 21.3 4.3 21.3 46.8
29 — 16–18 16.7 13.8 27.6 6.9 27.6 24.1
— 48 16–19 16.8 18.8 25.0 2.1 20.8 33.3
20 8 30–59 48.5 7.1 42.9 0 28.6 21.4
31 14 60–69 64.9 20.5 18.2 2.3 29.6 29.6
34 7 70–81 74.2 7.7 38.5 0 25.6 28.2
85 — 31–81 64.9 12.1 38.6 1.2 31.3 16.1
— 29 30–79 62.3 14.3 10.7 0 17.9 57.1
38 — 14–16 15.6 7.9 23.7 21.1 15.8 31.6

— 43 15–16 15.5 9.8 17.1 19.5 12.2 41.5
15 — 17–18 17.1 20.0 33.3 6.7 20.0 20.0
— 14 17–18 17.2 21.4 14.3 7.1 21.4 37.7
9 — 13–15 13.9 22.2 11.1 0 11.1 55.6
— 14 11–15 13.5 21.4 0 7.1 7.1 64.3
44 — 16–18 16.5 13.6 20.5 9.1 34.1 22.7
— 40 16–18 16.6 15.0 15.0 10.0 22.5 37.5
56 18 18–59 44.8 22.2 25.0 1.4 27.8 23.6
16 2 60–83 71.1 12.5 37.5 0 6.3 43.8
72 — 19–83 50.8 18.8 30.4 1.4 2.7 27.5
— 20 26–69 46.9 27.8 16.7 0 33.3 22.2
*20 *5 26–80 57.5 32.0 12.0 0 28.0 28.0
A = It is much lighter than air; B = It is much heavier than air; C = It has a strong odor of burning
materials; D = It is about the same density as air; E = I don’t know.
dangerous in terms of CO. Certainly they too can be dangerous because even though
they have much smaller engines, they also lack catalytic converters. The basis for
CO risk posed by electric water heaters, which as many as 27% of one student group
chose, is unclear. It may represent a total misunderstanding of how CO is generated.
School teaching programs should take note of this abberation.
15.8 THE WORST POISON (QUEST. 7)

This question makes it clear that while people of all ages may have confusion on
other issues relating to CO, high percentages know that CO debilitates and kills
people. Instances of this are frequently reported in the newspapers and on TV. Some
groups, especially young people chose carbon dioxide. Confusion between these two
carbon oxides could be serious. The one we breathe out is not toxic until very high
8417: “8417_c015” — 2007/9/11 — 12:15 — page 335 — #11

TABLE 15.11
Have YOU Experienced the Toxic Effects of Carbon Monoxide Poisoning?
Percent of Total

Michigan
38 — 14–15 14.4 5.3 5.3 60.5 5.3 23.7
— 49 14–15 14.4 4.3 4.3 59.6 14.9 17.0
29 — 16–18 16.7 13.8 0 55.2 69.9 24.1
— 48 16–19 16.8 4.2 2.1 75.0 8.3 10.4
20 8 30–59 48.5 10.7 3.6 64.3 7.1 14.3
31 14 60–69 64.9 11.1 4.4 55.6 11.1 17.8
34 7 70–81 74.2 0 5.1 79.5 5.1 10.3
85 — 31–81 64.9 6.0 4.8 63.9 8.4 16.9
— 29 30–79 62.3 10.3 3.5 72.4 6.9 6.9
38 — 14–16 15.6 7.9 2.6 76.3 0 13.2

— 43 15–16 15.5 0 0 71.4 4.8 23.8
15 — 17–18 17.1 0 0 66.7 0 33.3
— 14 17–18 17.2 0 0 100 0 0
9 — 13–15 13.9 11.1 0 55.6 11.1 22.2
— 14 11–15 13.5 7.1 0 64.3 0 28.6
44 — 16–18 16.5 4.6 0 77.3 9.1 9.1
— 40 16–18 16.6 7.5 0 52.5 7.5 32.5
56 18 18–59 44.8 4.1 6.9 67.1 5.5 16.4
16 2 60–83 71.1 6.3 0 75.0 6.3 12.5
72 — 19–83 50.8 5.8 5.8 65.2 7.3 15.9
— 20 26–69 46.9 0 5.3 79.0 0 15.8
*20 *5 26–80 57.5 8.0 4.0 64.0 16.0 8.0
A = Yes, once; B = Yes, more than once; C = No; D = No, but I know of someone who has; E = I don’t
know.
levels are reached, while its sibling with one less oxygen atom is deadly at very low
concentrations. Science classes in school could do better in teaching these important
distinctions.
15.9 THE BEST ADVICE (QUEST. 8)

The majority of respondents recognized that the best action to take when CO expos-
ure is suspected, is prompt presentation at an emergency center for evaluation and
treatment by a health professional. A small but substantial percentage of respondents
believed that simply going outside and breathing fresh air was adequate to conter-
act the effects of CO. This is the general wisdom of many older individuals in the
medical community, even though it is wrong. CO poisoning is not to be trifled with.
8417: “8417_c015” — 2007/9/11 — 12:15 — page 336 — #12

TABLE 15.12
Sources of Carbon Monoxide for Survey Respondents
Michigan Florida
Automobile—7 Automobile—3
In trunk—1 Repair facility—1
In garage with door shut—1 In garage—1
Exhaust—2 Exhaust—1
Suicide attempt—2 Tractor—1
In house—1 Gas oven—1
Furnace—4 Gas heater—1
Heater—2 Coal heater—1
Pilot light failure—1 Heating system—1
Kerosene heater—2 Camp fires—1
Water heater (electric)—1 Fire—1
Fire (electrical)—1 Fireplace leakage—2
Plugged chimney—1 Lawn mower—1
Lawn mower—1 In garage—1
Pool heater/AC—1 Boat (anchored)—2
Gas refrigerator—1 Inside cabin—1
Golf cart—1 Generator—1
Boat—1 Generator during
Exhaust gases—1 hurricane—1
Race car with bad ventilation system—1 Commercial diving—1
Drag race—1
U.S. Army tank + 2-1/2 tn truck—1 In chemistry class—1
TABLE 15.13
Rental and Ownership of Selected Combustion Devices by Survey Respondents
in Michigan and Florida, Adults and Juveniles∗
Rental Ownership
Chain Press, Kerosene Chain Press, Kerosene

Saw Washer Heater Generator Saw Washer Heater Generator
Michigan
Men (85) 4.7 16.5 1.2 5.9 65.9 27.1 17.7 21.2
Women (29) 0 6.9 0 0 48.3 20.7 20.7 20.7
Boys (67) 16.4 11.9 7.5 16.4 68.7 35.8 28.4 37.3
Girls (97) 9.3 11.3 8.3 8.3 54.6 21.7 18.6 22.7
Florida
Men (72) 25.0 34.7 6.9 13.9 44.4 31.9 6.9 27.8
Women (20) 0 25.0 0 5.0 25.0 20.0 0 25.0
Boys (106) 22.6 34.9 7.6 21.7 50.0 45.3 15.1 40.6
Girls (111) 12.6 36.0 1.8 17.1 45.1 33.3 8.1 36.0
∗ It is assumed that this reflects rental and ownership by the juvenile’s families.
8417: “8417_c015” — 2007/9/11 — 12:15 — page 337 — #13

A substantial fraction of those poisoned develop long-term health damage. It must

be treated immediately and correctly! Likewise, ignoring it as a few respondents
thought, will not guarantee that it will go away. Finally, action must be prompt—
calling a physician for an appointment days to weeks hence is of no value relative to
maintenance of health after CO poisoning, and is equivalent to doing nothing.
15.10 DURATION OF TIME IN THE BODY (QUEST. 9)

Probably the greatest confusion related to the length of time CO remains in the body.
Being a very smart poison, as it has been characterized, it “washes out” quickly when
one breathes clean air. The half-life of carboxyhemoglobin (COHb) is approximately
4–5 h, so within 24 h (5–6 half-lives) body CO concentration (i.e., COHb) is back at
or very near background level (0.4–1.4% for nonsmokers). The healthcare-scientific
subgroup did only slightly better on this question than lay people in any age group.
Fifty percent of the people in that group didn’t know the answer. This is consistent
with my experience with physicians, nurses, and so forth. Physicians will often ask
patients to come to their office or a hospital on Monday to have blood drawn for
COHb measurement, when the fact of CO exposure was discovered the previous
Friday. Consequently, measurements made in this way are useless, and when used
as a forensic tool may be directly injurious to the CO victim, since the finding of
a normal COHb value may cause the patient to be sent back to the same site for
continued CO exposure, injury, and even death.
15.11 PROPERTIES OF CO (QUEST. 10)

This question gave respondents almost as much trouble as number 9. Only 15–30%
knew that CO has about the same density as air (actually 4% lighter). An approxim-
ately equal or even larger percentage indicated they didn’t know CO’s properties. Of
the healthcare-scientific subgroup, 28% got the question correct and 28% said they
didn’t know the answer. Significant fractions of people thought CO was either much
lighter or much heavier than air. An elementary knowledge of chemistry (molecular
weights of the gases) allows one to easily compute density without recourse to pencil
and paper, providing the information necessary to answer this question. Again, the
educational establishment must take note. Of the select group of healthcare-scientific
respondents, 32% believed CO was much lighter than air. CO mixed with air only
rises when the air is warmer than surrounding air—that is, hot air rises! As stated
elsewhere, CO cannot separate from air and rise as cream may separate from milk.
That would violate the Second Law of Thermodynamics. Gases containing CO may
have a strong or a weak odor, but CO plays no part in that. It is odorless and tasteless.
15.12 EXPERIENCE WITH CO POISONING?

(QUEST. 11)
As many as 10–15% of respondents believed they had experienced CO poisoning
once or more than once. Up to another 15% knew of someone who had. That means
8417: “8417_c015” — 2007/9/11 — 12:15 — page 338 — #14

that as many as one-fifth of respondents had some direct or indirect experience with
CO poisoning. Another significant fraction of respondents didn’t know whether they
had or not. Table 15.12 lists the sources of CO poisoning that were written in by
respondents. Vehicles and heaters were prominent sources, both in Michigan and
in Florida. Other sources included water heaters, fires, lawn mowers, boats, gen-
erators, and so forth. The percentages of respondents indicating poisoning at least
once was not different between Michigan and Florida, and almost equal numbers
of sources were given in both areas of the country. This tends to repudiate asser-
tions that CO poisoning is much more prevalent in the cold north as opposed to the
warm south.
15.13 USE OF CO EMITTING DEVICES

Table 15.13 provides data about rental and ownership of CO emitting devices by
respondents in Michigan and in Florida. Chain saws were owned by the largest
percentages of respondents, both adults and juveniles or their parents, whether in
Michigan or in Florida. Ownership was higher in the north than in the south. In all
cases, a higher percentage of males owned chain saws than females. Ownership of
generators was higher in Florida than in Michigan. A male–female difference was
not clear-cut. Ownership of pressure washers was somewhat higher in the south than
the north, but for ownership of kerosene heaters this difference was reversed. Chain
saw, pressure washer, and generator rental tended to be higher in the south than in
the north. Clearly, whether in the north or in the south, a significant percentage of
people own or rent one or more small CO emitting devices that potentially expose
them to CO. Improper use of these devices because of misperceptions and/or lack of
education could result in injury or death.
ACKNOWLEDGMENTS
In Benzie County Michigan, we wish to thank the Rotary club of Frankfort, the Benzie
Sunrise Rotary Club, and the Benzie Bicycle Club. Our many thanks go to Mr. Gary
Waterson, science teacher at Benzie County Central High School in Benzonia.
In St. Johns County Florida, we wish to thank the St. Augustine Rotary Club,
the St. Augustine Sunrise Rotary, and the Ancient City Road Runners. Many thanks
also to St. Augustine science teachers Karen Ford, Ph.D. at Pedro Menendez High
School (Public), Becky Melton, M.D. at St. John’s Academy (Christian School), and
Mr. Peter Bugnet at St. Joseph Academy (Catholic High School).
References
1. Hampson, N.B., Zmaeff, J.L. Carbon monoxide poisoning from portable electric
generators. Am. J. Prev. Med., 28, 123–125, 2005.
2. Centers for Disease Control and Prevention. Use of carbon monoxide alarms to prevent
poisonings during a power outage—North Carolina, December, 2002. Morb. Mortal.
Wkly. Rept., 53, 189–192, 2004.
8417: “8417_c015” — 2007/9/11 — 12:15 — page 339 — #15

8417: “8417_c015” — 2007/9/11 — 12:15 — page 340 — #16
16 Treatment of Carbon
Monoxide Poisoning
Suzanne R. White
CONTENTS
16.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341

16.2 Historical Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342
16.3 Mechanisms of Carbon Monoxide Toxicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344
16.4 Currently Available Neuroprotective Treatments . . . . . . . . . . . . . . . . . . . . . . . . . 348
16.4.1 Normobaric Oxygen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
16.4.2 Hyperbaric Oxygen Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
16.4.3 Allopurinol and N-acetylcysteine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
16.4.4 Insulin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
16.4.5 NMDA Receptor Antagonists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354
16.4.6 Brain-Derived Neuropeptides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
16.4.7 Hypothermia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
16.5 Approach to the Patient with Carbon Monoxide Poisoning . . . . . . . . . . . . . . 355
16.5.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
16.5.1 Neuroimaging. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
16.5.2 Pregnancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
16.5.3 Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
16.6 Management of the Sequelae of Carbon Monoxide Poisoning . . . . . . . . . . . 361
16.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
16.1 INTRODUCTION
Carbon monoxide (CO) poisoning is the leading cause of toxicologic death in the
United States of America, with 5600 fatalities reported annually.1 Worldwide CO
remains the most lethal toxin in every community in which it has been studied.2
While mortality rates associated with acute exposure to CO may have declined over
the past two decades, the total public health burden has not decreased.3 Most not-
ably, delayed neuropsychiatric sequelae in a significant percentage of CO-poisoning
survivors continues to pose an enormous challenge.4−9
The following chapter will focus primarily on the various treatment aspects of
acute CO poisoning. It should be kept in mind that our present knowledge regarding
341
8417: “8417_c016” — 2007/9/11 — 12:15 — page 341 — #1

therapy for CO poisoning is limited for several reasons. First, effective medical
treatment is ideally guided by diagnosis of either positive or negative outcomes
following exposure to toxic substances. For example, blood or urine levels of toxins,
combined with characteristic signs or symptoms of toxicity often aid in the institu-
tion of appropriate therapy. Unfortunately, symptoms relating to CO exposure are
notoriously vague, and some studies estimate that the diagnosis is missed in 30%
of cases presenting to the emergency department.10 For instance, when all neurolo-
gic admissions over a 5-month period were screened, it was determined that three
out of 29 patients admitted with impaired consciousness and no lateralizing neuro-
logical signs had serious CO intoxication.11 A further toxicological challenge is that
carboxyhemoglobin (COHb) levels neither correlate with toxicity nor predict the risk
for development of long-term effects.12,13 Although other predictors of long-term
neuropsychiatric sequelae have been proposed (i.e., loss of consciousness,14 cerebral
edema on brain computed tomography,15 elevated blood glucose16 or a history of a
“soaking” type exposure,17 )their sensitivity and specificity are largely unproven. As
a result, how best to treat patients with these clinical warning signs and symptoms
remains controversial. Second, appropriate therapy for poisoned patients is ideally
guided by an understanding of the toxic mechanisms of that poison. Unfortunately,
even though CO has most likely been present since the beginning of time, and has
been studied clinically for over 100 years, an adequate understanding of its toxic
mechanisms continues to elude us. Lastly, treatment guidelines should ideally be on
the basis of prospective, well-controlled, peer-reviewed studies. There is a dearth of
such studies relating to CO-poisoning treatment in the literature.
Despite these limitations, a general approach to treatment will be described. As
an overview, treatment is on the basis of cessation of tissue hypoxia, the removal of
CO from the body, the consideration of potential neuroprotective interventions, and
management of the long-term sequelae of CO poisoning. First, a review of historical,
often failed treatments for CO poisoning will be presented, followed by a discussion
of promising neuroprotective agents. Finally, a clinical approach to the CO-poisoned
patient will be outlined.
16.2 HISTORICAL PERSPECTIVE

Oxygen therapy has been the mainstay of treatment for CO poisoning since it was
first used therapeutically by Linas and Limousin in 1868.18 Haldane subsequently
was able to experimentally demonstrate that mice exposed to “carbonic oxide” were
unaffected if oxygen was provided during the exposure. In this seminal work, Haldane
concluded that “the higher the oxygen tension the less dependent an animal is on its
red corpuscles as oxygen carriers, since the oxygen simply dissolved in the blood
becomes considerable when the oxygen tension is high.”19 Indeed, sea level oxy-
gen decreases the half-life of CO from 320 to 80 min. Unfortunately, sea level
oxygen alone has not been entirely effective in the treatment of CO poisoning, par-
ticularly with regard to the prevention of delayed neuropsychiatric sequelae. This
realization has prompted researchers and physicians to search for yet other treatment
modalities.
8417: “8417_c016” — 2007/9/11 — 12:15 — page 342 — #2

Treatment of Carbon Monoxide Poisoning 343
The use of resuscitative gasses other than oxygen has been proposed. Studies by
Killick demonstrated more rapid clearance of COHb with 5% carbogen (5% carbon
dioxide, plus 95% oxygen), which was thought to be related to increased ventilatory
drive.20−21 Schwerma et al.,22 exposed dogs to 0.3% CO until near-respiratory arrest
occurred. Upon removal from exposure, 36% survived with fresh air alone. The
survival rate increased to 50% when mechanical ventilation was used, 69% when
ventilation with 100% oxygen was applied, and 66% when mechanical ventilation
was combined with 7% carbogen. There was no clinical advantage to the use of
carbogen in terms of improved survival, normalization of pH or lactate, or decreased
incidence of neurologic sequelae in animals relative to breathing 100% oxygen alone.2
Thus, this method of treatment has subsequently fallen out of favor.
Several other fascinating drug therapies for CO poisoning have not proven to
be effective, and are mentioned here only for historical interest. Methylene blue,
succinic acid, persantine, iron and cobalt preparations, and ascorbic acid have all been
tried, without benefit.23 In animals, cytochrome c, theorized to activate cytochrome
oxidase upon supplementation, has not been associated with clinical improvement.24
Hydrogen peroxide infusions do reduce COHb content in experimental animals, but
the absence of human experience with this chemical and the danger of air embolism
preclude its clinical use.25 While ultraviolet radiation was proposed to facilitate the
dissociation of COHb from red blood cells during transit through skin capillaries and
to decrease mortality in animals,26 these results were not able to be duplicated in a
subsequent animal trial.27 Intravenous procaine hydrochloride does not improve the
anoxia of CO poisoning in humans.28 Intravenous lidocaine was advocated, on the
basis of its facilitation of neuronal recovery after cerebral ischemia in experimental
animals, but has not yet been employed in CO-poisoned humans.29 Dipyridamole
pretreatment in rats with inhalational CO toxicity inhibited ultrastructural changes
in capillary endothelial cells, myocardial mitochondria, and myocardial myofilament
arrangement,30 but follow-up studies were never peformed.
Exchange transfusion has been reported to improve survival following CO poison-
ing in an animal model.31 Despite the fact that this method has been utilized in only a
single patient,32 it is still promoted by some clinicians as an alternative to hyperbaric
oxygen therapy (HBOT).33 While exchange transfusion does in fact lower COHb
levels, given the complex mechanisms for CO toxicity that extend well beyond the
toxicity of COHb, this technique is not likely to be effective as a sole therapy. Fur-
thermore, given the potential for exchange transfusions to deplete valuable blood
product resources and place the patient at risk for blood-borne pathogen infections,
this treatment modality can no longer be recommended.
Perfluorochemical infusions have been used in animal models as treatment for
CO toxicity.34,35 Recently, pyridoxalated hemoglobin-polyoxyethylene conjugates
(PHPs) have been developed. These agents act as blood substitutes capable of trans-
porting oxygen through chemical modification of hemoglobin derived from human
blood erythrocytes whose shelf-life has expired. Affinity of PHP for oxygen is almost
identical to that of whole blood. PHP use in rabbits poisoned by CO was associated
with prolonged survival time, temporary recovery of PO2 and PCO2 , and elevations
in pH and blood pressure in comparison with animals treated with saline.36 Beyond
the fact that human use of this product has not yet been reported, its efficacy as
8417: “8417_c016” — 2007/9/11 — 12:15 — page 343 — #3

a sole therapy is unlikely, for the same reasons as discussed above with exchange
transfusion.
HBOT was first suggested as treatment for CO poisoning in 1901 by Mosso.37
The first clinical use of HBOT in the treatment of human CO poisoning, occurred in
1960.38,39 This modality has subsequently become the mainstay of therapy for severe
CO poisoning in an average of nearly 1500 patients treated annually.40
16.3 MECHANISMS OF CARBON MONOXIDE

TOXICITY
To gain an understanding of the available methods of treatment for CO poisoning,
a review of what is known about mechanisms for CO toxicity, albeit an incomplete
comprehension of the problem, is presented here.
Hypoxic ischemia plays a significant role in the neurotoxicity of CO, which binds
slowly to hemoglobin, but with extremely high affinity (240 times that of oxygen).
Oxygen binding sites are occupied by CO at very low partial pressures of the gas,
decreasing the oxygen carrying capacity of the blood and subsequently decreasing the
usual facilitation phenomenon for further unloading of oxygen at the tissue level. The
net result is an abnormally hyperbolic oxygen dissociation curve that is shifted to the
left. Those tissues most susceptible to the hypoxic effects of CO are those that are the
most metabolically active. Oxygen delivery may further be impaired through the alter-
ation of erythrocyte diphosphoglycerate concentration.41 In adults, COHb half-life is
dependent upon the concentration of inspired oxygen, and is most commonly reported
to be approximately 4.5 h on room air, 90 min on normobaric 100% oxygen, and 20
min with oxygen applied at hyperbaric concentrations. It should be noted that reported
half-lives are extremely varied in the literature. In children, the half-life of COHb has
not been well-studied, but is reported by one author42 to be 44 min on 100% oxygen
at normobaric pressure, on the basis of measurements performed on 26 school-aged
children. The half-life of COHb in the fetus is approximately 7 h.43 Over and above
hypoxia, CO induces ischemia secondary to hypotension. Hypotension may be medi-
ated through CO binding and activation of guanylyl cyclase which increases cGMP
and triggers cerebral and peripheral vasculature smooth muscle dilatation or myocar-
dial suppression. It may further be compounded by CO triggered release of nitric
oxide from platelets with subsequent central and peripheral vasodilation. The degree
of central nervous system (CNS) damage observed following poisoning correlates
well with the degree of hypotension noted.44
Additional mechanisms of toxicity have been proposed given observations that
(1) COHb levels did not correlate with toxicity, (2) COHb formed by noninhalational
routes did not produce the same lethal consequences as inhalational exposure to CO,
and (3) that delayed neuropsychiatric sequelae were common after apparent complete
recovery from the initial CO insult. Early researchers such as Haldane,45 Brabkin,46
and Goldbaum47−49 suggested intracellular uptake of the gas as a possible mechanism
for toxicity. In competition with oxygen, CO will bind iron or copper scontaining pro-
teins such as myoglobin, mixed-function oxidases, and cytochrome c oxidase in vitro.
The binding to cytochrome c oxidase (a, a3 ) has been proposed as the mechanism
8417: “8417_c016” — 2007/9/11 — 12:15 — page 344 — #4

for intracellular CO toxicity and has been demonstrated in animals.50 It has also been
demonstrated that during recovery, the ultimate restoration of mitochondrial function
lags behind clearance of COHb.51 However, the Warburg constant for cytochrome oxi-
dase is unfavorable for CO binding relative to the other hemeproteins.52 Furthermore,
only reduced cytochrome a, a3 binds CO. It is likely, then, that other hemeproteins
act as “CO buffers,” thus preventing significant binding to cytochrome c oxidase at
COHb levels of less than 50%. At high levels of COHb, depletion of high energy
stores and intracellular neuronal acidosis occurs, which may favor CO-cytochrome
binding. On the other hand, in vivo data from Rivers53 supports hemoglobin binding
with impaired oxygen delivery, rather than mitochondrial poisoning as the etiology
of the metabolic acidosis in CO poisoning. In this model, even at extremely high
COHb levels, dogs were able to fully extract and utilize oxygen, indicating a lack
of mitochondrial effect. In addition work by Ward54 demonstrated that expression
of the heat shock proteins 72 and 32 (sensitive markers of acute neuronal stress)
did not occur following CO poisoning in rats who were maintained normotensive
throughout the exposure. This caused the authors to question the role of CO as a dir-
ect neurotoxin, and to suggest that neuronal injury results from hypotension-induced
ischemia. The role of iron as a promoter or attenuator of CO toxicity is not clear.
Iron deficiency, results in lowered hemoglobin, cytochrome and myoglobin levels
in the animal model.55 These combined effects could potentially predispose to CO
toxicity. Conversely, neuronal tissues high in iron content, such as the basal ganglia
seem particularly vulnerable to the effects of CO. In fact, limiting iron availability
confers neuroprotection from CO in the developing auditory system.56
Myoglobin binding may play a role in CO-mediated toxic effects. Myoglobin is
a hemeprotein with similar three-dimensional configuration to hemoglobin that can
bind CO reversibly. Myoglobin binds CO more slowly and with greater affinity than
does hemoglobin in vivo. Normally, myoglobin is an O2 carrier protein that facilitates
oxygen diffusion into skeletal or cardiac muscle cells and serves to place oxygen
stores in close proximity to mitochondria. Cardiac and skeletal muscle injury could
theoretically result from impaired myoglobin function. While the clinical significance
of COHb formation is not yet clear, there is increasing emphasis on cardiac injury
related to CO poisoning in the literature. Cardiac injury has historically been observed
at COHb levels of 20–40%.57 Recently, Aslan reported on 83 young, healthy patients
with severe CO poisoning. These victims had loss of consciousness in 63% and
an average COHb level 34.4%. They were evaluated with echocardiogram (ECHO)
and myocardial perfusion single-photon emission computed tomography (SPECT)
scanning. Findings included diagnostic ischemic electrocardiogram (EKG) changes
in 14.4% and abnormal SPECT results in 11%. Six of the latter group had confirmatory
and corresponding ECHO abnormalities.58 Henry and colleagues noted ischemic
EKG changes in 30% of 230 patients with moderate to severe CO poisoning. Cardiac
biomarkers were elevated in 35% of these patients and in-hospital mortality was 5%.59
In a subsequent outcome study, these investigators noted an association between
moderate to severe CO exposure and myocardial injury, finding this injury to be a
strong predictor of mortality. It was further suggested that patient subsets include those
with a “stunned myocardium” and others with “unmasking” of underlying coronary
artery disease.60 Longer durations of CO exposure predisposed people to myocardial
8417: “8417_c016” — 2007/9/11 — 12:15 — page 345 — #5

injury, but not mortality in this study. Defining the earliest and most sensitive cardiac
markers of injury is clearly an important area of future investigation.61,62
The mechanisms underlying the delayed effects of CO poisoning have been
longstanding toxicological conundrums. An increasingly complex body of liter-
ature suggests that brain ischemia/reperfusion injury, lipid peroxidation, vascular
oxidative stress, neuronal excitotoxicity, apoptosis, and immunotoxicity all play
significant roles. After removal from the CO environment, animal models demon-
strate marked changes in neutrophil structure and function. Abnormal adherence to
brain endothelial cell receptors quickly occurs, possibly as a result of endothelial
damage. Up-regulation of endothelial intercellular adhesion molecules (ICAMs) is
demonstrated on endothelial cell surfaces as a result of activation by inflammat-
ory mediators. ICAMs bind beta 2 integrins located on PMN surfaces, resulting
in aggregation of polymorphonuclear cells (PMNs) onto endothelial cells in the
neuromicrovasculature. Subsequent degranulation of PMNs results in release of
destructive proteases, which cause oxidative injury and trigger further inflammatory
responses. Thom’s work63−65 has been instrumental in elucidating the CO-induced
perivascular oxidative changes that occur during recovery from CO poisoning and
ultimately lead to superoxide formation, prolonged lipid peroxidation reactions, react-
ive oxygen species (ROS) generation, vascular injury, and neuronal death. Even
lower level CO exposure can produce vascular oxidative stress as evidenced by
platelet-mediated nitric oxide release and deposition of peroxynitrate, a highly oxid-
ative substance.66 Peroxynitrite, which forms from NO released from platelets and
endothelial cells, can further inactivate mitochondrial enzymes and damage vascu-
lar endothelium of the brain.67,68 ROS generation can be attributed to several other
sources including mitochondria and cycloxygenase. ROS production increases not-
ably during CO hypoxia, with the highest oxidative stress occurring in the most
vulnerable brain regions.69 This stress may result from lower antioxidant capacity
or higher tissue concentrations of iron. In fact, limiting iron availability confers
neuroprotection from CO.56 In mitochondria, decreased ratios of reduced oxidized
glutathione are seen following CO poisoning, and may reflect decreased ability to
detoxify ROS.70
As in animal models, acute CO poisoning in humans has resulted in intravascu-
lar platelet-neutrophil interactions and neutrophil activation. Thom71 demonstrated
these phenomena in 50 patients by measuring actual aggregates and myleoperoxidase
(MPO) concentrations. It was noted that patients with exposures of greater than 3 h
duration had increased neutrophil expression of CD18 surface receptors and MPO. In
animal models, MPO was deposited along the brain vascular lining and colocalized
with nitrotyrosine. Changes did not occur in thrombocytopenic models or those using
platelet–neutrophil interaction inhibitors such as tirofiban. Lastly, theses changes
were not noted when l-nitroargninine methyl ester, a nitric oxide synthesis inhibitor,
was given or in knock-out mice lacking MPO.71
Reactive products of lipid peroxidation like malonylaldehyde can cause adduct
formation with neuronal myelin basic protein (MBP). The altered cationic state
MBP triggers an adaptive immunological cascade that includes antibody-mediated
degradation of MPB over the course of days. This triggers a secondary influx
of macrophages and CD-4 lymphocytes that exhibit an autoreactive, proliferative
8417: “8417_c016” — 2007/9/11 — 12:15 — page 346 — #6

response to the altered MPB. Brain microglial activation ultimately occurs. These
neuropathological changes are associated with decrements in learning that are not
seen in rats immunologically tolerant to MBP.72
CO-mediated oxidative stress leads to the release of excitatory amino acids
(EAA).73 Subsequent neuronal excitation leading to cell death may also play a role
in the development of delayed toxicity following CO poisoning. These effects have
been extensively reviewed by Piantadosi,74 and are paraphrased here. Excitatory
amino acids such as glutamate, accumulate in synaptic clefts during neuronal depol-
arization owing to both excessive presynaptic release and failure in ATP-dependent
reuptake mechanisms.75 Interstitial glutamate concentration increases in the hippo-
campus during and after CO exposure. Postsynaptic binding to at least three glutamate
receptors including N-methyl-d-aspartate (NMDA), Kainic acid (KA), and α-amino-
3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) occurs with a secondarily
increased influx of calcium into postsynaptic neurons. This hypercalcemia is associ-
ated with neuronal death. ROS production follows increased excitatory amino acid
(EAA) release as well. NMDA receptor antagonism attenuates delayed neuronal
degeneration in the hippocampus after CO poisoning, and is discussed further below.
Physiological amounts of CO generated by heme breakdown seem to act to
impact neurologic processes such as long-term potentiation of memory and neur-
otransmitter release. Conversely, in toxic doses, CO alters the modulating influence
of local NO on soluble guanylate cyclase, an effect that is most evident at 7 days
after exposure.76 Despite its role in vasodilation and peroxynitrite formation, NO
has therefore been proposed to also mediate the toxic effects of CO.77 Given these
disparate data, the exact role of NO as an attenuator or mediator of CO toxicity is still
under investigation.78
Catecholamine excess may also be detrimental following CO exposure. EAA
release causes excessive surges of norepinephrine and dopamine release and synaptic
accumulation of these neurotransmitters. This effect appears to somehow be linked
to NO production, as both events can be prevented by nitric oxide synthase (NOS)
inhibition. CO induces both heme oxygenase and NOS in cortical pyramidal neurons.
While it is unclear whether the resulting altered cerebral blood flow is a pathological or
protective response, worsened outcome in sheep treated with hemoxygenase (HO) and
NOS blockers suggests the latter.79 Therefore, EAA release, catecholamine release,
and NO production are under tight regulation in vivo and can be influenced by CO-
induced oxidative stress. Furthermore, auto-oxidation or oxidative deamination of
catecholamines occurs during ischemia and reperfusion by type B monoamine oxidase
and contribute to further ROS formation.80 ROS production after CO exposure can
be inhibited by partially blocking type B monoamine oxidase, located predominantly
in glia.81 Gliosis, a known neuronal response to injury, and a condition found in
Alzheimer’s disease, develops and may play a role in the delayed toxicity seen after
CO hypoxia.
ROS are capable of triggering programmed cell death or apoptosis. Apoptotic
cell death requires activation and/or expression of specific cellular processes, some
of which may act through oxidant pathways. In animals, CO-induced neuronal loss
was slight at Day 3, increased at Day 7 and persistent at Day 21 following exposure.
Neuronal apoptosis was observed to be present upon histopathological examination
8417: “8417_c016” — 2007/9/11 — 12:15 — page 347 — #7

of the animals in this model.82 Others have not demonstrated apoptosis in rats, des-
pite moderately severe poisoning,83 but have observed gliosis. Caspase-1 and NOS
inhibitors both block CO induced apoptosis.84
As eloquently summarized by Piantadosi,74 “impaired mitochondrial energy
provision in CO hypoxia/ischemia leads to neuronal depolarization, EAA and cat-
echolamine release, and failure of re-uptake until energy metabolism is restored
during reoxygenation. These processes, normally modulated by NO production, could
contribute to degeneration of neurons in vulnerable regions, possibly by enhancing
mitochondrial ROS generation which can initiate apoptosis.”
16.4 CURRENTLY AVAILABLE NEUROPROTECTIVE

TREATMENTS
16.4.1 NORMOBARIC OXYGEN
There are limited data to suggest that normal neuropsychiatric outcome is possible
after treatment with normobaric oxygen (NBO). In one study, 33 patients with acute
CO poisoning (mean COHb 29.4%, 10 patients above 40%, 7 comatose on arrival)
were treated with 100% NBO. Recovery was reportedly rapid, with no neurological
deficits at discharge. Formal neuropsychiatric testing and follow-up were not per-
formed, however.85 Similarly, four patients presenting comatose from CO poisoning
who did not receive HBOT were identified retrospectively, but then evaluated with
formal neuropsychological testing at 6 and 12 months after exposure. All had normal
neuropsychiatric examinations.86 Meert looked retrospectively at the outcome of chil-
dren treated with NBO, and concluded that acute neurologic manifestations resolve
rapidly without HBOT.87 However, neurologic outcome was assessed from nursing
and physician records, physical and occupational therapy evaluations, and unspecified
neuropsychological examinations in an unspecified number of patients. Therefore, the
major limitation of this study is the lack of detailed neurologic assessment both at
presentation and at follow-up that would allow detection of those specific neuropsy-
chiatric changes known to result from CO poisoning. Nonetheless, the authors noted
“gross” neurologic abnormalities in nine (8.5%) survivors, with seven of those per-
sistent at various stages of follow-up (2 months to 3.3 years). Three patients developed
delayed neurologic syndromes including tremors, hallucinations, seizures, occipital
lobe infarctions, defects in cognitive, and interpersonal skills. The presence of seri-
ous comorbidities, such as smoke inhalation, burns, need for mechanical ventilation,
and need for surgical procedures certainly confounds the outcomes reported by these
investigators, which are not reflective of pure CO exposure. Overall, given the small
size of these trials and their significant methodological limitations, they do not signi-
ficantly add to our understanding of predictors for good outcome after treatment with
NBO therapy.
16.4.2 HYPERBARIC OXYGEN THERAPY

In addition to the aforementioned effects of increasing both the amount of dissolved
oxygen in the blood and the rate of displacement of CO from hemoglobin, HBO may
8417: “8417_c016” — 2007/9/11 — 12:15 — page 348 — #8

have other beneficial effects. In animal models, HBO at 3 atmospheres absolute (ATA)
also prevents functional neurological impairment88 and HBO at 2.5–3.0 ATA revers-
ibly inhibits PMN CD18 beta 2 integrin activation therefore decreasing adherence of
PMNs to endothelial cells.89,90 This effect has also recently been noted in patients with
acute severe CO poisoning, in whom HBO modulated neutrophil generation of ROS
and surface expression of CD18 receptors.91 Moreover, HBO is known to regenerate
inactivated cytochrome oxidase, and may thereby restore mitochondrial function.92
An astroglial structural protein S100B that is a proposed marker for neuronal injury
is elevated in CO-poisoned rats treated with ambient air or NBO, but not in those
treated with HBO.93 Other proposed beneficial effects of HBOT include decreased
production of ROS,94 protection against cerebral edema and increased cerebrospinal
fluid (CSF) pressure,95,96 induction of production of protective stress proteins (SP-
72), and antagonism of NMDA excitotoxic neuronal injury. Conversely, HBO has not
been effective in animal models as a treatment for nonCO mediated acute cerebral
ischemia with reperfusion.97
Anecdotal human case reports suggest significant clinical improvement from CO
poisoning during HBOT.98−105 In addition, numerous other case series report on the
beneficial effects of HBOT, but these are limited by either their retrospective nature, or
prospective design without the use of randomization, double-blinding, or controlled
methodology.106−112
Three pioneering research efforts attempted to discern potential benefit from
HBOT, through randomization of CO-poisoned patients into HBO and NBO treat-
ment groups. Raphael113 studied 343 mildly poisoned CO-patients (those who had
not lost consciousness) and found no difference at 1 month follow-up in neuro-
logic outcome between HBO and NBO-treated groups, and no benefit to multiple
HBO treatments in a more severely poisoned group (those who had lost conscious-
ness), randomized to either one or two HBO treatments. Criticisms of this study
regarding the method of neurologic evaluation used, the application of inadequate
doses of oxygen therapy, and potential delays in HBO treatment have been raised.114
Ducasse115 randomized noncomatose patients to HBO and NBO groups and found
significantly improved differences in quantitative electroencephalogram (EEG) and
cerebral vascular responsiveness to acetazolamide in the HBO group during the first
24 h. Differences at 3 week follow-up are not reported consistently. Early improve-
ments in clinical signs and symptoms, such as headache, reflex impairment, and
asthenia were significant in the HBO group. Thom116 randomized patients with mild
to moderate CO poisoning into NBO versus HBO groups. The incidence of delayed
neuropsychiatric sequelae (DNS) was 23% after treatment with oxygen at ambient
pressures and 0% in the HBO group. DNS persisted for an average of 41 days fol-
lowing exposure. The author concluded that HBOT decreased the incidence of DNS.
Limitations of this study117 are the lack of double-blinding, inconsistent NP testing
methods used, and exclusion of severely poisoned patients.
Recently, several additional more rigorously designed trials have been carried out.
All are prospective, randomized, double-blinded, controlled clinical trials assessing
the benefit of HBO versus NBO in the treatment of CO poisoned patients. Two have
been completed and one has resulted in publication of the interim data analysis. These
are reviewed below.
8417: “8417_c016” — 2007/9/11 — 12:15 — page 349 — #9

Scheinkestel et al. peformed the first large prospective, randomized, double

blinded, controlled clinical trial investigating the neurologic sequelae in 191 patients
with all grades of CO poisoning after treatment with HBO and NBO.118 Sham HBO
treatments were given to those randomized to the NBO group. Pregnant women, chil-
dren, and burn victims were excluded. Higher doses of oxygen were utilized than
reported in most previous studies, averaging approximately 37 COHb-dissociation
half-lives in the HBO group and 28.5 in the NBO group (up to three daily treatments
in those not improving). Neuropsychiatric evaluations were performed at completion
of treatment and at discharge. No benefit and possible adverse effects of HBO were
found. Overall mortality was 3%, with persistent neurologic deficits in 71% at hos-
pital discharge and 62% at one-month follow-up. All five patients with delayed onset
of neurologic deficit occurred in the HBO group. Limitations of the study are that
44% of victims had ingested other drugs, there was a mean treatment delay of 7.1 h,
56% of patients were lost to 1-month follow-up, and 76% were suicidal, which could
impact neuropsychological testing. These results led the investigators to conclude
that HBOT should no longer be recommended for CO poisoning treatment.
A similarly designed longitudinal follow-up study employing sham treatments,
found approximately 30% of patients with acute CO poisoning have neurocognitive
problems 1 year after poisoning. Of these patients, approximately one-third have the
delayed neuropsychiatric syndrome and two-thirds have persistent neurocognitive
problems, mainly difficulties with memory and executive function. 119 Patients in the
treatment arm received three HBO sessions. This randomized control trial demon-
strated a 46% reduction in cognitive sequelae in HBO-treated patients at 6 weeks
following poisoning, which was maintained at 1 year after poisoning.120 It should be
noted that nonspecific symptoms were the primary determinant of a statistical differ-
ence between treatment groups. A third randomized controlled trial (RCT) performed
on noncomatose CO-poisoned patients showed no benefit between HBO and NBO
at 1 month, 6 months, or 1 year postexposure. Since the study is published only in
abstract form, details are minimal.121 Interim results from a 4th RCT trial employ-
ing one HBOT session in patients with moderate to severe CO poisoning showed no
difference in outcome.122 Assessments were made by questionnaire and a blinded
neurology examination. A high rate of symptoms in both groups was noted (39–42%)
at 1 month after poisoning.
The use of repetitive HBOT is controversial but common with 23% of US
HBOT facilities automatically giving more than one HBO treatment per CO-poisoned
patient.40 Some investigators perform additional treatments if lack of improvement is
noted after the 1st HBO treatment.104,123−125 Others report no benefit to multiple HBO
treatments.113 Turner et al.124 recently proposed the use of the initial hydrogen ion
concentration (degree of metabolic acidosis) as a marker for repetitive HBO treatment
requirement, on the basis of a retrospective analysis of 48 patients. McNulty et al.126
found impairment of short-term memory for verbal material to be predictive of the
number of HBO treatments needed. Finally, the use of peak alpha frequency on EEG as
an indicator of need and efficacy of repetitive HBO treatments has been proposed.127
While generally safe, unusual risks of HBOT include complications arising from
transport to an HBOT facility, barotrauma, oxygen toxicity with resultant seizures in
1–3% of CO-poisoned patients128 and fire or explosion hazard. Historically, the lack of
definitive results from RCTs have rendered the indications for HBOT in CO poisoning
8417: “8417_c016” — 2007/9/11 — 12:15 — page 350 — #10

arbitrary and tremendously varied. Recommendations have primarily been derived

from patient history, presenting neurologic and cardiovascular signs and symptoms,
and laboratory data such as glucose, lactate, arterial blood gases, electrocardiography,
and COHb levels. Unfortunately, none of these clinical or laboratory findings at
presentation are entirely predictive of long-term outcome following CO exposure.
Those that show future promise, such as SB100 and neutrophil response, are not yet
clinically available. Of note, recent preliminary work suggests that HBOT benefit
may be limited to reducing the incidence of persistent neuropsychiatric sequelae
(PNS) but not DNS, perhaps through modulating early but not later mechanisms for
brain injury.129 Overall, the evidence favoring HBOT as protective against DNS and
PNS remains under fire as experts continue to debate the results and limitations of
each of the randomized controlled trial performed to date.130−138 These uncertainties
definitely suggest that greater future research emphasis be placed on non-HBOT
methods of therapy.
16.4.3 ALLOPURINOL AND N-ACETYLCYSTEINE

As described above, there is considerable evidence that reactive oxygen metabol-
ites mediate neurologic injury in models of CO poisoning. Lipid peroxidation is
documented in rats after CO exposure at concentrations sufficient to cause uncon-
sciousness. Products of lipid peroxidation are increased by 75% over the baseline
values 90 min after CO exposure. Unconsciousness is associated with a brief period
of hypotension, so brief that in itself it causes no apparent insult. Lipid peroxid-
ation occurs only after the animals are returned to CO-free air; and there is no
direct correlation between the degree of lipid peroxidation and COHb level.63 This
suggests that ischemia and reperfusion may play a role in the ultimate neurologic
injury. Xanthine oxidase also has a central role in this toxicity. During the CO-
induced PMN degranulation described above, released proteases convert xanthine
dehydrogenase to xanthine oxidase. Xanthine oxidase generates superoxide free
radicals and lipid peroxidation occurs.65 The xanthine oxidase enzyme is an NAD-
dependent dehydrogenase that under ischemic conditions converts to an oxidase,
utilizing molecular oxygen rather than nicotinamide adenine dinucleotide (NAD) as
an energy source and generates the superoxide radical and hydrogen peroxide. These
products in turn cause tissue injury, the brain being particularly susceptible with its
low content of catalase and glutathione peroxidase.139 The restoration of xanthine
dehydrogenase functional activity is accomplished through the use of xanthine oxi-
dase inhibitors (allopurinol)140 and sulfhydryl donors [N-acetylcysteine (NAC)]141
in nonCO-mediated neuronal injury. Fechter et al.142 noted that acute CO poisoning
produces preferential high-frequency hearing impairment, noted to be a consequence
of other types of anoxic exposure. These investigators also discovered that either
allopurinol or phenyl-n-tert-butyl-nitrone (PBN), a free radical scavenger blocked the
formation of characteristic compound action potential threshold elevation and coch-
lear microphonic amplitude. Therefore, both agents were effective in blocking loss
of CO-mediated auditory threshold if given prophylactically in the guinea pig model.
Allopurinol and N-acetylcysteine have been used in the treatment of CO-induced
neuronal injury. Thom63 demonstrated decreased conversion of xanthine dehydro-
genase to xanthine oxidase with decreased lipid peroxidation in rats pretreated with
8417: “8417_c016” — 2007/9/11 — 12:15 — page 351 — #11

allopurinol. Only one human case report demonstrated effectiveness of such com-
bined therapy for the treatment of CO poisoning. A 26-year-old male with a COHb
level of 25%, 40 h postexposure, who was comatose for 4 days with cerebral edema
on computed tomography (CT) was treated with both a xanthine oxidase inhibitor
(allopurinol) and a sulfhydryl donor, NAC. NAC was given intravenously over a
20-h period and allupurinol was given orally for 2 weeks. Eight hours after the com-
pletion of this regimen, the patient became responsive and gradually improved over
the next three weeks. “Neurological and mental examination at six weeks follow-
up were normal.”143 Although no formal neuropsychiatric testing was reported, this
type of therapy may perhaps provide a basis for further study. Since allopurinol the-
oretically prevents the formation of free radicals, it remains to be seen whether any
benefit exists for postexposure administration. Other antioxidants such as dimethyl
sulfoxide and disulfiram have been shown to prevent learning and memory deficits
in CO-poisoned mice in preliminary reports.144 Such agents may someday serve as
useful adjuncts in limiting free radical mediated injury.
16.4.4 INSULIN
In humans and animals numerous studies have shown that elevated blood glucose is
associated with worsened neurologic outcome after brain ischemia caused by stroke
or cardiac arrest. Acute severe CO poisoning is characterized by hyperglycemia
and this elevation has been linked to increased severity of brain dysfunction in the
rat.145 Indeed, animal studies show that CO exposure raises blood glucose in a dose-
dependent manner, and is an independent predictor of neurologic outcome. A few
similar observations have been made in CO-poisoned patients. Penney146 observed
that elevated admission blood glucose was associated with worse neurologic outcome
after CO poisoning in patients. Leikin found elevated blood glucose in most patients
presenting with COHb saturation above 25%.147 Furthermore, anecdotal evidence in
the human literature suggests that the neurologic outcome in diabetics poisoned with
CO is generally worse than in nondiabetics.148
Considerable basic research has been directed at identifying molecular mechan-
isms of tissue injury and potential interventions to allow the preservation or rescue
of neurons after stroke and cardiac arrest. On the basis of the aforementioned asso-
ciation between elevated blood glucose and poor outcome, insulin has recently been
investigated as a potential therapeutic agent in various models of brain and spinal
cord ischemia, and indeed appears to substantially ameliorate neuronal death induced
by ischemia in these studies. Surprisingly, this neuron-sparing effect is known to be
independent of insulin-induced reductions in blood glucose, and is hypothesized to be
mediated through cell signal transduction mechanisms, in common with other growth
factors.
The question of whether insulin ameliorates neuronal injury secondary to CO
toxicity was investigated in rats who were exposed to a CO LD50 of 2400 ppm for
90 min. Survivors were treated for 4 h with (1) normal saline infusion (2) continuous
infusion of glucose to clamp blood glucose levels at 250–300 milligrams per deciliter
(mg/dL) and (3) continuous infusion of glucose to maintain blood glucose levels at
250–300 mg/dL with intraperitoneal (IP) injections of 4 units/kg of regular insulin.
8417: “8417_c016” — 2007/9/11 — 12:15 — page 352 — #12

Neurologic scoring was performed at time 0, 5.5, 24, 48, 72, and 96 h. It was noted
that significant neurologic deficit occurred in all groups after the CO exposure and
treatment period. Induced hyperglycemia after CO exposure was associated with sig-
nificantly worsened neurologic scores as compared to saline-treated controls. Insulin
therapy simultaneous with induced hyperglycemia significantly improved neurologic
scores at all times despite maintenance of comparable hyperglycemia with respect to
the group treated only with glucose. No significant difference in mortality was found
between treatment groups.149
Several theories have evolved regarding the postreceptor binding protective effect
of insulin on the neuron. Insulin has been shown to provide neuromodulatory inhib-
ition of synaptic transmission in vivo and in vitro. As an inhibitor of glial uptake of
gamma amino butyric acid (GABA), insulin may increase the availability of this inhib-
itory neurotransmitter and may decrease neuronal firing, beneficially reducing cell
metabolism.150 The additional effect of sodium extrusion from the cell which affords
subsequent protection against water accumulation may prevent neuronal swelling.151
Furthermore, it has been suggested that an insulin-induced elevation of brain cat-
echolamines through both inhibition of catecholamine uptake and stimulation of
release might be a contributory neuroprotective mechanism, since catecholamines
have been found to attenuate ischemic brain damage.
Of these theories regarding insulin’s neuroprotective activity, however, the most
recent highlights its role in stimulating second messengers, and emphasizes its poten-
tial genomic effects, that is, the regulation of protein synthesis, enzymatic activity,
and the signaling of cell proliferation. It is well established that the neonatal brain is
rich in insulin-like growth factor receptors. Indeed, insulin is similar in structure to
other growth factors such as platelet-derived growth factor (PDGF), epidermal growth
factor (EGF), insulin-like growth factor-1 (IGF-1). Such peptides are involved in basic
neuron development and differentiation. Once bound to its receptor, like other growth
factors, insulin triggers signal transduction by internal autophosphorylation of tyr-
osine on the insulin receptor, which subsequently enhances further phosphorylation
reactions of other tyrosine containing substrates by tyrosine kinase, located on the
insulin receptor.
This type of tyrosine phosphorylation is important in signaling pathways for such
growth factors and products of proto-oncogenes. Insulin is a progression growth factor
in replication G0G1 phase and works synergistically with other growth factors to gen-
erate both competence and progression of cells. Through tyrosine phosphorylation of
phosphokinase C, other second messengers such as diacyglycerol are formed, which
increase intracellular calcium, activate the sodium–hydrogen pump, and increase
intracellular pH. This pH change in turn activates the sodium potassium ATPase
pump, which signals cell proliferation.152 Insulin also regulates specific mRNA
levels through diacylglycerol153 and may increase mRNA efflux from the nucleus
through nuclear triphosphatase activation.154,155 More importantly, insulin stimu-
lates lipid neogenesis. It appears therefore, that the effects of insulin are fundamental
with regard to cell signaling, proliferation, replication, and repair following injury.
These processes are crucial to cells such as neurons which normally are terminally
differentiated and contain little if any capacity to replicate or to synthesize repair
lipids.
8417: “8417_c016” — 2007/9/11 — 12:15 — page 353 — #13

Certainly, anatomic correlates to the above proposed mechanisms are in place.

For example, it has been demonstrated that the location of insulin and IGF-1 receptors
correlates with phosphotyrosine products in the brain.156 Moreover, the basal ganglia,
areas typically found to be damaged by CO, possess low levels of insulin receptors.
Although the initial results of animal studies such as those above may provide building
blocks for clinical work, any recommendations regarding the use of insulin in humans
as treatment for CO poisoning will of course await further studies.
16.4.5 NMDA RECEPTOR ANTAGONISTS

Recent evidence implicates the endogenous excitatory amino acids such as NMDA
in ischemic neurodegeneration.157−159 Moreover, the NMDA receptor antagonist,
MK801, prevents nonCO induced ischemic neurodegeneration in animal models.160
Successive CO exposures induce a consistent pattern of degeneration of hippo-
campal CA1 pyramidal neurons, a selective neuronal death that resembles that
seen with other models of cerebral ischemia. This observation has prompted the
study of NMDA receptor antagonists in CO poisoning in mice. Ishimaru pre-
treated animals with a competitive NMDA antagonist, CPP; a noncompetitive
NMDA antagonist, dizocilpin (MK-801); a glycine binding site antagonist, 7-CK;
a polyamine binding site antagonist, ifenprodil; glycine; and saline. Seven days
postexposure the number of hippocampal CA1 pyramidal cells was quantified
using an image analyzer. A decrement of 20% in the number of hippocampal
CA1 pyramidal cells was noted relative to the control group. Those animals
receiving high doses of MK-801, 7-CK, and CPP had significant reductions in
neuronal damage. No clear protective effect was obtained with ifenprodil. Inter-
estingly, glycine, a facilitory neurotransmitter at the NMDA receptor complex
did not exaggerate the CO-induced neuronal damage as might be expected.161
Although no neurologic outcome correlates or survival data are reported, this work
may provide valuable mechanistic and possibly future therapeutic insights. Sim-
ilar work by Lui162 suggested beneficial effects of MK 801 when administered
either systemically or directly to the cochlea in protecting against CO-induced
ototoxicity. Glutamate is another excitatory neurotransmitter acting at the NMDA
receptor complex. Postexposure treatment of mice with glutamate antagonists pre-
vents CO-induced learning and memory deificits.163 Finally, NOS inhibitors prevent
NMDA receptor activation and were protective of learning deficits in CO poisoned
mice.73
Ketamine is a widely used dissociative anesthetic agent, which is known to have
NMDA receptor-blocking properties.164 It has been shown to be neuroprotective
in various animal models of ischemic and anoxic neuronal injury. It has also been
observed to blunt hypotension, a condition known to worsen CO-mediated neuropath-
ologic changes. Promising work by Penney165 demonstrated significantly reduced
cerebral edema, more rapid recovery from hypotension, and suppressed lactate form-
ation following CO-poisoning when 40 mg/kg ketamine was administered to rats
before and during CO exposure.165 This same study did not yield positive results with
the use of verapamil, which could theoretically block NMDA-mediated postsynaptic
calcium uptake in neurons.
8417: “8417_c016” — 2007/9/11 — 12:15 — page 354 — #14

16.4.6 BRAIN-DERIVED NEUROPEPTIDES

Cerebrolysin, a drug,produced by enzymatic breakdown of lipid free proteins of
porcine brain is a putative neuroprotective agent of unknown mechanism. Pro-
posed neurotrophic effects are supported by reports that cerbrolysin-treated rats had
increased brain protein synthesis,166 prevention of neuronal degeneration,167 and
enhanced neuronal growth in tissue culture.168 Effects on the blood brain barrier
have also been noted. Interestingly, cerebrolysin increases expression of the blood
brain barrier glucose transported gene in brain endothelial cell cultures. It is hypo-
thesized that cerebrolysin may accelerate repair of the blood brain barrier in regions
compromised by hypoxia.169 Recently, a model of acute CO poisoning combined
with spreading depression-induced metabolic stress was used to examine the protect-
ive effects of cerebrolysin on the development of electrophysiological, behavioral,
and morphological signs of hypoxic damage in rats. Spreading depression waves
reflect the recovery of cerebral cortex in the peri-ischemic areas, or penumbra zone.
After a 90 min exposure to 0.8–5% CO, microinjections of 5% KCl into the cortical
and hippocampal areas were performed and the duration of spreading depression was
noted. At 9 and 18 day follow-up, repeat spreading depression measurements were
taken, and a decrease in amplitude was used as an index of brain damage. Postexpos-
ure cerebrolysin-treated animals had significantly improved hippocampal recovery.
Better performance was also noted on behavioral testing, and no apparent histolo-
gical damage was apparent in the hippocampus as compared to controls.170 This very
promising neuroprotective agent, which appears to be effective even if given postex-
posure, certainly deserves further study to elucidate any possible beneficial role in
humans.
16.4.7 HYPOTHERMIA
Hypothermia was found to be beneficial in the management of CO poisoning by
Sluijter,171 an effect that was thought to be secondary to increased dissolved oxygen
in the blood at lower temperatures. Peirce et al.,172 howeve, was unable to demonstrate
any synergistic effect when hypothermia was used in conjunction with HBO in a dog
model.172 An interesting report of the use of mechanical ventilation and hypothermia
in patients with abnormal motor activity or coma to treat CO toxicity, noted complete
reversal of these manifestations in three patients when therapy was initiated within
the first 24 h. No beneficial effects were noted in a fourth patient who did not receive
hypothermic treatment until 5 days after exposure. HBO was not available to these
patients.173
16.5 APPROACH TO THE PATIENT WITH CARBON

MONOXIDE POISONING
16.5.1 GENERAL
The most critical step in managing the patient poisoned with CO is the cessation
of tissue hypoxia. This involves supplementation with 100% oxygen, delivered
8417: “8417_c016” — 2007/9/11 — 12:15 — page 355 — #15

by either a tight-fitting continuous positive airway pressure (CPAP) mask or by

endotracheal intubation. Intubation may be necessary in the patient with chronic
obstructive pulmonary disease (COPD), to avoid carbon dioxide retention secondary
to high concentrations of oxygen. NBO should be initiated as soon as the diagnosis is
entertained, and should not be delayed for confirmatory COHb levels. As discussed
above, the use of 5% carbon dioxide mixed with 95% oxygen (carbogen) has been
proposed by some to facilitate the release of CO from hemoglobin by increasing
ventilatory response. This therapy is of questionable value, and has fallen out of
favor. Even though not demonstrated in animals,174 the potentially life-threatening
possibility of carbogen-induced carbon dioxide retention and subsequent worsening
of an already existing acidosis would contraindicate it use in the patient with COPD,
concurrent poisoning with respiratory depressants, or altered mental status. The dur-
ation of oxygen therapy is guided by a knowledge of the COHb half-life and allows
for a margin of safety. Generally this would involve at least 6 h of therapy on 100%
oxygen, longer if the patient is gravid or an infant. An early chest x-ray is mandat-
ory to assess for evidence of pulmonary edema resulting from CO or other inhaled
toxins.
Once airway control and oxygenation are assured, attention should be direc-
ted toward the cardiovascular system. Continuous cardiac monitoring is advisable
and a 12 lead EKG should be obtained to assess for subclinical cardiac ischemia.
Myocardial enzymatic changes with or without EKG changes, are increasingly
described in adults with CO poisoning as outlined above. It is prudent therefore
to perform a cardiac evaluation on patients with CO poisoning. It is not clear,
however, whether CO-induced myocardial injury, particularly the “stunned myocar-
dium” is a predictor of poor clinical outcome or mortality. Should arrhythmias,
ischemia, or hemodynamic instability occur despite therapy with 100% oxygen,
the patient could be considered a candidate for HBOT. Myocardial depression and
arrhythmias may occur secondary to extremely low arterial pH, such as has been
noted in patients with severe lactic acidosis. Severe acidosis should therefore be
treated aggressively. However, correction of mild acidosis with sodium bicarbonate
is not advisable as this could result in a further shift of the oxyhemoglobin disso-
ciation curve to the left, and impair the unloading of oxygen to hypoxic tissues.
A novel calcium sensitizer, levosimendan, improves myocardial contractility and
increases coronary artery flow. Its use in a single patient with cardiogenic shock
from CO-induced myocardial stunning was associated with improved hemodynam-
ics relative to dobutamine. Improvement was assessed by cardiac magnetic resonance
imaging.175
In those patients with altered mental status, causes of rapidly reversible causes
of coma should be considered and treated by the bedside assessment of a fingerstick
glucose and the administration of thiamine and naloxone, a narcotic antagonist. Sup-
plemental glucose may be needed to correct hypoglycemia, but every attempt should
be made to maintain euglycemia, and avoid iatrogenic hyperglycemia. A thorough
physical assessment for burns, odors, toxidromes, skin findings, and signs of smoke
inhalation, trauma, or abuse is indicated. Acareful history regarding the circumstances
surrounding the exposure must be obtained once the patient is stabilized. Consider-
ations should be made to gastric or skin decontamination and activated charcoal
8417: “8417_c016” — 2007/9/11 — 12:15 — page 356 — #16

administration in the setting of suspected intentional drug abuse, suicide attempt, or

dermal chemical exposure (methylene chloride).
Removal of CO from the body is best accomplished through displacement by
oxygen, either normobaric or hyperbaric. Other less conventional therapies have been
used anecdotally with favorable outcomes, and include both exchange transfusion
and extracorporeal oxygenation.176 These invasive approaches however, would be
recommended only in unusual circumstances, for example if HBOT was not available
in a deteriorating or moribound patient.
Should the patient be CO-poisoned from smoke inhalation, numerous other
products of combustion may be contributing to the metabolic and pulmonary derange-
ments seen. Of particular concern is cyanide, a lethal combustion product, that is
commonly elaborated when plastics or synthetic materials burn. The patient with CO
poisoning and evidence of smoke inhalation who remains significantly acidotic des-
pite treatment with oxygen should be suspected to have concomitant cyanide toxicity.
Specifically, cyanide poisoning is associated with enclosed space fires, the presence
of soot in mouth or sputum, altered consciousness, hypotension, and an elevated
lactate >8–10 mmol/L without significant burns.177
Some authors advocate empiric treatment of fire victims with suspected cyanide
poisoning with the sodium thiosulfate component of the cyanide antidote kit. The
other components, methemoglobin-forming agents such as amyl nitrite and sodium
nitrite, have been traditionally withheld if CO poisoning is suspected in order to
avoid further hemoglobinopathy and worsened hypoxia. Moore et al.178 demonstrated
a 25% increase in mortality in sodium nitrite-treated animals with CO-poisoning
compared to the untreated controls. Despite this, a human study demonstrated that
five of seven patients with CO poisoning were safely treated with the antidote kit
in its entirety.179 These patients however, had only moderate COHb levels, with
a mean level of 26%. Coupled with the fact that only a small number of patients
were studied, this form of empiric treatment for cyanide poisoning in victims of
smoke inhalation cannot yet be widely recommended. Another cyanide antidote,
hydroxycobalamin is under Food and Drug Administration (FDA) consideration for
approval. While not yet readily available in the United States, this new antidote
will not harbor the risk of methemoglobinemia found with nitrites and will ulti-
mately offer a safer option for victims of smoke inhalation. For now, given the
relative safety of the sodium thiosulfate component of the cyanide antidote kit, its
sole use may be advisable for the treatment of the patient dually poisoned with CO and
cyanide.
Once the patient has been stabilized, consideration of the use of possible neuropro-
tective agents including HBO should be made. If the patient is awake, a mental status
examination should be performed. Abbreviated neuropsychologic tests (CONSB)
have been developed specifically for the CO poisoned patient, but often are not prac-
tical for use in the emergency department or in those with moderate or severe acute
intoxication. Examples of tasks performed by the patient during administration of the
CONSB include placing pegs in a board, complete rapid finger tapping, memoriz-
ation, construction, number processing, and subjective stress response.180 Memory
impairments are the most frequent cognitive impairment noted following CO pois-
oning, some improving over time.181 Other common deficits include visual tracking,
8417: “8417_c016” — 2007/9/11 — 12:15 — page 357 — #17

visuomotor skills, visuospatial planning, and abstract thinking. Impaired executive

function, information processing speed, attention and concentration are also common.
16.5.1 NEUROIMAGING
Neuroimaging studies can be valuable adjuncts to the neuropsychiatric assessment.
They should be considered in CO-poisoned patients with altered mental status,
abnormal, or lateralizing neurologic examinations, or a history of head trauma.
Increasingly, correlates are described between cognitive impairment and neuroima-
ging findings (see other chapters in book), particularly relative to basal ganglia
atrophy, fornix atrophy, and white matter hyperintensities.182 Moreover in adults,
CT abnormalities have been prognostic with regard to neurologic outcome in sev-
eral studies.183−185 Pathologic lesions seen on CT owing to CO intoxication are
variable, including cerebral edema, symmetrical low density areas in the basal
ganglia, symmetrical and diffuse white matter low density areas, and, as late changes,
ventricular dilation, and sulcal widening. The classic finding of bilateral symmetrical
hypodensities in the basal ganglia, especially the globus pallidus, most typically
becomes evident within 24–48 h of exposure. However, such abnormalities have
been reported to appear anywhere from the first day to 5 years following CO expos-
ure. Remarkably, such basal ganglia lesions have been reported to occur in 32%
to 86% of CO-poisoned patients.186 These lesions are not pathognomic for CO
poisoning, however, and when encountered, the differential diagnosis includes meth-
anol, cyanide, or hydrogen sulfide toxicity; hypoxia; hypoglycemia; the hemolytic
uremic syndrome; osmotic myelinolysis; encephalitis; inborn errors of metabol-
ism; and Huntington’s disease.187 Likewise, white matter lesions are frequent and
include hyperintesities in the periventricular and centrum semiovale or deep regions,
generalized white matter degeneration, and generalized atrophy. Several studies sug-
gest that white matter lesions occur even more commonly than do basal ganglia
lesions.188
Magnetic Resonance Imaging (MRI) may be superior to CT in detecting white
matter, cerebral, cerebellar, substantia nigral, and basal ganglia lesions following
CO poisoning.189−195 Quantitative MRI may be more sensitive in evaluating the
hippocampal regions in patients with DNS following CO poisoning. Some authors
report a good correspondence between MRI and memory deficits on neuropsy-
chological evaluation in adults.196−198 Conversely, Prockop199 noted a significant
percentage of patients with normal MRI examinations had intellectual impairment on
neuropsychological testing.
Functional imaging such as SPECT scanning provides an indicator of the severity
of cerebral damage and correlates with outcome.200 The combination of EEG with
SPECT scanning may provide greater sensitivity for detecting anomalies than EEG
alone.201 In a cohort of adult patients with acute severe CO poisoning, treated with
HBO, positron emission tomography (PET) scan findings of globally increased oxy-
gen extraction ratios and decreased blood flow in the frontal and temporal cortex
were most severe in those patients with DNS or PNS. These changes are temporary
in patients who appear normal following CO exposure and in those with temporary
neurological and psychiatric deficits. This suggests that ischemia is ongoing after
8417: “8417_c016” — 2007/9/11 — 12:15 — page 358 — #18

CO intoxication, even after apparent normalization of the clinical status.202 Others

have noted prolonged perfusion and metabolic abnormalities in patients with neur-
ologic deficits following exposure.203 In a small case series, PET abnormalities in
the basal ganglia of CO victims were undetected by CT or MRI, suggesting possible
greater sensitivity with PET.204 Newer techniques in SPECT scanning may allow for
even earlier detection of regional CO-induced anomalies.205 Specifically, decreased
perfusion in the basal ganglia and cortex correlated with parkinsosian symptoms
and cognitive deficits, respectively. More recently, magnetic resonance spectro-
scopy (MRS) detected decreased n-acetyl aspartase in the basal ganglia bilaterally in
one-third of CO-poisoned patients studied.199
Should the patient perform abnormally on the CO neuropsychiatric screening
battery (CONSB), have a history of a soaking-type exposure or loss of conscious-
ness, have abnormal neurologic findings (particularly cerebellar findings,119 exhibit
cerebral edema on CT scan, or have evidence of cardiac ischemia, HBOT should
be recommended. Despite this general practice, clinical predictors of DNS or PNS
remain elusive and controversial. Even syncope unreliably predicts the need for
HBOT.119 Similarly, laboratory markers such as COHb level, lactate level, or base
deficit are unreliable factors in predicting DNS or PNS. Preliminary animal evidence
points us toward the potential future use of laboratory markers such as peripheral
lymphocyte cytochrome c oxidase, cyclic GMP, cholinergic muscarinic receptors
and S100 B protein for determining patient prognosis.206,207
As suggested above, the use of COHb levels to guide therapy is controversial. A
survey of medical directors of US and Canadian facilities indicated that 62% use a
specific COHb level as the sole criterion for asymptomatic patients. The same survey
found that when a specific COHb level was used as the indication for HBOT, 25%
was the most common level chosen.208 See Table 16.1. Others suggest that HBOT is
prescribed on the basis of COHb level in 40% of patients and in 60% on the basis of
central nervous system or cardiac dysfunction.209 The patient’s clinical findings and
history are of equal importance relative to COHb in determining the need for HBOT.
Patients who could be considered candidates for HBOT include the pregnant patient
with a COHb level greater than 10–15%, the patient with a history of coronary artery
disease and a COHb level greater than 20%, the asymptomatic patient with a level
TABLE 16.1
Proposed Indications for Hyperbaric Oxygen Therapy in
Pregnancy
1. Abnormal CONSB (CO neuropsychiatric screening battery)

2. Neurologic abnormalities (particularly cerebellar findings)
3. Loss of consciousness (syncope)
4. COHb > 25–40%
5. Ongoing myocardial ischemia
6. Worsened, recurrent,or refractory symptoms on NBO
7. Relative considerations: soaking exposure, cerebral edema on CT scan
8417: “8417_c016” — 2007/9/11 — 12:15 — page 359 — #19

greater than 25% to 40% COHb, or the patient with recurrent or persistent symptoms
despite 6 hours of therapy with NBO. If HBOT is indicated, treatment within 6 h
is desirable.210 Patients should undergo a full neuropsychiatric evaluation prior to
discharge. Close follow-up is necessary with repeat neuropsychiatric examinations
at 6 weeks, 6 months, and 12 months.
16.5.2 PREGNANCY
The effects of CO on the fetus has been extensively reviewed by Penney.211 The
fetus is particularly vulnerable to the effects of CO, which readily crosses the pla-
centa, and, in animal models, is even more tightly bound to fetal hemoglobin than
adult hemoglobin. The fetus also reaches higher peak COHb levels than does the
mother. Fetuses that survive a significant CO poisoning may be left with limb mal-
formation, hypotonia, areflexia, persistent seizures, mental and motor disability, and
microcephaly.212,213 The only prospective, multicenter study of acute CO poisoning
in pregnancy recently reported adverse outcomes in 60% of children whose mothers
suffered severe CO toxicity. Of those babies born to mothers with mild to mod-
erate CO exposure, normal physical exams and neurobehavioral development were
reported.214 Since CO elimination from the fetus is prolonged (7–10 h), it is generally
accepted that HBOT is indicated at lower maternal COHb levels than would be acted
upon in the nongravid patient. In addition, surface oxygen therapy should be exten-
ded to four to five times the normal duration. Although controversial, HBO has been
reported to be safe in the pregnancy,215 despite theoretical dangers of fetal hyper-
oxia in animal models.216−218 (Such animal models exceeded the time and pressure
routinely used in clinical therapy). A recent report of 44 women undergoing HBOT
during pregnancy for CO exposure suggests that it is safe and should be considered,
although miscarriages did occur, and six patients were lost to follow-up.219 It should
be noted that HBO was implicated in the induction of labor in one pregnant patient,
the pregnancy however, was near term when the CO exposure occurred.220 Proposed
indications for HBOT in the pregnant patient are listed in Table 16.2, although these
are not well-studied.
16.5.3 CHILDREN
Pediatric CO poisoning has been reviewed by White.221 Younger children have tra-
ditionally been viewed to be more susceptible to CO poisoning on the basis of more
TABLE 16.2
Proposed Indications for Hyperbaric Oxygen Therapy in Pregnancy
1. Maternal COHb level > 10–15% at any time during the exposure
2. Any neurological signs or symptoms other than headache
3. Evidence of fetal distress (fetal tachycardia, decreased beat-to-beat variability, late decelerations)
4. If maternal neurologic symptoms or fetal distress persist 12 h after initial therapy, additional HBO
treatments may be necessary
8417: “8417_c016” — 2007/9/11 — 12:15 — page 360 — #20

rapid metabolic rate and higher oxygen demands. They may also have more atypical
presentations relative to adults. Both persistent and delayed sequelae are described in
the children, however, formal neuropsychiatric testing is generally difficult and not
well-documented in such case series.222 The use of HBOT in the treatment of pediat-
ric CO poisoning is controversial, and recommendations vary, even among pediatric
toxicology experts. Until further knowledge and experience is gained in this area,
children are likely be treated as aggressively as adults who are CO-poisoned.
16.6 MANAGEMENT OF THE SEQUELAE OF CARBON

MONOXIDE POISONING
Delayed sequelae from CO poisoning is devastating and occurs in 10–43% of per-
sons recovering from the acute exposure. Parkinsonism, the most dramatic long-term
neurologic complication has a grim prognosis. Fortunately, most cases of DNS asso-
ciated with mild CO poisoning resolve within two months.119 Unfortunately, only
one-third of severely CO-poisoned patients surviving to HBOT have resolution of
DNS. Conventional therapy of DNS-related parkinsonism with l-dopa has been dis-
appointing. Use of another centrally acting dopaminergic agonist, bromocriptine
has been reported. Nine patients (mean age 61 years) suffering from CO-induced
parkinsonism who were given bromocriptine (5–30 mg/day), displayed improve-
ment in Webster’s scores while under treatment.223 Clearly no definitive conclusions
regarding bromocriptine therapy can be made on the basis of small study, but per-
haps it will provide a basis for future investigations. Treatment with ziprasidone,
a newer atypical antipsychotic agent resulted in improved neuropsychiatric symp-
toms and cognitive function in a patients with CO-induced severe DNS refractory to
HBOT, bromocriptine, conventional antipsychotics, and other atypical antipsychot-
ics, risperidone and quetiapine.224 Similar success was noted using aripiprazole in
managing CO-induced psychotic symptoms and parkinsonism.225
One report involving hyperpyrexia and muscle rigidity as sequelae of CO pois-
oning was treated successfully with a prolonged course of dantrolene sodium, a
peripheral skeletal muscle relaxant.226 Given that the patient manifested signs char-
acteristic of severe hypoxic/ischemic encephalopathy, this therapy was symptomatic
for that condition, and not specific to CO poisoning. Dantrolene would not likely
provide any benefit beyond other safer sedatives, such as benzodiazepines, in treating
such complications.
Another common sequelae from CO poisoning is memory impairment. Recent
work by Hiramatsu et al.227 focused on treating delayed amnesia in mice. The
investigators treated mice with documented amnesia 5 days postexposure with dyn-
orphin A (1–13). They found this treatment regimen to be effective in reversing
CO-induced memory impairment. Nor-binaltrophimine (kappa opioid receptor ant-
agonist) blocked the effect of dynorphin A (1–13), suggesting that kappa receptors
mediated the reversal of impairment in memory seen from CO poisoning in this
animal model.227 The authors reported similar findings with a second kappa receptor
agonist, U-50488H, which appeared to additionally activate the cholinergic neuronal
system, known also to play an important role in cognitive deficits associated with
8417: “8417_c016” — 2007/9/11 — 12:15 — page 361 — #21

other conditions such as aging and neurodegenerative diseases.228 These agents may
hold promise for the future in treating the persistent or delayed detrimental effects of
CO on acquisition and consolidation of memory.
Delayed, sometimes repetitive, HBOT has been advocated by some to improve
the long-term neurologic deficits from CO toxicity, even if instituted weeks after the
initial CO exposure.229−235 Such practice which is advocated by several treatment
centers in the United States lacks validation by well-controlled, blinded clinical stud-
ies that utilize neuropsychiatric testing data. Interestingly, behavioral treatment has
been successful when guided by formal neuropsychiatric testing. In certain patients,
indirect measures of learning are better predictors of treatment efficacy.236
Patients who present to health care facilities late or have suffered recurrent or
chronic lower level CO exposures pose particular treatment challenges to the clinician.
Any proposed therapeutic approach to such patients should be considered carefully
given the fact that no definitive clinical or animal studies in this area exist.
16.7 CONCLUSION
Much remains to be learned about CO, including the mechanisms of toxicity, pre-
dictors of outcome after poisoning, and best treatments. Further research is needed
to formulate clear-cut clinical indications for the use of potentially neuroprotective
agents (i.e., insulin, sulfhydryl donors, allopurinol, ketamine, brain-derived peptides,
kappa receptor agonists). Given that multiple pathways are involved in the ultimate
neuronal injury, how to best use these agents, perhaps synergistically as a “cock-
tail” approach, remains to be seen. Other areas that deserve further study are the
clarification of risk factors for adverse fetal outcome following CO exposure during
pregnancy, the delineation of the true incidence of PNS and DNS in children, and
the best test for indicators of risk for these sequelae. The future for HBOT in CO
poisoning remains to be seen. Given the disparate results from randomized clinical
trials using HBOT, we are compelled to continue to carefully select patients for this
therapy, and to promote further study to delineate subpopulations such as children
and pregnant women who may potentially benefit.
References
1. Cobb, N., Etzel, R.A. Unintentional CO-related deaths in the United States, 1979
through 1988, JAMA, 266, 659,1995.
2. Runciman, W.W., and Gorman, D.F. Carbon monoxide poisoning: from old dogma
to new uncertainties, Med. J. Aust., 158, 439, 1993.
3. Hampson, N.B. Trends in the incidence of carbon monoxide poisoning in the United
States, Am. J. Emerg. Med., 23, 838, 2005.
three-year follow up, Br. Med. J., 1, 279, 1983.
5. Lugaresi, A., Montagna, P., and Morreale, A. "Psychic akinesia" following carbon
monoxide poisoning, Eur. Neurol., 30, 167, 1990.
6. Klawans, H.L., Stein, B.W., and Tanner, C.M. A pure parkinsonian syndrome
following acute carbon monoxide intoxication, Arch. Neurol., 39, 302, 1982.
8417: “8417_c016” — 2007/9/11 — 12:15 — page 362 — #22

7. Werner, B., Back, W., Akerblom, H., Barr, P.O. Two cases of acute carbon monoxide
poisoning with delayed neurologic sequelae after a ‘free’ interval, Clin. Toxicol., 23,
249, 1985.
8. Hart, I.K., Kennedy, P.G., Adams, J.H., et al. Neurological manifestations of carbon
monoxide poisoning, Postgrad. Med. J., 64, 213, 1988.
9. Sauk, G.M., Watson, C.P., Tebragge, K., et al. Delayed encephalopathy following
carbon intoxication, Can. J. Neurol. Sci., 8, 77, 1981.
10. Barret, L., Danel, V., and Faure, J. Carbon monoxide poisoning, a diagnosis frequently
overlooked, Clin. Toxicol., 23, 309, 1985.
11. Balzan, M.A., Agius, G., and Debono A.G. Carbon monoxide poisoning, easy to treat
but difficult to recognise, Post grad. Med. J., 72,470,1996.
12. Sokal, J.A., Majka, J., and Palus, J. The content of carbon monoxide in the tissues of
rats intoxicated with carbon monoxide in various conditions of acute exposure, Arch
Toxicol, 56, 106, 1984.
13. Lasater, S.R. Carbon monoxide poisoning, Can. Med. Assoc. J., 134, 991, 1986.
14. Olson, K.R. Carbon monoxide poisoning: mechanisms, presentation, and controver-
sies in management, J. Emerg. Med., 1, 233, 1984.
15. Ikeda, T., Kondo, T., Mogami, H., Miura, T., Mitomo, M., Shimazaki, S., and
Sugimoto, T. Computerized tomography in cases of acute carbon monoxide poisoning,
Med. J. Osaka Univ., 29, 253, 1978.
16. Penney, D.G. Hyperglycemia exacerbates brain damage in acute severe carbon
monoxide poisoning, Med. Hypotheses, 27, 241, 1988.
17. Bogusz, M., Cholewa, L., Pach, J., and Mlodkowska, K. A comparison of two types
of acute carbon monoxide poisoning, Arch. Toxicol., 33, 141, 1975.
18. Linas, A.J., and Limousin, S. Bul. Mem. Soc. Ther., 2, 32, 1868.
19. Haldane, J. The relation of the action of carbonic oxide to oxygen tension,
J. Physiology, 18, 201, 1895.
20. Killick, E.M., and Marchant, J.V. The effect of barbiturates on the resuscitation of
dogs from severe acute CO poisoning, J. Physiol., 180, 80, 1965.
21. Killick, E.M., and Marchant, J.V. Resuscitation of dogs from severe, acute carbon
monoxide poisoning, J. Physiol., 147, 274, 1959.
22. Schwerma, H., Ivy A.C., Friedman, H., and Brosse, E.L. A study of resuscita-
tion from the juxtalethal effects of exposure to carbon monoxide, Occup. Med., 5,
24, 1948.
23. Jain, K.K. Carbon Monoxide Poisoning, Warren H. Green, Inc., St Louis, 1990, 140.
24. Gros, J.F., and Leandri, P. Traitment de l’intoxication oxycarbonée par le cytochrome,
Presse Méd, 64, 1356, 1956.
25. Bentolila, P., Tran, G., and Olive, G. Essai de traitment de l’intoxication oxycarbonée
par perfusion de solutions diluées de paroxyde d’hysrogene, resultats obtnus chez le
lapin, Thérapie, 28, 1043, 1973.
26. Koza, F. Die kohlenmonoxidevergiftungund deren neurartige therapie mit
bestrahlung, Med. Klinik, 26, 422, 1930.
27. Estler, W. Experimentelle untersuchungen uber die anwendung der bestrahlung mit
ultravoilettem licht zur behandlung der kohlenoxidvergiftung, Arch. Hyg. Bakt., 115,
152, 1935.
28. Amyes, E.W., Ray, J.W., and Brockman, N.W. Carbon monoxide anoxia; intravenous
administration of procaine hydrochloride in the treatment of acute and chronic effects,
JAMA, 142, 1054, 1950.
29. Evans, D.E., Catron, P.W., McDermott, J.J., et al. Effect of lidocaine after
experimental cerebral ischemia induced by air embolism, J. Neurosurg., 70, 97, 1989.
8417: “8417_c016” — 2007/9/11 — 12:15 — page 363 — #23

30. Ojeda-Sahagun, J.L. Accion protectora del dipiridamol frente a la intoxicacion por
gas del alumbrado, Rev. Esp. Fisiol., 27, 305, 1971.
31. Agostini, J.C., Ramirez, R.G., Albert, S.N., Goldbaum, L.R., Absolon, K.B. Suc-
cessful reversal of lethal carbon monoxide intoxication by total body asanguineous
hypothermic infusion, Surgery, 75, 213, 1974.
32. Yee, L.M., and Brandon, G.K. Successful reversal of presumed carbon monoxide-
induced semicoma, Aviat. Space Envir. Md., 54, 641, 1983.
33. Shumate, M.J. Carbon monoxide poisoning, Chest, 107, 1474, 1995.
34. Geyer, R.P., Haggard, H.W. Review of perfluorochemical-type blood substitutes, In
Proc. 10th Int. Cong. Nutrition-Symposium on Perfluorochemical Artificial Blood,
Igakushobe, Osaka Japan, 1976, 3.
35. Levine, E.M., and Tremper, K.K. Perfluorochemical emulsions: potential clinical uses
and new developments, Int. Anesthesiol. Clin., 23, 211, 1985.
36. Uchiyama, T., and Harafuji, K. Life-saving effect of pyridoxalated hemoglobin-
polyoxyethylene conjugate on carbon monoxide intoxication of rabbits, Artif.
Organs,18, 576, 1994.
37. Mosso, A. La mort apparente du coer secours l’empoisonneement par l’oxyd de
carbonee, Arch. Ital. Biol., 35, 75, 1901.
38. Smith, G., and Sharp, G.R. Treatment of carbon monoxide poisoning with oxygen
under pressure, Lancet, ii, 905, 1960.
39. Sluitjer, M.E. The treatment of carbon monoxide poisoning by administration of
oxygen at high pressure, P. Roy. Soc. Med., 56, 123, 1963.
40. Hampson, N.B., and Little, C.E. Hyperbaric treatment of patients with carbon
monoxide poisoning in the United States, Undersea Hyper. Med., 32, 21, 2005.
41. Dinman, B., Eaton, J., and Brewer, G. Effects of carbon monoxide on DPG
concentrations in the erythrocyte, Ann. NY Acad. Sci., 1974, 246, 1970.
42. Klasner, A.E., Smith, S.R., Thompson, M.W., and Scalzo, A.J. Carbon monoxide
mass exposure in a pediatric population, Acad. Emerg. Med., 5, 992, 1998.
43. Decker, J., and Wiley J. F. Carboxyhemoglobin elimination half-life in children,
presented at the North American Congress of Clinical Toxicology, Salt Lake City,
Utah, September, 1994.
44. Okeda R., Funata, N., Song, S.J., et al. Comparative study of selective cerebral lesions
in carbon monoxide poisoning and nitrogen hypoxia in cats, Acta. Neuropathol., 56,
265, 1982.
45. Haldane, J.B.S. Carbon monoxide as a tissue poison, Biochem. J., 21, 1086, 1927.
46. Brabkin, D.L., Lewey, F.H., Bellet, S., et al. The effect of replacement of nor-
mal blood by erythrocytes saturated with carbon monoxide, Am. J. Med. Sci., 205,
755, 1943.
47. Goldbaum, L.R. Studies on the relationship between carboxyhemoglobin concentra-
tion and toxicity, Aviat. Space Envir. Med., 48, 969, 1977.
48. Goldbaum, L.R. Mechanism of the toxic action of carbon monoxide, Ann. Clin. Lab.
Sci., 6, 372, 1976.
49. Goldbaum, L.R., Orellano, T., and Dergal, E. What is the mechanism of Carbon
monoxide toxicity? Aviat. Space Envir. Med., 46, 1289, 1975
50. Brown, S., and Piantadosi, C. In vivo binding of CO to cytochrome oxidase in rat
brain, J. Appl. Physiol., 68, 604, 1990.
51. Brown, S., and Piantadosi, C.A. Recovery of energy metabolism in rat brain after CO
hypoxia, J. Clin. Invest., 89, 666, 1992.
52. Coburn, R.F., and Forman, H.J. Carbon monoxide toxicity, In: Fishman, A.P., Farki,
L.E., and Geiger S.R. eds., Handbook of Physiology, The Respiratory System, Sect 3,
Vol. IV, Chapt. 21, 439, Baltimore: Williams & Wilkins, 1987.
8417: “8417_c016” — 2007/9/11 — 12:15 — page 364 — #24

53. Smithline, H.A., Rivers, E.P., Chiulli, D.A., Rady, M.Y., Baltarowich, L.L., Blake,
H.C., et al. Systemic hemodynamic and oxygen transport response to graded carbon
monoxide poisoning. Ann. Emerg. Med., 5, 203, 1993.
54. Ward, K.R., Lin, J., Zhang, L., and Chopp, M. Carbon monoxide poisoning does
not cause neuronal heat shock protein 72 or 32 expression, Acad. Emerg. Med., 5,
487, 1998.
55. Dallman, P., and Schwartz, H. Distribution of cytochrome C and myoglobin in rats
with dietary iron deficiency, Pediatrics, 35, 677, 1965.
56. Webber, D.S., Lopez, I., Korsak, R.A., Hirota, S., Acuna, D., and Edmond, J. Lim-
iting iron availability confers neuroprotection from chronic mild carbon monoxide
exposure in the developing auditory system of the rat, J. Neurosci. Res., 80, 620,
2005.
57. Wittenberg, B.A., and Wittenberg, J.B. Effects of carbon monoxide on isolated heart
muscle cells, Research Report-Health Effects Institute, 62, 1, 1993.
58. Aslan, S., Uzkeser, M., Seven, B., Gundogdu, F., Acemoglu, H., Akasakal, E., and
Varoglu, E. The evaluation of myocardial damage in 83 young adults with carbon
monoxide poisoning in the East Anatolia region in Turkey, Hum. Exp. Toxicol., 2006,
25, 439, 2006.
59. Satran, D., Henry, C.R., Adkinson, C., Nicholson, C.I., Bracha, Y., and Henry, T.D.
Cardiovascular manifestations of moderate to severe carbon monoxide poisoning,
J. Am. Coll. Cardiol., 45, 1513, 2005.
60. Henry, C.R., Satra, D., Lindgren, B., Adkinson, C., Nicholson, C.I., and Henry,
T.D. Myocardial injury and long-term mortality following moderate to severe carbon
monoxide poisoning, JAMA, 295, 398, 2006.
61. Yucel, M., Avsarogullari, L., Durukan, P., Akdur, O., Ozkan, S., Sozuer, E.M.,
Muhtaroglu, S., and Ikizceli, I. The relationship between the severity of myocar-
dial injury and the levels of brain natriuretic peptide and cardiac troponin-I in
rats experimentally exposed to carbon monoxide, Eur. J. Emerg. Med., 13, A11
(abstract), 2006.
62. Aurora, T., Chung, W., Dunne, R., Martin, G., Ward, K., Rivers, E., Knoblich, B.,
Nguyen, H.B., and Tomlanovich M.C. Occult myocardial injury in severe carbon
monoxide poisoning, Acad. Emerg. Med., 6, 394 (abstract), 1999.
63. Thom, S.R. Carbon monoxide-mediated brain lipid peroxidation in the rat, J. Appl.
Physiol., 68, 997, 1990.
64. Thom, S.R. Leukocytes in carbon monoxide-mediated brain oxidative injury, Toxicol.
Appl. Pharmacol., 123, 234, 1993.
65. Thom, S.R. Dehydrogenase conversion to oxidase and lipid peroxidation in brain after
carbon monoxide poisoning, J. Appl. Physiol. ,73, 1584, 1992.
66. Thom, S.R., Garner S., and Fisher D. Vascular oxidative stress from carbon monoxide
exposure, Undersea Hyperbar. Med., 25 (Suppl), 47, 1998.
67. Thom, S.R., Xu, Y.A., and Ischiropoulos, H. Vascular endothelial cells generate
peroxynitrite in response to carbon monoxide exposure, Chem. Res. Toxicol., 10,
1023, 1997.
68. Ischiropoulos, H., Beers, M.F., Ohnishi, S.T., Fisher, D., Garner, S.E., and Thom,
S.R. Nitric oxide production and perivascular nitration in brain after carbon monoxide
poisoning in the rat. J. Clin. Invest., 97, 2260, 1996.
69. Hall, E.D., Andrus, P.K., Althaus, J.S., et al. Hydroxyl radical production and lipid per-
oxidation parallels selective post-ischemic vulnerability in gerbil brain, J. Neurosci.
Res. 1993, 34, 107.
70. Zhang, J., and Piantadosi, C.A. Mitochondrial oxidative stress after carbon monoxide
hypoxia in the rat, J. Clin. Invest., 90, 1193, 1992.
8417: “8417_c016” — 2007/9/11 — 12:15 — page 365 — #25

71. Thom, S.R., Bhopale, V.M., Han, S-T., Clark, J.M., and Hardy, K.R. Intravascular
neutrophil activation due to carbon monoxide poisoning, Am. J. Respir. Crit. Care
Med., 2006, 174, 1239–1248.
72. Thom, S.R., Bhopale, V.M., Fisher, D., Zhang, J., and Gimotty, P. Delayed neuro-
pathology after carbon monoxide poisoning is immune-mediated, PNAS, 101, 13660,
2004.
73. Thom, S.R., Fisher, D., Zhang, J., et al. Neuronal nitric oxide synthase and
N-methyl-D-aspartate neurons in experimental carbon monoxide poisoning, Toxicol.
Appl. Pharmacol., 1914, 280, 2004.
74. Piantadosi, C.A. Toxicity of carbon monoxide, hemoglobin vs. histotoxic mechan-
isms, In Carbon Monoxide, Penney, D.G., ed., Boca Raton, FL, CRC Press, Inc, 1996,
Chapt. 8, pp. 163–186.
75. Amara, S.G. Neurotransmitter transporters, Nature, 360, 420, 1992.
76. Hernandez-Viadel, M., Castoldi, A.F., Coccini, T., Manzo, L., Erceg, S., and Felip,
V. In vivo exposure to carbon monoxide causes delayed impairment of activation
of soluble guanylate cyclase by nitric oxide in rat brain cortex and cerebellum,
J. Neurochem., 89, 1157, 2004.
77. Meilin, S., Rogatsky, G.G., Thom, S.R., Zarchin, N., Guggenheimer-Furman, E., and
Mayevsky, A. Effects of carbon monoxide on the brain may be mediated by nitric
oxide, J. Appl. Physiol., 81, 1078, 1996.
78. Shinomura, T., Nakao, S., and Mori, K. Reduction of depolarization-induced glutam-
ate release by hemoxygenase inhibitor, possible role of carbon monoxide in synaptic
transmission, Neurosci. Lett., 166, 131, 1994.
79. Gorman, D., Lin, H.Y., and Williams, C. Blockade of haemoxygenase and nitric
oxide synthetase causes cortical dysfunction in sheep exposed to carbon monoxide,
Toxicology, 209, 237, 2005.
80. Zhang, J., and Piantadosi, C.A. Prevention of H2 O2 generation by monoamine oxidase
protects against CNS O2 toxicity, J. Appl. Physiol., 71, 1057, 1991.
81. Piantadosi, C.A., Tatro, L., and Zhang, J. Hydroxyl radical production in the brain
after CO hypoxia in rats, Free Radical Bio. Med., 18, 603, 1995.
82. Pianttadosi, C.A., Zhang, J., and Levin, E.D., et al. Apoptosis and delayed neuronal
damage after carbon monoxide poisoning in the rat, Exp. Neurol., 147, 103, 1997.
83. Ward, K.R., Junmin, L, Zhang, L, and Chopp, M. Moderately severe carbon monoxide
poisoning does not cause neuronal apoptosis, Acad. Emerg. Med., 5, 487, 1998,
abstract.
84. Thom, S.R., Fisher, D., Xu, Y.A., et al. Adaptive responses and apoptosis in
endothelial cells exposed to carbon monoxide, P. Natl. Acad. Sci. USA, 97,
1305, 2000.
85. Van Der Hoeven, J.G., Compier, E.A., and Meinders, A.E. Goede resultaten van
behandeling met 100% zuurstop wegens acute koolmonoxide-intoxicatie; voorlopig
geen indicatie, voor hyperbare zuurstoftoediening, Ned. Tijdschr. Genees., 137,
864, 1993.
86. Weaver, L.K., Hopkins, R.O., and Larson-Lohr, V. Neuropsychologic and functional
recovery from severe carbon monoxide poisoning without hyperbaric oxygen therapy,
Ann. Emerg. Med., 27, 736, 1996.
87. Meert, K.L., Heidemann, S.M., and Sarnaik, A.P. Outcome of children with carbon
monoxide poisoning treated with normobaric oxygen, J. of Trauma, 44, 149, 1998.
88. Tomaszewski, C., Rudy, J., Wathen, J., Brent, J., Rosenberg, N., and Kulig, K.
Prevention of neurologic sequelae from carbon monoxide by hyperbaric oxygen in
rats, Ann. Emerg. Med., 21, 631–632, 1992.
8417: “8417_c016” — 2007/9/11 — 12:15 — page 366 — #26

89. Thom, S. Functional inhibition of neutrophil beta 2 integrins by HBO in CO-mediated

brain injury, Toxicol. Appl. Pharmacol., 123, 248–256, 1993.
90. Thom, S.R. Mendiguren, I., Hardy, K., et al. Inhibition of human neutrophil beta2-
integrin-dependent adherence by hyperbaric oxygen, Am. J. Physiol., 272, C770,
1997.
91. Schnittger, V., Rosendahl, K., Lind, F., Pamblad, J. Effects of carbon monoxide
poisoning on neutrophil response in patients treated with hyperbaric oxygen, J. Invest.
Med., 52, 523, 2004.
92. Brown, S.D., and Piantadosi, C.A. Reversal of carbon monoxide-cytochrome c oxi-
dase binding by hyperbaric oxygen in vivo, In Adv. Exp. Med. Biol., K. Rakusan, G.D.
Biro, T.K. Goldstick, Eds, Plenum Press, New York, Vol. 248, 1989, pp. 747–754.
93. Brvar, M., Finderle, Z., Suput, D., and Bunc, M. S100B protein in conscious carbon
monoxide-poisoned rats treated with normobaric or hyperbaric oxygen, Crit. Care
Med., 34, 2228, 2006.
94. Thom, S.R., Antagonism of carbon monoxide-mediated brain lipid peroxidation by
hyperbaric oxygen, Toxicol. Appl. Pharmacol., 105, 340, 1990.
95. Sukoff, M.H., and Ragatz, R.E. Hyperbaric oxygen for the treatment of acute cerebral
oedema, Neurosurgery, 10, 29, 1982.
96. Jiang, J., and Tyssebotn, I. Cerebrospinal fluid pressure changes after acute carbon
monoxide poisoning and therapeutic effects of normobaric and hyperbaric oxygen in
conscious rats, Undersea Hyperbar. Med., 24, 245, 1997.
97. Roos, J.A., Jackson-Friedman, C., and Lyden, P. Effects of hyperbaric oxygen on
neurologic outcome for cerebral ischemia in rats, Acad. Emerg. Med., 5, 18, 1998.
98. Kokame, G.M., and Shuler, S.E. Carbon monoxide poisoning, Arch. Surg., 96, 211,
1968.
99. Myers, R.A.M., Snyder, S.K., Linberg, and S., Cowley, R.A. Value of hyperbaric
oxygen in suspected carbon monoxide poisoning, JAMA, 246, 2478, 1981.
100. Winter, A., and Shatin, L. Hyperbaric oxygen in reversing carbon monoxide coma,
New York State J. Med., 1, 880, 1970.
101. Thomson, L.F., Mardel, S.N., Jack, A., and Shields, T.G. Management of the moribund
carbon monoxide victim, Arch. Emerg. Med, 9, 208, 1992.
102. Welsh, F., Matos, L., and DeTreville, R.T.P. Medical hyperbaric oxygen therapy: 22
cases, Aviat., Space Envir. Md. , 51, 611, 1980.
103. Gibson, A.J., Davis, F.M., and Ewer, T. Delayed hyperbaric oxygen therapy for carbon
monoxide intoxication- two case reports, New Zeal. Med. J., 104, 64, 1991.
104. Dean, B.S., Verdile, V.P., and Krenzelok, E.P. Coma reversal with cerebral dysfunc-
tion recovery after repetitive hyperbaric oxygen therapy for severe carbon monoxide
poisoning, Am. J. Emerg. Med., 11, 61, 1993.
105. Ersanli, D., Yildiz, S., Togrol, E., Ay, H., and Qyrdedi, T. Visual loss as a late com-
plication of carbon monoxide poisoning and its successful treatment with hyperbaric
oxygen therapy, Swiss Med. Wkly., 134, 650, 2004.
106. Roche, L., Bertoye, A., Vincent, P., et al. Comparison de deux groupes de vingt
intoxications oxycarbonees traitees par oxygene normobare et hyperbare, Lyon Med.,
49, 1483, 1968.
107. Mathieu, D., Nolf, M., Durocher, A., Saulnier, F., Frimat, P., Furon, D., and Wattel, F.
Acute carbon monoxide poisoning risk of late sequelae and treatment by hyperbaric
oxygen, Clin. Toxicol., 23, 315, 1985.
108. Gorman, D.F., Clayton, D., Gilligan, J.E., and Webb, R.K. A longitudienal study of
100 consecutive admissions for carbon monoxide poisoning to the Royal Adelaide
hospital, Anaesth. Intens. Care, 20, 311, 1992.
8417: “8417_c016” — 2007/9/11 — 12:15 — page 367 — #27

109. Myers, R.A.M., Snyder, S.K., and Emhoff, T.A. Subacute sequelae of carbon
monoxide poisoning, Ann. Emerg. Med., 14, 1163, 1985.
110. Hsu, L.H., and Wang, J.H. Treatment of carbon monoxide poisoning with hyperbaric
oxygen, Chin. Med. J-Peking, 58, 407, 1996.
111. Gozal, D., Ziser, A., Shupak, A., and Melamed, Y. Accidental carbon monoxide
poisoning, Clin. Pediatr. 24, 132, 1985.
112. Norkool, D.M., and Kirkpatrick, J.N. Treatment of acute carbon monoxide poisoning
with hyperbaric oxygen: a review of 115 cases, Ann. Emerg. Med., 14, 1168, 1985.
113. Raphael, J.,C., Elkharrat, D., Jars-Guincestre, M.C., et al. Trial of normobaric
and hyperbaric oxygen for acute carbon monoxide intoxication, Lancet, 1, 414,
1989.
114. Weaver, L.K. Carbon monoxide poisoning, Crit. Care Clin., 15, 297, 1999.
115. Ducasse, J.L., Izard, P.H., Celsis, P., et al. Moderate carbon monoxide poisoning:
hyperbaric or normobaric oxygenation? Proceedings of the Joint Meeting of the 2nd
European Conference and the 2nd Swiss Symposium on Hyperbaric Medicine, Basel,
Switzerland, 1990, pp. 289–297.
116. Olson, K.R., and Seger, D. Hyperbaric oxygen for carbon monoxide poisoning: does
it really work, Ann. Emerg. Med., 25, 535, 1995.
117. Weaver, L.K., Hopkins, R.O., and Larson-Lohr, V. Hyperbaric oxygen and carbon
monoxide poisoning, Ann. Emerg. Med., 2693, 390, 1995.
118. Scheinkestel, C.D., Bailey, M., Myles, P.S., Jones, K., Cooper, D.J., Millar, I.L., and
Tuxen, D.V. Hyperbaric or normobaric oxygen for acute carbon monoxide poisoning:
a randomised controlled clinical trial, MJA, 170, 203, 1999.
119. Weaver, L.K., Hopkins, R.O., Howe, S., Larson-Lohr, V., and Churchill, S. Outcome
at 6 and 12 months following acute CO poisoning, Undersea Hyperbar. Med., 23S,
9, 1996.
120. Weaver, L.K., Hopkins, R.O., Chan, K.J., et al. Hyperbaric oxygen for acute carbon
monoxide poisoning, N. Engl. J. Med., 347, 1057, 2002.
121. Mathieu, D., Wattel, F., Mathiew-Nolf, M., et al. Randomized prospective study
comparing the effect of HBO versus 12 hours NBO in non-comatose CO poisoned
patients: results of the interim analysis, Undersea Hyperbar. Med., 23 (Suppl.), 7,
1996.
122. Raphael, J.C., Chevret, S., Driheme, A., and Annane, D. Managing carbon monox-
ide poisoning with hyperbaric oxygen (abstract), J. Toxicol. Clin. Toxicol., 42, 455,
2004.
123. Gorman, D.F., Clayton., D., and Gilligan, J.E. A longitudinal study of 100 consecutive
admissions for carbon monoxide poisoning to the Royal Adelaide Hospital, Anesth.
Intens. Care, 20, 311, 1992.
124. Turner, M., Esaw, M., and Clark, R.J. Carbon monoxide poisoning treated with hyper-
baric oxygen: metabolic acidosis as a predictor of treatment requirements, J. Accid.
Emerg. Med. 16, 96, 1999.
125. Tibbles, P.M., and Edelsberg, J.S. Hyperbaric oxygen therapy, N. Engl. J. Med., 334,
1642, 1996.
126. McNulty, J.A., Maher, B.A., Chu, M., and Sitnikova, T. Relationship of short-term
verbal memory to the need for hyperbaric oxygen treatment after carbon monoxide
poisoning, Neuropsy., Neuropsy. Be., 10, 174, 1997.
127. Murata, M., Suzuki, M., Hasegawa,Y., Nohara, S., and Kurachi, M. Improvement of
occipital alpha activity by repetitive hyperbaric oxygen therapy in patients with car-
bon monoxide poisoning: a possible indicator for treatment efficacy, J. Neurological
Sciences, 235, 69, 2005.
8417: “8417_c016” — 2007/9/11 — 12:15 — page 368 — #28

128. Hampson, N.B., Simonson, S.G., and Kramer, C.C. Central nervous system oxygen
toxicity udring hyperbaric treatment of patients with carbon monoxide poisoning,
Undersea Hyperbar Med., 23, 215, 1996.
129. Weaver, L.K., Hopkins, R.O., Valentine, K.J., and Churchill, S. Persistent and
delayed carbon monoxide cognitive sequelae, Abstract presented at the Under-
sea and Hyperbaric Medical Society Annual Scientific Meeting, June, 2006, Las
Vegas, NV.
130. Scheinkestel, C.D., Jones, K., Myles, P.S., Cooper, J., Miller, I.L., and Tuxen, D.V.
Where to now with carbon monoxide poisoning? Emerg. Med. Australas., 16, 151,
2004.
131. Weaver, L.K., Hopkins, R.O., Chan, K.J., Thomas, F., Churchill, S.K., Elliott, C.G.,
and Morris, A. Carbon monoxide research group, LDS Hospital, Utah in reply to
Scheinkestel et al. and Emerson: The role of hyperbaric oxygen in carbon monoxide
poisoning, Emerg. Med. Australas., 16, 394, 2004.
132. Emerson, G. The dilemma of managing carbon monoxide poisoning, Emerg. Med.
Australas., 16, 101, 2004.
133. Juurlink, D.N., Buckley, N.A., Stanbrook, M.B., Isbister, G.K., Bennett, M., and
McGuigan, M.A. Hyperbaric oxygen for carbon monoxide poisoning (review),
Cochrane Db. Syst. Rev. 2005, Issue 1, Art. No.:CD002041.DoI:10.1002/14651858.
CD002041. pub 2.
134. Buckley, N.A., Isbister, G.K., and Juurlink, D.N. Hyperbaric oxygen for carbon
monoxide poisoning: evidence versus opinion, Toxicol. Rev., 24, 159, 2005.
135. Thom, S.R. Hyperbaric oxygen therapy for carbon monoxide poisoning: is it time to
end the debates, Toxicol. Rev., 24, 157, 2005.
136. Bentur, Y. Hyperbaric oxygen for carbon monoxide poisoning, Toxicol. Rev., 24,
153, 2005.
137. Olson, K.R. Hyperbaric or normobaric oxygen? Toxicol. Rev., 24, 151, 2005.
138. Henry, J.A. Hyperbaric oxygen therapy for carbon monoxide poisoning, To treat or
not to treat, that is the question, Toxicol. Rev., 24, 149, 2005.
139. Marklund, S.L., Westman, G., Lundgren, E., and Roos, G. CuZn superoxide dis-
mutase, Mn superoxide dismutase, catalase and glutathione peroxidase in normal and
neoplastic human cell-lines and normal human tissues, Cancer Res., 42, 1955–1961,
1982.
140. Toledo-Pereyra, L.H., Simmons, R.L., and Najarian, J.S. Effect of allopurinol on the
preservation of ischemic kidneys perfused with plasma or plasma substitutes, Ann.
Surg., 180, 780–782, 1974.
141. Stewart, J.R., Blackwell, W.H., Crute, S.L., Loughlin, V., Hess, M.L., and Greenfield,
L.J. Prevention of myocardial ischemia / reperfusion injury with oxygen-free radical
scavengers, Surg. Forum, 33, 317–320, 1982.
142. Fechter, L.D., Liu, Y., and Pearce, T.A. Cochlear protection from carbon monox-
ide exposure by free radical blockes in the guinea pig, Toxicol. Appl. Pharm., 142,
47, 1997.
143. Howard, R.J., Blake, D.R., Pall, H., Williams, A., and Green, I.D. Allopurinol/n-
acetylcysteine for carbon monoxide poisoning, Lancet, 2, 628–629, 1987.
144. Gilmer, B., Tomaszewski, C., and Watts, J.A. The neuroprotective effects of dimethyl
sulfoxide on memory following actue carbon monoxide poisoning in mice, Ann.
Emerg. Med., 35, S69, 2000.
145. Penney, D.G., Helfman, C.C., Hull, J.C., Dunbar, J.C., and Verman, K. Elevated
blood glucose is associated with poor outcome in the carbon monoxide poisoned rat,
Toxicol. Lett., 54, 287, 1990.
8417: “8417_c016” — 2007/9/11 — 12:15 — page 369 — #29

146. Penney, D.G. Hyperglycemia exacerbates brain damage in acute–severe carbon

monoxide poisoning, Med. Hypoth., 27, 241–244, 1988.
147. Leikin, J.B., Goldenberg, R.R., Edwards, D., and Zell-Kantor, M. Metabolic
predictors of carbon monoxide poisoning, Vet. Hum. Toxicol., 30, 40–42, 1988.
148. Pulsinelli, W., Waldman, S., Sigsbee, B., Rawlinson, D., Schere, P., and Plum, F.
Experimental hyperglycemia and diabetes mellitus worsen stroke outcome, T. Am.
Neurol. Assoc., 105, 21–24, 1980.
149. White, S.R., and Penney, D.G. Effects of insulin and glucose treatment on neurologic
outcome after carbon monoxide poisoning, Ann. Em. Med., 23, 606, 1994.
150. Bouhaddi, K., Thomopoulos, P., Fages, C., Khelil, M., and Tardy, M. Insulin effect
on GABA uptake in astroglial primary cultures, Neurochem. Res., 13, 1119, 1988.
151. Stahl, W. The Na-K-ATPase of nervous tissue, Neuro. Int., 8, 449, 1988.
152. MaCara, I.G. Oncogenes, ions, and phospholipids, Am. J. Physiol., 248, C3–C11,
1985.
153. Standaert, M.L., and Pollet, R.J. Insulin-glycerolipid mediators and gene expression,
FASEB J., 2, 2453, 1988.
154. Goldfine, I.D. The insulin receptor: molecular biology and trans membrane signaling,
Endocrinology, 8, 235, 1987.
155. Goldfine, I.D., Purrello, F., Vigneri, R., and Clawson, G.A. Direct regulation of
nuclear functions by insulin: relationship to mRNA metabolism, in Molecular Basis
of Insulin Action, Czech, M.P., ed., Plenum, New York, 1985, 329.
156. Moss, A.M., Unger, J.W., Moxley, R.T., and Livingston, J.N. Location of
phosphotyrosine-containing proteins by immunocytochemistry in the rat forebrain
corresponds to the distribution of the insulin receptor, Proc. Natl. Acad. Sci., 87,
4453, 1990.
157. Beneviste, H., Drejer, J., Schousboe, A., and Diemer, N. H. Elevation of extracellular
concentrations of glutamate and aspartate in rat hippocampus during transient cerebral
ischaemia monitored by intracerebral microdialysis, J. Neurochem., 43, 1369, 1984.
158. Jorgensen, M.B., and Diemer, N.H. Selective neuron loss after cerebral ischaemia in
the rat: possible role of transmitter glutamate, Acta. Neurol. Scand., 66, 536, 1982.
159. Simon, R.P., Swan, J.H., Griffith, T., and Meldrum, B.S. Blockade of N-methyl-D-
aspartate receptors may protect against ischaemic damage in the brain, Science, 226,
850, 1984.
160. Gill, R., Foster, A.C., and Woodruff, G.N. Systemic administration of MK-801
protects against ischemia-induced hippocampal neurodegeneration in the gerbil,
J. Neurosci., 3343, 1987.
161. Ishimaru, H., Katoh, A., Suzuki, H., Fukuta, T., Kameyama, T., and Nabeshima, T.
Effects of N-methyl-D-aspartate receptor antagonists on carbon monoxide induced
brain damage in mice, J. Pharm. Exp. Ther., 261, 349–352, 1991.
162. Liu, Y., and Fechter, L.D. MK-801 protects against carbon monoxide induced hearing
loss, Toxicol. Appl. Pharm., 132, 196, 1995.
163. Gilmer, B., Thompson, C., Tomaszewski, C., and Watts, J.A. The protective effects of
experimental neurodepressors on learning and memory following carbon monoxide
poisoning, J. Toxicol. Clin. Toxicol. 36, 606, 1999.
164. Anis, N.A., Berry, S.C., Burton, N.R., and Lodge, D. The dissociative anaesthetics,
ketamine and phencyclidine, selectively reduce excitation of the central mammalian
neurons by N-methyl-aspartate, Br. J. Pharmacol., 79, 565, 1983.
165. Penney, D.G., and Chen, K. NMDA receptor-blocker ketamine protects during acute
carbon monoxide poisoning, while calcium channel-blocker verapamil does not,
J. Appl. Toxicol., 16, 297, 1996.
8417: “8417_c016” — 2007/9/11 — 12:15 — page 370 — #30

166. Piswanger, A., Paier, B., and Windisch, M. Modulation of protein synthesis in a cell
free system from rat brain by Cerebrolysin during development and aging, In Amino
Acids, Lubec, G., and Rosenthal, G.A., eds., ESCOM Science Publishers, 1990,
p. 651.
167. Akai, F., Himura, S., Sato, T., Iwamoto, N., Fufimoto, M., Ioku, M., and Hashimoto,
S. Neurotrophic factor-like effect of FPF 1070 on septal cholineric neurons after
transections of fimbria-fornix in the rat, Histol. Histopathol., 7, 213, 1992.
168. Satou, T., Imano, F., Akai, S., Himura, S., Hashimoto, T., Itoh, T., and Fujimoto,
M. Morphological obeservation of effects of Cerebrolysin on cultured neural cells, In
Alzheimer’s Disease and Related Disorders, Nicolini, M., Zatta P.F.,and Coraine, B.,
eds., Pergamon, Oxford, 1993, pp. 195–96.
169. Boado, R.J. Brain-derived peptides increase the expression of a blood-brain barrier
GLUT1 glucose transporter reporter, Neurosci. Lett., 220, 53, 1996.
170. Koroleva, V.I., Korolev, O.S., Mares, V., Pastalkova, and E., Bures, J. Hippocampal
damage induced by carbon monoxide poisoning and spreading depression allevi-
ated by chronic treatment with brain derived polypeptides, Brain Res., 816, 618,
1999.
171. Sluitjer, M.E. The treatment of carbon monoxide poisoning by administration of
oxygen at high pressure, P. Roy. Soc. Med., 56, 1002–1008, 1963.
172. Peirce, E.C., Zacharias, A., Alday, J.M., Hoffman, B.A., and Jacobson, J.H. Carbon
monoxide poisoning: experimental hypothermic and hyperbaric studies, Surgery, 72,
229–237, 1972.
173. Boutros, A.R., and Hoyt, J.L. Management of carbon monoxide poisoning in the
absence of hyperbaric oxygenation chamber, Crit. Care Med., 4, 144–147, 1976.
174. Norman, J.N., and Ledingham, I.McA. Carbon monoxide poisoning: investigations
and treatment, P. Roy. Soc. Med., 56, 101, 1963.
175. Rocco, M., Carbone, I., Morelli, A., Palantonio, P., Rossi, S., Spadetta, G., Passariello,
R., and Pietropaoli, P. The calcium sensitizer levosimendan improves carbon monox-
ide poisoning related stunned myocardium: a cardiac magnetic resonance study, Acta.
Anaesth. Scand., 50, 897, 2006.
176. Radushevich, V.P., and Koroteeva, E.L. Parallel blood circulation with oxygenation
of blood in severe poisoning with carbon monoxide fumes, Vestn. Khir., 116,
131–134, 1976.
177. Baud, F.J. Value of lactic acidosis in the assessment of the severity of acute cyanide
poisoning, Crit. Care Med., 30, 2044, 2002.
178. Moore, S.J., Norris, J.C., Walsh, D.A., et al. Antidotal use of methemoglobin-forming
cyanide antagonists in concurrent carbon monoxide/cyanide intoxication, J. Pharm.
Exp. Ther., 242, 70–73, 1987.
179. Kirk, M.A., Gerace, R., and Kulig, K.W. Cyanide and methemoglobin kinetics in
smoke inhalation victims treated with the cyanide antidote kit, Ann. Emerg. Med., 22,
1413–1418, 1993.
180. Seger, D., and Welch, L. Carbon monoxide controversies: neuropsychologic testing,
mechanism of toxicity, and hyperbaric oxygen, Ann. Emerg. Med., 24, 242–248,
1994.
181. Hopkins, R.O., and Woon, F.L.M. Neuroimaging, cognitive, and neurobehavi-
oral outcomes following carbon monoxide poisoning, Behavioral and Cognitive
Neuroscience Reviews, 5, 141, 2006.
182. Pulsipher, D.T., Hopkins, R.O., and Weaver, L.K. Basal ganglia volumes follow-
ing CO poisoning: a prospective longitudinal study, Undersea Hyperbar. Med., 33,
245, 2006.
8417: “8417_c016” — 2007/9/11 — 12:15 — page 371 — #31

183. Sawada, Y., Ohashi, N., Maemura, K., et al. Computerized tomography as an
indication of long-term outcome after carbon monoxide poisoning, Lancet, 1,
783, 1980.
184. Pracyk, J.B., Stolp, B.W., Fife, C.E., et al. Brain computerized tomography after
hyperbaric oxygen therapy for carbon monoxide poisoning, Undersea Hyperbar.
Med., 22, 1, 1995.
185. Lee, M.S., and Marsden, C.D. Neurological sequelae following carbon monoxide
poisoning clinical course and outcome according to the clinical types and brain
computed tomography, Movement Disord., 9, 550, 1994.
186. Hopkins, R.O., Fearing, M., Weaver, L., and Foley, J. Basal ganglia lesions following
carbon monoxide poisoning, Brain Injury, 20, 273, 2006.
187. Ho, V.B., Fitz, C.R., Chuang, S.H., et al. Bilateral basal ganglia lesions, Pediatric
differential considerations, Radiographics, 13, 269, 1993.
188. Parkinson, R.B., Hopkins, R.O., Cleavinger, H.B., Weaver, L.K., Victoroff, J., Foley,
J.F., et al. White matter hyperintensities and neuropsychological outcome following
carbon monoxide poisoning, Neurology, 58: 1525, 2002.
189. Horowitz, A.L., Kaplan, R., and Sarpel, G. Carbon monoxide toxicity, MR imaging
in the brain, Radiology, 162, 787, 1987.
190. Taverni, N., Dal Pozzo, G., Bartolozzi, C., Caramelli, L., and Boddi, P. Magnetic
resonance imaging in the study of brain changes due to carbon monoxide poisoning,
Radiol. Med., 76, 289–292, 1988.
191. Silverman, C.S., Brenner, J., and Murtagh, F.R. Hemorrhagic necrosis and vas-
cular injury in carbon monoxide poisoning , MR demonstration, AJNR, 14,
168, 1993.
192. Uchino, A., Hasuo, K., Shida, K., Matsumoto, S., Yasumori, K., and Masuda, K. MRI
of the brain in the chronic phase of carbon monoxide poisoning, Neuroradiology, 36,
399, 1994.
193. Mascalchi, M., Petruzzi, P., and Zampa, V. MRI of cerebellar white matter damage
due to carbon monoxide poisoning, case report, Neuroradiology, 38, S73, 1996.
194. Kawanami, T., Kato, T., Kurita, K., and Sasaki, H. The pallidoreticular pattern of brain
damage on MRI in a patient with carbon monoxide poisoning, J. Neurol. Nuerosur.
Ps., 64, 282, 1998.
195. Tuchman, R.F., Moser, F.G., and Moshe, S.L. Carbon monoxide poisoning: bilateral
lesions in the thalamus on MR imaging of the brain, Pediatr. Radiol., 20, 478, 1990.
196. Gale, S.D., Hopkins, R.O., Weaver, L.K., Bigler, E.D., Booth, E.J., and Blatter, D.D.
MRI, quantitative MRI, SPECT, and neuropsychological findings following carbon
monoxide poisoning, Brain Injury, 13, 229, 1999.
197. Hopkins, R.O., Weaver, L.K., and Kesner, R.P. Long-term memory impairments and
hippocampal magnetic resonance imaging in carbon monoxide poisoned subjects,
Undersea Hyperbar. Med., 20 (Suppl.), 15 (abstract), 1993.
198. Hopkins, R.O., Weaver, L.K., and Kesner, R.P. Quantitative MRI analysis of the
hippocampus corresponds with persistent memory impairments in carbon monoxide
poisoned subjects, Brain Cognition, 28, 215 (abstract), 1995.
199. Prockop, L.D. Carbon monoxide brain toxicity: clinical, magnetic resonance ima-
ging, magnetic resonance spectroscopy, and neuropsychological effects in 9 people,
J. Neuroimaging, 15, 144, 2005.
200. Turner, M., and Kemp, P.M. Isotope brain scanning with Tc-HMPAO: a predictor of
outcome in carbon monoxide poisoning? J. Accid. Emerg. Med., 14, 139, 1997.
201. Denays, R., Makhoul, E., Dachy, B., Tondeur, M., Noel, P., Ham, H.R., and Mols,
P. Electroencephalopgrahic mapping and 99mTc HMPAO single-photon emission
8417: “8417_c016” — 2007/9/11 — 12:15 — page 372 — #32

computed tomography in carbon monoxide poisoning, Ann. Emerg. Med., 24,

947, 1994.
202. DeReurck, J., Decoo, D., Lemahieu, I., et al. A positron emission tomography study
of patients with acute carbon monoxide poisoning treated by hyperbaric oxygen,
J. Neurol., 240, 430–434, 1993.
203. Shimosegawa, E., Hatazawa, J., Nagata, I., and Okudera, T. Cerebral blood flow
and glucose metabolism measurements in a patient surviving one year after carbon
monoxide intoxication, J. Nucl. Med., 33, 1696, 1992.
204. Hon, K.E., Yeung, W., Assunta, C., Leung, W.A., Li, A.M., Chu ,W.C., and Chan,
Y. Neurologic and radiologic manifestations of three girls surviving acute carbon
monoxide poisoning, J. Child Neurol., 21: 737, 2006.
205. Kao, C.H., Hung, D.Z., ChangLai, S.P., Liao, K.K., and Chieng, P.U. HMPAO
brain SPECT in acute carbon monoxide poisoning, J. Nucl. Med., 39, 769–772,
1998.
206. Castoldi, A.F., Coccini, T., Randine, G., Hernandez-Viadel, M., Felipo, V., and
Manzo, L. Lymphocyte cytochrome c oxidase, cyclic GMP and cholinergic muscar-
inic receptors as peripheral indicators of carbon monoxide neurotoxicity after acute
and repeated exposure in the rat, Life Sci., 78, 1915, 2006.
207. Brvar, M., Mozina, M., Osredkar, J., Suput, D., and Bunc, M. Prognostic value
of S100B protein in carbon monoxide-poisned rats, Crit. Care Med., 32, 2128,
2004.
208. Hampson, N.B., Dunfored, R.G., Kramer, C.C., et al. Selection criteria utilized for
hyperbaric oxygen treatment of carbon monoxide poisoning, J. Emerg. Med., 13,
227, 1995.
209. Sloan, E.P., Murphy, D.G., Hart, G., Cooper, M.A., Turnbull, T., Berreca, R.S., and
Ellerson, B. Complications and protocol considerations in carbon monoxide-poisoned
patients who require hyperbaric oxygen therapy: report from a ten-year experience,
Ann. Emerg. Med., 18, 629, 1989.
210. Goulon, M., Barois, A., Rapin, M., et al. Carbon monoxide poisoning and acute
anoxia due to breathing coal gas and hydrocarbons, J Hyperbaric Med., 1, 23, 1986.
211. Penney, D.G. Effects of carbon monoxide exposure on developing animals and
humans, In Carbon Monoxide, Penny, D.G., ed., Boca Raton, FL, CRC Press, Inc,
1996, Chap 6, pp. 109–144.
212. Ginsberg, M.D., and Myers, R.E. Fetal brain injury after maternal carbon monoxide
intoxication, Neurology, 26, 15–23, 1976.
213. Ginsberg, M.D., and Myers, R.E. Fetal brain damage following maternal carbon
monoxide intoxication, Acta. Obstet. Gynecol., 53, 309–317, 1974.
214. Koren, G., Sharav, T., Pastuszak A., Garrettson, L.K., Hill, K. Samson, I., Rorem, M.,
King A., and Dolgin J.E. A multi-center, prospective study of fetal outcome following
accidental carbon monoxide poisoning in pregnancy, Reprod. Toxiol., 5, 397–403,
1991.
215. Brown, D.B., Mueller, G.L., and Golich, F.C. Hyperbaric oxygen treatment for car-
bon monoxide poisoning in pregnancy: a case report, Aviat. Space Envir. Md., 63,
1011–1014, 1992.
216. Ferm, V.H. Teratogenic effects of hyperbaric oxygen, P. Soc. Exp. Biol. Med., 116,
975–976, 1964.
217. Fujikura, T. Retrolental fibroplasia and prematurity in newborn rabbits induced by
maternal hyperoxia, Am. J. Obstet. Gynecol., 90, 854–858, 1964.
218. Miller, P.D., Telfored, I.D., and Haas, G.R. Effects of hyperbaric oxygen on
cardiogenesis in the rat, Biol. Neonatorum, 17, 44–52, 1971.
8417: “8417_c016” — 2007/9/11 — 12:15 — page 373 — #33

219. Elkaharrat, D., Raphael, J.D., Korach, J.M., Jars-Guincestre, M.C., Chastang, C.,
Harborn, C., and Gajdos, P. Acute carbon monoxide intoxication and hyperbaric
oxygen in pregnancy, Intens. Care Med., 17, 289–92, 1991.
220. Farrow, J.R., Davis, G.J., Roy, T.M., et al. Acute carbon monoxide intoxication and
hyperbaric oxygen in pregnancy, Intens. Care Med., 17, 289–292, 1991.
221. White, S.R. Pediatric Carbon Monoxide Poisoning. In: Carbon Monoxide Toxicity,
Penney, D.G., ed., CRC Press LLC, Boca Raton 2000, Chapt. 21.
mechanism of toxicity, and hyperbaric oxygen, Ann. Emerg. Med., 24, 242, 1994.
223. De Pooter, M.C., Leys, D., Godefroy, O., DeReuck, J., and Peter, H. Parkinsonian
syndrome caused by carbon monoxide poisoning. Preliminary results of the treatment
with bromocriptine, Rev. Neurol., 147, 399–403, 1991.
224. Hu, M.C., Shiah, I.S., Yeh, C.B., Chen, H.K., and Chen, C.K. Ziprasidone in the
treatment of delayed carbon monoxide encephalopathy, Prog. Neuro-psychoph. , 30,
755, 2006.
225. Pae, C.U., Kim, T.S., Lee, C., and Paik, I.H. Effect of aripiprazole for a patient with
psychotic symptoms and parkinsonism associated with delayed sequelae of carbon
monoxide intoxication, J. Neuropsych. Clin. Neurosci., 18, 130, 2006.
226. Ten Holter, J.B.M, and Schellens, R.L.L.A.M., Dantrolene sodium for treatment of
carbon monoxide poisoning, Brit. Med. J., 296, 1772, 1988.
227. Hiramatsu, M., Sasaki, M., Nabeshima, T., and Kaemyama, T. Effects of dyn-
orphin A(1–13) on carbon monoxide-induced delayed amnesia in mice. Pharmacol.
Biochem., 56, 73, 1997.
228. Hiramatsu, M., Hyodo, T., and Kameyama T. U-50488H a selective k-opioid
receptor agonist, improves carbon monoxide-induced delayed amnesia in mice, Eur.
J. Pharmacol., 315, 119, 1996.
229. Myers, R.A., Snyder, S.K., and Emhoff, T.A. Subacute sequelae of carbon monoxide
poisoning, Ann. Emerg. Med., 14, 1163–1167, 1985.
three-year follow up, Brit. Med. J., 1, 318–321, 1973.
231. Myers, R.A., Mitchell, J.T., and Cowley, R.A. Psychometric testing and carbon
monoxide poisoning, Disaster Med., 1, 279–281, 1983.
232. Myers, R.A., Snyder, S.K., Linberg, S., et al. Value of hyperbaric oxygen in suspected
carbon monoxide poisoning, JAMA, 246, 248, 1981.
233. Coric, V., Oren, D.A., Wolkenberg, A., and Kravitz, R.E. Carbon monoxide poisoning
and treatment with hyperbaric oxygen in the subacute phase, J. Neurol. Neurosur. Ps.,
65, 245, 1998.
234. Maeda, Y., Kawasaki, Y., Jibiki, I., Yamaguchi, N., Matsuda, H., and Hisada, K.
Effect of therapy with oxygen under high pressure on regional cerebral blood flow
in the interval form of carbon monoxide poisoning: observation from subtraction of
Technetium-99m HMPAO SPECT brain imaging, Eur. Neurol., 31, 380, 1991.
235. Samuels, A.H., Vamos, M.J., and Taikato, M.R. Carbon monoxide, amnesia and
hyperbaric oxygen therapy, Aust. NZ. J. Psychiat., 26, 316, 1992.
236. Heinrichs, R.W. Relationship between neuropsychological data and response to beha-
vioral treatment in a case of carbon monoxide toxicity and dementia, Brain Cognition,
14, 213–219, 1990.
8417: “8417_c016” — 2007/9/11 — 12:15 — page 374 — #34

17 The Case for the Use of
Hyperbaric Oxygen
Therapy in Carbon
Monoxide Poisoning
Christian Tomaszewski
CONTENTS
17.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375

17.2 Carbon Monoxide Effects at the Cellular Level . . . . . . . . . . . . . . . . . . . . . . . . . . . 376
17.2.1 HBOT Reverses the Cellular Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
17.3 Clinical Efficacy of HBOT in Carbon Monoxide Poisoning . . . . . . . . . . . . . 378
17.3.1 Negative Trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378
17.3.2 Positive Trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379
17.4 Indications for HBO Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382
17.5 Delayed Administration of HBOT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
17.6 Repeated Treatment with HBOT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
17.7 Future Directions in Better Targeting HBO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
17.8 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
17.1 INTRODUCTION
Carbon monoxide (CO) is a serious but complex poison. If one is lucky enough to
survive the acute hypoxic event from avid binding of hemoglobin, the patient still has
to contend with a potential of up to 40% chance of delayed and/or persistent neur-
ological deficits.1–6 These effects can be debilitating including dementia, amnestic
syndromes, parkinsonism, movement disorders, and cortical blindness.7–9 The prob-
lem is that patients may appear initially well, and following several days to weeks,
develop delayed neurological sequelae (DNS).9 These problems can last for a year
or longer.10
The main issues in treating CO poisoning is identifying those that are at risk for
neurological sequelae and referring these patients to a treatment modality that will
prevent such sequelae. These issues are important once the patient has reached the
375
8417: “8417_c017” — 2007/9/11 — 12:15 — page 375 — #1

hospital, having been resuscitated and stabilized. Traditionally 100% oxygen was
advocated as initial first aid to enhance removal of CO from hemoglobin and reverse
any concomitant hypoxia. But starting in the 1960s, especially in the United States,
hyperbaric oxygen therapy (HBOT) had started to be advocated for treating all such
poisonings.11 Anywhere between 450 and 2500 HBO treatments are done for CO
poisoning each year in the US alone.12,13
According to recent review articles, the weight of evidence does not appear to
support the use of HBO for CO poisoning.14–17 But most of these critical review
articles give equal weighting to all controlled studies. Almost all the studies failing to
show benefit from HBO in CO poisoning have critical flaws in treatment regimen or
follow-up. They also contradict consistent studies showing benefit in animal models,
in which the pathophysiology for improvement is now being elucidated. All these
studies, along with some recent randomized trails, suggest that a safe modality like
HBO has the potential to offer hope in a poisoning with often unforeseen devastating
neurological consequences.
17.2 CARBON MONOXIDE EFFECTS AT THE

CELLULAR LEVEL
The simplest explanation for the utility of HBO is that it accelerates the removal of CO
from hemoglobin. Normally, the half-life of carboxyhemoglobin (COHb) with 100%
oxygen averages 74–92 min, on the basis of three studies.6,18,19 With HBO, COHb
half-lives are reduced to as low as 20 min.20,21 Should someone be hypoxic, HBO
also increases dissolved oxygen by about ten times, sufficient to supply metabolic
needs.22 But on entry into the HBO chamber, in Weaver’s trial of ill CO-poisoned
patients, the mean COHb was less than 5%.6 So in most cases of CO poisoning,
COHb clearance is not an issue because of aggressive treatment with normal pressure
oxygen prior to ability to arrange HBO treatment.
The consensus is that there must be an alternative mechanism for HBOT effic-
acy in CO poisoning. This is why the percentage of COHb does not always correlate
with degree of acute toxicity and with eventual neurological outcome.2,10,23,24 Once a
patient is stabilized, the real target organ of interest is the brain. CO is delivered intra-
cellularly, where it can bind to heme proteins other than hemoglobin. At high levels,
CO interferes with cellular respiration by binding to and inhibiting mitochondrial
cytochrome oxidase.25 This is particularly exaggerated during episodes of hypoxia
and hypotension.
The inhibition of cellular respiration may contribute to the ischemic reper-
fusion that occurs in rat brains during CO poisoning ultimately leading to lipid
peroxidation.26 These free radicals cause endothelial damage, which allows lymph-
ocytes to attach and release proteases that promote more production of oxygen free
radicals.27 The end result of this process is delayed central nervous system (CNS)
damage that is accompanied by learning decrements.
Part of the neuronal damage may also be mediated by excitatory amino acids.28,29
Glutamate binds to N-methyl-d-aspartate (NMDA) receptors causing intracellular
calcium release, thus causing delayed neuronal cell loss. Accompanying this may
8417: “8417_c017” — 2007/9/11 — 12:15 — page 376 — #2

The Case for the Use of Hyperbaric Oxygen Therapy 377
be apoptosis.30,31 Some of the most sensitive areas for neuronal cell loss from these
processes are the basal ganglia and hippocampus, resulting in impaired learning and
memory.32
All these rodent studies demonstrate that CO is not a simple chemical asphyxiant.
The CO molecule sets in a motion a cascade of cytochemical events that days later
result in CNS cell loss. So in spite of COHb clearance from the bloodstream early
on with normal pressure, or normobaric oxygen (NBO), HBO may still have an
essential role in most cases of CO poisoning in preventing these delayed neurological
events.
17.2.1 HBOT REVERSES THE CELLULAR EFFECTS

HBO has been shown to reverse many of the biochemical effects of CO in the
same rodent models that elucidated CO’s neurochemical and immune effects. First,
HBO accelerates regeneration of inactivated cytochrome oxidase.25 This reduction in
potential for oxidative stress is accompanied by prevention of lipid peroxidation. Rat
models show a dose response effect from HBO, maximal at 3 atm. absolute (ATA),
in decreasing the lipid peroxidation products in the brains of rats exposed to CO.26
The critical event that precedes lipid peroxidation is the vascular abnormalit-
ies that CO induces in cerebral endothelium. Adherence of neutrophils amplifies
CO-mediated oxidative stress.27,33 The end result of this is necrosis and apop-
tosis of critical areas of the brain for learning and memory, particularly the
hippocampus.28,30,34 HBOT is able to prevent neutrophil adherence to the brain
microvascular endothelium, an essential step for amplification of CNS damage from
CO.35 This blockage of lipid peroxidation prevents the precipitation of abnormal-
ities in myelin basic proteins. Therefore, the CO-mediated oxidative stress that
causes alteration in myelin basic protein and leads to immunologic effects is blocked
by HBO.36
The final outcome is that HBO reduces mortality in rodent models of serious CO
poisoning. This appears to be due to protection against cerebral edema. All control rats
died in one study from cerebellar herniation as opposed to 100% survival in those that
received HBO.38 The end result is that cognitive decrements from CO are prevented by
HBO, as demonstrated in radial maze performance.36 The histological manifestation
of this protection is lack of lesions in the globus pallidus and hippocampus in rats,
which are seen after CO poisoning.32 One recent study does not support this, but there
were issues with excessive toxicity from concomitant hypoxic stress.39
In conclusion, animal models show that the mechanism for CO toxicity is more
than hypoxia. The poisoning initiates an immune cascade in the brain that includes
changes in vascular endothelium, adherence of leukocytes, and lipid peroxidation.
The end result of these events is neurological deficits in learning and memory. HBO,
but not NBO, prevents all of these events, usually in a dose-related fashion. None of
these mechanisms are related to the continued presence or clearance of COHb. This
lends support to the use of HBO in symptomatic CO-poisoned patients once they
have been stabilized regardless of the lack of remaining COHb. These patients are
still at risk for delayed or persistent neuropsychological sequelae and therefore could
benefit from HBO.
8417: “8417_c017” — 2007/9/11 — 12:15 — page 377 — #3

17.3 CLINICAL EFFICACY OF HBOT IN CARBON

MONOXIDE POISONING
HBOT has been advocated as the treatment of choice for patients with significant
CO exposures.11,40 However, clinical studies have not consistently demonstrated
efficacy for HBO in preventing neurological damage from CO, as basic science studies
discussed above suggest. Initially, interest in HBO for CO poisoning was fueled
by uncontrolled human clinical series. In such studies, the incidence of persistent
neuropsychiatric symptoms, including memory impairment, ranged from 12% to 43%
in patients treated with 100% oxygen, and was as low as 0–4% in patients treated
with HBO.4,5,40–42 Such early studies were obviously biased. So although HBO was
advocated widely for CO poisoning, no randomized controlled trials (RCTs) existed
to support such a stance.
17.3.1 NEGATIVE TRIALS

The first truly randomized study in acutely poisoned CO patients did not come until
1989.2 The study failed to show the beneficial effect of HBO in over 300 patients.
Patients who presented without loss of consciousness (LOC), n = 343, were random-
ized to one HBO session or 100% oxygen at room pressure by mask. Outcome criteria
were not strict and were mainly based on questionnaire responses regarding symp-
toms. At 1 month, there was no difference in both groups with respect to self-reported
neurological symptoms: 32% in the HBO group, 34% in the room pressure oxygen
group. Another outcome, return to prior occupation, was 97% in both groups. Major
flaws in this study, besides the lack of blinding and poor endpoints, were suboptimal
HBO pressure used, 2.0 ATA, and almost half of the patients receiving HBO over 6
h from end of CO poisoning, which was a major delay.43
After the Raphael trial described above2 , a larger, double-blinded trial was hailed
by many as conclusive proof that HBO was not useful in CO poisoning.44 This trial
involved 191 patients who up to 24 h after poisoning were randomized to HBO (2.8
ATA maximum) versus sham HBO treatments. Patients were treated aggressively at a
maximal pressure of 2.8 ATA. They received daily treatments for 3 days, and up to 6
additional daily sessions if they did not recover. HBO provided no benefit in this trial
with the majority of each group having adverse neurological outcomes at 1 month:
74% in HBO-treated patients, and 68% in controls (reported odds ratio (OR) = 1.7;
95% confidence interval (CI) = 0.8–4.0; P = 0.19, not significant (NS)). The major
flaw in this study was a mean delay of over 6 h to treatment with HBO. In addition
54% of the original subjects were lost to follow-up. Disproportionate numbers of
suicide cases (about two-thirds) and drug toxicity (44%), with accompanying excess-
ive neuropsychological defects, could have confounded any beneficial effect from
HBOT. Other work shows that depression can contribute to poor neuropsychological
testing.45 Because of these multiple flaws, it is not surprising that HBO failed to show
any benefit in what was initially a well designed study.
The most recent study showing no beneficial effect from HBO in CO poisoning
is only available in abstract form.3 One hundred and seventy-nine patients that had
transient LOC, but who were not comatose, were randomized to one session of HBO
8417: “8417_c017” — 2007/9/11 — 12:15 — page 378 — #4

at 2 ATA versus 6 h of normal pressure oxygen. Outcome was self-assessed symptoms

and blinded neurological examination at 1 month. Recovery was the same in both
groups: HBO 46/79 (58%) versus control 45/74 (61%). As in Raphael’s previous
study2 , lack of sufficient pressure, at 2.0 ATA, may have undermined the efficacy
of HBO.
There are consistent themes in the negative studies failing to show beneficial
effects of HBO in CO. The first is delay in treatment. Animal studies show that the
biochemical effects begin within hours after exposure to CO.36 The issue is whether
delayed treatment can reverse this once it has started.43 The other problem is that
treatment may have not been optimal. Three ATA is the pressure at which animal
studies show the most benefits on tissue effects and neurological outcomes after CO
poisonings.25,46,47 In fact, to prevent human neutrophil adherence, only marginal
inhibition occurs at 2.0 ATA; 2.8 ATA gives almost complete inhibition.46 Other
issues with these studies include poor outcome data due to both quality and quantity
of follow-up.
17.3.2 POSITIVE TRIALS

As the critics of HBO mounted, studies began to show the beneficial effects of
HBO with CO poisoning. The first randomized positive study used mildly poisoned
patients.1 Sixty-five patients without LOC were randomized to one session of HBO
or 100% oxygen by face mask at room pressure. A decrease in any one of six neuro-
psychological tests immediately after poisoning versus at 4 weeks defined outcome.
No patient [95% C.I. 0–12%] in the HBO group versus 23 [95% C.I. 10–42%] in the
control group showed deterioration. Critics point out that the neurological deterior-
ation was seen in only one test, Trail-Making; but neurological sequelae persisted
for a mean of 41 days, and 10% of the controls had difficulties with daily activities.
The success of this trial can be partially attributed to the adequate pressure used,
maximum 2.8 ATA, and that all patients were treated within 6 h of discovery.
A subsequent study by Mathieu et al.48 examined the efficacy of HBO at 2.5 ATA
for 90 min versus NBO for 12 h. All patients had to be within 12 h of poisoning
termination and noncomatose on presentation. At one and three months there was
a lower incidence of persistent neurologic manifestations in the HBO group. This
was significant at 3 months: HBO 9.5% versus NBO 15%, p < .02. This differ-
ence resolved over the following year. This study suffers from lack of detail and the
fact that it is an interim analysis with no final analysis published. Regardless, early
aggressive treatment, using pressures above 2.0 ATA, confirm the efficacy of HBO
in CO poisoning.
The landmark randomized control trial showing efficacy of HBO in CO pois-
oning differs from the other positive trials, in that all patients, including intubated
cases (13%), were enrolled. Patients, n = 152, who presented with symptoms from
acute CO poisoning were randomized to three sessions of HBO, with the first hour
at 3.0 ATA, versus 100% oxygen at room pressure.6 Rather than looking at the abso-
lute difference between the two groups on neurological testing, neurological sequelae
were defined a priori. Aggregate performance on six neuropsychological tests at least
one standard deviation below published norms, or an aggregate score greater than
8417: “8417_c017” — 2007/9/11 — 12:15 — page 379 — #5

or equal to two standard deviations below expected, was defined as neurological

sequelae. At 6 weeks, HBO caused an absolute reduction of 21% [95% C.I. 6–34%]
in the proportion of neurological sequelae: 24% in the HBO group; 46% in control,
which equates to a number needed to treat 5. At 1 year, there was still some benefit,
albeit less, with an absolute reduction of 15% [95% C.I. 1–28%]. As stated earlier,
the patients were seriously ill on presentation, with a mean initial COHb level of
25% and half having suffered LOC. Although patients could be entered up to 24
h post CO-poisoning, the mean time to treatment was less than 2 h. The combin-
ation of high pressure and early treatment probably contributed to the success of
this trial.
The main criticisms with Weaver’s study6 lie in the statistical definition of
neurological sequelae.14 Although there were individual differences in the neuropsy-
chological test of Trail-Making, there was no difference at 6 weeks in the parametric
comparison of mean neuropsychological score. In addition, there was no difference
in activities of daily living at 6 weeks and 12 months. In defense of the study and
HBO, patients had decreased self-reported memory problems at 6 weeks (28% versus
51%), and the beneficial effect on cognitive sequelae lasted up to one year. Another
criticism leveled on the Weaver study was that patients in the control group were more
ill, with a higher incidence of cerebellar dysfunction on presentation, which ended
up being a predictor for cognitive sequelae.17 But, when doing a logistic regression
to adjust for this incidental finding, HBO was still protective (OR=0.45, 95% C.I.
0.22–0.92).6 In conclusion, Weaver’s study appears methodologically sound, espe-
cially when considering that outcome measures were determined a priori and the
fact that the trial was double-blinded including sham HBO treatments in the control
group.
The Cochrane Review recently examined all six controlled studies published as of
2006 that examined the effect of HBO versus NBO in CO poisoning.15 (Table 17.1).
The common outcome was the presence of neurological symptoms or signs at time
of primary analysis 4–6 weeks postpoisoning. With a collective 1335 participants,
the incidence of persistent signs and symptoms was 29% (202/691) in the HBOT
group versus 34% (219/444) in the normobaric group. The overall odds ratio favored
HBO at 0.78 [95% C.I. 0.54–1.12]. Because of the lack of statistical significance
their conclusion was that existing trials do not support use of HBO in CO-poisoned
patients in order to reduce neurological sequelae. Similarly, the American College of
Emergency Physicians, on the basis of the same data, formulated a similar conclusion
in their 2007 Clinical Policy for the management of acute CO poisoning.49 They
concluded that although HBO should not be mandated for CO poisoning, it “remains
a therapeutic option to potentially reduce the incidence of neurological sequelae.” The
consensus among all these expert reviews is that further RCTs are needed, particularly
with regard to identifying which CO-poisoned patients are most likely to benefit
from HBO.
Using the same studies, the Underwater and Hyperbaric Medical Society
recommends HBO treatment for any CO-poisoned patients with signs of serious
intoxication.11 With little risk,50 almost 1500 patients are treated with HBO for CO
poisoning in the United States each year.51 The main risk with HBO is related to baro-
traumas such as tympanic perforation.6 Less likely, and not usually associated with
8417: “8417_c017” — 2007/9/11 — 12:15 — page 380 — #6

TABLE 17.1
Patients with Signs or Symptoms at 4–6 Weeks Post Carbon Monoxide Poisoning
Maximum HBO
Study Design Pressure Time To RX Treatment Control Odds Ratio∗ Comment
Mathieu 199648 HBO 90 min vs. 2.5 ATA <12 h 69/299 73/276 0.83(0.57– Abstract. Excluded coma. 3
12 h NBO (23%) (26%) 1.22) months outcome favorable
for treatment: 9.5 vs. 15%
(P = 0.016)
Raphael 19892 HBO 2.0 h vs. 6 h 2.0 ATA Mean 7.1 h 51/159 50/148 0.93(0.57– Subjective outcome.
NBO (32%) (34%) 1.49) Subgroup data on transient
loss of consciousness group.
Thom 19951 HBO 2 h vs. 2.8 ATA Mean 2.0 h 0/30 7/30 0.05(0.00– No patient with LOC entered
100% NBO until (0%) (23%) 0.95)
asymptomatic
The Case for the Use of Hyperbaric Oxygen Therapy
Scheinkestel 199944 HBO 1 h 2.8 ATA Mean 7.1 h 30/48 25/40 1.00(0.42– 69% attempted suicide, 56%
(min × 3) vs. (63%) (63%) 2.38) lost to follow up
NBO 100 min
Weaver 20026 HBO 2 h (×3) vs. 3.0 ATA Mean 5.6 h 19/76 35/76 0.39(0.20– No difference in overall
NBO 2 h (×1) (25%) (46%) 0.78) neuropsychological scores
or daily activities
Raphael 20043 HBO at 2.0 ATA 2.0 ATA < 12 h 33/79 29/74 1.11(0.58– Abstract. Subgroup data on
60 min vs. 6 h (42%) (39%) 2.12) transient loss of
NBO consciousness group
8417: “8417_c017” — 2007/9/11 — 12:15 — page 381 — #7

∗ Odds ratio less than 1.00 favors treatment with HBO.
381
any demonstrable neuronal damage, are oxygen induced seizures.40,52 On the basis
that at least three randomized trials that used early and adequate pressure demonstrated
a benefit for HBO, and the fact that it is a relatively safe modality, it is surprising
that the consensus of opinion is not stronger in support of HBO for serious CO
poisoning.
17.4 INDICATIONS FOR HBO THERAPY

The Cochrane review pointed out that studies are needed with regard to the partic-
ular role of HBO in CO poisoning.15 Although specific indications for HBO after
acute CO poisoning are widely published (Table 17.2), none have been prospectively
evaluated.11,53 The patients with the most to gain from this procedure are those with
the greatest potential for persistent or DNS.
Traditionally COHb levels have been advocated, with specific cut offs, for decid-
ing if HBO is needed.53 However, COHb levels do not correlate with clinical findings
acutely nor with final outcome.10,40,54–56 The target organ, once a patient has been
resuscitated and stabilized, at the time at which one would consider HBO, is no
longer the blood but the CNS. A single COHb level does not describe the actual area
under the dosage curve, which more likely represents the true degree of poisoning.
Before all the intricacies in the cellular action of CO were known, many authors
advocated treating all patients with COHb levels of 40% or greater with HBOT.
Many HBO centers arbitrarily use a more conservative level of 25% as an indic-
ation for HBO. In Weaver’s recent trial, post hoc analysis predicted that patients
with COHb levels greater than 25% were one of the subgroups to benefit most from
HBO.57,58 However, other studies suggest that no absolute COHb level on presenta-
tion has been found to be necessarily predictive of outcome, and therefore, the need for
HBO.2,10,24
Another traditional marker for serious CO poisoning, and therefore the need for
HBO, has been syncope, or LOC.59 This may represent the episode of hypotension
that is necessary for causing neuronal damage from CO-induced ischemic-reperfusion
TABLE 17.2
Suggested Indications for Hyperbaric Oxygen Therapy After Carbon Monoxide
Poisoning
• Syncope (loss of consciousness)
• Coma
• Seizure
• Persistent altered mental status (GCS < 15) or confusion
• Abnormal cerebellar examination
• Carboxyhemoglobin level > 25%
(Adapted from Tomaszewski, C., Goldfrank’s Toxicologic Emergencies, Goldfrank, L.R., Flomenbaum,
N.E., Hoffman, R.S., Howland, M.A., Lewin. N,A., Nelson, L.S (Eds), McGraw-Hill, New York, 2006.)
8417: “8417_c017” — 2007/9/11 — 12:15 — page 382 — #8

injury in animal models.60,61 However, syncope is not entirely predictive for cognitive
sequelae.6
Patients with long exposures, or “soaking” periods, are also at greater risk for
neurological sequelae.62,63 Animal models and human cases suggest that “soaking,”
that is, prolonged exposure to high levels of CO, is a factor that actually predicts final
neurological outcome, rather than any particular COHb level.64,65 The presence of
a significant metabolic acidosis may be a surrogate marker for this. A base excess
lower that −2 mmol/L was an independent predictor for cognitive sequelae and the
potential for benefit from HBO in Weaver’s study.6,58 Other studies confirm the utility
of metabolic acidosis as a predictor of HBO requirement.10,66,67
It is still unclear if mild neurological symptoms (e.g., confusion, headache, dizzi-
ness, visual blurring) or abnormal mental status testing on initial presentation after
CO poisoning is prognostic for cognitive sequelae. These symptoms simply represent
CO poisoning, which, at COHb levels approaching 10% in volunteers, can cause tem-
porary impairment of learning and memory.68 To date, neuropsychological screening
tests have not been found to be reliable indicators of the need for HBO because they
do not consistently predict neurological sequelae.1 In a recent prospective clinical
trail of CO poisoning, the incidence of cerebellar dysfunction portended a higher
incidence of cognitive sequelae (odds ratio 5.7 [95% C.I. 1.7–19.3]).6 Therefore, dif-
ficulty with finger-to-nose, heel-to-shin, rapid alternating hand movements, or even
ataxia, should all be considered indications for HBO.
Some authors recommend selective use of HBO because of cost and difficulties
in transport if the primary facility lacks a chamber.69 However, complications that
may make such transfers and treatment unsafe are rare.50 Although HBO cannot be
recommended for every patient with CO poisoning, it is a relatively safe treatment
that should be considered in all serious exposures. Post hoc analysis of Weaver’s
data showed that the patients most likely to benefit were those who presented with
a base deficit greater than 2 mmol/L, unconsciousness, age ≥50 years, and COHb
level greater than 25%.6,70 Therefore, most clinicians refer any CO-poisoned patient
with any of these criteria for HBO (Table 17.2). One group of patients who could
probably be excluded from HBOT are those who have had a cardiac arrest from CO;
these cases have been universally fatal.71
17.5 DELAYED ADMINISTRATION OF HBOT

The optimal timing of HBO treatment for CO poisoning is unclear. Patients treated
later than 6 h after exposure tend to have worse outcomes in terms of delayed sequelae
(30% versus 19%) and mortality (30% versus 14%).72 Randomized trials with longer
times to treatment have generally failed to show benefit from HBO.73 In the recent
randomized trial showing beneficial effects from HBO, with a number needed to treat
of 5 in order to prevent one case of neurological sequelae, patients were entered
into treatment up to 24 h post-poisoning.6 In fact 38% of the patients were treated
later than 6 h. Studies purporting beneficial effects much later are anecdotal and lack
controls.4 Therefore, at this time it seems reasonable to attempt treatment up to 24 h
post-exposure.
8417: “8417_c017” — 2007/9/11 — 12:15 — page 383 — #9

17.6 REPEATED TREATMENT WITH HBOT

Repeated HBO treatments have been advocated for patients that do not initially
improve from CO poisoning, particularly those in coma.74 In the recent randomized
study showing beneficial effects from HBO, all patients received three HBO treat-
ments within 24 h of presentation.6 A retrospective review of records on patients who
received two versus one treatment showed a reduction in DNS from 555 to 18%.75
Prospective studies comparing single versus multiple courses of HBOT have failed
to confirm any benefit from repeated HBO treatment.2,3 Therefore, multiple HBO
treatments cannot be recommended at this time. The most recent clinical guidelines
from the Underwater and Hyperbaric Society state that the optimal number of HBO
treatments for CO poisoning is unknown, reserving multiple treatments for patients
who fail to recover after the initial treatment.11
17.7 FUTURE DIRECTIONS IN BETTER TARGETING

HBO
A search has been made for plasma markers for CO exposure that predict degree of
toxicity and therefore, perhaps, the potential for neurological sequelae and the need
for HBO. Many plasma markers of oxidative stress, such as glutathione release from
erythrocytes, increase within hours of CO poisoning.76 Multiple peripheral vascular
cells have been implicated in CO poisoning: erythrocytes, platelets, leukocytes, and
endothelial cells.27,46,77 They actually appear to mirror what is going on in the target
organ of concern, the CNS.78,79 One recent study showed that CO exposure in rats
resulted in a 50% decrease in cyclic guanosine monophosphate (GMP) activity, as
measured in leukocytes, that started at 24 h post-exposure.80 No human series of
plasma markers are available to show if they are predictive of final outcome and
therefore a potential need for HBOT.
Plasma markers are still only surrogates for the real target organ of CO poison-
ing, the brain. S100B is a neurobiochemical marker of brain damage. It is found in
astroglial cells and is released into the blood with brain injury. Rat studies show that
it is elevated after severe CO poisoning and is a better predictor of death than level
of consciousness.81 In patients presenting with Glasgow Coma Scores (GCS) of less
than eight after CO poisoning, but who did not experience LOC, immediate blood
samples show elevated peripheral blood levels of SB100B.82 HBO prevents this rise
in SB100B when given immediately after CO poisoning in rats.83 A recent clinical
study, however, showed no increase in S100B levels after CO poisoning.84 Therefore,
it is unclear if SB100B can be used as a consistent marker of CO poisoning and the
need for HBO.
As an alternative to serum markers, various types of neurological imaging are
available that show early changes from CO poisoning. Except for xenon enhancement,
computed tomography does not show early changes from CO.85 More sensitive is
Magnetic Resonance Imaging (MRI), which can show changes within the first day
post-exposure.86 Diffusion-weighted is even more sensitive, detecting changes in
subcortical white matter within hours of CO poisoning.87 However, MRI changes
have not been shown to correlate with eventual outcome.57,87
8417: “8417_c017” — 2007/9/11 — 12:15 — page 384 — #10

A more sensitive radio imaging modality is single-photon emission computed

tomography (SPECT), which gauges regional blood flow noninvasively. It shows
perfusion changes within watershed regions that are associated with delayed neuro-
psychological impairment months after CO poisoning.88,89 With HBO, an improve-
ment in oxygen extraction of the frontal and temporal cortices was demonstrated
on positron emission tomography (PET) scan.90 Because of practicality considera-
tions, along with lack of prospective data, no neurological imaging technique can be
advocated at this time to predict who needs HBO treatment.
17.8 SUMMARY
HBOT may originally have been a “therapy in search of a disease” with respect to
CO poisoning. But now several controlled studies show early benefits in preventing
cognitive sequelae from CO poisoning. Although these studies are not perfect, the
alternative negative clinical controlled studies suffer from serious flaws. In addition,
the positive clinical results corroborate the findings of animal studies that show that
HBO can prevent the cascade of neurochemical events that occur during recovery
from acute poisoning. The biggest challenge though will be to decide who can really
benefit from this therapy. To date, there is no consistent marker to predict who will
suffer DNS and therefore, who has the most to gain. Until more studies are available
confirming HBO’s utility, and owing to the inherent delay with such, we probably
should not be subjecting patients to extraordinary transports to receive such therapy.
But based on the safety of the procedure, and relative low resource utilization, there is
no reason not to at least attempt one HBO treatment in any seriously ill CO-poisoned
patients.
References
1. Thom, S.R., and Taber, R.L., Mendiguren II et al. Delayed neuropsychologic sequelae
after carbon monoxide poisoning: prevention by treatment with hyperbaric oxygen.
Ann. Emerg. Med. 25, 474–480, 1995.
2. Raphael, J.C., Elkharrat, D., Jars-Guincestre, M-C et al. Trial of normobaric and
hyperbaric oxygen for acute carbon monoxide intoxication. Lancet 2, 414–419, 1989.
3. Raphael, J.C., Chevret, S., Driheme, A., and Annane, D. Managing carbon monoxide
poisoning with hyperbaric oxygen. J. Toxicol. Clin. Toxicol. 42, 455–456. 2004.
(Abstract)
4. Myers, R.A.M., Snyder, S.K., and Emhoff, T.A. Subacute sequelae of carbon
monoxide poisoning. Ann. Emerg. Med. 14, 1167, 1985.
5. Mathieu, D., Nolf, M., and Durocher, A. Acute carbon monoxide poisoning: risk
of late sequelae and treatment by hyperbaric oxygen. J. Toxicol. Clin. Toxicol. 23,
315–324, 1985.
6. Weaver, L.K., Hopkins, R.O., Chan, K.J et al. Hyperbaric oxygen for acute carbon
monoxide poisoning. NEJM 347, 1057–1067, 2002.
7. Lee, M.S., and Marsden, C.D. Neurological sequelae following carbon monoxide
poisoning: clinical course and outcome according to the clinical types and brain
computed tomography scan findings. Movement Disorders 9, 550–558, 1996.
8417: “8417_c017” — 2007/9/11 — 12:15 — page 385 — #11

8. Ginsberg, M.D. Carbon monoxide intoxication: clinical features, neuropathology, and

mechanisms of injury. J. Toxicol. - Clin. Toxicol. 23, 281–288, 1985.
9. Choi, I.S. Delayed Neurological Sequelae in Carbon Monoxide Intoxication. Arch.
Neurol. 40, 433–435, 1983.
10. Sokal, J.A., and Kraldowska, E. The relationship between exposure duration, carboxy-
hemoglobin, blood glucose, pyruvate and lactate and the severity of intoxication in
39 cases of acute carbon monoxide poisoning in man. Arch. Toxicol. 57, 196–199,
1985.
11. Thom, S.R., and Weaver, L.K. Carbon monoxide poisoning, Hyperbaric Oxygen
2003 Indications and Results: The Hyperbaric Oxygen Therapy Committee Report,
Feldmeier, J.J., Ed, Chap 2. Dunkirk, M.D., Undersea Hyperb. Med. Soc., 2003, pp.
11–17.
12. Hampson, N., Dunford, R.G, Kramer, C.C et al. Selection criteria utilized for hyper-
baric oxygen treatment of carbon monoxide poisoning. J. Emerg. Med. 13, 227–231,
1995.
13. Watson, W.A., Litovitz, T.L., Rodgers, G.C et al. 2004 Annual Report of the American
Association of Poison Control Centers Toxic Exposure Surveillance System. Am. J.
Emerg. Med. 23, 589–666, 2005.
14. Buckley, N.A., Isbister, G.K., Stokes, B et al. Hyperbaric oxygen for carbon monoxide
poisoning: a systematic review and critical analysis of the evidence. Toxicol. Rev. 24,
75–92, 2005.
15. Juurlink, D.N., Buckley, N.A., Stanbrook, M.B., Isbister, G.K., Bennett, M., and
McGuigan, M.A. Hyperbaric oxygen for carbon monoxide poisoning. Cochrane
Database of Systemic Reviews 4(2), CD002041, 2006. (Abstract)
16. Silver, S., Smith, C., and Worster, A. Should hyperbaric oxygen be used for carbon
monoxide poisoning? Can. J. Emerg. Med. 8, 43–46, 2006.
17. Judge, B.S., and Brown, M.D. To dive or not to dive? Use of hyperbaric oxygen
therapy to prevent neurological sequelae in patients acutely poisoned with carbon
monoxide. Ann. Emerg. Med. 46, 462–464, 2005.
18. Levasseur, L., Galliot-Guilley, M., Richter, F et al. Effects of mode of inhalation
of carbon monoxide and of normobaric oxygen administration on carbon monoxide
elimination from the blood. Hum. Exp. Toxicol. 15, 898–903, 1996.
19. Weaver, L.K., Howe, S., Hopkins, R et al. Carboxyhemoglobin half-life in carbon
monoxide-poisoned patients treated with 100% oxygen at atmospheric pressure. Chest
117, 801–808, 2000.
20. Myers, R.A., Jones, D.W., and Britten, J.S. Carbon monoxide half-life study, In:
Proceedings of the 9th International Congress on Hyperbaric Medicine, Kindwall,
E.P., Ed, Flagstaff, AZ, Best Publishing, 1987, pp. 263–266.
21. Sasaki, T. On half-clearance of carbon monoxide hemoglobin in blood during
hyperbaric oxygen therapy. Bull. Tokyo Med. Dent. Univ. 22, 63–77, 1975.
22. Boerema, I., Meyne, I., Brummelkamp, W.H et al. Life without blood. Arch. Chir.
Neer. 11, 70–83, 1959.
mechanism of toxicity, and hyperbaric oxygen. Ann. Emerg. Med. 24, 242–248, 1994.
24. Myers, R.A.M., and Britten, J.S. Are arterial blood gases of value in treatment
decisions for carbon monoxide poisoning? Crit. Care. Med. 17, 139–142, 1989.
25. Brown, S.D., and Piantodosi, C.A. Recovery of energy metabolism in rat brain after
carbon monoxide hypoxia. J. Clin. Invest. 89, 666–672, 1991.
26. Thom, S.R. Antagonism of carbon monoxide-mediated brain lipid peroxidation by
hyperbaric oxygen. Toxicol. Appl. Pharmacol. 105, 340–344, 1990.
8417: “8417_c017” — 2007/9/11 — 12:15 — page 386 — #12

27. Thom, S.R. Leukocytes in carbon monoxide-mediated brain oxidative injury. Toxicol.
Appl. Pharmacol. 123, 234–247, 1993.
28. Ishimaru, H., Katoh, A., Suzuki, H et al. Effects of N-methyl-D-aspartate receptor
antagonists on carbon monoxide-induced brain damage in mice. J. Pharmacol. Exp.
Ther. 261, 349–352, 1992.
29. Thom, S.R., Fisher, D., Zhang, J et al. Neuronal nitric oxide synthase and N-methyl-D-
aspartate neurons in experimental carbon monoxide poisoning. Tox. Appl. Pharmacol.
194, 280–295, 2004.
30. Piantadosi, C.A, Zhang, J., Levin, E.D et al. Apoptosis and delayed neuronal dam-
age after carbon monoxide poisoning in the rat. Exper. Neurology 147, 103–114,
1997.
31. Thom, S.R., Fisher, D., Xu, Y.A et al. Adaptive responses and apoptosis in endothelial
cells exposed to carbon monoxide. Proc. Natl. Acad. Sci., USA 97, 1305–1310,
2000.
32. Thom, S.R. Learning dysfunction and metabolic defects in globus pallidus and hip-
pocampus after CO poisoning in a rat model. Undersea Hyperb. Med. 23 (Suppl.),
20. 1997. (Abstract)
33. Ischiropoulos, H., Beers, M.F., Ohnishi, S.T et al. Nitric oxide and perivascular
tyrosine nitration following carbon monoxide poisoning in the rat. J. Clin. Invest. 97,
2260–2267, 1997.
34. Nabeshima, T., Katoh, A., Ishimaru, H et al. Carbon monoxide-induces delayed
amnesia, delayed neuronal Death and change in acetylcholine concentrations in mice.
J. Pharmacol. Exp. Ther. 256, 378–384, 1991.
35. Thom, S.R. Functional inhibition of leukocyte B2 integrins by hyperbaric oxygen
in carbon monoxide-mediated brain injury in rats. Toxicol. Appl. Pharmacol. 123,
248–256, 1993.
36. Thom, S.R., Bhopale, V.M., and Fisher, D. Hyperbaric oxygen reduces delayed
immune-mediated neuropathology in experimental carbon monoxide toxicity. Tox.
Appl. Pharmacol. 213, 152–159, 2006.
37. Jiang, J., and Tyssebotn, I. Normobaric and hyperbaric treatment of acute carbon
monoxide poisoning in rats. Undersea Hyperb. Med. 24, 107–116, 1997.
38. Jiang, J., and Tyssebotn, I. Cerebrospinal fluid pressure changes after acute carbon
monoxide poisoning and therapeutic effects of normobaric and hyperbaric oxygen in
conscious rats. Undersea Hyperb. Med. 24, 245–254, 1997.
39. Gilmer, B., Kilkenny, J., Tomaszewski, C et al. Hyperbaric oxygen does not prevent
neurologic sequelae after carbon monoxide poisoning. Acad. Emerg. Med. 9, 1–8,
2002.
40. Norkool, D.M., and Kirkpatrick, J.N. Treatment of acute carbon monoxide poisoning
with hyperbaric oxygen: a review of 115 cases. Ann. Emerg. Med. 14, 1168–1171,
1985.
41. Goulon, M., Barios, A., Raphin, M et al. Carbon monoxide poisoning and acute
anoxia due to breathing coal gas and hydrocarbons. Ann. Med. Intern. 120, 335–349,
1969.
42. Smith, G.I., and Sharp, G.R. Treatment of carbon monoxide poisoning with oxygen
under pressure. Lancet 2, 905–906, 1960.
43. Brown, S.D., and Piantadosi, C.A. Hyperbaric oxygen for carbon monoxide
poisoning. Lancet 2, 1032, 1989.
44. Scheinkestel, C.D., Bailey, M., Myles, P.S et al. Hyperbaric or normobaric oxygen
for acute carbon monoxide poisoning: a randomised controlled clinical trial. Med. J.
Aust. 170, 203–210, 1999.
8417: “8417_c017” — 2007/9/11 — 12:15 — page 387 — #13

45. Schiltz, K.L. Failure to assess motivation, need to consider psychiatric variables,
and absence of comprehensive examination: a skeptical review of neuropsycholo-
gic assessment in carbon monoxide research. Undersea Hyperb. Med. 27, 48–50,
2000.
46. Thom, S.R., Mendiguren, I., Hardy, K.R et al. Inhibition of human neutrohil B2
integrin-dependent adherence by hyperbaric oxygen. Am. J. Physiol. (Cell Physiol)
272, C770–C777, 1997.
47. Thom, S.R. Carbon monoxide-mediated brain lipid peroxidation in the rat. J. Appl.
Physiol. 68, 997–1003, 1990.
48. Mathieu, D., Wattel, F., Mathieu-Nolf, M., Durak, C., Tempe, J.P., Bouachour, G.,
and Sainty, J.M. Randomized prospective study comparing the effect of HBO versus
12 hours of NBO in non comatose CO poisoned patients. Undersea Hyperb. Med.
23(Suppl.), 7–8. 1996. (Abstract)
49. Wolf, S.J., Lavonas. E.J., Sloan. E.P et al. Clinical Policy: critical issues in the
management of adult patients presenting to the emergency department with acute
symptomatic carbon monoxide poisoning. Ann. Emerg. Med. 49, in press, 2007.
50. Sloan. E.P., Murphy. D.G., Hart, R et al. Complications and protocol considerations
in carbon monoxide-poisoned patients who require hyperbaric oxygen therapy: report
from a ten-year experience. Ann. Emerg. Med. 18, 629–634, 1989.
51. Hampson, N.B., Little, C.E. Hyperbaric treatment of patients with carbon monoxide
poisoning in the United States. Undersea Hyperb. Med. 32, 21–26, 2005
52. Hampson, N.B., Simonson, S.G., Kramer, C.C et al. Central nervous system oxygen
toxicity during hyperbaric treatment of patients with carbon monoxide poisoning.
Undersea Hyperb. Med. 23, 215–219, 1996.
53. Tomaszewski, C. Carbon Monoxide, In: Goldfrank’s Toxicologic Emergencies, Gold-
frank, L.R., Flomenbaum, N.E., Hoffman, R.S., Howland, M.A., Lewin. N,A.,
Nelson, L.S (Eds): Chap 120, New York, McGraw-Hill, 2006, pp. 1689–1704.
54. Jones, J.S., Lagasse, J., Zimmerman, G. Computed tomography findings after acute
carbon monoxide poisoning. Am. J. Emerg. Med. 12, 448–451, 1994.
55. Benignus, V.A., Kafer, E.R., Muller, K.E et al. Absence of symptoms with
carboxyhemoglobin levels of 16–23%. Neurotoxicol. Teratol. 9, 345–348, 1987.
56. Davis, S.M., Levy, R.C. High carboxyhemoglobin level without acute or chronic
findings. J. Emerg. Med. 1, 539–542, 1984.
57. Parkinson, R.B., Hopkins, R.O., Cleavinger, H.B et al. White matter hyperintensities
and neuropsychological outcome following carbon monoxide poisoning. Neurology
58, 1525–1532, 2002.
58. Thom, S.R. Hyperbaric-oxygen therapy for acute carbon monoxide poisoning. NEJM
347, 1105–1106, 2002.
three year follow-up. BMJ 1973, 318, 1973.
60. Thom, S.R., Bhopale, V.M., Fisher, D et al. Delayed neuropathology after carbon
monoxide poisoning is immune-mediated. Proc. Natl. Acad. Sci., USA 101, 13660–
13665, 2004.
61. Okeda, R., Runata, N., Takano, T et al. The pathogenesis of carbon monoxide
encephalopathy in the acute phase—physiological and morphological conditions.
Acta. Neuropathol. 54, 1–10, 1981.
62. Bogusz, M., Cholewa, L., Pach, J et al. A comparison of two types of acute carbon
monoxide poisoning. Arch. Toxicol. 33, 141–149, 1975.
63. Wara-Wasoweki, J., Myslak, Z., Graczyk, M et al. An attempt at comparing the results
of carboxyhemoglobin level in blood and gasometric determinations in capillary blood
8417: “8417_c017” — 2007/9/11 — 12:15 — page 388 — #14

in cases of carbon monoxide poisoning when treatment began at the place of the
accident. Anaesthesia, Resusc. Inten. Therapy 4, 249, 1976.
64. Burney, R.E., Wu, S.C., and Nemiroff, M.J. Mass carbon monoxide poisoning: clin-
ical effects and results of treatment in 184 victims. Ann. Emerg. Med. 11, 399,
1982.
65. Piantadosi, C.A. Diagnosis and treatment of carbon monoxide poisoning. Respir.
Care. Clin. N. Am. 5, 183–202, 1999.
66. Turner, M., Esaw, M., and Clark, R.J. Carbon monoxide poisoning treated with
hyperbaric oxygen: metabolic acidosis as a predictor of treatment requirements. [see
comments.]. J. Accid. Emerg. Med. 16, 96–98, 1999.
67. Larkin, J.M., and Moylan, J.A. Treatment of carbon monoxide poisoning: prognostic
factors. J. Trauma 16, 111–114, 1976.
68. Amitai, Y., Zlotogorski, Z., Golan-Katzav, V et al. Neuropsychological impairment
from acute low-level exposure to carbon monoxide. Arch. Neurol. 55, 845–848, 1998.
69. Olson, K.R., Carbon monoxide poisoning: mechanisms, presentation, and controver-
sies in management. J. Emerg. Med. 1, 233–243, 1984.
70. Thom, S.R., Hyperbaric Oxygen, In: Goldfrank’s Toxicologic Emergencies, Flomen-
baum, N.E., Goldfrank, L.R., Hoffman, R.S., Howland, M.A., Lewin, N.A., Nelson,
L.S., Eds, Chap A34. New York, McGraw-Hill, 2006, pp. 1705–1711.
71. Hampson, N.B., and Zmaeff, J.L. Outcome of patients experiencing cardiac arrest
with carbon monoxide poisoning treated with hyperbaric oxygen. Ann. Emerg. Med.
38, 36–41, 2001.
72. Goulon, M., Barios, A., and Rapin, M. Carbon monoxide poisoning and acute anoxia
due to breathing coal gas and hydrocarbons. J. Hyperbar. Med. 1, 23–41, 1986.
73. Scheinkestel, C.D., Bailey, M., Myles, P.S et al. Hyperbaric or normobaric oxygen for
acute carbon monoxide poisoning: a randomized controlled clinical trial. Undersea
Hyperb. Med. 27, 163–164, 2000.
74. Dean, B.S., Verdile, V.P., and Krenzelok, E.P. Coma reversal with cerebral dysfunc-
tion recovery after repetitive hyperbaric oxygen therapy for severe carbon monoxide
poisoning. Am. J. Emerg. Med. 11, 616–618, 1993.
75. Gorman, D.F., Clayton, D., Gilligan, J.E et al. A longitudinal study of 100 consecutive
admissions for carbon monoxide poisoning to The Royal Adelaide Hospital. Undersea
Hyperb. Med. 20, 311–316, 1992.
76. Thom, S.R., Kang, M., Fisher, D et al. Release of glutathione from erythrocytes and
other markers of oxidative stress in carbon monoxide poisoning. J. Appl. Physiol. 82,
1424–1432, 1997.
77. Thom, S.R., Ohnishi, S.T., and Ischiropoulos, H. Nitric oxide release by platelets
inhibits neutrophil B2 integrin function following acute carbon monoxide poisoning.
Toxicol. Appl. Pharmacol. 128, 105–110, 1994.
78. Costa, L.G., and Manzo, L.Biomarkers in occupational neurotoxicology, In: Occu-
pational Neurotoxicology, Costa LG, Manzo L., Eds, Boca Raton, CRC Press, 1998,
pp. 75–100.
79. Minana, M.D., Corbalan, R., Montoliu, C et al. Chronic hyperammonemia in rats
impairs activation of soluble guanylate cyclase in neurons and in lymphocytes:
a putative peripheral marker for neurological alterations. BBRC 257, 405–409,
1999.
80. Castoldi, A.F., Coccini, T., Randine, G et al. Lumphocyte cytochrome c oxidase,
cyclic GMP and cholinergic muscarinic receptors as peripheral indicators of carbon
monoxide neurotoxicity after acute and repeated exposure in the rat. Life Sciences 78,
1915–1924, 2006.
8417: “8417_c017” — 2007/9/11 — 12:15 — page 389 — #15

81. Brvar, M., Mozina, H., and Osredkar, J. Prognostic value of S100B protein in carbon
monoxide-poisoned rats. Crit. Care Med. 32, 2128–2130, 2004.
82. Brvar, M., Mozina, H., Osredkar, J et al. S100B protein in carbon monoxide poisoning:
a pilot study. Resuscitation 61, 357–360, 2004.
83. Brvar, M., Finderle, Z., Suput, D et al. S100B protein in conscious carbon monoxide-
poisoned rats treated with normobaric or hyperbaric oxygen. Crit. Care Med. 34,
2228–2230, 2006.
84. Rasmussen, L.S., Poulsen, M.G., Christiansen, M et al. Biochemical markers for brain
damage after carbon monoxide poisoning. Acta. Anaesthes. Scand. 48, 469–473,
2004.
85. Sesay, M., Bidabe, A.M., Guyot, M et al. Regional cerebral blood flow measure-
ments with Xenon-Ct in prediction of delayed encephalopathy after carbon monoxide
intoxication. Acta. Neurol. Scand. 166 (Suppl.), 22–27, 1996.
86. Kanaya, N., Imaizumi, H., Nakyama, M et al. The utility of MRI in acute stage of
carbon monoxide poisoning. Intensive Care Med. 18, 371–372, 1992.
87. Kawada, N., Ochiai, N., and Kuzuhara, S. Diffusion MRI in acute carbon monoxide
poisoning. Intern. Med. 43, 639–640, 2004.
88. Gale, S.D., Hopkins, R.O., Weaver, L.K et al. MRI, quantitative MRI, SPECT, and
neuropsychological findings following carbon monoxide poisoning. Brain Inj. 13,
229–243, 1999.
89. Choi, I.S., Kim, S.K., Lee, S.S. Evaluation of outcome of delayed neurologic sequelae
after carbon monoxide poisoning by technetium-99m hexamethylpropylene amine
oxime brain single photon emission computed tomography. Eur. Neurol. 35, 137–142,
1995.
90. De Reuck, J., Decoo, D., Lemahieu, I et al. A positron emission tomography study
of patients with acute carbon monoxide poisoning treated by hyperbaric oxygen. J.
Neurology 240, 430–434, 1993.
8417: “8417_c017” — 2007/9/11 — 12:15 — page 390 — #16

18 Hyperbaric Oxygen for
Acute Carbon Monoxide
Poisoning: Useful
Therapy or Unfulfilled
Promise?
Carlos D. Scheinkestel and Ian L. Millar
CONTENTS
18.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392

18.2 Hyperbaric Oxygen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393
18.2.1 Hyperbaric Oxygen Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
18.2.2 Basic Science Overview in Use of Hyperbaric Oxygen . . . . . . . . . . 395
18.2.3 HBO for Carbon Monoxide Poisoning: Human Evidence. . . . . . . . 398
18.2.3.1 “Trial of Normobaric and Hyperbaric Oxygen for
Acute Carbon Monoxide Intoxication” by Raphael . . . . 400
18.2.3.2 “Non-Comatose Patients with Acute Carbon
Monoxide Poisoning: Hyperbaric or Normobaric
Oxygenation?” By Ducasse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
18.2.3.3 “Delayed Neuropsychologic Sequelae after Carbon
Monoxide Poisoning: Prevention by Treatment With
Hyperbaric Oxygen” by Thom . . . . . . . . . . . . . . . . . . . . . . . . . . 402
18.2.3.4 “Randomized Prospective Study Comparing the Effect
of HBO Versus 12 h NBO in Noncomatose CO
Poisoned Patients: Results of the Interim Analysis” by
Mathieu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404
18.2.3.5 “Managing Carbon Monoxide Poisoning with
Hyperbaric Oxygen” by Raphael . . . . . . . . . . . . . . . . . . . . . . . . 405
18.2.3.6 “Hyperbaric or Normobaric Oxygen for Acute Carbon
Monoxide Poisoning: a Randomised Controlled
Clinical Trial” by Scheinkestel . . . . . . . . . . . . . . . . . . . . . . . . . . 406
18.2.3.7 “Hyperbaric Oxygen for Acute Carbon Monoxide
Poisoning” by Weaver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
18.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
391
8417: “8417_c018” — 2007/9/19 — 13:22 — page 391 — #1

18.4 Where to Now? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426
18.1 INTRODUCTION
Superficial consideration suggests that hyperbaric oxygen (HBO) should be the
optimal antidote for acute carbon monoxide (CO) poisoning. It is a well-established
form of therapy, albeit with a limited distribution of facilities, and the side-effect
and complication profile is well established, manageable and rarely associated with
longterm sequelae. By generating the highest tolerable intravascular partial pressure
of oxygen, HBO provides the most rapid means available of simultaneously reversing
cellular hypoxia and accelerating the elimination of CO, not only from its binding
with hemoglobin, but also from intracellular binding sites.
Human clinical use of HBO for this purpose appears to have first occurred in
19421 and was proposed again by Pace in the publication “Acceleration of CO elim-
ination in man by high pressure oxygen” in 1950. This US Navy study measured
acceleration of elimination of CO in volunteers treated with 2.5 atmospheres absolute
(ATA) pressure oxygen after a brief loading exposure with CO, producing carboxy-
hemoglobin (COHb) in the range of 20–30% 2 . CO poisoning subsequently became
established as one of the principal indications for HBO therapy (HBOT) as clinical
use of HBO grew in the 1960s. Acute CO poisoning has subsequently been listed as an
indication for HBO in the guidelines produced by multiple international hyperbaric
medicine societies.
Despite this, HBO has failed to become the standard of care for CO poisoning. For
instance, in the US northwest, it was estimated in 1994 that only 6.9% of CO poisoning
patients presenting to emergency departments received HBO.3 The historical reasons
for this were probably related to the limited availability of hyperbaric chambers,
limited awareness of this form of therapy, a reluctance to expose patients to the
hazards of transfer for treatment, as well as a degree of general scepticism regarding
the therapy.
In recent times, clinical trial results and debate surrounding these have ques-
tioned the effectiveness of HBO in producing significant improvements in outcome.
In the US, the number of CO-poisoned patients treated with HBO has remained rel-
atively constant since 1992 at about 1500 per annum4 despite a more than doubling
of hyperbaric facilities in that time. The lack of increase in patients treated in this
environment may be a result of increasing doubt as to the effectiveness of HBOT,
although any change in usage must be interpreted in the light of changes to the rates
of poisoning. In the US, survey results and toxicology service data indicate that while
the CO related death rate has fallen, nonfatal CO emergency calls have remained
relatively stable. This correlates with the situation regarding HBO use.4 In Australia,
by contrast, the number of CO poisoning patients treated by hyperbaric units fell from
240 per annum to 60 over the last 10 years. In Australia the most common cause of
CO poisoning has been automobile exhaust suicide attempt. The incidence of this has
fallen in association with depression recognition, suicide prevention campaigns and
the reduction in CO production mandated for newer model vehicles.5,6 At the same
8417: “8417_c018” — 2007/9/19 — 13:22 — page 392 — #2

Hyperbaric Oxygen for Acute Carbon Monoxide Poisoning 393
time, it is likely that public health measures have reduced occupational and domestic
exposures. The decline in CO poisoning numbers treated with HBO may thus be
as much a result of reduced incidence of poisoning as it is of reduced provision of
treatment arising from implementation of evidence from clinical trials.
Meanwhile, our understanding of the biology, pharmacology, and toxicology of
both CO and HBO has become much more complex, and it is clear that both these
low molecular weight gases (i.e., CO and oxygen) are two edged swords—essential
elements of normal physiology but toxic at higher doses, and thus potentially useful
therapeutically yet capable of harm at doses which overlap with those required for
therapeutic effect.7−13 Alongside the sometime impassioned debate about interpret-
ation of clinical trials, there is thus a fascinating and rapidly evolving stream of basic
science research which will hopefully soon deliver us a better basis for designing
future clinical trials aimed at answering the question of whether HBO has a place in
the routine treatment of CO poisoning, and if so, in what clinical settings, and at what
dose.
18.2 HYPERBARIC OXYGEN

HBO provides a means of elevating oxygen concentration in the body to the maximum
tolerable, in order to seek therapeutic effects not achievable with administration of
oxygen in the normal ward environment, with its 1 ATA pressure at sea level, or less
at altitude. Hyperbaric chambers are used to expose patients to pressures that are
normally 2–3 times atmospheric, that is, 2–3 ATA. Combined with 100% inspired
oxygen breathing, this can deliver arterial partial pressures as high as 15–20 times
normal. As a result, not only is the volume of oxygen delivered to tissues increased
but more importantly, intracellular partial pressures of oxygen increase well beyond
normal, providing potentially useful effects both directly and by inducing a response
to the oxidant stress generated.
The technology for providing HBOT is well established. Pressure vessels for
human occupancy have been built for clinical use for over 150 years, although the
early proponents of HBOT mostly believed that it was pressure that was therapeutic
and increased levels of oxygen were not used, limiting the therapeutic effects possible.
During the 1930s, Cunningham used an air pressurized hyperbaric chamber for
patients with severe acute respiratory disease, probably maintaining life through
a crisis period by effectively increasing the partial pressure of oxygen available
to patients, albeit in a much more complicated fashion than achievable now via
administration of facemask (normobaric) oxygen (NBO) therapy.
The clinical combination of increased oxygen with increased pressure was first
promoted in the 1950s as a means of increasing the time window for cardiovascular
surgery requiring circulatory arrest and in the 1960s a significant number of large
hyperbaric chambers were constructed in hospitals around the world. Many of these
are still in use today although surgery is now rarely performed under pressure. These
operating theater chambers and smaller chambers evolved from commercial diving
industry recompression chambers. They have been the basis for the development of
the modern “multiplace” chamber. Over a similar period, single person, oxygen filled
8417: “8417_c018” — 2007/9/19 — 13:22 — page 393 — #3

“monoplace” chambers have evolved as an alternative, originally used to enable co-

administration of HBO and radiotherapy, but now widely used for both hospitalized
and ambulatory patients.
Ideally, hyperbaric facilities should be located inside the building structure of a
hospital, close to relevant patient care areas to minimize transport difficulties, and
they should have good critical care support. The most versatile multiplace chambers
have multiple compartments, are quiet and well lit, have doors that will readily admit a
trolley or even better a full sized patient bed and have a floor level matching that of the
surrounding building. In recent years, a number of manufacturers have provided rect-
angular chambers that have further advanced the aim of making hyperbaric chambers
as close to a normal clinical room as possible.
Multiplace chambers are pressurized with air to minimize fire risk and cost.
Oxygen must be administered to patients via a mask or hood, or in the case of intub-
ated patients, via a suitable ventilator. In most cases a hyperbaric nurse, doctor or
paramedic attends patients undergoing treatment which typically lasts from 1.5 to 2 h
overall. Although there are safety and functional limitations on what equipment can
be used inside a chamber, most major hospital facilities can provide intensive care
including positive pressure ventilation, arterial pressure monitoring, and inotrope
infusion.
There are many variations in chamber size and configuration but the predom-
inant alternative type of chamber in use is the single person, horizontal acrylic
cylinder style monoplace chamber. These are much smaller, cheaper, and easier to
install than properly designed clinical multiplace chambers and are thus in wide-
spread use in many parts of the world, especially in facilities that concentrate on
using HBO for problem wound cases and the late side effects of radiotherapy. While
most hyperbaric centers with a routinely used critical care capability have multiplace
chambers, monoplace chambers can be used for critical care patients by experienced
teams.14
18.2.1 HYPERBARIC OXYGEN EFFECTS

HBO provides it’s effects via multiple mechanisms. These can be grouped into those
related to gas dynamics, to restoration of function by normalization of oxygena-
tion and those that are a therapeutic result of the oxidative challenge achieved by
very high pO2 . Some, such as elimination of gas bubbles and those underlying
HBO’s anti-infective and wound healing promotion effects, are probably not dir-
ectly relevant to CO poisoning. It is interesting to note, however, that in many cases,
effects previously thought to be brought about merely by reversing hypoxia are now
understood to result from HBO triggering intracellular signaling via various oxid-
ative and nitric oxide related biochemistry.15−18 Other mechanisms of HBO that
would seem to have particular potential in CO poisoning include accelerating elimin-
ation of CO,2,19 reversing cellular hypoxia,19,20 reducing edema,21−25 up-regulating
antioxidant systems,26,27 inhibiting reperfusion injury18,28 and possibly modulating
other elements of the secondary injury cascade.29,30 Some models show reduced
cellular necrosis and apoptosis,31−37 although the mechanisms for these effects are
still emerging.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 394 — #4

A key element of HBO therapeutics is that hyperbaric sessions are of limited

duration—usually a maximum of around 2 h. The limited duration and intermittent
nature of HBO is an essential factor in the safety profile of this therapy. Pulmonary,
optic, and central nervous system (CNS) toxicity are well recognized complications
of excess exposure to oxygen at pressure and the threshold of toxicity overlaps with
the therapeutic dose range.38,39 Of particular interest is acute CNS oxygen toxicity,
manifested most frequently by short duration loss of consciousness (LOC) and tonic-
clonic seizure activity which self-terminates as toxic levels of oxygen fall in the
temporary absence of respiration. The incidence of acute CNS oxygen toxicity can be
very low in comparatively well, ambulatory patients, with Yildiz40 reporting only two
cases in 80,000 treatments,40 although most recent author’s estimates are in the order
of 3/10,000.39,41 The incidence rises significantly with higher pressure treatments
and results of this are seen in the treatment of acute patients such as those suffering
decompression illness.42 In CO poisoning patients suffering neurological injury, the
rate of acute CNS toxicity can be significant,43 with Sloan38 reporting a 5% incidence
in 297 patients seen over 10 years.38
As applies to CO poisoning, the rapidly evolving field of free-radical and
oxidant/antioxidant research has major implications for our understanding of the
therapeutic and toxic effects of HBO. Some years ago, there were significant con-
cerns that short-term benefits seen with HBO might be associated with at least some
burden of oxidative stress related problems such as acceleration of cardiovascular
disease or even premature aging. Reassuringly, the balance of findings to date sug-
gests that although HBO is an oxidative stressor, it up-regulates antioxidant systems
sufficiently to avoid significant net damage or even, in some cases, to provide a
paradoxical and beneficial net antioxidant effect. An important caution must remain,
however, that the net result of HBO probably depends upon the physiology of the host
receiving HBO. Without adequate nutritional substrates for antioxidant production,
when faced with excessive total oxidative stress or when specific processes are in
train such as lipid peroxidation, HBO might well exacerbate the problem rather than
ameliorate it, at least in some doses or at some time points.
18.2.2 BASIC SCIENCE OVERVIEW IN USE OF HYPERBARIC

OXYGEN
The history of using HBO for CO poisoning is sometimes claimed to date back
to Haldane’s classic 1895 paper: “The relation of the action of carbonic oxide to
oxygen tension”.44 This paper in fact demonstrates that, in a mouse model, the high
levels of oxygen achievable in the hyperbaric environment can make normally lethal
levels of CO exposure tolerable without mortality. Confusing this demonstration
of the competitive inhibition of CO uptake with the therapeutic use of HBO after
CO poisoning underscores one of the major challenges to date in trying to draw
clinically useful conclusions from animal studies. Most human CO poisoning patients
experience a delay of many hours between the termination of CO poisoning and the
start of HBO, while most animal researchers have initiated HBO soon after termination
of CO exposure, and often immediately. The effectiveness of HBO and indeed the
entire basis for interaction between HBO and CO poisoning pathology are likely to
8417: “8417_c018” — 2007/9/19 — 13:22 — page 395 — #5

be very different at different time points after exposure: from the acute rescue and
resuscitation phase, through the period when it may be possible to prevent secondary
injury, to modulating established secondary injury processes, attempting to prevent
delayed neurological sequelae (DNS) and even using HBO to accelerate recovery or
for delayed treatment for established residuae.
Although there is some debate regarding the degree to which the acute manifesta-
tions of CO poisoning are mediated by a shortfall in intracellular oxygen availability,
there is no doubt that hypoxia is a significant factor, at least in severe cases.45,46 . It may
be that the level of hypoxia that is critical for any particular individual’s physiology is
a key determinant of susceptibility to toxicity at any given level of CO exposure. The
elderly, collapsed patients and those affected by certain drugs and co-toxins will have
limited capacity for compensatory cardiac output and cerebral blood flow increase
as the oxygen carrying-capacity of blood falls. This may underlie the susceptibil-
ity of these populations to CO poisoning. High altitude (hypoxic/hypoxia) exposure
increases the toxicity of CO.47−52
HBO dissolves sufficient oxygen in plasma to meet physiological demands, mean-
ing that oxygen can be immediately delivered to cells despite the presence of high
COHb levels. Clinically this can be demonstrated by resolution of abnormal electro-
cardiogram (ECG) activity and rapid recovery of consciousness, which are sometimes
but not always seen when HBO is used for CO poisoning. HBO also reduces COHb
levels much more rapidly than can be achieved by NBO. Although there are inter-
individual and inter-study differences in actual numbers reported, the human half-life
reduction is in the order of 4.5 h on air, to 90 min on 100% oxygen, to 20 min
with HBO at 2.5–3 ATA.53 In the patient with significant cellular hypoxia result-
ing from CO poisoning, HBO can therefore expedite resolution of this problem
and this would be expected to not only provide immediate benefit but to reduce
the risk of ongoing hypoxia compounding existing injury. Because the uptake of
CO within cells is slower than the association of CO with hemoglobin, it could
be hoped that early HBO might prevent the onset of cellular toxicity.46 HBO will
also speed the dissociation of CO from intracellular sites, although the proportional
acceleration is different for different biochemical processes and varies with species
studied.
All of this would seem to provide a case for the immediate use of HBO for CO
poisoning where this is can be made available. Some high risk industrial site and
paramedic rescue systems have been established using portable hyperbaric chambers
for this purpose.54,55 Whether very early HBO actually produces more survivors
and better outcomes than NBO has not been definitively established, given that all
major clinical trials to date have had insufficient numbers of patients treated within,
say, the first hour after CO poisoning. Of the randomized human studies, only that
of Ducasse56 came near to this with all patients treated within 2 h of CO exposure.
However, this study was small and used outcome measures that are difficult to interpret
as will be discussed.
Most animal studies to date do seem relevant to the question of early therapy
benefit, however with therapy usually provided to exposed animals commencing
either immediately after cessation of poisoning or after a relatively short delay—most
usually between 15 and 60 min.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 396 — #6

In many animal models of CO poisoning, HBO has shown benefit on a range of

measures including survival, recovery of consciousness, recovery of brain function
and reduction in postmortem neuropathology, including reductions in cellular necrosis
and apoptosis. Despite early initiation of treatment, it is noteworthy that not all animal
studies show benefit of HBO over NBO, however, or even of any kind of oxygen
therapy over air breathing. Amongst recent studies looking at mechanisms underlying
longterm sequelae, Brvar et al.57 did demonstrate that in rats poisoned without loss of
consciousness (LOC), 30 min of 3 ATA HBO, but not 30 min of NBO, significantly
reduced an immediate post-CO rise in blood levels of S100B, an astroglial structural
protein that shows promise as a marker of CO poisoning severity.57 The CO exposure
was 3000 ppm for 60 min. The same group has also reported however, in the same
model, that both NBO and HBO resulted in a dramatic but effectively equal reduction
in pyknotic cells in the hippocampus when the rats were sacrificed two weeks later.58
By contrast, in a mouse model of more severe poisoning with LOC, Gilmer et al.59
found that neither NBO nor HBO commenced 15 min after poisoning provided any
significant protection against learning dysfunction or hippocampal pyknosis.59 LOC
was induced in this model with a 4–9 min exposure to 50,000 ppm of CO after 40 min
at 1000 ppm. It could be argued that this produces a sufficiently severe injury to be
irreversible. This experimental poisoning regimen can be compared with Thom’s well
established rat model which uses 3000 ppm for up to 20 min in order to induce LOC
after 40 min of exposure to 1000 ppm.60 Treatment commences after a more clinically
relevant interval of 45 min and the HBO regimen used is 45 min at 2.8ATA. This model
was used to demonstrate that another structural protein, myelin basic protein (MBP) is
released after CO poisoning and triggers delayed neuropathology via a cell mediated
immune response.61 Most recently, Thom62 has shown that intervention with HBO,
but not NBO, reduces learning dysfunction and the sensitization of lymphocytes to
MBP, but this effect is only partial.62 This study adds to work showing that early
HBO can block leukocyte adherence to endothelium and reduces leukocyte mediated
oxidative damage that compounds CO-related injury upon reoxygenation.63
Interpreted together, the mechanisms of action of HBO and the studies referred
to above suggest there is potential for benefit from early HBO. It is not clear,
however, to what extent reversal of acute toxicity is responsible for outcome benefit as
opposed to secondary injury reduction via mechanisms such as modulation of delayed
immune-response and up-regulation of antioxidant systems. It is also not clear whether
these effects are useful if HBO is applied much later after termination of CO exposure.
Oxygen dosing is another major variable, with various pressures between 1.5
and 3.0 ATA used in therapeutic research, for durations between 45 min and 2 h.
Single dose therapy and multiple dose therapy at varying intervals, have been used.
In clinical practice, therapeutic oxygen is used in variable amounts and durations
before and during hospital admission and therapeutic and toxic effect in animals
should ideally be studied with varying doses of HBO and both with and without NBO
in the periods before and after HBO, if findings are to be translated into human clinical
practice or research design.
It is also important to remember that developing a satisfactory animal model is
critical to laboratory research. An experimental poisoning regimen must be found
that produces sufficient toxicity to be associated with measurable pathology but the
8417: “8417_c018” — 2007/9/19 — 13:22 — page 397 — #7

animals must survive and the damage should not be so severe as to be irreversible.
Such controlled exposures are not the reality for humans and even in the animal
models, demonstrated benefit may only apply in certain degrees of injury severity
and both the optimal HBO dose and level of poisoning amenable to therapy seem
likely to be species specific.
The issue of animal model generalizability to humans is particularly important in
CO poisoning. Most CO poisoning studies are done in rodents, which have the advant-
ages of being relatively cheap, easy to handle and importantly, they are now available
in a wide range of genetically altered strains which can be used to test biochemical
mechanism hypotheses. Although rodents are mammals, they have critical differences
from larger species. Their small body mass and high metabolic rate result in much
more rapid uptake and distribution of gases than is the case for humans.52,64 More
importantly, rodents are relatively hypoxia and carbon dioxide resistant, presumably
evolved traits associated with survival in the rodent environment but ones which
may have critical impacts on the interpretability of findings regarding CO poisoning,
given the importance of hypoxia in the pathology of CO poisoning. This is not to
argue against rodent research; it has generated and will continue to generate signi-
ficant advances in our knowledge regarding the mechanisms of both CO and HBO.
Generating clinical outcome predictions, case selection criteria or dosing require-
ments from rodent work has significant potential for error however. Gorman argues
this in reporting results from his instrumented sheep model of CO poisoning. In this
model, animals lose consciousness with an exposure to 1% CO for 120 min and seem
to tolerate LOC-inducing levels of CO poisoning without the same neuropathology
described in other models, although some peri-ventricular white matter infarcts and
gliosis do occur. In addition to species differences, tolerance almost certainly depends
upon cardiovascular and cerebrovascular reflex responses being intact and sufficient
to maintain ongoing cerebral oxygenation. This has required chronic instrumentation
of the animals, as if CO poisoning is provided too early after cannulation of the carotid
vessels, more severe neuropathology was seen, localized to the side of instrumenta-
tion possibly related to impairment in the normal compensatory increase in cerebral
blood flow.45,65−68
In summary, most animal studies suggest benefit should arise from using HBO
as a therapy for CO poisoning, but much caution is needed in translating findings
from pure and highly controlled CO exposure in healthy animals, treated rapidly with
HBO, to the messy and complicated realities of human CO poisoning.
18.2.3 HBO FOR CARBON MONOXIDE POISONING: HUMAN

EVIDENCE
In October 2006, the American College of Emergency Medicine released its draft
“Clinical Policy: Critical Issues in the Management of Adult Patients Presenting to the
Emergency Department with Acute Symptomatic CO Poisoning”.69 This document
uses an evidence-based approach. In response to the question “Should HBOT be used
for the treatment of patients with CO poisoning; and can clinical or laboratory criteria
identify CO-poisoned patients who are most likely to benefit from this therapy?”—the
best it could do was “Level C” recommendations, that is, recommendations based
8417: “8417_c018” — 2007/9/19 — 13:22 — page 398 — #8

on preliminary, inconclusive or conflicting evidence. The Level C recommendations

were:
1. HBOT should not be mandated for the treatment of CO poisoning.

2. HBOT remains a therapeutic option to potentially reduce the incidence of
neurological sequelae.
3. Do not use COHb levels alone to choose therapy in CO poisoning.
4. The available evidence does not identify either a subgroup of patients for
whom HBOT is clearly indicated, or a subgroup of patients who clearly
have no potential to benefit from HBO.
Prior to this, Phin,70 in reviewing the evidence for therapy of CO poisoning,

concluded that there was no evidence to support HBOT being of clear effectiveness.
Evidence—based on call: Acute Medicine in 200171 recommended avoiding HBO in
CO therapy, classifying this as an A class recommendation.
In November 2000, The Medicare Services Advisory Committee of the
Department of Health and Aged Care of the Australian Government,72 released its
Report on HBOT. It concluded that there should be no support for “public fund-
ing for HBOT in either a multi-place or mono-place chamber for CO poisoning.”
On February 9, 2001, the Minister for Health and Aged Care accepted this
recommendation and such funding was withdrawn.
This recommendation was based on a systematic review by Juurlink et al.73 in 2000
on behalf of the Cochrane Database of Systematic Reviews. “The review collected six
reports of randomized controlled trials involving nonpregnant adults acutely poisoned
with CO. Only three studies scored ≥ 3/5 on the Jadad quality scale (assessment based
on randomization, double blinding and withdrawals and dropouts),74 and these three
by Raphael,75 Thom,76 and Scheinkestel77 were analyzed.
Juurlink et al.73 found that the severity of CO poisoning varied between trials and
each trial employed different doses of HBO. The results for a total of 455 patients were
available for analysis. Nonspecific neurological symptoms were present in 81/237
patients (34.1%) in the HBO group compared to 81/218 (37.2%) in the NBO group
[odds ratio (OR) = 0.82; 95% Confidence Interval (CI) = 0.4, 1.66]. This systematic
review failed to demonstrate a significant reduction in neurologic sequelae following
HBOT for CO poisoning.”
The German government has similarly recommended discontinuation of HBOT
for CO poisoning78 stating “no studies could be identified, which could justify a
continuation of HBOT for CO poisoning.”
The review by Juurlink et al.73 underwent a substantive amendment in November,
2004 and was subsequently published in the Cochrane Library 2006.73 In this review,
six trials were evaluated, the previous three plus a further one from Raphael,79 one
from Mathieu,80 and one from Weaver.81 Of these, the latest one from Raphael is in
abstract form only and that from Mathieu is an interim analysis.
Using the Jadad scale again,74 Mathieu’s trial scored 2/5, the two from Raphael
and the one by Thom scored 3/5 with those by Weaver and Scheinkestel scoring 5/5.
Juurlink et al.73 state: “Of the six trials included, four found no benefit of HBO
for the reduction of neurologic sequelae, while two did. While pooled analysis does
8417: “8417_c018” — 2007/9/19 — 13:22 — page 399 — #9

not suggest a benefit for HBOT, (OR for neurological deficits = 0.78; 95% CI = 0.54,
1.12, p = 0.18), significant methodological and statistical heterogeneity means that
this result must be interpreted with caution. Design and analysis flaws were evident
in all the trials and importantly, the conclusion of one positive trial may have been
influenced by failure to adjust for multiple hypothesis testing while the other positive
trial is hampered by apparent changes in the primary outcome during the course of
the trial.” Juurlink et al.73 again concluded that: “there was no evidence to support
the use of HBO for treatment of patients with CO poisoning.”
It is worth carefully reviewing all these studies, and also including for review that
by Ducasse,82 which is also a prospective, randomized trial, but not usually included
because of the use of surrogate outcome measures.
Most of the discussion will concentrate on the two main studies: Weaver’s and
Scheinkestel’s. Case series, retrospective reviews, nonrandomized studies, animal
work and manuscripts based on theory have not been included in this review.
They are discussed in chronological order below.
18.2.3.1 “Trial of Normobaric and Hyperbaric Oxygen for

Acute Carbon Monoxide Intoxication” by Raphael
Raphael’s study published in 1989,75 enrolled only patients poisoned in the 12 h prior
to hospital admission. Three hundred and forty-three (343) patients with mild CO
poisoning (no impairment of consciousness) were randomized to receive either NBO
or HBO. Two hundred and eighty six (286) severely poisoned patients (with impair-
ment of consciousness) were randomized to either one or two sessions of HBO, 12 h
apart. Critically ill patients were excluded. Patients refusing the allocated treatment
after randomization (n = 19), were still retained in the study and analyzed accord-
ing to treatment intended. NBO therapy consisted of 6 h of 100% inspired oxygen
by facemask or endotracheal tube. HBOT was 2 h of HBO in a mono-place cham-
ber (0.5 h for compression, 1 h at 2.0 ATA and 0.5 h for decompression), plus 4 h
of NBO.
Eleven percent (11%) of patients were lost to follow-up.
Thirty-nine patients (39) were intolerant of HBO and five had confirmed baro-
trauma. Raphael reported persistent neurological sequelae (PNS) in 32–34% of the
patients with no LOC, and in 46–48% of those with LOC. The diagnosis of PNS
was based on gross signs and symptoms, as neuropsychological testing was not
performed.
Patient assessment at 1 month consisted of a self-assessment questionnaire and a
physical examination performed by the patients’ own doctor. If there was no response
to the questionnaire, patients were telephoned.
Raphael concluded that in patients without LOC, HBO had no advantage over
NBO. At the 1-month review, recovery occurred, respectively, in 66% of 170 NBO
patients and 68% of 173 HBO patients. Ninety-seven percent of patients resumed
their usual occupation and social activities irrespective of treatment.
In patients with transient LOC, he found no difference in outcome at 1 month
between those patients having one or two HBO treatments (54% versus 52% recovery,
p = 0.42).
8417: “8417_c018” — 2007/9/19 — 13:22 — page 400 — #10

The problems identified with this study include:
• Individuals were excluded on the basis of COHb measurement alone.

• Those patients who refused randomization were retained in the study and
included in the final analysis as members of intended treatment group.83
• NBO was given by facemask only, making the exact percentage of inspired
oxygen unclear.
• Lack of true controls. Only less severely CO poisoned patients were ran-
domized to HBO versus NBO groups.78 All severely poisoned patients
received HBO.
• The HBO regimes used are considered by some to be ineffective (2.0 ATA
instead of 2.8 ATA).84
• The times from poisoning to treatment entry criteria (up to 12 h) were too
long.85,86 Oxygen treatment did not begin until after more than 6 h after
poisoning in approximately half the cases.75
• Patients were not stratified according to interval between exposure and
therapy.83
• Neither the investigators nor the patients were blind to treatment group.
• The use of insensitive outcome measures (self-assessment questionnaire
by telephone or mail one month after poisoning, discussion with the
patients’ personal physicians),87 with recovery being determined by a
lack of symptoms and/or resumption of former activities and delayed
neurologic sequelae (DNS) being diagnosed when patients reported any
of a variety of complaints.88
• Not using neuropsychometric tests to assess outcome resulted in an absence
of objective or quantitative evaluation of cortical function.
• The physical examination was performed by the patients’ own physi-
cian, rather than a physician experienced in CO-poisoning. Inexperienced
physicians, not used to assessing these patients, may miss the more subtle
signs and symptoms.
• The use of multiple physicians precluded consistency.
• 70 of 629 (11.1%) patients were lost to follow-up at 1 month.
18.2.3.2 “Non-Comatose Patients with Acute Carbon

Monoxide Poisoning: Hyperbaric or Normobaric
Oxygenation?” by Ducasse
Ducasse’s56 series comprised 26 noncomatose patients with a glasgow coma score
(GCS) of 15 on admission; 13 patients were randomized to HBO and 13 patients
were randomized to NBO. They started treatment at a mean time after CO exposure
of 53 min. The HBO group received HBOT for 120 min at 2.5 ATA pressure, followed
by 100% NBO for 4 h and a further 6 h of 50% NBO. The NBO group received oxygen
through a face-mask at 100% for 6 h, then at 50% for 6 h. Patients were assessed on
clinical signs and symptoms, electroencephalogram (EEG) and cerebral blood flow
response to acetazolamide. No neuropsychological assessments were performed. The
reported incidence of both persistent neurologic sequelae (PNS) and DNS was zero.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 401 — #11

Clinical assessment at 2 and 12 h favored the HBO group. Two patients in the
NBO group were changed to HBO at 12 h and were asymptomatic at completion
of treatment. All 26 patients were discharged home well, an average of 28 h after
presentation.
There was no difference in the EEG at 24 h between HBO and NBO groups. Eight
of twenty-six patients (31%) were lost to follow-up (three in the NBO group and five
in the HBO group). Follow-up EEG in the remainder was worse in the NBO group
(p ≤ 0.02), but all patients were clinically normal.
Cerebral blood flow was assessed in 20 patients (four HBO, six NBO, and ten
controls). There were no differences between HBO, NBO and controls with regard
to perfusion of the basal ganglia or in cerebral blood flow values. Reactivity to
acetazolamide was similar in the controls and the HBO group, and were said to
be statistically significantly different from the NBO group (p ≤ 0.04).
Adverse events due to HBO are not mentioned.
Ducasse concluded that HBO reduced the time to initial recovery and the number
of delayed functional abnormalities in noncomatose patients with acute CO poisoning.
He attributed some of his success to the rapidity of treatment.
The problems identified with this study include:
• Small study with few patients.26

• Only patients with mild CO poisoning were enrolled.
• The study was nonblinded.
• The use of surrogate outcome measures which are of questionable
significance.85,87,89
• The significance of an abnormal EEG in clinically normal patients is
unclear.
• No statistical significance values were reported for the EEG results.
• Thirty-one percent (31%) of subjects we lost to follow-up.
• While controls were part of the 3-week evaluation, the authors did not
define this subgroup’s characteristics.
• There were no measures of cognitive function, as standardized neuropsy-
chiatric testing was not performed.
• The test data appear to conflict with one of the author’s conclusions: that
HBO treated patients showed a better cerebral blood flow response and
greater improvement in the 3 week EEG studies.83
• Inadequate allocation concealment.89
• Statistical significance of reactivity to acetazolamide is questionable, given
the small numbers and large range of results.
18.2.3.3 “Delayed Neuropsychologic Sequelae after Carbon

Monoxide Poisoning: Prevention by Treatment With
Hyperbaric Oxygen” by Thom
Thom’s study76 also included only mild poisonings. Patients with a history of LOC
were excluded. Patients presented within 6 h of exposure and usually commenced
8417: “8417_c018” — 2007/9/19 — 13:22 — page 402 — #12

treatment in about 1 h. Thirty-two patients were allocated to NBO and thirty-three to

HBO. HBO patients were treated once for 30 min at 2.8 ATA, followed by 90 min
at 2.0 ATA. NBO patients received 100% oxygen through a nonrebreather face-mask
until all symptoms resolved (4.2 ± 3 h).
The presence or absence of five signs and symptoms together with a COHb level
were used to assume the two groups were of similar severity of poisoning. Thom
used a CO neuropsychological screening battery (CONSB) designed by Myers after
completion of treatment.
Formal neuropsychological testing was performed at 1 month. The 3-month
review consisted of a telephone interview only.
Twelve of sixty-five patients (18%) were lost to follow-up.
Thom reported no DNS in the HBO group, while 7 of 30 in the NBO group
developed problems (p < 0.05). No specific treatment was given to those who
developed DNS. Three of these patients refused follow-up. In the remaining four,
neuropsychometric testing was repeated at intervals of 2–3 weeks until scores returned
to baseline. It would appear that all DNS resolved.
No adverse events resulted from HBO treatment.
Thom concluded that HBOT decreased the incidence of DNS after CO poisoning.
Problems reported with this study include:
• Neither the patients nor the investigators were blinded to treatment.

• The randomization process is unclear.
• The consent procedures are unclear.
• Only patients with mild poisoning were included, sick patients were
excluded.90
• While a COHb level was one of the measures used to assume that the
two groups were of similar severity of poisoning, the delay taken in
measurement of COHb is not mentioned.
• Baseline neuropsychometric testing was not performed to ensure that the
two groups were similar.
• There was greater comorbidity in the NBO group at randomization.85,87 .
NBO patients were slightly older and had a higher incidence of cardiovas-
cular and respiratory disease.
• Neuropsychological tests were repeated at intervals of 2–3 weeks until
scores returned to baseline. The effect of repetition and learning of tests
on outcome was controlled by comparing to the effects achieved by prac-
tice on learning in a control group of eight patients. This is an insufficient
number of control subjects to assist with the interpretation of neuropsycho-
logical tests86,91 and no details of the characteristics of the control group
were provided, in particular how they matched with the HBO and NBO
groups.90,91
• The location and conditions for neuropsychometric testing were incon-
sistent. Some were performed immediately on completion of treatment,
but if patients were “fatigued”, the tests were performed in the patients’
homes within the next 12 h. No details are provided as to how the
neuropsychometric testing was performed on the controls.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 403 — #13

• The paper does not state the number of clinicians involved in performing
the neuropsychometric testing, nor their experience.
• The neuropsychological tests used (CONSB) are said not to adequately
measure memory.
• The definition of DNS was a deterioration in one or more subtest scores on
the neuropsychological test battery, but this “deterioration” is not defined
in the paper.91
• The incidence of delayed sequelae may have been significantly altered
in the normobaric group if normal psychometric testing had been a
prerequisite for discharge.88
• At the 4 week follow-up, the NBO group had a worse score in one subtest
only, Trail-Making, which Thom states may reflect the presence of a
subtle impairment of learning ability when these patients first took the
psychometric test and hence a lack of familiarity on retesting.76
• The 3-month review consisted of a telephone interview only.
• 18% of patients were lost to follow-up.86
• All patients reported complete resolution of symptoms.
Hampson89 states that this study was stopped early due to a treatment advantage
in the HBO group, but the actual paper makes no mention of this. In fact, Juurlink73
points out that in an interim analysis describing the outcome of 58 patients, pub-
lished in 1992, there was no difference in symptoms between patients in the NBO
versus HBO groups (4 of 29 patients versus 0 of 29 patients, respectively). Juurlink
notes that seven additional patients were recruited after this interim analysis, three
to the NBO arm (all of these patients experienced neurological sequelae), and four
to the HBO arm (none of these experienced neurological sequelae). He states that
although the recruitment of seven additional patients with this distribution of alloca-
tion and outcomes could be due to chance (p = 0.014 by Fisher’s exact test), it may
reflect premature termination of the trial after recruitment of only seven more patients,
greatly exaggerating the treatment effect. A statistical penalty to adjust for inflation of
the type I error rate was not introduced, and would have rendered the final result stat-
istically insignificant.73 Had adjustment for multiple comparisons been performed,
no significant difference between treatments would have been identified.92
18.2.3.4 “Randomized Prospective Study Comparing the

Effect of HBO Versus 12 h NBO in Noncomatose CO
Poisoned Patients: Results of the Interim Analysis” by
Mathieu
Mathieu80 reported on an interim analysis after 3 years of a 5-year multicenter study.
Only patients noncomatose on hospital admission, and poisoned in the preceding 12 h
were enrolled. Delay to treatment is not specified.
Treatment was either one HBO session of 90 min at 2.5 ATA (299 patients), or
12 h of 100% NBO (276 patients). Patients were neurologically normal at the time of
hospital discharge, but at 1 month approximately one quarter had sequelae with no
difference between HBO and NBO (23% versus 26%). At three months, the incidence
8417: “8417_c018” — 2007/9/19 — 13:22 — page 404 — #14

fell and there was a statistically significant difference between HBO and NBO (9.5%
versus 15%; p = 0.016), but this was no longer evident at 6 months (6.4% versus
9.5%; p = 0.09) or 12 months (4.3% versus 5%).
The limitations of this report are:
• It is only available in abstract form.
• There was no blinding.
• The definitions of the primary outcome are missing.92
• There was a long entry time of 12 h from poisoning to treatment.85
• No details of the neurological assessment are provided.
• Neurological manifestations are referred to as PNS, but given that the
patients were neurologically normal at the time of hospital discharge, it is
not clear why they are PNS and not DNS.
• Neuropsychological assessments were not performed either pre- or post-
treatment.
• Mathieu provides no details of attrition rates for follow-up.92
• Recovery was considered complete if the patient had no complaints.88
• Had the investigators adjusted their analysis for multiple comparisons, no
significant difference between treatments would have been identified at
any interval.92
18.2.3.5 “Managing Carbon Monoxide Poisoning with

Hyperbaric Oxygen” by Raphael
Raphael79 in 2004, published in abstract form the results of a second randomized
controlled trial. Patient recruitment took place between 1996 and 2000 and involved
385 victims of accidental, domestic CO poisonings presenting within 12 h of CO
exposure. One hundred and seventy nine (179) patients had transient LOC and were
randomized to receive either 6 h of NBO (A0 group, 86 patients) or one treatment
with HBO (at a plateau of 2 ATA for 60 min), plus 4 h of NBO (A1 group, 93 patients).
Two hundred and six (206) comatose patients were randomized to receive either
one (B1 group-101 patients) or two (B2 group-105 patients) HBO treatments (at a
plateau of 2 ATA for 60 min) plus 4 h of NBO.
The primary end-point was the proportion of patients who recovered at 1 month,
with recovery being defined by a normal self-assessment questionnaire and normal
blinded neurological examination.
The trial was stopped prematurely because the interim analysis showed that giving
two HBO sessions rather than one, was associated with a poorer outcome in comatose
patients. Furthermore, in patients with transient LOC, the recovery rate was not
modified by addition of HBO to NBO.
The percentage who had recovered at 1 month were as follows:
A0 61% 45/74
A1 58% 46/79 (OR = 0.90, 95% CI = 0.47–1.71, p = 0.87)
B1 68% 54/80
B2 47% 42/90 (OR = 0.42, 95% CI = 0.23–0.79, p = 0.007)
8417: “8417_c018” — 2007/9/19 — 13:22 — page 405 — #15

There were four deaths and four survivors with severe neurological sequelae in
the B2 arm, (two HBO treatments). All but one of the deaths and severe neurological
sequelae were in the B group, that with comatose patients receiving HBO treatment.
Severe neurological sequelae occurred only in those with coma at presentation.
Seventeen (17%) percent of patients were lost to follow-up.
The authors concluded “There is no evidence supporting the routine use of HBO
in patients with acute CO poisoning”
The limitations of this report are:
• All the ones of his previous study, as the trials are essentially the same:
• Lack of true controls. Only less severely poisoned patients were ran-
domized to HBO versus NBO.73 All severely poisoned patients received
HBO.
• The HBO regimes used are considered by some to be ineffective (2.0
ATA instead of 2.8 ATA).84
• The times from poisoning to treatment entry criteria (up to 12 h) were
too long.85,86
• Patients were not stratified according to interval between exposure and
therapy.83
• Neither the investigators nor the patients were blind to treatment group.
• The use of insensitive outcome measures (self-assessment question-
naire by telephone or mail 1 month after poisoning, discussion with the
patients’ personal physicians),87 with recovery being determined by a
lack of symptoms and/or resumption of former activities and DNS being
diagnosed when patients reported any of a variety of complaints.88
• Not using neuropsychometric tests to assess outcome resulted in an
absence of objective or quantitative evaluation of cortical function.
• The physical examination was performed by the patients’ own phys-
ician, rather than a physician experienced in CO-poisoning and
inexperienced physicians, not used to assessing these patients may miss
the more subtle signs and symptoms.
• The use of multiple physicians precluded consistency.
In addition,
• The study is only available in abstract form.

• 17% of patients were lost to follow-up.
18.2.3.6 “Hyperbaric or Normobaric Oxygen for Acute

Carbon Monoxide Poisoning: A Randomised
Controlled Clinical Trial” by Scheinkestel
We98 commenced a trial in 1993 in which we aimed to overcome the shortcomings of
the previous trials. We randomized patients with all grades of CO poisoning, including
severely poisoned patients, within 24 h of poisoning, used sham treatments in the
multiplace chamber for the NBO group with both patients and the outcome assessor
8417: “8417_c018” — 2007/9/19 — 13:22 — page 406 — #16

blinded to treatment group. Our primary end-point was outcome at completion of

treatment. Our secondary end-point was outcome at 1 month.
In the absence of universally accepted recommendations for depth or duration of
pressurization for HBO treatments, we used what was at the time, the best available
data to determine the HBO treatment protocol, which consisted of three treatments
of 100 min with 60 min at 2.8 ATA over 3 days, based on the conclusions of
Gorman and Runciman93 that such protocols were associated with the lowest mortal-
ity and neurological deficits and would provide the maximum potential advantage for
HBOT. In between treatments, patients received high-flow oxygen by nonocclusive
facemask.
In order for the treatment and control groups to be identical in every way except
for the hyperbaric component, the NBO group had exactly the same treatment as the
HBO group but had NBO in the chamber.
Patient assessment at entry included the clinical effects of poisoning and a mini-
mental examination. After the third treatment, patients were reassessed medically
and underwent full neuropsychological testing. Patients with a poor outcome had
a further three treatments (NBO or HBO as previously allocated) and continued
with high flow oxygen between treatments. We used a pretreatment, baseline mini-
mental examination, an extended series of neuropsychological tests to assess both
persistent and delayed neuropsychological deficits, a clinical psychologist trained
in neuropsychological assessment of brain injured patients to perform all the tests
(at completion of treatment and at follow-up), and computerized testing to standardize
administration and increase objectivity of these tests.
Following consent, patients were stratified into four groups: suicide versus acci-
dental, then mechanically ventilated versus non-ventilated prior to randomization to
HBO (104 patients) or NBO (87 patients).
In our prospective randomized controlled trial of 191 patients, in which both
groups received high doses of oxygen, the HBO regimen used, did not benefit and
may have worsened outcome.
More patients in the HBO group required additional treatments (28% versus
15%, p = 0.01 for all patients; 35% versus 13%, p = 0.001 for severely poisoned
patients). HBO patients had a worse outcome in the learning test at completion of
treatment (p = 0.01 for all patients; p = 0.005 for severely poisoned patients) and a
greater number of abnormal test results at completion of treatment (3.4 versus 2.7,
p = 0.02 for all patients; 3.7 versus 2.6, p = 0.008 for severely poisoned patients).
A greater percentage of severely poisoned patients in the HBO group had a poor
outcome at the completion of treatment (85% versus 65%, p = 0.03). DNS were
restricted to the HBO patients (p = 0.03). No outcome measure was worse in the
NBO group.
The comprehensive assessment of all patients at completion of treatment showed
no benefit for HBO.
Our study has attracted a number of criticisms:
• Cluster randomization
In the accompanying editorial, Moon and DeLong,99 while acknowledging that “the
study design is amongst the most rigorous yet published”, expressed concern about
8417: “8417_c018” — 2007/9/19 — 13:22 — page 407 — #17

cluster randomization. Kao87 and Weaver100 expressed similar concerns. To min-

imize the impact of the trial on daily practice, cluster randomization was used for
simultaneously presenting patients from the same exposure: the treatment group
assigned to the first patient was allocated to the other simultaneously presenting
patient(s). Randomization took place only after the group was assembled. Cluster
randomization accounted for the difference in numbers between the HBO and NBO
groups.
We used cluster randomization, allocating more than one person simultaneously to
the same treatment on 22 occasions, (2 patients on 12 occasions, 3 patients on 5 occa-
sions and 4 patients on 5 occasions). Overall 14 clusters (40 patients) were allocated
to HBO and 8 clusters (19 patients) to NBO. As patients presenting simultaneously
could be uniquely identified by having identical measurements for three continuous
baseline severity measurements (exposure time, time to COHb measurement and
time to treatment), any effects due to cluster randomisation could be controlled and
adjusted for by including these variables in the generalized linear model.
Continuous outcome variables were also analyzed by the mixed procedures in
statistical analysis software (SAS)94 which allows a repeated measures analysis of
variance, with the variable cluster being treated as a random repeated measurement,
thus “adjusting for” within cluster variation. We also repeated the analysis excluding
all patients who were allocated as part of a cluster. It has been shown95,96 that the
effect of cluster randomization is to increase the size of standard errors and p-values.
By including these three variables we found that the standard errors and the p-values
were increased in comparison to models excluding these variables. To further val-
idate results, sensitivity analysis was performed for the binomial outcome variables
by comparing results with those obtained by excluding all participants who were
randomized into a cluster.
Original OR Results Excluding

Variable (95% CI) All Clusters
Required > 3 treatments 2.8 (1.3–6.2) 3.3 (1.3–8.2)

Required > 3 treatments (severe) 5.4 (2.0–14.8) 5.3 (1.7–16.0)
Poor outcome 1.7 (0.8–4.0) 2.2 (0.8–5.9)
Poor outcome (severe) 3.6 (1.1–11.9) 3.2 (0.9–11.9)
This analysis produced consistent results for all variables which suggests that our
results have not been biased by cluster randomisation.97,98
• Allocation Concealment
The Cochrane review73 scores our study a “B” for allocation concealment because
of concern of failure of concealment related to fixed block sizes if the treating team
were aware of previous assignment. We specifically addressed this by having blocks
of various sizes so that you could not predict the next treatment. Randomization
(HBO or NBO) was performed by a hyperbaric technician who opened progressively
numbered, sealed, opaque envelopes (from random blocks of 4, 6, 8 and 10 envelopes,
each block containing equal numbers of HBO and NBO selections).
8417: “8417_c018” — 2007/9/19 — 13:22 — page 408 — #18

• Delay to Treatment
Both Moon99 and Weaver100 also express concern about our delay to treatment (geo-
metric mean of 7.1 h (95% CI = 1.9–26.5 h). Firstly this is not dissimilar from some
of the other studies. Raphael’s inclusion criteria was up to 12 h in both his studies74,79
with a mean time to treatment of 6.4 h74 . Mathieu’s80 entry criteria was also within
12 h of exposure. Weaver’s study81 included patients within 24 h of exposure and the
mean times in his study were only marginally different from ours: for HBO 5.8 ± 2.9
versus 7.5 (6.6–8.6), NBO 5.7 ± 2.9 versus 6.6 (5.7–7.5), respectively, for Weaver’s
and Scheinkestel’s patients.
As Tighe points out,101 “these time delays are representative of most clinical
practice because of late presentation and the need for stabilization and transport to a
remote hyperbaric facility.”
Although the geometric mean treatment delay was 7.1 h, we performed subgroup
analysis of patients treated within 4 h (all patients and just severely poisoned patients).
There were 44 patients treated within 4 h (22 HBO and 22 NBO), 33 of which were
severely poisoned (15 HBO and 18 NBO) with no outcome measure favoring the use
of HBO. We further analyzed time to treatment in quartiles (<3, 3–6, 6–12, >12 h)
and found no difference in outcome between HBO and NBO. Further multivariable
analysis did not identify delay in treatment as a predictor of poor outcome. Thus there
was no evidence that delay in treatment could have explained the lack of benefit of
HBO.97
• Concomitant depression, suicide attempt and use of psycho-active drugs
Weaver,100 Shank,85 and Moon99 express concern that we included patients whose
exposure was due to suicidal intent, had consumed cointoxicants and had a history of
depression.
They question whether this might have influenced the outcome of the neuropsy-
chological tests and therefore the results of the study. While it is true that depression
and the use of medication may have resulted in a higher incidence of poor outcome
overall, this would not in any way have biased the comparison between normobaric
and hyperbaric groups as patients were specifically stratified for suicide attempt prior
to randomization to therapy.
As we also stated, analysis of accidental poisonings (excluding suicide attempts)
also showed no differences between HBO and NBO groups.77 Furthermore the
incidence of self-administration of drugs and alcohol was identical in both groups
(44%).77
In a subsequent letter to the British Medical Journal (BMJ)102 Weaver states
“I agree that attempted suicide probably did not bias the outcome between the two
arms”.
• Lack of pretreatment neuropsychological assessment
Our lack of pretreatment neuropsychological assessment has been considered a
problem by Moon,99 Schiltz103 and Denson and Hay.104
In order to address our specific research question, whether HBOT is superior to
NBO therapy in preventing residual cognitive impairment following CO poisoning,
8417: “8417_c018” — 2007/9/19 — 13:22 — page 409 — #19

two alternative research designs were considered: a cross sectional design wherein
subjects are randomly assigned to two groups (HBO versus NBO), and a lon-
gitudinal design where within group comparisons are made (pre-treatment verses
post-treatment). When employing neuropsychological tests to evaluate patient out-
come within a longitudinal design, both practice effects and spontaneous recovery
may over-estimate treatment benefits. As Dr. Schiltz points out,103 it would be
pertinent within a longitudinal design to take into account premorbid intelligence
and psychiatric status. In selecting a randomized cross-sectional design (where both
assessor and data analyzer were blind to treatment membership) however, statistical
comparisons are not confounded by practice effects nor spontaneous recovery.
Additionally, while we acknowledge that we could not clearly establish that our
two groups (NBO and HBO) were equal with respect to premorbid intellectual activ-
ity and level of depression, there was no difference in initial mini-mental score
between groups (27.0 (26.1–27.9) versus 26.4 (25.4–27.4), p = 0.27). There was
similar improvement in mini-mental score in both groups, for all patients (p = 0.53),
and for severely poisoned patients (p = 0.71). The score improved further to normal
levels in both NBO and HBO groups in those attending follow-up for all patients as
well as severely poisoned ones.
Clinical considerations were equally important in selecting the most appropriate
research design and methodology. Firstly, there is the issue of obtaining meaningful
data. In an acutely ill, disoriented, agitated and distressed patient, prolonged psy-
chometric testing is not practical or meaningful. In more cases than not, the patient
would not be able to sustain concentration for a 1.5-h assessment.
Further, availability of a trained psychologist 24 h a day, 7 days a week to perform
pretreatment full neuropsychological assessments and delaying treatment by a further
90 min was not practical. We therefore performed the mini-mental test on admission,
which is quick and easy to administer, rather than comprehensive neuropsychological
assessment which was performed at completion of treatment once patient’s mental
status had stabilized.
• Type of tests
There has been criticism that the neuropsychological tests we used were not standard
ones.87 Neuropsychological assessment of the CO population is challenging in light
of the confounding psychiatric variables. A further challenge to researchers is the
collection of data in a clinical setting that needs to be sensitive to patient motivation,
level of cooperation and imminent discharge. For these reasons, highly sensitive
cognitive tests were selected that sampled the pertinent neuropsychological realms
(attention, new learning, visuo-construction and executive functioning) in 90 min
in order to maximize cooperation and minimize fatigue. The choice reaction time
(RT) measure we used, required participants to respond to stimuli that appeared at
various positions around the periphery of the screen and as such required rapid visual
scanning. RT is considered sensitive to cerebral lesions of any localization. Should
specific visuo-spatial deficits occur in the CO-poisoned population, it is likely that
they would be reflected in the RT. In addition to being sensitive to diffuse cerebral
lesions, the choice RT task can be argued to be sensitive specifically to visuospatial
deficits.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 410 — #20

• Multitude of tests
Moon99 and Denson and Hay104 raise the matter that we performed a multitude of
tests and that only one showed a statistically significant result (in favor of NBO).
Although we performed a multitude of tests, not one showed a benefit in favor of
HBO. We were not trying to prove a statistically significant advantage for NBO,
rather we could not demonstrate a benefit in favor of HBO.
As multiple tests were considered, there is an increased likelihood of a type I
error. But, given that an outcome measure based on combining all tests was used,
in conjunction with death and > 3 treatments, the chance of a spurious result was
minimized.
We agree that there were two major limitations to our study:
• Low follow-up rate
While our study was inclusive with 83% of potential patients being entered (4% were
excluded and 13% refused consent), only 46% of patients entered attended the follow-
up at 1 month. Thus, longterm follow-up was only available in 38% of all potential
patients. This unfortunate result occurred despite significant effort. Patients were
requested to attend for review at 1 month. The review appointment was confirmed by
mail and if required, patients were actively pursued by telephone. Despite repeated
efforts, only 46% of patients attended follow-up. This low rate of attendance at
follow-up is indeed a major problem in interpreting our patients’ outcomes. Our
different patient population, with characteristics associated with suicide attempts and
depression, many referrals from distant locations and lack of incentive, probably
contributed to the low follow-up rate, which was however, equal in both groups, and
evenly distributed across subgroups.
For those attending follow-up, our assessment was rigorous (as opposed to a
telephone survey) and failed to show any benefit for HBO.
The previously published randomized studies have experienced lower but signi-
ficant nonattendance rates at delayed review of 11.1%,75 31%,82 18%,76 17%,79 with
Mathieu’s study not quoting the attrition rate.
• The treatment regimens
We have been criticized for using nonstandard HBO and NBO therapy46,87,99 because
this is not representative of usual practice. The treatment regimens in our study were
not conventional. It should be noted however, that established practice continues to
vary very widely.4 In the absence of universally accepted recommendations for depth
or duration of pressurization for HBO treatments, our protocol for the management of
CO poisoning consisted of three treatments of 100 min with 60 min at 2.8 ATA over
3 days, based on the conclusion of Gorman and Runciman93 that this achieved the
lowest mortality and neurological deficits and would provide the maximum potential
advantage for HBOT.
Our hyperbaric treatment regimen for CO poisoning included interval NBO. To
retain blinding, we also provided the control group with NBO for 3 days. As a result,
our normobaric group received more NBO than used in previous studies and this may
have been a factor in the lack of outcome difference between the treatment groups.105
8417: “8417_c018” — 2007/9/19 — 13:22 — page 411 — #21

As Tighe comments,101 “previous studies have been flawed by a failure to optimize

treatment in the normobaric group, and Scheinkestel et al’s study clearly shows that
such cheap, available, safe treatment is also effective.”
Moon99 raised the concept that repetitive treatments at 2.8 ATA could have been
neurotoxic and offset any potential benefit from HBO. This proposal of oxygen
neurotoxicity is a valid one, but it is important to note that repeated daily treat-
ments at 2.8 ATA is standard therapy for CO poisoning in many centers and also for
cerebrally-impaired patients with decompression illness.
We set out to demonstrate an advantage for HBO and selected the optimum treat-
ment regime to achieve this. We have now demonstrated that there is no advantage
in outcome for HBO over 3 days of high flow oxygen therapy. What we cannot now
answer is whether the same outcomes can be achieved with less days of high flow
oxygen therapy.
Piantadosi46 states that the difference in O2 doses between the groups was neg-
ligible. We calculated that the HBO group received oxygen therapy equating to
approximately 35.7 COHb dissociation half-lives, while the NBO group received
the equivalent of 28.5 COHb dissociation half-lives, a difference of 7.2 COHb dis-
sociation half-lives. Most other studies have used total oxygen doses of less than 7.0
COHb dissociation half-lives.74,75,80,82 The difference in our two groups is greater
than the difference in doses used in these other studies. We just used a higher baseline
oxygen dose.
18.2.3.7 “Hyperbaric Oxygen for Acute Carbon Monoxide

Poisoning” by Weaver
Weaver81 in 2002 reported on the outcome of 152 patients enrolled within 24 h
of exposure to CO in his double-blind randomized trial. Seventy-six patients were
assigned to have three hyperbaric sessions within a 24-h period and 76 had one NBO
treatment and two sessions of exposure to normobaric room air. Patients were stratified
according to whether or not they had lost consciousness, the interval between end of
exposure and entry into the chamber (<6 h or >6 h) and age (<40 years or >40 years).
Treatments took place in a Sechrist monoplace chamber. The first HBO session
was with 100% O2 for 2.5 h with 55 min at 3.0 ATA, the second and third were with
100% O2 for 2 h with 90 min at 2.0 ATA.
It is not clear from the paper as to the exact NBO treatment given. In one section it
states: “patients in the NBO group were exposed to air at one ATA for all three chamber
sessions.” In another section it states: “100% oxygen was delivered to those in the
NBO group during chamber session 1. During NBO sessions 2 and 3, patients were
exposed to 1 ATA and breathed air” unless the patients were intubated and ventilated
or had arterial oxygen saturations <90%, in which case they received 100% oxygen.
Neuropsychological assessments were administered after chamber sessions 1 and
3, at 2 weeks, 6 weeks, 6 months, and 12 months. The primary outcome was cognitive
sequelae at 6 weeks.
The trial was stopped after the third of four scheduled interim analyses.
Cognitive sequelae at 6 weeks were less frequent in the hyperbaric group (25%)
than in the normobaric group (46%), p = 0.007, even after adjusting for cerebellar
8417: “8417_c018” — 2007/9/19 — 13:22 — page 412 — #22

dysfunction and for stratification variables (adjusted OR = 0.45, 95% CI = 0.22 −

0.92, p = 0.03.
The presence of cerebellar dysfunction before treatment was associated with the
occurrence of cognitive sequelae (odds ratio:5.71 (95%, CI 1.69–19.31), p = 0.005,
and was more frequent in the normobaric group (15% versus 4%, p = 0.03). Cognitive
sequelae were less frequent in the hyperbaric group at 12 months, according to
intention to treat analysis (p = 0.04).
The authors concluded that three HBO treatments within a 24-h period appeared
to reduce the risk of cognitive sequelae at 6 weeks and 12 months after acute CO
poisoning.
An abstract by Weaver117 further analyzed the data and concluded that HBO
improved outcome if any of the following were present: LOC, COHb > 25%, age >
50 and metabolic acidosis. In patients with none of these four criteria, HBO did not
improve outcome.
While proponents of HBO for CO-poisoning consider this to be a landmark paper,
and Dr Weaver believes his paper to be the only one worth considering and the
definitive paper on HBO in CO-poisoning,78 we believe there are some fundamental
problems with the Weaver study:
• Primary Outcome analysis
Dr. Juurlink summarizes this eloquently.78 “In 1995, Weaver and colleagues published
the first interim analysis of their study.106 In that report, the only test of statistical
significance was applied to DNS, and the investigators indicated that enrollment
would continue because the p value had not achieved the threshold required for
premature termination of the trial.
That same year,107 Weaver presented a description of his ongoing trial and gave
an explicit definition of its primary outcome: “Our major question is, does HBO2
reduce the incidence of DNS?”
He also wrote: “During the course of the trial, it became evident that operational
definitions of DNS and PNS were needed . . . Our definition for DNS is: development
of a new neurologic abnormality not present at day 1, and/or decrement of neuropsy-
chologic subtest score of more than two SDs below the mean or two subtest scores
more than one SD below the mean compared to standardized norms (prior normal
neuropsychologic test). If the prior neuropsychologic test is abnormal, then we use a
decrement of an abnormal subtest of more than one SD compared to the prior score
or more than 0.5 SDs below each of at least two abnormal subtests.”
In his letter to the editor,107 written with the study underway and 47 patients
enrolled, neither the definition of DNS nor that of PNS includes patient symptoms.
They are both clearly defined and these definitions are solely based on the outcome
of neuropsychological tests.”
DNS appears to be the original intended outcome of the trial. This outcome has
never been subsequently reported in any forum.
“In 2001, a discussion of his soon-to-be published study89 contained no men-
tion of DNS. Indeed, a very different outcome (“cognitive sequelae”) had been
defined: “Cognitive sequelae were considered present if any 6-week neuropsycho-
logical subtest score was >2 standard deviations below the mean (or if at least two
8417: “8417_c018” — 2007/9/19 — 13:22 — page 413 — #23

subtest scores were each more than one SD below the mean) of demographically
corrected standardized scores. Cognitive sequelae were present if a neuropsycholo-
gical subtest score was >1 SD below the mean or if two subtest scores each were
>0.5 SD below the mean and the patient complained of memory, attention, and/or
concentration difficulties.”
In the final publication,81 the cut-off values used to define abnormal results
changed yet again.
It is not clear why the statistical analysis on the T-scores for the neuropsychological
subtests is performed on the aggregate of the three scores obtained after chamber
session 3, at 2 weeks and at 6 weeks, given that “the primary outcome was cognitive
sequelae at 6 weeks”. The other two scores (after chamber session 3 and at 2 weeks)
are not relevant to this end-point.”
In summary, the original intended outcome of this trial was DNS defined by rig-
orous clinical criteria. In contrast, the final publication reports “cognitive sequelae,”
a distinctly different outcome. PNS and DNS have been bundled together. The defin-
ition of a poor outcome was changed to include a subjective component of patient
symptoms and the results of neuropsychological tests after treatment, at 2 weeks and
at 6 weeks have also been bundled together.
Juurlink states: “Dr. Weaver and his coinvestigators have obviously collected
the data necessary to examine DNS as an outcome, and we urge them to present this
analysis.73 Juurlink states, “doing so would help settle the present debate. While HBO
enthusiasts may argue that ‘cognitive sequelae’ is a meaningful outcome, skeptics
may legitimately wonder if the revised outcome was simply that which cast the most
favorable light on HBO once all the data were collected.”
Buckley108 asserts that a significant difference between HBO and NBO would
almost certainly not have been demonstrated if the originally intended outcome had
been analyzed.
Other problems identified with Weaver’s study include:
• Early termination of the study
In 1999, Weaver wrote109 “Blinded interim analysis showed no difference in outcome
between the two groups after 50 and 100 patients. Yet after 52 additional patients, the
results were so clear-cut, that the trial was terminated early, “after the third of four
scheduled interim analyses”. We interpret this to mean that there was one more interim
analysis scheduled and presumably 100 more patients still to enroll. When Weaver
adjusted for differences in baseline severity (cerebellar dysfunction—see section on
failure of randomization), the difference between groups only just makes statistical
significance at p = 0.05, well below the predetermined “stopping rules.”
• Enrollment and follow-up
Only 33% of potential patients were entered: 28% were excluded and 39% refused
consent, thus introducing the possibility of significant selection bias. The 95% follow-
up rate of these highly selected patients resulted in only 31% of all potential patients
being assessed. Weaver argues110 that he has followed up 91 of the 180 who declined
to be in the trial. They had a lower suicide rate (16% versus 30%) and a lower COHb
(20% versus 25%). Weaver concludes that they were similar to those enrolled in the
8417: “8417_c018” — 2007/9/19 — 13:22 — page 414 — #24

study. Further he states that the incidence of 6 week cognitive sequelae of 43% was
similar to that of the NBO group in the trial (46.1%), but it is clear from his reasons
for “declined enrollment in the trial” group,81 that at least some of these declined
because the referring physician insisted on HBO.
• Failure of randomization
This is the only study to have a large difference in baseline severity. Despite random-
ization, Weaver et al.’s normobaric group appeared to have suffered more severe CO
poisoning. The normobaric group had a CO exposure of 22 ± 64 h, almost double
that of the hyperbaric group (13 ± 41 h), and the COHb saturation at first entry to the
chamber was significantly higher (p = 0.02). Furthermore, the incidence of pretreat-
ment cerebellar dysfunction was 15% in the normobaric group and only 4% in the
hyperbaric group, an almost fourfold difference that was also statistically different
(p = 0.03). The presence of cerebellar dysfunction before treatment was associated
with cognitive sequelae (p = 0.005). Shorter exposure to CO, a lower level of COHb,
and a lower incidence of cerebellar dysfunction would be expected to favor a better
outcome in the hyperbaric group.
It is also worth noting that the COHb was actually only available in 83 patients
(in 36 HBO, less than half of the 76 HBO patients enrolled, and in 47 NBO). In the
other 69, missing values were imputed on the basis of the data in the other 83 patients.
The duration of exposure to CO has been presented as a mean with a standard
deviation. From these results (13 ± 41 versus 22 ± 64) it is clear that the distribu-
tion of duration is skewed. Variables measuring duration are often well characterized
by a log-normal distribution and would have been more appropriately presented as
geometric means with 95% confidence intervals, or if the distribution was unknown
(nonparametric), as medians with interquartile ranges. It is likely that there are signi-
ficant outliers included and this may have had a detrimental effect on the significance
of baseline differences and therefore on the multivariate analysis as a whole.
Weaver argues that the CO exposure was not statistically significant and produces
the requested statistics:110
• Median exposures: 4.2 (range 0.2–308 h) for HBO versus 5.0 (0.3–397 h)
for NBO
• Geometric means: 4.0 h (95% CI = 2.9–5.4) for HBO versus 5.4 (95% CI
= 3.8–7.6) for NBO, p = 0.2
• Interquartile range 7.3 h (1.6–8.9) for HBO versus 9.5 (2.0–13.5) h for
NBO, p = 0.2
These all confirm a trend, albeit non-significant, towards increased exposure in
the NBO group, which is confirmed by the statistically, significantly higher COHb
in the NBO group and is in keeping with the increased cerebellar signs in the NBO
group. Weaver argues however that the difference in COHb is of no consequence as
there is no relationship between COHb and outcome.110
This still leaves the fourfold increase in prechamber cerebellar signs in the NBO
group compared to the HBO group. If these patients are not excluded, his intention
to treat analysis finds a highly significant difference in six weeks outcome (25.0%
versus 46.1%, p = 0.007) in favor of HBO.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 415 — #25

If only patients with normal cerebellar function are analyzed, the benefit for HBO
only just reaches statistical significance (23.2% versus 39.0%, p = 0.05) and the trial
would not we assume have been terminated early.
• Inadequate oxygen dose to normobaric group
The amount of oxygen therapy given to the normobaric group may have been sub-
optimal. In all studies, patients received variable amounts of NBO therapy prior
to hospital arrival. Current recommendations for in-hospital NBO administration
most commonly involve occlusive face-mask and reservoir bag for 6–12 h or until
symptoms resolve. In this study, unless the patient’s oxygen saturation was less
than 90%, or they were intubated, the normobaric group only received 135 min of
oxygen therapy with a reservoir and a nonocclusive face-mask while in the cham-
ber. Thus in-hospital oxygen treatment was, by conventional criteria, short. The four
other randomized studies that have been published all utilized longer duration of
oxygen therapy in the normobaric group: 6–12 h80 6 h75 12 h,82 and an average
of 4.2 h.76
Weaver argues110 that the sum of oxygen therapy for the NBO group was 6.9 h,
135 min in the chamber and 4.5 ± 2.2 h before chamber session 1. The range was
3.9–18.8 h. However, he is including the prehospital time, whereas all other studies
provide at least 6 h once at the hospital. The median duration of oxygen therapy in
Weaver’s NBO group was 6.2 h. By definition 50% of patients had less than 6.2 h
of oxygen therapy. Further, Weaver states109 that referring physicians gave NBO
therapy to CO patients, but there is no statement to guarantee that these patients were
given high flow oxygen from a reservoir through a non-rebreathing face mask prior
to arrival at LDS Hospital.
Conventional teaching is that hyperoxia is required in CO poisoning to help “off-
load” CO. Not providing oxygen therapy to patients whose oxygen saturations were
>90% would be considered inappropriate by most centers. Weaver argues110 that they
treated all CO-poisoned patients with NBO until the COHb was less than 5%. This
is not stated in the paper. In fact, the paper is quite clear in several sections, that
supplemental oxygen was only provided if the oxygen saturation was <90%. Nor is
there a statement that COHb was remeasured at intervals to establish there was no
rebound rise in COHb postcessation of treatment, as is common.
Further, he does not state how oxygen saturation was monitored to ensure it was
>90%. If this was done using pulse oximetry, as is the norm, then there is considerable
literature as to the inadequacy of pulse oximetry to monitor oxygen saturation as this
cannot differentiate between COHb and oxyhemoglobin, and therefore over-estimates
oxyhaemoglobin.111
• Non-conventional HBO regime
Kao raised concerns that Weaver used a nonstandard HBO protocol.87 The first HBO
session was with 100% O2 for 2.5 h with 55 min at 3.0 ATA, the second and third
were with 100% O2 for 2 h with 90 min at 2.0 ATA. These are not the normally used
treatments and are not in keeping with The HBOT Committee Report 2003,112 which
states as follows: “the optimal number of hyperbaric treatments, the time following
poisoning after which therapy is no longer effective and the optimal treatment pressure
8417: “8417_c018” — 2007/9/19 — 13:22 — page 416 — #26

will require additional study . . . all patients at high risk deserve a single treatment . . .
subsequent treatments may be performed within 6–8 h and continued once or twice
daily until there is no further improvement . . . the optimal dose cannot be clearly
stated . . . between 2.5 and 3.0 ATM seems appropriate.”
Weaver’s HBO regime was less conventional than Scheinkestel’s. Weaver only
gave three treatments, 18% of the HBO group did not complete the three treatments
and the treatments were at lower ATA than recommended in the HBO Committee
Report. This compares with Scheinkestel’s treatment regime that gave a further three
treatments (total of six) if abnormalities persisted and all were at the recommended
ATA.
• Unjustified assumptions made in interpreting six week data
The intention to treat analyzes used by Weaver as “patients with missing data for
neuropsychological tests at six weeks were assumed to have cognitive sequelae.”
This is contrary to the standard intention to treat (ITT) approach of carrying the last
observation forward. Further, only one patient was lost to follow-up in the HBO
group whereas four were lost in the NBO group, thus this assumption favored the
HBO group.92
The impact of such arbitrary definitions on outcomes can be seen in Weaver’s
2002 abstract113 where he took the opposite approach and if data were missing, the
outcome was deemed to be “favorable.” Data from the 6 and 12 month follow-up
are apparently combined. Weaver states “a favorable outcome was found in 62/76
(82%) of HBO patients compared with 50/76 (66% treated with NBO (p = 0.027).
If data from patients with unknown 6 and 12 month outcome data were excluded, a
favorable outcome was present in 49/58 (84%) treated with HBO compared to 42/60
(70%) treated with NBO (p = 0.061).”
• Unjustified assumptions made in interpreting 6 and 12 month data
In the final publication,81 the definitions change again. With respect to the outcomes
at 6 and 12 months, these were performed on the basis that ‘if patients had cognitive
sequelae at 6 weeks, and missing data at 6 or 12 months, they were assumed to
have cognitive sequelae at those times. This is invalid, as it does not allow for the
improvement with time, which was demonstrated in patients with complete data. Such
a definition couples the first outcome to later events and creates a spurious outcome
dependent on the first. The statistical differences reported at 6 and 12 months merely
reflect the results at 6 weeks, not necessarily the true longterm outcome. If the analysis
is restricted to those patients with complete data, the statistically significant difference
in late outcome is lost. Dr Weaver disputes110 that statistical significance is lost at 12
months but a p = 0.08 is not statistically significant.
• Soft outcome measures
The entire positive outcome of this study is based on reported symptoms. These were
the primary determinant of a statistical difference between treatments. In the final
publication,81 neuropsychological testing identified no difference between HBO and
NBO; indeed, the mean neuropsychological testing scores for patients treated with
NBO were within the normal range.78,92
8417: “8417_c018” — 2007/9/19 — 13:22 — page 417 — #27

HBOT patients reported fewer difficulties with memory (p = 0.004) and this was
the only significant symptom difference between the two groups. However, there were
no statistically significant differences between groups on any of the memory-related
neuropsychology test scores.
• Interpretation of neuropsychological tests
Disproportionate numbers of patients with cerebellar problems entered one arm of
the Weaver study.
Neurological sequelae occurred more commonly if there were cerebellar signs
at the time of enrollment into the study. This was particularly so because two of
the six neuropsychiatric tests involved “Trail-Making,” and this would be affected
by even minor degrees of cerebellar dysfunction.78 This imbalance alone could have
accounted for half the actual observed difference between groups because the absolute
difference between arms was 16 individuals, yet there were 8 more individuals with
cerebellar dysfunction and neurological sequelae in the NBO group.92
Olsen is more critical114 and states: “The neuropsychological data presented by
Weaver are clinically underwhelming. The raw scores show a statistically significant
difference between treatment groups in only one of six subtests (Trail-Making, Part
1), and even in this subtest, the normobaric group was at the mean demographically
corrected score for a normal population at 6-week follow-up.”
Buckley also comments that the mean performance of patients in the NBO
group was normal for five of the six neuropsychiatric tests and above normal in
the sixth.92,115 He questions how a meaningful outcome could label 46% of patients
in the control group as having “cognitive sequelae,” when in fact, five of six of the
mean test scores in that group were actually normal or above average.
Weaver 110 responds that the neuropsychological test scores of patients with dys-
function are obscured by those without sequelae and that therefore the group mean
scores do not detect a difference between groups, but no data are provided to support
this.
It is also of interest to note that the frequency of cognitive sequelae amongst
patients who completed three HBO sessions was not different from those who did not
complete the three sessions, and 18.4% of the hyperbaric group did not complete the
required number of chamber sessions.
Weaver et al.’s outcome measures were performed by any of ten different psy-
chologists. Weaver states that the examiners were all psychology Ph.D. candidates
with proper training, with inter-rater reliability being well established for these tests
at 0.9 or higher. He also states that as the trial took 6 years, this necessitated multiple
examiners.
• Interpretation of 6 and 12 month data
Weaver claims115 that “Cognitive sequelae at six months and 12 months were less
frequent in the HBO group than in the NBO group, both according to the intention-
to-treat analysis (p = 0.02 at 6 months, p = 0.04 at 12 months) and according to the
efficacy analysis, (p = 0.03 at 6 months, p = 0.08 at 12 months).”
Interestingly, in his presentation at the Undersea and Hyperbaric Medicine Society
Annual Scientific Meeting July, 2002, and presented in abstract form, Weaver states
“if data from patients with unknown 6 and 12 month data were excluded, a “favorable
8417: “8417_c018” — 2007/9/19 — 13:22 — page 418 — #28

outcome” was present in 49/59 (84%) treated with HBO compared to 42/60 (70%)
treated with NBO (p = 0.061).”
• Lack of pretreatment neuropsychological assessment
Weaver’s study performed no baseline neuropsychological assessment. As Raphael
states, baseline information on abnormal cognitive tests is not provided. Given the
heterogeneity of the population and the rather small sample size, one cannot rule
out an imbalance between the treatment groups with respect to abnormal results of
cognitive tests just as there was an imbalance with respect to cerebellar signs.116
• Outcomes of other tests
There were no statistically significant differences between groups on the Geriatric
Depression Scale, the Katz index of activities of daily living, nor in scores on the
subscales of the SF 36 (social function, physical role, mental health, and energy).
Weaver argues110 that these tests do not test cognition and confirm that the cog-
nitive impairment is not due to depression. He also states that the “activities of daily
Living” measures gross abilities to perform daily activities not cognition. He also
states “the SF36 measures health related quality of life. It does not measure cogni-
tion but rather the patients’ perception as to whether the CO poisoning resulted in
decreased quality of life for physical and psychological functioning.”
However, in regard to the latter, his paper states: “We found no treatment-related
differences in scores on the subscales (social function, physical role, mental health
and energy) of the SF36. Clearly then, the conclusion is that the patients did not have
the perception that the CO-poisoning results in decreased quality of life for physical
and psychological functioning.
• Adverse events
Weaver reports a significant difference in the incidence of nystagmus post-treatment,
with hyperbaric patients having a 12% incidence compared to 2.7% for the nor-
mobaric group (p = 0.05). The reason for this adverse effect in the hyperbaric group
is not clear. It is also worth noting that 18.4% of the hyperbaric group did not com-
plete the required number of chamber sessions for reasons including anxiety and
middle ear problems. In an abstract presented to the ASM of UHMS117 in 2001,
Weaver states that the NBO group tolerated chamber therapy better (96% versus 82%,
p = 0.002).
While these adverse events are not major, they must be taken in the context of the
degree of benefit obtained from the treatment.
• Cost analysis
In his 1995 letter to the editor of the Annals of Emergency Medicine,107 Weaver
states that other questions his trial may answer include differences between the two
therapies (HBO and NBO) related to cost (including transport). This analysis is yet
to be seen.
Other criticisms include the following:
• Small number of intubated patients (12) prevents interpretation of what

HBO may or may not have to offer seriously injured patients.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 419 — #29

• Having stratified for time to treatment less than or greater than 6 h, Weaver
does not report if there was a difference between these two groups, yet he
is so critical of those studies where there was a delay of >6 h.
• Weaver’s subsequent analysis concluded that: “HBO improved outcome if
any of the following: LOC, COHb >25%, age > 50 and metabolic acidosis
were present. In patients with none of these four criteria, HBO did not
improve outcome.
This conclusion uses an absolute value of COHb at the time this was first sampled
and the recommended selection criteria could therefore exclude patients with signi-
ficant exposure but delay in COHb measurement. It is worth noting that the actual
COHb levels were only available in 55% of his patients. This statement is in con-
tradiction with the literature and Weaver’s previous and subsequent declaration that
“the difference in COHb between the groups is of no consequence as there is no
relationship between COHb and outcome.110
Further it is difficult to understand the choice of age > 50, when his study stratified
for age > 40 years and did not present the data for age greater than or less than 40.
• The outcome should not be derived, as Weaver’s was, from complex inter-
pretations of pooled differences in test scores, especially when those tests
are not routinely conducted in clinical practice.115
Buckley92 concludes overall that the unbalanced recruitment, changed primary

outcome, and the intention to treat (ITT) assumption when considered with the very
small number of patients (16) resulting in the finding in favor of HBO, could easily
mean that the trial outcome would have changed from significant to nonsignificant
had these factors been otherwise.
Given all the above, Weaver et al.’s conclusion of benefit arising of HBOT is not
convincing. The benefit of HBOT demonstrated in this study, if there is one, may not
be clinically significant.
18.3 CONCLUSIONS
There are significant difficulties with comparing the outcomes of these investigations.
Variations in study design, HBO and NBO protocols used, outcomes measured
and patient populations included, make it difficult to draw firm conclusions.87 In
addition, many of the studies show bias towards use of HBO in the more severely
affected patients, follow-up is incomplete and overall, the numbers of patients studied
is low.
No reliable method to identify patients at high risk for neurologic sequelae
has been identified. The efficacy of one HBO protocol over another has not been
determined.
Timing of evaluation (discharge, 1 month, 6 weeks, 1 year) has also not been
determined.
The ongoing debate about the efficacy of HBO is driven largely by these discrepant
results.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 420 — #30

The best, most sophisticated, unbiased and comprehensive analysis of published

human trial work to date, without conflicts of interest, and no pre-existing finan-
cial or reputational bias either pro or against HBO, was performed by Buckley92
who concluded that “The role of HBO remains unclear and the weight of the avail-
able evidence neither confirms nor refutes a clinically meaningful net benefit.”
Buckley assessed the effectiveness of HBO compared with NBO for the preven-
tion of neurological sequelae in patients with CO poisoning. Eight randomized
controlled trials were identified. Two had no evaluable data and were excluded.
A pooled OR for the presence of neurological symptoms at 1-month follow-up was
calculated.
At 1 month follow-up after treatment, sequelae possibly related to CO poisoning
were present in 242 of 761 patients (36.1%) treated with NBO compared with 259
of 718 patients (31.8%) treated with HBO. The OR for neuropsychiatric symptoms
with HBO was 0.77 (95% CI = 0.51, 1.14).
A further analysis was performed by removing the results of trials with a Jadad
score of less than 3/5 (the two trials published in abstract form only), leaving a
total of 146 of 383 patients (38%) randomized to HBO with neurological sequelae at
one month compared with 152 of 368 patients (41%) randomized to NBO [OR for
neurological sequelae 0.70 (95% CI = 0.34, 1.47)].
When restricting the analysis to the two studies that enrolled more severely
poisoned patients (the only two with a Jadad score of 5/5) the results remained
inconclusive [OR for neurological sequelae 0.73 (95% CI = 0.22, 2.48)].
Buckley states that there were methodological shortcomings in all trials and
empiric evidence of bias in some, particularly those suggesting benefit from HBO.
The trials enrolled patients with CO poisoning of varying severity, employed different
regimens of HBO and NBO. Only two were conducted with double-blinding through
sham treatment. Pooled analysis of such inconsistent studies should be interpreted
with caution.
In the following issue of Toxicological Reviews, four knowledgeable clinical
toxicologists from different parts of the world were asked to provide editorials on the
subject. Brent, the editor, comments: “Given six relevant randomized clinical trials
involving 1479 patients, if the effect of HBO were real and large, it is difficult to
imagine that the trials would not be more definitive.118 Even the positive ones can be
interpreted as having only marginal benefit.” “There are clearly insufficient data to
consider HBO as a standard of care and it should be considered to be a therapy of as
yet unproven benefit. Further, any benefit deriving from HBO is likely to be small.
Thus, no physician or poison center should be held liable for withholding HBO given
the uncertainty and even possibility of harm associated with this treatment.”
Henry119 states: “Most of us would probably choose HBO because we know
there is no real evidence of harm and “because it might do some good.” However,
the benefit is not likely to be great.”
Olsen114 concludes: “There is no recognized standard of care mandating the use
of HBO.”
Bentur120 was the most positive in favor of HBO, stating “it is impossible to state
that HBOT should not be offered.” He goes on to provide an algorithm for treatment
of CO poisoning based on the results of Weaver’s and Scheinkestel’s studies.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 421 — #31

Seger121 states: “believers believe that the studies that support HBO have fewer
limitations than the studies that do not support HBO (the believers are very fervent,
and tend to lump the atheists and agnostics together and suggest both be put to the
sword). The more sceptical nonbelievers feel that the studies demonstrating no benefit
have fewer limitations. The camps are divided,” with many in each camp seeming to
have an investment in maintaining their belief system, regardless of the data.
Lastly, there is in fact a downside to HBO. There is a cost both for providing the
service and transport to the facility. This has not to-date been quantified. Transports of
critically ill patients are associated with risk and while this has not specifically been
quantified for CO-poisoning, there is considerable evidence for this in critically-ill
patients. Further, there is documented morbidity to patients due to the treatment in a
hyperbaric chamber.
A report by the Department of Health and Human Services, Office of Inspector
General, June Gibbs Brown, in October 2000 on HBOT: Its Use and
Appropriateness,122 states: “According to our review, 18% of beneficiaries exhibit
side effects (significantly greater than the literature suggests). The most common side
effect is ear-related trauma, representing 63% of all observed side effects. While side
effects are generally not severe, two individuals within our sample showed signs of
oxygen toxicity. This relates to 1.3% of the population which also is significantly
greater than the expected value cited in the literature. These statistics were based
on our analysis of the 1998 National Claims History file maintained by Health Care
Financing Administration (HCFA).”
18.4 WHERE TO NOW?

While independent reviewers have, predictably, called for large, multi-center random-
ized, controlled and blinded, clinical trials, many in the hyperbaric community appear
to believe this is either unnecessary, unethical or impractical, if not all three. Plans
are therefore emerging for further randomized studies comparing different doses of
HBO, selected from existing regimens and a pilot has been conducted.123
The optimal dose of HBO is certainly an important question for all indications for
HBO, but it could be argued that at the present time we know too little about HBO
for CO to select either the optimal doses or the optimal patient eligibility criteria for
an optimal study.
It will be apparent from the preceding that we agree that it would be absolutely
necessary for further randomized trials to provide proof of benefit before HBO could
be generally and unconditionally recommended for CO poisoning. We also believe
that it would not in any way be unethical to randomize patients, given the lack of
clear-cut evidence for the superiority of any one form of treatment over any other. It
is not clear to us, however, that another randomized controlled trial is the appropriate
next step.
Large multicenter studies carry a large financial cost however and sometimes
there can be a significant opportunity cost as well. If one study consumes the limited
numbers of eligible patients, capable centers and research funding available for sev-
eral years, this can be to the detriment of potentially better studies conceived after
8417: “8417_c018” — 2007/9/19 — 13:22 — page 422 — #32

new basic science or epidemiologically generated information becomes available.

It is therefore most important that any future studies have an optimal design and
conduct, and are designed to answer a reasonable hypothesis. Herein arises a sig-
nificant issue. Simple and broad hypotheses applied to heterogeneous populations
with many confounding variables require very large studies to ensure matching of
active and control groups, to ensure that relatively small but significant outcome
improvements are detected and to allow for the subgroup analyses necessary to
identify response in important subpopulations. Another factor that mandates large
sample sizes is low incidence of the outcome measure of importance and this applies
to the CO poisoning outcomes of death after rescue and development of delayed onset
neurological sequelae (DNS). As an example, a highly important 50% reduction in
mortality or DNS rate from, say, 10% to 5% would require a two armed randomized
study sample size of nearly 1000 subjects to provide 80% likelihood of detection at
p = 0.05 (2-tailed T test).
The alternative trial design strategy is to utilize tight enrollment criteria to identify
a specific subgroup for a test of therapy thought to provide significant benefit, based
upon pilot data or coherent and generalizable large animal research. This could allow
for smaller study sizes if the therapeutic effect is powerful and the outcome measure
is relatively close to evenly distributed in the control group. Caution is subsequently
needed in generalizing the results of any such study to different populations, doses
or treatment timings.
Patient selection for any therapy is clearly important and this could be a critical
issue in the case of CO poisoning. To date, recommendations regarding the optimal
population for treatment with HBO have generally used combinations of age, COHb
level, history of LOC and presenting signs and symptoms. Consideration of age and
indicators of cardiovascular or cerebral disease do seem very logical given the way
that cerebral or cardiac hypoxia can exacerbate the severity of clinical poisoning
that results from any given COHb level. It is not clear from the studies published to
date whether such criteria are valid. While outcome for patients who have suffered
CO-poisoning-related cardiac arrest is universally bad,124 there may or may not be a
definable larger group of patients so severely brain injured as to have no reasonable
prospect of recovery. It is clear that long-term harm can occur with and without LOC
and recovery can follow severe as well as mild poisoning. It is not yet clear, however,
whether there is a minimum threshold for injury and most significantly, whether there
is better response to HBO amongst any particular CO poisoning population.
In attempting to generate potential clinical study designs, a number of relevant
hypotheses can usefully be generated from the existing human and animal data and
from basic principles:
• The value of HBO may vary significantly with the stage of injury.
• The optimal dose of HBO may vary with stage of injury.
• Optimal dose, especially in very early stage treatment, may be variable with
degree of poisoning and this would indicate the need for therapy tailored
to some measurable variable.
• HBO and even NBO might be harmful at certain stages of the second-
ary injury process, perhaps only in susceptible individuals, by increasing
8417: “8417_c018” — 2007/9/19 — 13:22 — page 423 — #33

lipid peroxidation or other manifestations of oxidative injury rather than

reducing them.
• Unconsciousness or gross neurological deficit on admission is correl-
ated with poorer outcomes and this is usually taken to be an indicator
of severity of brain injury but this may not mean that this is the most
treatment-responsive group.
• If one could measure and allow for the effects of residual intracellular CO
and/or cellular hypoxia this seems likely to correlate with injury severity,
but again might not select for the most treatment responsive group.
• LOC may be possible without harm, provided that cerebral oxygen delivery
is maintained by greatly increased cerebral blood flow, and this might be
a group of patients who can recover without any active therapy.
• Overall outcomes and response to therapy may be different depending
upon the presence or absence of cotoxins such as volatile hydrocarbons,
cyanides and other products of combustion.
• There may be significant inter-individual variability in susceptibility to CO
mediated secondary cerebral injury and in the degree of recovery possible.
• There may be a significant inter-individual variability in response to HBO.
• Such variabilities may have both genetically determined and acquired
elements.
• If HBO does accelerate recovery and reduce hospital stay, this may be
sufficiently valuable to be worthwhile even if there is no net change in
longterm outcome. If so, therapy could be most valuable to populations
selected for characteristics other than susceptibility, for instance working
age patients, parents, and carers or the psychiatrically disturbed.
The size and complexity of the studies likely to be needed and the extent of the
above and, no doubt, other unknowns would suggest that the time is not yet right for
large and costly human randomized controlled trials.
Given that the provision of HBO for CO poisoning is established practice in some
centers, it is not unreasonable for these centers to undertake studies comparing differ-
ent but commonly used doses of HBO provided adequately powered, well governed,
collaborative studies can be achieved at moderate cost, with study designs that address
the limitations of previous work.
Meanwhile, a huge natural experiment is continuing, unfortunately with little
analyzable data being collected; CO poisoning is common and patients currently
receives a wide range of different oxygen doses in both the normobaric and hyperbaric
treatment environment. A well-designed clinical registry has the potential to generate
specific hypotheses for testing with clinical trials and even to answer many questions
outright. Meaningful data must be collected, however, and this will require agreement
on markers of poisoning severity and outcome measures. It would also be much more
valuable if any registry drew data from centers that do not use HBO as well as from
the hyperbaric community’s patients. In addition to enabling comparison of outcomes
between patients receiving NBO and HBO, the optimal initial duration of NBO for
different poisoning severity is unknown, as is the place of increased inspired oxygen
fraction during the postpoisoning, secondary brain injury phase.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 424 — #34

Animal and cell based research will continue to produce information that should
be taken into consideration and of particular note here is the field of CO therapeutics.
In recent years, identification of the role of CO as an intracellular signalling molecule
and element of the hemoxygenase stress response system has led to two important
conclusions: some elements of the pathology of CO poisoning may be related to
disturbances of the physiological role of CO45,125,126 and exogenous CO might be a
useful therapeutic substance.9,127−139 . Both of these ideas have implications for CO
poisoning therapeutics. Exogenous CO can inhibit ischemia-reperfusion injury and
appears to have neuro-protective, cardio-protective and lung protective properties in
some animal models and this has already led to some human clinical trials. The doses
used in these trials have been as high as 500 ppm,140 an order of magnitude above cur-
rent occupational health limits and a level that has been associated with pathological
outcomes, a fact which has been pointed out in critical commentaries.11,12 Neverthe-
less, the potential for exogenous CO to modulate immune and oxidative stress-related
processes raises the question as to whether residual CO may have some protective
effects after the bulk of excessive CO is eliminated following CO poisoning. It is
clear that both CO and HBO biochemistry are in a state of evolution that should be
closely monitored by toxicologists and clinical researchers.
Work in other types of neurological injury is also likely to yield important informa-
tion, both with regard to the potential for HBO use in the later and delayed onset stages
of CO-related injury and with respect to alternative therapies to minimize CO related
brain injury. Strategies which show promise for diffuse and hypoxic brain injury, such
as moderate hypothermia and various anti-oxidant drugs,141,142 could well provide
benefit for CO poisoning patients. If alternative neuroprotective strategies show sig-
nificant benefit for CO, it would be appropriate to compare these with HBO and to
test combined therapy.
CO-related cardiac injury is also be a field needing further investigation.
Recent work indicates that measurable injury is far more common than previously
considered143 and HBO may moderate this, although this needs investigation. Cardiac
injury might also prove a useful and more easily measured outcome than neuro-
toxicity for research aiming to select optimal dosing of oxygen and might even
prove to be a usable surrogate enabling early assessment of overall severity of
injury.
Given the lack of predictive power of clinical signs and current biochemical
markers, new markers of neurological injury severity would be useful if they were
shown to correlate with either outcome or response to therapy. Cleaved tau, non-
specific enolase, MBP and S100B57,144−146 are amongst markers being considered.
Serum could be readily collected from CO poisoning patients to produce a database
which could yield useful information, provided good follow up and outcome measures
are available to enable correlations to be sought.
Medical imaging is advancing rapidly and various markers of severity have
been proposed based upon Computed Tomography (CT), Magnetic Resonance
Imaging (MRI) and Tc-Hexamethylpropyleneamine Oxime (HMPAO) Single Photon
Emission Computed Tomography scanning (SPECT).147−157 To date none have
become established as reliable determinants of response to treatment and any such
markers would need to be readily available with rapid turn-around if they are to be
8417: “8417_c018” — 2007/9/19 — 13:22 — page 425 — #35

useful in deciding therapeutic directions. Quantitative EEG monitoring has also been
tested as a marker of response, no doubt, will other measures.158
A final critical factor is that it seems likely that there are variations in innate
susceptibility or resistance to injury and responsiveness or resistance to therapy for
CO poisoning, as is increasingly seen as important in many other areas. Gender
is almost certainly significant, as is the case in trauma and stroke.159−161 As this
may be based upon antioxidant protective mechanisms of female hormones, it is
conceivable that response to exogenous oxygen therapy could be different and any
future research should look for potential gender differences. Recently, Hopkins162 has
reported in abstract that ApoE typing is a significant predictor of outcome in patients
followed-up at his institution. The ApoE e4 allele, associated with poor outcomes
in many other forms of degenerative and acute brain conditions, was associated
with higher incidence of cognitive sequelae after CO poisoning and benefit from
HBO appeared to be confined to patients with ApoE4—in those without the e4 allele
there was no difference in outcome between HBO and non HBO treated groups.
At present this measure is not available as a point of care test but this and other
markers, including those drawn from the fields of genomics and proteomics may
hold the future of case selection and progress monitoring for specific treatments
such as HBO.
Despite over four decades of hope and persistence, HBO has, as yet, failed to
live up to it’s promises as a valuable therapy for CO poisoning. While there are
sufficient indications of potential benefit to support ongoing research, there is insuf-
ficient evidence to support promotion of HBO use outside of clinical trials or for
the development of new HBO facilities specifically for CO poisoning treatment.
Given an ongoing problem with CO poisoning and an increasing availability of high
standard hyperbaric facilities incorporated into major hospitals, the infrastructure
for next generation clinical trials exists. Hopefully lessons will be taken from trials
to date in developing future trials but it should not be assumed that existing proto-
cols are optimal and an open mind should be kept to alternatives. Meanwhile, those
facilities treating CO poisoning regularly will hopefully contribute data to registries
which will also gather data on patients treated with NBO, which continues to be
the standard of care, albeit without any good data regarding the optimal duration
of therapy.
References
1. End E, Long CW. Oxygen under pressure in carbon monoxide poisoning. J. Ind. Hyg.
Toxicol. 1942; 24: 302–306.
2. Pace N, Strajman E, Walker EL. Acceleration of carbon monoxide elimination in man
by high pressure oxygen. Science 1950; 111(2894): 652–654.
3. Hampson NB. Emergency department visits for carbon monoxide poisoning in the
Pacific Northwest. J. Emerg. Med. 1998; 16(5): 695–698.
4. Hampson NB. Trends in the incidence of carbon monoxide poisoning in the United
States. Am. J. Emerg. Med. 2005; 23(7): 838–841.
5. Brennan C, Routley V, Ozanne-Smith J. Motor vehicle exhaust gas suicide in Victoria,
Australia 1998–2002. Crisis 2006; 27(3): 119–124.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 426 — #36

6. Morrell S, Page AN, Taylor RJ. The decline in Australian young male suicide. Soc.
Sci. Med. 2007; 64(3): 747–754.
7. Ryter SW, Morse D, Choi AM. Carbon monoxide and bilirubin: potential therapies for
pulmonary/vascular injury and disease. Am. J. Respir. Cell. Mol. Biol. 2007; 36(2):
175–182.
8. Hill-Kapturczak N, Agarwal A. Carbon monoxide: from silent killer to potential
remedy. Am. J. Physiol. Renal. Physiol. 2006; 290(4): F787–788.
9. Hoetzel A, Schmidt R. [Carbon monoxide - poison or potential therapeutic]. Der
Anaesthetist 2006; 55(10): 1068–1079.
10. Mannaioni PF, Vannacci A, Masini E. Carbon monoxide: the bad and the good side
of the coin, from neuronal death to anti-inflammatory activity. Inflamm. Res. 2006;
55(7): 261–273.
11. Hampson NB, Weaver LK, Piantadosi CA. "Low-level" carbon monoxide adminis-
tration may carry risk. Am. J. Respir. and Crit. Care Med. 2005; 172(6): 784.
12. Thom SR, Weaver LK, Hampson NB. "Therapeutic" carbon monoxide may be toxic.
Am. J. Respir. Crit. Care Med. 2005; 171(11): 1318.
13. Choi AM, Dolinay T. "Therapeutic" carbon monoxide may be a reality soon. Am. J.
Respir. Crit. Care Med. 2005; 171(11): 1318–1319.
14. Weaver LK. Operational use and patient care in the monoplace hyperbaric chamber.
Resp. Care Clinics of N. Am. 1999; 5(1): 51–92.
15. Hunt TK, Ellison EC, Sen CK. Oxygen: at the foundation of wound healing -
introduction. World J. Surg. 2004; 28(3): 291–293.
16. Thom SR, Bhopale VM, Velazquez OC, Goldstein LJ, Thom LH, Buerk DG. Stem
cell mobilization by hyperbaric oxygen. Am. J. Physiol. Heart. Circ. Physiol. 2006;
290(4): H1378–1386.
17. Buras J. Basic mechanisms of hyperbaric oxygen in the treatment of ischemia-
reperfusion injury. Internatl. Anesthesiology Clin. 2000; 38(1): 91–109.
18. Buras JA, Stahl GL, Svoboda KK, Reenstra WR. Hyperbaric oxygen downregulates
ICAM-1 expression induced by hypoxia and hypoglycemia: the role of NOS. Am. J.
Physiol. Cell. Physiol. 2000; 278(2): C292–302.
19. Brown SD, Piantadosi CA. Recovery of energy metabolism in rat brain after carbon
monoxide hypoxia. J. Clin. Invest. 1992; 89(2): 666–672.
20. Brown SD, Piantadosi CA. Reversal of carbon monoxide-cytochrome c oxidase
binding by hyperbaric oxygen in vivo. Adv. Exper. Med. Biol. 1989; 248: 747–754.
21. Sukoff MH, Ragatz RE. Hyperbaric oxygenation for the treatment of acute cerebral
edema. Neurosurgery 1982; 10(1): 29–38.
22. Al-Waili NS, Butler GJ, Beale J, et al. Hyperbaric oxygen in the treatment of patients
with cerebral stroke, brain trauma, and neurologic disease. Adv. In. Therapy. 2005;
22(6): 659–678.
23. Veltkamp R, Siebing DA, Sun L, et al. Hyperbaric oxygen reduces blood-brain bar-
rier damage and edema after transient focal cerebral ischemia. Stroke 2005; 36(8):
1679–1683.
24. Yang ZJ, Camporesi C, Yang X, et al. Hyperbaric oxygenation mitigates focal cerebral
injury and reduces striatal dopamine release in a rat model of transient middle cerebral
artery occlusion. Eur. J. Appl. Physiol. 2002; 87(2): 101–107.
25. Hills BA. A role for oxygen-induced osmosis in hyperbaric oxygen therapy. Med.
Hypotheses. 1999; 52(3): 259–263.
26. Freiberger J, Coulombe K, Suliman H, Carraway M, Piantadosi C. Superoxide dis-
mutase responds to hyperoxia in rat hippocampus. Undersea Hyperb. Med. 2004;
31(2): 227–232.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 427 — #37

27. Speit G, Dennog C, Eichhorn U, Rothfuss A, Kaina B. Induction of heme oxygenase-1

and adaptive protection against the induction of DNA damage after hyperbaric oxygen
treatment. Carcinogenesis 2000; 21(10): 1795–1799.
28. Thom SR, Mendiguren I, Hardy K, et al. Inhibition of human neutrophil beta2-
integrin-dependent adherence by hyperbaric O2 . Am. J. Physiol. 1997; 272(3 Pt 1):
C770–777.
29. Wilson HD, Wilson JR, Fuchs PN. Hyperbaric oxygen treatment decreases inflamma-
tion and mechanical hypersensitivity in an animal model of inflammatory pain. Brain
Res. 2006; 1098(1): 126–128.
30. Sumen G, Cimsit M, Eroglu L. Hyperbaric oxygen treatment reduces carrageenan-
induced acute inflammation in rats. Eur. J. Pharmacol. 2001; 431(2): 265–268.
31. Lou M, Chen Y, Ding M, Eschenfelder CC, Deuschl G. Involvement of the
mitochondrial ATP-sensitive potassium channel in the neuroprotective effect of
hyperbaric oxygenation after cerebral ischemia. Brain Res. Bull. 2006; 69(2):
109–116.
32. Vlodavsky E, Palzur E, Soustiel JF. Hyperbaric oxygen therapy reduces neuroinflam-
mation and expression of matrix metalloproteinase-9 in the rat model of traumatic
brain injury. Neuropath. and Appl. Neurobiol. 2006; 32(1): 40–50.
33. Vlodavsky E, Palzur E, Feinsod M, Soustiel JF. Evaluation of the apoptosis-
related proteins of the BCL-2 family in the traumatic penumbra area of the rat
model of cerebral contusion, treated by hyperbaric oxygen therapy: a quantitative
immunohistochemical study. Acta. Neuropathologica. 2005; 110(2): 120–126.
34. Ostrowski RP, Colohan AR, Zhang JH. Mechanisms of hyperbaric oxygen-induced
neuroprotection in a rat model of subarachnoid hemorrhage. J. Cereb. Blood Flow
Metab. 2005; 25(5): 554–571.
35. Yu Y, Matsuyama Y, Yanase M, et al. Effects of hyperbaric oxygen on GDNF
expression and apoptosis in spinal cord injury. Neuroreport 2004; 15(15): 2369–2373.
36. Li Y, Zhou C, Calvert JW, Colohan AR, Zhang JH. Multiple effects of hyperbaric
oxygen on the expression of HIF-1 alpha and apoptotic genes in a global ischemia-
hypotension rat model. Exper. Neurol. 2005; 191(1): 198–210.
37. Yin D, Zhang JH. Delayed and multiple hyperbaric oxygen treatments expand thera-
peutic window in rat focal cerebral ischemic model. Neurocritical care 2005; 2(2):
206–11.
38. Sloan EP, Murphy DG, Hart R, et al. Complications and protocol considerations in
carbon monoxide-poisoned patients who require hyperbaric oxygen therapy: report
from a ten-year experience. Annals Emerg. Med. 1989; 18(6): 629–634.
39. Plafki C, Peters P, Almeling M, Welslau W, Busch R. Complications and side effects
of hyperbaric oxygen therapy. Aviation, Space, Environ. Med. 2000; 71(2): 119–124.
40. Yildiz S, Aktas S, Cimsit M, Ay H, Togrol E. Seizure incidence in 80,000 patient
treatments with hyperbaric oxygen. Aviation, Space, Environ. Med. 2004; 75(11):
992–994.
41. Hampson N,Atik D. Central nervous system oxygen toxicity during routine hyperbaric
oxygen therapy. Undersea Hyperb. Med. 2003; 30(2): 147–153.
42. Smerz RW. Incidence of oxygen toxicity during the treatment of dysbarism. Undersea
Hyperb. Med. 2004; 31(2): 199–202.
43. Hampson NB, Simonson SG, Kramer CC, Piantadosi CA. Central nervous sys-
tem oxygen toxicity during hyperbaric treatment of patients with carbon monoxide
poisoning. Undersea Hyperb. Med. 1996; 23(4): 215–219.
44. Haldane J. The Relation of the Action of Carbonic Oxide to Oxygen Tension. J.
Physiol. 1895; 18(3): 201–217.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 428 — #38

45. Gorman D, Drewry A, Huang YL, Sames C. The clinical toxicology of carbon
monoxide. Toxicology 2003; 187(1): 25–38.
46. Piantadosi CA. Carbon monoxide poisoning. Undersea Hyperb. Med. 2004; 31(1):
167–177.
47. Weaver L. Carbon monoxide (CO) poisoning in 12 young, college educated skiiers.
Undersea Hyperb. Med. 2006; 33(5): 338.
48. McGrath JJ. Effects of altitude on endogenous carboxyhemoglobin levels. J. Toxicol.
Environ. Health 1992; 35(2): 127–133.
49. Kleinman MT, Leaf DA, Kelly E, Caiozzo V, Osann K, O’Niell T. Urban angina in the
mountains: effects of carbon monoxide and mild hypoxemia on subjects with chronic
stable angina. Archives. Environ. Health 1998; 53(6): 388–397.
50. Foutch RG, Henrichs W. Carbon monoxide poisoning at high altitudes. Am. J. Emerg.
Med. 1988; 6(6): 596–598.
51. Lipman GS. Carbon monoxide toxicity at high altitude. Wilderness Environ. Med.
2006; 17(2): 144–145.
52. Purser D. Interactions among carbon monoxide, hydrogen cyanide, low oxygen hyp-
oxia, carbon dioxide, and inhaled irritant gases. In: Penney DG, Ed. Carbon Monoxide
Toxicity. Boca Raton: CRC Press; 2000: 157–191.
53. White S. Update on the clinical treatment of carbon monoxide poisoning. In: Penney
DG, Ed. Carbon Monoxide Toxicity. Boca Raton: CRC Press; 2000: 264.
54. Shimada H, Morita T, Kunimoto F, Saito S. Immediate application of hyperbaric
oxygen therapy using a newly devised transportable chamber. Ame. J Emerg. Med.
1996; 14: 412–415.
55. Rogatsky GG, Mayevsky A. Mobile hyperbaric chamber: an essential need in
emergency medicine? Ann. Emerg. Med. 2003; 42(6): 849–850.
56. Ducasse JL, Celsis P, Marc-Vergnes JP. Non-comatose patients with acute carbon
monoxide poisoning: hyperbaric or normobaric oxygenation? Undersea Hyperb. Med.
1995; 22(1): 9–15.
57. Brvar M, Finderle Z, Suput D, Bunc M. S100B protein in conscious carbon monoxide-
poisoned rats treated with normobaric or hyperbaric oxygen. Crit. Care Med. 2006;
34(8): 2228–30.
58. Bunc M, Luzar B, Finderle Z, Suput D, Brvar M. Immediate oxygen therapy prevents
brain cell injury in carbon monoxide poisoned rats without loss of consciousness.
Toxicology 2006; 225(2–3): 138–141.
59. Gilmer B, Kilkenny J, Tomaszewski C, Watts JA. Hyperbaric oxygen does not prevent
neurologic sequelae after carbon monoxide poisoning. Acad. Emerg. Med. 2002; 9:
1–8.
60. Thom SR. Carbon monoxide-mediated brain lipid peroxidation in the rat. J. Appl.
Physiol. 1990; 68(3): 997–1003.
61. Thom SR, Bhopale VM, Fisher D, Zhang J, Gimotty P. Delayed neuropathology
after carbon monoxide poisoning is immune-mediated. PNAS 2004; 101(37): 13660–
13665.
62. Thom SR, Bhopale VM, Fisher D. Hyperbaric oxygen reduces delayed immune-
mediated neuropathology in experimental carbon monoxide toxicity. Toxicol. Appl.
Pharm. 2006; 213(2): 152–159.
63. Thom SR. Functional inhibition of leukocyte B2 integrins by hyperbaric oxygen in
carbon monoxide-mediated brain injury in rats. Toxicol. Appl. Pharm. 1993; 123(2):
248–256.
64. Penney DG. Acute carbon monoxide poisoning: animal models: a review. Toxicology
1990; 62(2): 123–160.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 429 — #39

65. Gorman D, Lin HY, Williams C. Blockade of haem oxygenase and nitric oxide syn-
thetase causes cortical dysfunction in sheep exposed to carbon monoxide. Toxicology
2005; 209(3): 237–243.
66. Gorman D, Huang YL, Williams C. A narcotic dose of carbon monoxide induces
neuronal haeme oxygenase and nitric oxide synthetase in sheep. Toxicology 2002;
179(1–2): 79–84.
67. Gorman DF, Huang YL, Williams C. Prolonged exposure to one percent carbon
monoxide causes a leucoencephalopathy in un-anaesthetised sheep. Toxicology 2001;
165(2–3): 97–107.
68. Langston P, Gorman D, Runciman W, Upton R. The effect of carbon monoxide on
oxygen metabolism in the brains of awake sheep. Toxicology 1996; 114(3): 223–232.
69. American College of Emergency Physicians. Clinical policy: critical issues in the
management of adult patients presenting to the emergency department with acute
symptomatic carbon monoxide poisoning. 2006.
70. Phin N. Carbon monoxide poisoning (acute). Clin. Evid. 2005; (13): 1732–43.
71. Ray J, Ball C. Carbon monoxide poisoning. In: Ball CM and Phillips RS, Ed.
Evidence-Based on Call: Acute Medicine: Churchill Livingstone; 2001: 144–148.
72. Medicare Services Advisory Committee (MSAC). Hyperbaric Oxygen Therapy-
MSAC Applications 1018–1020 Assessment report. http://wwwhealyhgovau/
haf/msac 2001.
73. Juurlink DN, Buckley NA, Stanbrook MB, Isbister GK, Bennett M, McGuigan MA,
Hyperbaric oxygen for carbon monoxide poisoning (Review). The Cochrane Library
2006(2).
74. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized
clinical trials: is blinding necessary? Control. Clin. Trials. 1996; 17(1): 1–12.
75. Raphael JC, Elkharrat D, Jars-Guincestre MC, et al. Trial of normobaric and
hyperbaric oxygen for acute carbon monoxide intoxication. Lancet 1989; 2(8660):
414–419.
76. Thom SR, Taber RL, Mendiguren, II, Clark JM, Hardy KR, Fisher AB. Delayed
neuropsychologic sequelae after carbon monoxide poisoning: prevention by treatment
with hyperbaric oxygen. Ann. Emerg. Med. 1995; 25(4): 474–480.
77. Scheinkestel CD, Bailey M, Myles PS, et al. Hyperbaric or normobaric oxygen for
acute carbon monoxide poisoning: a randomised controlled clinical trial. Med. J. Aust.
1999; 170(5): 203–210.
78. Federal committee of the physicians and health insurance companies. Hyperbare
Sauerstofftherapie (HBO) Zusammenfassender Bericht des Arbeitsausschusses
"Ärztliche Behandlung" des Bundesausschusses der Ärzte und Krankenkassen
über die Beratungen der Jahre 1999 und 2000 zur Bewertung der Hyperbaren
Sauerstofftherapie gemäß §135 Abs.1 SGB V [Hyperbaric Oxygen Therapy (HBO)
Summary Report of the Working Committee on “Medical Treatment” of the
Federal Committee of Physicians and Health Insurance Companies Covering Years
1999 and 2000 for the Evaluation of Hyperbaric Oxygen Therapy in Accordance
with §135 Abs.1 SGB V] [German]. Köln, Germany:. In: Geschäftsführung des
Arbeitsausschusses; 2006.
79. Raphael JC, Chevret, S, Driheme, A, Annane, D,. managing carbon monoxide
poisoning with hyperbaric oxygen (Abstract). J. Toxicol.-Clin. Toxicol. 2004; 42:
455–456.
80. Mathieu D, Wattel F, Mathieu-Nolf M, Durak C, Tempe JP, Bouachour G, Sainty JM.
Randomised prospective study comparing the effect of HBO versus 12 hours NBO
8417: “8417_c018” — 2007/9/19 — 13:22 — page 430 — #40

in nin-comatose CO poisoned patients: Results of the interim analysis. Undersea

Hyperb. Med. 1996; 23(Suppl.): 7.
81. Weaver LK, Hopkins RO, Chan KJ, et al. Hyperbaric oxygen for acute carbon
monoxide poisoning. NEJM 2002; 347: 1057–1067.
82. Ducasse JL. Celsis P, Marc-Vergnes JP. Non-Comatose patients with acute carbon
monoxide poisoning: hyperbaric or normobaric oxygenation? Undersea Hyperb. Med.
1995; 22(1): 9–15.
83. Tibbles PM, Perrotta PL. Treatment of carbon monoxide poisoning: a critical review
of human outcome studies comparing normobaric oxygen with hyperbaric oxygen.
Ann. Emerg. Med. 1994; 24(2): 269–276.
84. Mark P. Carbon Monoxide poisoning: a review. South Pacific Underwater Med. Soc.
1992; 22(3): 127–135.
85. Shank ES, Muth CM. Decompression illness, iatrogenic gas embolism, and carbon
monoxide poisoning: the role of hyperbaric oxygen therapy. Internatl. Anesthesiol.
Clin. 2000; 38(1): 111–138.
86. Weaver LK. Carbon monoxide poisoning. Crit. Care Clin. 1999; 15(2): 297–317, viii.
87. Kao LW, Nanagas KA. Carbon monoxide poisoning. Emerg. Med. Clin. of N. Am.
2004; 22(4): 985–1018.
88. Seger D, Welch L. Carbon monoxide controversies: neuropsychologic testing, mech-
anism of toxicity, and hyperbaric oxygen. Ann. Emerg. Med. 1994; 24(2): 242–248.
89. Hampson NB, Mathieu D, Piantadosi CA, Thom SR, Weaver LK. Carbon monox-
ide poisoning: interpretation of randomized clinical trials and unresolved treatment
issues. Undersea Hyperb. Med. 2001; 28(3): 157–164.
90. Olson KR, Seger D. Hyperbaric oxygen for carbon monoxide poisoning: does it really
work? Ann. Emerg. Med. 1995; 25(4): 535–537.
91. Weaver LK, Hopkins RO, Larson-Lohr V. Hyperbaric oxygen and carbon monoxide
poisoning. Ann. Emerg. Med. 1995; 26(3): 390–392.
92. Buckley N.A. Isbister,GK, Stokes B. Hyperbaric oxygen for carbon monoxide pois-
oning: asystematic review and critical analysis of the evidence. Toxicol. Rev. 2005;
24: 75–92.
93. Gorman DF, Runciman WB. Carbon monoxide poisoning. Anaesth. Intens. Care
1991; 19(4): 506–511.
94. Littell RC Milliken GA, Stroup W.W. Wolfinger R.D. SAS (computer program). In:
Cary, NC: SAS Institute Inc., 1996.
95. Bland JMK, Sally M. Statistical Notes: Trials randomised in clusters. BMJ 1997;
315(6 September): 7108.
96. Simpson JM, Klar N, Donnor A. Accounting for cluster randomization: a review of
primary prevention trials, 1990 through 1993. Am. J. Public Health 1995; 85(10):
1378–1383.
97. Scheinkestel CD, Tuxen DV, Bailey M, et al. Hyperbaric oxygen in carbon monoxide
poisoning. Authors of study clarify points that they made. BMJ (Clinical research ed.)
2000; 321(7253): 109–111.
98. Scheinkestel CD, Bailey M, Myles PS, et al. Hyperbaric or normobaric oxygen for
acute carbon monoxide poisoning: a randomised controlled clinical trial. Med. J. Aust.
1999; 170(11): 563–565.
99. Moon RE, DeLong E. Hyperbaric oxygen for carbon monoxide poisoning. Med. J.
Aust. 1999; 170(5): 197–199.
100. Weaver LK. Hyperbaric oxygen in carbon monoxide poisoning. BMJ (Clinical
Research ed.) 1999; 319(7217): 1083–1084.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 431 — #41

101. Tighe SQ. Hyperbaric oxygen in carbon monoxide poisoning. 100% oxygen is best
option. BMJ (Clinical Research ed.) 2000; 321(7253): 110–111.
102. Weaver LK. Hyperbaric oxygen in carbon monoxide poisoning. BMJ (Clinical
research ed.) 2000; 321(8 July): 109.
103. Schiltz KL. Failure to assess motivation, need to consider psychiatric variables,
and absence of comprehensive examination: a skeptical review of neuropsycholo-
gic assessment in carbon monoxide research. Undersea Hyperb. Med. 2000; 27(1):
48–50.
104. Denson LA, Hay PJ. Was the neuropsychological assessment appropriate? MJA 1999;
170(7 June): 563.
105. Scheinkestel CD, Bailey M, Myles PS, Jones K, Cooper DJ, Millar I, Tuxen DV,
Hyperbaric oxygen for carbon monoxide poisoning. Med. J. Aust. 2000; 172:
141–142.
106. Weaver LK. Double-blind, controlled, Prospective, Randomized Clinical trial (RCT)
in patients With Acute Carbon Monoxide poisoning: Outcome of patients treated With
Normobaric Oxygen or hyperbaric Oxygen-An Interim Report. Undersea Hyperb.
Med. 1995; 22(suppl.): 14.
107. Weaver LK, Hopkins RO, Larson-Lohr V. Carbon monoxide poisoning: a review of
human outcome studies comparing normobaric oxygen with hyperbaric oxygen. Ann.
Emerg. Med. 1995; 25(2): 271–272.
108. Buckley NA. Isbister,GK, Juurlink DN, Hyperbaric oxygen for carbon monoxide
poisoning. Toxicol. Rev. 2005; 24(3): 159–160.
109. Weaver LK, Hopkins RO, Elliott G. Carbon monoxide poisoning. NEJM 1999;
340(16): 1290.
110. Weaver LK, Hopkins RO, Chan KJ, et al. Carbon Monoxide Research Group, LDS
Hospital, Utah in reply to Scheinkestel et al. and Emerson: the role of hyperbaric
oxygen in carbon monoxide poisoning. Emerg. Med. Aust. 2004; 16(5–6): 394–399;
discussion 481–482.
111. Blumenthal I. Carbon monoxide poisoning. J. Royal Soc. Med. 2001; 94(6): 270–272.
112. The Hyperbaric Oxygen Therapy Committee Report. Hyperbaric Oxygen 2003:
Indications and Results. Kensington, MD: Undersea Hyperb. Med. Soc. 2003.
113. Weaver LK. 6 & 12 month outcome of acute carbon monoxide poisoning treated with
hyperbaric or normobaric oxygen. UHMS, Annual Scientific Meeting 2002 (July
2002, published in UHM, 2002, Vol. 29, No. 2, p. 139).
114. Olson KR. Hyperbaric oxygen or normobaric oxygen? Toxicol. Rev. 2005; 24(3): 151;
discussion 9–60.
115. Juurlink DN. The Cochrane Library — Feedback. http://wwwcochranefeedbackcom/
cf/cda/citationdo?id=9531# 9531 2006.
116. Raphael JC, Annane D, Chevret S. Hyperbaric oxygen for acute carbon monoxide
poisoning. NEJM 2003; 348(6): 557–60.
117. Weaver LK. Outcome of carbon monoxide poisoning treated with hyperbaric or nor-
mobaric oxygen. UHMS 2001 (Abstracts of the UHMS Annual Scientific Meeting
June 14–16, 2001).
118. Brent J. What does the present state of knowledge tell us about the potential role of
hyperbaric oxygen therapy for the treatment of carbon monoxide poisoning? Toxicol.
Rev. 2005; 24(3): 145–147.
119. Henry JA. hyperbaric therapy for carbon monoxide poisoning. Toxicol. Rev. 2005;
24(3): 149–150.
120. Bentur Y. Hyperbaric oxygen for carbon monoxide poisoning. Toxicol. Rev. 2005;
24(3): 153–154; discussion 9–60.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 432 — #42

121. Seger D. WL. The Myth. Toxicol. Rev. 2005; 24(3): 155–156.
122. Gibbs Brown J, Inspector General. Hyperbaric oxygen therapy: its use and appro-
priateness. In: Services Department of Health and Human Services, ed.: Office Of
Inspector General; 2000.
123. Hampson NB, Dunford RG, Ross DE, Wreford-Brown CE. A prospective, random-
ized clinical trial comparing two hyperbaric treatment protocols for carbon monoxide
poisoning. Undersea Hyperb. Med. 2006; 33(1): 27–32.
124. Hampson NB, Zmaeff JL. Outcome of patients experiencing cardiac arrest with carbon
monoxide poisoning treated with hyperbaric oxygen. Ann. Emerg. Med. 2001; 38(1):
36–41.
125. Piantadosi CA, Carraway MS, Suliman HB. Carbon monoxide, oxidative stress,
and mitochondrial permeability pore transition. Free Rad. Biol. Med. 2006; 40(8):
1332–1339.
126. Cronje FJ, Carraway MS, Freiberger JJ, Suliman HB, Piantadosi CA. Carbon monox-
ide actuates O(2)-limited heme degradation in the rat brain. Free Rad. Biol. Med.
2004; 37(11): 1802–1812.
127. Wu L, Wang R. Carbon monoxide: endogenous production, physiological functions,
and pharmacological applications. Pharmacol. Rev. 2005; 57(4): 585–630.
128. Kirkby KA, Adin CA. Products of heme oxygenase and their potential therapeutic
applications. Am. J. Physiol. Renal. Physiol. 2006; 290(3): F563–571.
129. Nakao A, Choi AM, Murase N. Protective effect of carbon monoxide in transplanta-
tion. J. Cell. Mol. Med. 2006; 10(3): 650–671.
130. Zuckerbraun BS, Chin BY, Wegiel B, et al. Carbon monoxide reverses established
pulmonary hypertension. J. Exper. Med. 2006; 203(9): 2109–2119.
131. Kohmoto J, Nakao A, Kaizu T, et al. Low-dose carbon monoxide inhalation prevents
ischemia/reperfusion injury of transplanted rat lung grafts. Surgery 2006; 140(2):
179–185.
132. Li Volti G, Rodella LF, Di Giacomo C, et al. Role of Carbon Monoxide and Biliv-
erdin in Renal Ischemia/Reperfusion Injury. Nephron Exp. Nephrol. 2006; 104(4):
e135–e139.
133. Botros FT, Navar LG. Interaction between endogenously produced carbon monoxide
and nitric oxide in regulation of renal afferent arterioles. Am. J. Physiol. Heart Circ.
Physiol. 2006; 291(6): H2772–2778.
134. Rodella L, Lamon BD, Rezzani R, et al. Carbon monoxide and biliverdin prevent
endothelial cell sloughing in rats with type I diabetes. Free Rad. Biol. Med. 2006;
40(12): 2198–2205.
135. Chen YH, Tsai HL, Chiang MT, Chau LY. Carbon monoxide-induced early throm-
bolysis contributes to heme oxygenase-1-mediated inhibition of neointimal growth
after vascular injury in hypercholesterolemic mice. J. Biomed. Sci. 2006; 13(5):
721–730.
136. Bilban M, Bach FH, Otterbein SL, et al. Carbon monoxide orchestrates a protective
response through PPARgamma. Immunity 2006; 24(5): 601–610.
137. Li MH, Cha YN, Surh YJ. Carbon monoxide protects PC12 cells from peroxynitrite-
induced apoptotic death by preventing the depolarization of mitochondrial transmem-
brane potential. BBRC 2006; 342(3): 984–990.
138. Lee BS, Heo J, Kim YM, et al. Carbon monoxide mediates heme oxygenase 1
induction via Nrf2 activation in hepatoma cells. BBRC 2006; 343(3): 965–972.
139. Schober P, Koch A, Zacharowski K, Loer SA. [Carbon monoxide: toxic molecule
with antiinflammatory and cytoprotective properties]. Anasthesiol. Intensivmed.
Notfallmed. Schmerzther 2006; 41(3): 140–149.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 433 — #43

140. Mayr FB, Spiel A, Leitner J, et al. Effects of carbon monoxide inhalation during
experimental endotoxemia in humans. Am. J. Resp. and Crit. Care Med. 2005; 171(4):
354–360.
141. Giffard RG, Jaffe RA. Advances in understanding protection from cerebral ischemia.
Curr. Opin. Anaesthesiol. 2002; 15(5): 495–500.
142. Wang KK, Larner SF, Robinson G, Hayes RL. Neuroprotection targets after traumatic
brain injury. Curr. Opin. Neurol. 2006; 19(6): 514–519.
143. Henry CR, Satran D, Lindgren B, Adkinson C, Nicholson CI, Henry TD. Myocar-
dial injury and long-term mortality following moderate to severe carbon monoxide
poisoning. JAMA 2006; 295(4): 398–402.
144. Rasmussen LS, Poulsen MG, Christiansen M, Jansen EC. Biochemical markers for
brain damage after carbon monoxide poisoning. Acta Anaesthesiologica Scand 2004;
48(4): 469–73.
145. Berger RP, Adelson PD, Richichi R, Kochanek PM. Serum biomarkers after traumatic
and hypoxemic brain injuries: insight into the biochemical response of the pediatric
brain to inflicted brain injury. Develop. Neurosci. 2006; 28(4–5): 327–335.
146. Brvar M, Mozina H, Osredkar J, et al. S100B protein in carbon monoxide poisoning:
a pilot study. Resuscitation 2004; 61(3): 357–360.
147. Gale SD, Hopkins RO. Effects of hypoxia on the brain: neuroimaging and neuro-
psychological findings following carbon monoxide poisoning and obstructive sleep
apnea. J. Int. Neuropsychol. Soc. 2004; 10(1): 60–71.
148. Wu CI, Changlai SP, Huang WS, Tsai CH, Lee CC, Kao CH. Usefulness of 99mTc
ethyl cysteinate dimer brain SPECT to detect abnormal regional cerebral blood flow
in patients with acute carbon monoxide poisoning. Nucl. Med. Comm. 2003; 24(11):
1185–1188.
149. Sener RN. Acute carbon monoxide poisoning: diffusion MR imaging findings. AJNR
2003; 24(7): 1475–7.
150. Teksam M, Casey SO, Michel E, Liu H, Truwit CL. Diffusion-weighted MR
imaging findings in carbon monoxide poisoning. Neuroradiology 2002; 44(2):
109–113.
151. Kesler SR, Hopkins RO, Blatter DD, Edge-Booth H, Bigler ED. Verbal memory defi-
cits associated with fornix atrophy in carbon monoxide poisoning. J Int Neuropsychol.
Soc 2001; 7(5): 640–646.
152. Hurley RA, Hopkins RO, Bigler ED, Taber KH. Applications of functional imaging
to carbon monoxide poisoning. J. Neuropsychiatry and Clin. Neurosci. 2001; 13(2):
157–160.
153. Gale SD, Hopkins RO, Weaver LK, Bigler ED, Booth EJ, Blatter DD. MRI, quant-
itative MRI, SPECT, and neuropsychological findings following carbon monoxide
poisoning. Brain Inj. 1999; 13(4): 229–243.
154. Silver DA, Cross M, Fox B, Paxton RM. Computed tomography of the brain in acute
carbon monoxide poisoning. Clin. Radiology 1996; 51(7): 480–483.
155. Pracyk JB, Stolp BW, Fife CE, Gray L, Piantadosi CA. Brain computerized tomo-
graphy after hyperbaric oxygen therapy for carbon monoxide poisoning. Undersea
Hyperb. Med. 1995; 22(1): 1–7.
156. Murata T, Itoh S, Koshino Y, et al. Serial cerebral MRI with FLAIR sequences
in acute carbon monoxide poisoning. J. Computer-Assisted Tomogr. 1995; 19(4):
631–634.
157. Gotoh M, Kuyama H, Asari S, Ohmoto T, Akioka T, Lai MY. Sequential changes in
MR images of the brain in acute carbon monoxide poisoning. Comput. Med. Imaging
Graph. 1993; 17(1): 55–59.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 434 — #44

158. Murata M, Suzuki M, Hasegawa Y, Nohara S, Kurachi M. Improvement of occipital

alpha activity by repetitive hyperbaric oxygen therapy in patients with carbon monox-
ide poisoning: a possible indicator for treatment efficacy. J. neurol. Sci. 2005;
235(1–2): 69–74.
159. Vink R, Nimmo AJ, Cernak I. An overview of new and novel pharmacotherapies for
use in traumatic brain injury. Clin. Exper. Pharm. Physiol. 2001; 28(11): 919–921.
160. Roof RL, Hall ED. Gender differences in acute CNS trauma and stroke: neuropro-
tective effects of estrogen and progesterone. J. Neurotrauma 2000; 17(5): 367–388.
161. Choudhry MA, Bland KI, Chaudry IH. Gender and susceptibility to sepsis following
trauma. Endocrine, Metab. Immune Disorders Drug Targets 2006; 6(2): 127–135.
162. Hopkins RO. Effects of the ApoE4 allele on cognitive outcome in acute CO poisoning.
Undersea Hyperb. Med. 2006; 33(5): 337–338.
8417: “8417_c018” — 2007/9/19 — 13:22 — page 435 — #45

8417: “8417_c018” — 2007/9/19 — 13:22 — page 436 — #46
19 A Challenge to the
Healthcare Community:
The Diagnosis of Carbon
Monoxide Poisoning
David G. Penney
CONTENTS
19.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437

19.2 Problems in the Diagnosis of CO Poisoning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
19.3 Chronic Carbon Monoxide Poisoning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
19.4 Addendum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446
19.4.1 The RAD-57 cm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448
19.1 INTRODUCTION
Overview Statement:
If you were able to chose the kind of brain injury you were to incur, it would be better
in terms of the potential for recovery to have a stroke, a concussion in a motor vehicle
accident, etc. than carbon monoxide poisoning.
There are many problems with CO poisoning generally (Table 19.1). Although
CO is a very simple molecule, its mechanisms of action are complex and multiple.
There is more to its pathophysiology than simply the tying up of hemoglobin such
that oxygen cannot be transported. It traverses the blood-barrier and dissolves in the
protoplasm of cells, attaching there to a variety of molecules.
As stated elsewhere in this book, CO is a very “smart poison.” It is colorless,
tasteless, odorless, and is completely nonirritating to respiratory and mucous mem-
branes. Thus it is undetectable by humans with unaided senses. Unlike other “dumb”
poisons, it leaves the body quickly when the victim again breathes unpolluted air,
leaving behind just the damage it has done. CO doesn’t hang around to be detected
and identified weeks, months, or years later, like lead and mercury.
437
8417: “8417_c019” — 2007/9/11 — 12:14 — page 437 — #1

TABLE 19.1
Some of the Major Problems in Dealing with Carbon
Monoxide Poisoning1
1. Mechanisms of toxicity involve more than CO attaching to hemoglobin.
2. Lack of public awareness of the dangers of accidental CO poisoning.
3. Insufficient professional awareness in the diagnosis of CO poisoning.
4. Extent of suicidal use of motor exhaust and other fumes.
5. Assessment of the severity of the poisoning.
6. Criteria to be used in the treatment of poisoned patients.
7. Pathogenesis and prediction of the development of delayed sequelae.
8. Effectiveness of hyperbaric oxygen therapy is largely unproven.
9. Are there any effective alternative treatments to HBO?
10. Unknown morbidity and mortality of undiagnosed cases.
The public and indeed most members of the healthcare community are ignorant of
many of the risks involved in working and playing around equipment that emits CO
(see Chapter 15). While automobiles have enjoyed the reputation of being extremely
dangerous in terms of CO emissions, this is no longer the case when they are in
good repair and are used properly. Because of catalytic converters, virtually all of
the CO produced by the automobile engine is now converted to harmless carbon
dioxide and water vapor. Other vehicles such as high-powered, watercraft fueled by
gasoline and using engines similar to those used in cars, now pose the greatest risk of
deadly CO poisoning to individuals. Some of these boats, such as a single inboard,
twin-engine ski boat may emit as much CO as 250–300 automobiles. CO concen-
trations immediately behind such boats may range to over 20,000 ppm depending
on boat movement, wind direction, and speed, CO concentrations at least 40 times
the lethal dose for humans. People have difficulty recognizing and appreciating the
danger posed by these boats, since it usually occurs in the open air, not inside a
structure.
Medical personnel have traditionally underappreciated the frequency of CO pois-
onings, both the acute and the chronic types, but especially the latter, and have
been notorious in misdiagnosing CO poisoning. Because the symptoms produced are
mostly general and nonspecific, they are usually associated with other diseases that
lie more central to internal medicine, conditions that result from bacteria, viruses,
hormonal changes, physical trauma, and so forth. In fact, the very nature of the
multiplicity of symptoms that CO poisoning produces, tends to confuse physicians
and nurses. If the flu, hypothyroidism, stroke, and so forth are not suggested, then
psychosomatic or psychiatric conditions often are. Since few symptoms or signs are
produced that occur only with CO-poisoning (i.e., pathognomonic), it is often diffi-
cult to immediately identify a cause. Identification of CO poisoning as the cause is
made easier by the taking of a full situational history. When possible, this consists
of determining where the patient was during the past few hours. Who was with him
and were the others also sick? Was it a house, apartment, recreational vehicle, and so
forth? Were pets also sick or died? How was the space heated? Did the patient notice
8417: “8417_c019” — 2007/9/11 — 12:14 — page 438 — #2

A Challenge to the Healthcare Community 439
a pattern in that he/she was less sick when he/she left the site, and became sicker
again upon return? Had there been reports of heating devices malfunctioning, or the
use of internal combustion engines near or within the structure?
Assessment of the severity of CO poisoning has been a thorny problem. Far
too much reliance has been placed on carboxyhemoglobin (COHb) measured at the
emergency center, and too little on evaluation of the conscious state and behavior
of the victim. Recent studies suggest that gait and balance may be among the most
important prognostic signs, and should be used far more often in deciding the course
of treatment. Long or “soaking” exposures to CO invite more rapid and aggressive
therapy than shorter CO exposures. The early use of the new generation of pulse
CO-oximeters (see addendum and Chapter 33), possibly even at the site of discovery,
will aid in assessing poisoning severity and treatment regimen.
As long as the use of hyperbaric oxygen therapy (HBOT) has been available to
treat CO-poisoned patients, it remains unclear whether it is really effective in reduced
long-term health damage. Two other chapters in this book discuss in detail the pros and
cons of the half a dozen or so clinical studies that have tested the HBO hypothesis.
Little progress appears to have been made in the past decade in developing new
approaches to treating CO poisoning, or in understanding HBOT.
19.2 PROBLEMS IN THE DIAGNOSIS OF CO

POISONING
CO poisoning is difficult to diagnose for a number of reasons, and the rate of mis-
diagnosis has been in some cases, shockingly high. Part of the problem has to do
with its special properties, and some has to do with healthcare workers ignorance
of the symptoms by which it presents (Table 19.2). There has been overreliance
on COHb level, and also the fact that COHb is not routinely measured in patients
TABLE 19.2
Major Reasons for Failure to Diagnose Carbon Monoxide Poisoning in the
Emergency Room
1. Usually a low (or zero) index of suspicion for CO poisoning by physicians.
2. Lack of training in toxicology and myopic thinking by physicians.
3. Too many disparate, seemingly unrelated, multisystems symptoms involved.
4. Nonspecific symptoms confused physicians.
5. Incorrect clue given by patient or companion.
6. Failure to obtain complete situational histories.
7. Presentation in ER that appears not to require emergency measures.
8. Mistaken belief that COHb must always be greatly elevated.
9. Distraction by other traumatic or metabolic condition/disease.
10. Dependence on old style, two wavelength pulse-oximeter for measurement of O2 saturation.
11. Use of less than the best, even irrelevant clinical tests for CO, unless goal is R/O (i.e., rule out).
12. COHb is NOT routinely measured in the ER.
8417: “8417_c019” — 2007/9/11 — 12:14 — page 439 — #3

TABLE 19.3
Carbon Monoxide Exposure Types
• Acute, brief exposure Immediate presentation high COHb
Delayed presentation COHb may be low
+ 100% oxygen COHb may be low
• Acute, “soaking” Immediate presentation - high COHb
+100% oxygen COHb may be low
• Chronic (>24 h) Immediate presentation COHb high to low
+100% oxygen COHb may be low
TABLE 19.4
A Situational History
• Living abode - house, apartment, etc.
• Heating systems
• Other people sick there
• Pets sick/dead
• Feel better when away, worse when return
• Periodicity with work, weekends, vacations/season
• Car driven - year, engine
• Use of small gasoline engine/attached garage
• Prolonged illness/seen other docs previously/recently
• New vs. old building
entering a physician’s office or an emergency room (ER). Recent studies show that
there is virtually no relationship between reported COHb saturation and long-term
outcome. In some cases COHb was barely elevated or even normal, yet the patient
sustained brain damage. The grid of the various types of CO poisonings shown in
Table 19.3 indicates why COHb may be normal or well below levels considered
toxic in individuals who sustained serious CO poisonings. We see oxygen saturation
obtained by two-wavelength pulse oximeters continuing to be printed on the charts of
CO-poisoned patients although it has no reality in such cases. The falsely high num-
ber could even lead to tragic action by a healthcare worker who does not understand
that the number is not accurate.
There are numerous examples of healthcare workers failing to avail themselves
of all of the clues available to them when a patient presents with symptoms consistent
with CO poisoning. Many of these cases end tragically. A high index of suspicion
must be maintained about possible CO poisoning when seeing patients (Table 19.4).
One very important tool in diagnosing CO poisoning, and indeed other poisons, is
the “situational history”. Table 19.5 shows some of the components, that is, questions
asked when and if possible, when building a situational history.
8417: “8417_c019” — 2007/9/11 — 12:14 — page 440 — #4

TABLE 19.5
Index of Suspicion (for CO Poisoning) Should be
Raised by:
• Presentation during winter
• Several/many people presenting “ill” from same household
• Elevated plasma glucose
• c/o Memory problems
• c/o Furnace/heat problem at home
• Presentation after electrical power outage
• c/o Pets sick/dead
• c/o Headache, Nz, Dz, fatigue/weakness
• c/o “Illness” when going to/staying at, one place, not at others
• Flushed pink/cherry red skin
• Unexplained LOC/syncope
Very recently in Florida, a man and his adult son were staying at a motel resort.
Overnight both were poisoned by CO. The son died, while the father survived. It
was later learned that other motel patrons had been sick the week before, had gone
to the hospital, where they were eventually diagnosed with CO poisoning. With
an appropriate index of suspicion of CO, the use of proper diagnostic techniques
by emergency personnel, and effective communication, tragedies like this can be
averted. Clearly, getting the right diagnosis can be critical with possible CO poisoning.
Table 19.6 lists possible misdiagnoses of CO poisoning, and Table 19.7 lists the
frequency of reported incidences of major misdiagnoses of CO poisoning.
In another instance several years ago, a prominent urologist was attending a
meeting in the west with his wife, a surgical nurse. After the first night in a posh new
ski hotel, both became violently sick. They went to the local ER. Upon entering the
facility the urologist said he believed they had eaten some “bad” chicken sandwiches
on the plane coming from the east. The ER personnel, hearing this from a fellow
physician, accepted his diagnosis and treated him and his wife with antiemetics, and
so forth. The doctor and his wife then went back to the hotel to continue the urology
meeting, skiing, and dining. The next morning the urologist was dead, and his wife
was in a coma that lasted several days. She suffered permanent, irreversible brain
damage, involving changes in cognitive ability, personality, physical pain, and so
forth. If only COHb had been measured at the ER. Later it was learned through
testimony and examination of internal memos that employees and management of
the hotel had been aware of a “CO problem” there for over 6 months before the
doctor died.
In another case in a large eastern US city, an elderly woman was taken to a very
prominent local hospital known for its expertise and training programs in emergency
and trauma medicine. ER personnel there failed to diagnose the patient with CO
poisoning, apparently because the emergency medical technicians who went to her
house and transported the woman saw a bottle near the chair she was found in. She was
treated as an alcohol overdose case. Shortly afterward, it was discovered that there
8417: “8417_c019” — 2007/9/11 — 12:14 — page 441 — #5

TABLE 19.6
Some Possible Mistaken Diagnoses in Patients with Carbon Monoxide
Poisoning1
Misdiagnosis Cause
Neurological
Cerebrovascular accident Cerebral ischemic accident due to CO poisoning
Migraine, tension headache Headache
Epilepsy Anoxic convulsions
Meningitis, encephalitis Vomiting, headache, bizarre neurological symptoms
Parkinsonism Late-onset parkinsonian symptoms
Psychiatric
Depression Lethargy, somatic symptoms
Anxiety state Hyperventilation, headache, malaise
Hyperventilation syndrome Hyperventilation
Acute confusional state Confusion, hallucinations
Cardiac
Myocardial infarction A critical coronary artery lesion decompensated through hypoxia
Cardiac arrhythmias Conduction system hypoxia
Pharmacological and toxicological
Drug overdose Hypoxic coma, nontraumatic rhabdomyolysis
Ethylene glycol poisoning Coma and renal failure
Ethanol intoxication Vomiting, ataxia, slurred speech, coma
Drug abuse Agitation, confusion, hallucinations
Infections
Influenza and other viral infections Muscle aches, tachypnea, headache, exhaustion
Post viral syndrome Lethargy, myalgia
Gastroenteritis and food poisoning Nausea, vomiting
Pneumonia Dyspnea, delirium
Sinusitis Headache, malaise
Others
Cholecystitis and other Abdominal pain, nausea, vomiting
acute abdominal conditions
was no liquor of any kind in the house. The woman had been sitting, then slumping,
in the same chair for as long as 3 days. At the hospital the usual “tox screen” found
no alcohol or other drugs in her body. Shortly after that, substantial concentrations
of CO were discovered in her house by the woman’s daughter. The woman remained
comatose for many days in hospital, had a long drawn out period of recovery, and
sustained severe brain damage. This is the second example of comments by less
qualified personnel being allowed to influence the decision about the proper medical
work-up to achieve an accurate diagnosis. In this case the patient failed to receive the
proper treatment, which in this case would almost certainly have been HBOT.
8417: “8417_c019” — 2007/9/11 — 12:14 — page 442 — #6

TABLE 19.7
Common Misdiagnosis for Carbon Monoxide
Poisoning by Reported Frequency2
Food poisoning 38%
Psychiatric disease, hysteria, confusion, depression 18%
Heart disease presenting as angina or syncope 13%
Alcohol intoxication or delirium tremens 7%
Acute solvent intoxication 7%
Migraine headache 6%
Ischemic cerebral disease 4%
Cerebral hemorrhage 4%
Cerebral tumor (convulsions) 3%
TABLE 19.8
Characteristics of Chronic Carbon Monoxide Poisoning6
1 Lasts more than 24 h
2 Often goes long undetected
3 Masquerades as the flu, fatigue, and so forth.
4 Often many people “sick” simultaneously
5 The “sickness” may show a periodicity synchronous with season
6 May go away upon leaving poisoning site (to work, on vacation, etc.)
7 Nearly always misdiagnosed by physicians
8 May involve pets “sick,” dead at same time
9 Rarely involves sinus congestion, cough [when present, it may be due to other compounds
(e.g., NOx , SO2 ) in exhaust gases]
19.3 CHRONIC CARBON MONOXIDE POISONING

Working Assumption:
For every single case of chronic carbon monoxide poisoning reported / successfully
diagnosed, there are 10 cases that go unreported/undiscovered/undiagnosed.
Chronic CO poisoning as I define it is an exposure to CO of more than 24 h,

whether continuous or discontinuous.3 Beyond this, the characteristics of such CO
poisonings (Table 19.8) often go long undetected, sometimes for years. It may mas-
querade as the flu, chronic fatigue, fibromyalgia, chemical hypersensitivity (MCS),
and so forth. Often several or many people are “sick” simultaneously, and for extended
periods uncharacteristic of viral or bacterial infections. The condition often subsides
or disappears when the victim leaves the poisoning site, but reappears when the victim
returns. It may show a seasonal periodicity if it is the primary space heating appli-
ance that is malfunctioning. This type of CO poisoning has the highest misdiagnosis
8417: “8417_c019” — 2007/9/11 — 12:14 — page 443 — #7

Cycle of misdiagnosis
CO ignored CO regarded as rare
Patient ill and stays at home
Chronic exposure to CO
Misdiagnosis Effects not understood
FIGURE 19.1 Cycle of misunderstanding and misdiagnosis.
rate by healthcare professionals. In the CO Support Study in the UK it was the heat-
ing professionals who were most likely to diagnose the problem.4 This type of CO
poisoning may result in pets sick with symptoms not unlike those of the humans. It
rarely involves sinus congestion, cough or sneezing, although leaking exhaust gases
containing irritants could do this—CO alone will not.5
Figure 19.1 illustrates the cycle of misdiagnosis often encountered with chronic
CO poisoning. The presence of CO in a breathing space usually goes long undetected,
and even when it is detected may be denied by those in authority. The presence of the
CO and its possible health damaging effects are often misunderstood by those being
poisoned, and worse, by the medical personnel consulted. Finally, people feeling ill
from what eventually turns out to be CO, almost invariably spend increased time in
the contaminated breathing space attempting to recover, but only become worse or
possibly even die with severe poisoning.
Some of the comments that I hear from the public regarding such poisonings
are found in Table 19.9. I have condensed these comments into what I call “com-
monalities.” The symptoms consist of headache, fatigue, nausea, dizziness, and so
forth.3 Nearly everyone who contacts me is certain he/she was exposed to CO in one
way or another. There has to be a source, since CO doesn’t come out of thin air.
A dominant theme I’ve heard hundreds of times is that “my doctor won’t take my
complaints seriously,” that “CO comes, goes (once you are in fresh air), and then you
are ok,” and finally, “there are no medical people in my area who understand CO’s
effects!”. The public wants to know how it is diagnosed, when he/she will get better,
and what the treatments are. Diagnosis of CO poisoning is a problem as discussed
above, particularly if the victim has left the site of exposure for some days or weeks,
since CO quickly leaves the body. Making pronouncements about when people will
8417: “8417_c019” — 2007/9/11 — 12:14 — page 444 — #8

TABLE 19.9
Inquiries About Carbon Monoxide Poisoning—Commonalities in my
Experience6
• Complaints of continuous headache, fatigue, nausea, dizziness, and so forth.
• Certainty by complainant that CO exposure occurred.
• “My doctor won’t take my complaints seriously!”—“CO comes, goes, and . . . (he/she says) you
are ok!”
• “There are no medical people are in my area who understand CO’s effects!”
• How is it diagnosed? When will I get better? What are the treatments?
• What long-term damage might the CO have caused to my child/children?
TABLE 19.10
Clues to the Discovery of Chronic Carbon Monoxide Poisoning6
• Lethargy, headache, and so forth of long duration
• Long-standing “illness” intractable to medical solutions
• Multiple cases of similar illness at one location
• “Illness” that may suddenly improve when leaving site
• “Illness” that improves when combustion device(s) is turned off or taken away
• Morbidity/mortality of pets
• CO alarm sounding, once or repeatedly
• Presence of malfunctioning furnace, water heater, and so forth.
• Measurement of CO by fireman, service personnel, and so forth at the presumed site of poisoning.
be well again is very risky even under circumstances of thorough knowledge of a

case, and telling people it is unlikely they will get well or that their health damage
is permanent is very hard to do. I do make recommendations about treatment, but as
we all know, brain damage currently remains irreversible. Finally, many people ask
what long-term damage might the CO have caused their children, born and unborn.
Many effects of CO on both children and fetuses have been described by Penney,7
but identifying damage caused by CO in any one case is difficult.
There are many clues to discovery of chronic CO poisoning (Table 19.10).
Headache and lethargy of long duration should raise the suspicion of CO expos-
ure. Similar to the comments above, a long-standing “illness” intractable to medical
solutions should have CO poisoning placed high on the differential diagnosis list.
Likewise, the incidence of multiple cases of similar “illness” at one location is an
important clue, and an “illness” that suddenly improves when the victim leaves the
site, or when a key combustion device is rendered nonfunctional or is removed,
should cause the investigator to suspect a site-specific poisoning. As above, morbid-
ity or even mortality of pets is important. “CO is an equal opportunity poison,” not
discriminating on the basis of skin color, gender, religion, or even species. If the
organism is warm-blooded and uses hemoglobin as its circulating oxygen carrier, it is
subject to injury or death from CO poisoning. The degradation of a heating appliance
8417: “8417_c019” — 2007/9/11 — 12:14 — page 445 — #9

TABLE 19.11
Common Misdiagnoses for Chronic Carbon
Monoxide Poisoning6
• The “flu”
• Other viral or bacterial pulmonary or gastrointestinal infections
• Bad/tainted/poisoned food
• Psychosomatic problem, malingering
• General “run-down” condition
• A “female” problem
• Allergy/asthma
• Psychiatric condition, for example, depression
• Chronic fatigue syndrome
• Chemical hypersensitivity (i.e., MCS)
• Fibromyalgia
• Multiple sclerosis (MS)
• Lymes disease
• Endocrine problem (e.g., hyper- or hypo-thyroid condition)
• Immune deficiency
such as furnace through corrosion of one or another part, often produces chronic CO
poisoning, while discovery of the problem usually occurs when the CO problem is
particularly severe. A CO detector/alarm that sounds intermittently over a period of
days or weeks is probably identifying a site of chronic CO poisoning.
Table 19.11 lists common misdiagnoses made in cases of chronic CO poison-
ing. The “flu” is probably the most common misdiagnosis. Lack of a fever does not
guarantee that the condition is not being caused by a “bug.” “Bad or tainted” food
may be pronounced the cause, but this curse is more commonly encountered as the
misdiagnosis in cases of acute CO poisoning, as revealed in the tragic story above. A
third large area of false cause identification centers around accusations of mental ill-
ness, depression, psychosomatic condition, malingering/faking it, and exaggerating
minor irritations. This usually stems from the confusion that CO poisoning causes in
medical personnel owing to the large number of symptoms generated. Other classic
misdiagnoses for chronic CO poisoning include multiple sclerosis, lyme’s disease,
fibromyalgia, chronic fatigue syndrome, hypothyroidism, chemical hypersensitivity,
allergic reaction, immune deficiency, a “female” problem, general run-down con-
dition, and endocrine problem. Many of these can be easily ruled out by doing the
appropriate clinical test.
19.4 ADDENDUM
19.4.1 THE RAD-57 CM
Masimo is the innovator of Signal Extraction Technology (SET)® pulse oximetry and
the inventor of pulse CO-oximetry. This technology is capable of continuously and
noninvasively measuring COHb and methemoglobin (MetHb) in the blood.
8417: “8417_c019” — 2007/9/11 — 12:14 — page 446 — #10

FIGURE 19.2 Masimo Rad-57, hand-held pulse CO-oximeter. It allows clinicians and others
to continuously and noninvasively measure carbon monoxide levels in the blood, reducing
the need for invasive and expensive arterial blood gas analysis. (From Masimo Corporation
website. Rad-57 Pulse CO-oximeter. Available at: http://www.masimo.comlrad-57/index.htm.
Accessed September 26, 2005.)
Masimo Rainbow SET™ was developed out of Masimo SET® . It informs the
operator about the vascular oxygen status of patients, while continuously and non-
invasively monitoring other species of hemoglobin, such as COHb (SpCO%) and
MetHb (SpMet%). Rainbow extends Masimo SET® through the addition of signal
analysis algorhithms that run in parallel with it, to reveal the presence and levels of
these hemoglobin species (see Figure 19.2)
Because the Masimo parallel engines and adaptive filters receive more than seven
discrete wavelengths of light (note incoming signal λ1 . . . λn ), multiple constituents
of hemoglobin are detectable and can be quantitated, as compared to conventional
pulse-oximeters that employ only two wavelengths. The wavelengths used span those
necessary to recognize oxygenated hemoglobin as well as COHb and MetHb. For
example, when CO is bound to hemoglobin, a conventional red/infrared oximeter
misreads COHb essentially as oxygenated hemoglobin, producing a falsely high SpO2
value that may have disastrous immediate, delayed, or even chronic effects on brain
and cardiac function.
Pulse CO-oximetry relies on the same principles of spectrophotometry used to
determine blood oxygen saturation in the laboratory. The underlying physical prin-
ciple in pulse oximetry is the absorption of specific light wavelengths while passing
8417: “8417_c019” — 2007/9/11 — 12:14 — page 447 — #11

through tissues. In pulse CO-oximetry the percent of total hemoglobin found as either
COHb (SpCO%) and/or MetHb (SpMet%) is measured.
The Masimo Rainbow sensor emits multiple wavelengths of light. Output of the
adjoining photo detector is sent to the instrument, where Rainbow SET™ technology
employs parallel algorhithms and adaptive digital filters to process the data. The Rain-
bow device displays as percentages: oxyhemoglobin (SpO2 %), COHb (SpCO%), and
MetHb (SpMet%), plus the core parameters derived from the Masimo SET technology
platform, pulse rate (PR), perfusion index (PI), and signal IQ® (SIQ).
A number of single-use, latex-free, adhesive sensors are available for patients
. . . from adults to neonates. Reusable adult and slender finger clip sensors are also
available for short-term monitoring and spot checks. Note that a Rainbow-empowered
device provides a choice of two wavelength sensors for SpO2 , pulse rate, and perfusion
index measurements, or Rainbow SET® multiple-wavelength sensors to add COHb
(SpCO%), MetHb (SpMet%), or both to the pulse-oximeter.
Editors note: This section has been included in this book for informational pur-
poses only. The editor is not an employee of Masimo Corporation and has taken no
money, gifts, whatever, and will not do so in the future, for providing this inclusion.
References
1. Lowe-Ponsford, F.L., Henry, J.A. Clinical aspects of carbon monoxide poisoning.
Adverse Drug Reactions and Acute Poisoning Reviews, 8, 217–240, 1989.
2. Bartlett, R. Carbon monoxide poisoning. In: Clinical Management of Poisoning and
Drug Overdose, 3rd ed., Haddad, L.M., Shannon, M.W., Winchester, J.F., ed., W.B.
Saunders Co., Philadelphia, Chapt.70, pp. 885–898. 1998.
3. Penney, D.G. Chronic carbon monoxide poisoning. In: Carbon Monoxide Toxicity,
D.G. Penney, Ed., CRC Press, NY, 2000, Chapt. 18, pp. 393–418.
4. Hay, A.W.H., Jaffer, S., Davis, D. Chronic carbon monoxide exposure: The CO
Support study. In: Carbon Monoxide Toxicity, D.G. Penney, ed., CRC Press, NY,
2000, Chapt. 19, pp. 419–438.
5. Dwyer, B., Leatherman, Manclark, Kimball, K., Rasmussen, R. Carbon Monoxide:
A Clear and Present Danger, ESCO Press, USA, 3rd ed., 2003 (www.escoinst.com).
6. Available at: www.coheadquarters.com/CO1.htm
humans. In: Carbon Monoxide, D.G. Penney, Ed., CRC Press, NY, 1996, pp. 109–144.
8417: “8417_c019” — 2007/9/11 — 12:14 — page 448 — #12

20 Neuroimaging after
Carbon Monoxide
Exposure
Gunnar Heuser
CONTENTS
20.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449

20.2 Sequential Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450
20.3 Location and Symmetry of Lesions, and Hypofrontality and
Disinhibition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450
20.4 Review of the Recent Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
20.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 454
20.1 INTRODUCTION
Neuroimaging was previously reviewed by I.S.S. Choi in the book Carbon Monoxide
Toxicity, by Penney, published in 2000.1 His review included representative
Computed Tomography (CT), Magnetic Resonance Imaging (MRI), Single Photon
Emission Computed Tomography (SPECT), and Magnetic Resonance Spectroscopy
(MRS) scans. My review is meant to update and complement that of Choi’s.
Neuroimaging helps to understand the clinical picture and to correlate findings
with other evaluations. The field has advanced in recent years in that the techniques
are in some respects vastly improved. The reader is invited to compare the SPECT
scans presented in Choi’s chapter with the ones presented here.
Present-day imaging of anatomical (MRI, CT) and functional [(SPECT, Positron
Emission Tomography (PET)] impairment provides a significant visual aid to the
assessment of brain anatomy and function, especially when presented in color. Brain
scans now also lend themselves to statistical analysis, comparing a given scan to a
control population or analyzing regions of interest (ROI) with other regions in the
same brain scan.2
449
8417: “8417_c020” — 2007/9/19 — 13:23 — page 449 — #1

20.2 SEQUENTIAL STUDIES

CO exposure, acute and chronic, can result in long lasting (i.e., years) impairment of
neurological and other functions. The need for sequential follow-up exams is obvious
in view of the potential long-term effects of CO poisoning.
Follow-up neuroimaging and other testing is often mandatory for assessment of a
given patient as the clinical picture evolves. It is also important if one wants to assess
the efficacy of the course of treatment.
Sequential neuropsychological testing is a very useful assessment tool. However,
it is expensive. Also, some tests administered suffer from potential learning effects,
thus giving the false impression of improvement at the time of follow-up, because the
patient has learned to perform better in the follow-up test since he or she is repeating
the test.
Examinations by a neurologist are not always comparable, especially if the patient
is not followed by the same physician.
In view of the above limitation, sequential neuroimaging studies appears to be
the procedure of choice in assessing a patient over time, provided the patient returns
to the same imaging facility which uses the same imaging protocol.
20.3 LOCATION AND SYMMETRY OF LESIONS, AND

HYPOFRONTALITY AND DISINHIBITION
From a diagnostic point of view, abnormalities after CO exposure would be expected
to be present in similar locations and to be symmetrical. This is not the case. One
would also expect clinical observations and neuroimaging abnormalities to show a
correlation. Often, this is actually not the case.
Finally, one would naively assume that damage in different brain regions should
be viewed as a focal abnormality which is not functionally connected to other areas
of the brain. This is also not true since frontal hypofunction (hypofrontality) may
result in hyperfunction in the posterior areas of the brain and also in subcortical areas.
Since the frontal brain exerts an inhibitory function, frontal lobe damage may result
in disinhibition.3,4
While neurotoxic exposure (e.g., solvents, pesticides, mold toxins) in general may
not be obvious on the MRI at all, or result in nonspecific small high intensity (demy-
elinating/vascular) lesions (foci), MRI studies in CO-exposed patients very frequently
show abnormal MRIs with multiple and significant abnormalities, especially in the
white matter and the globus pallidus. This is in contrast to other neurotoxic insults
and therefore, allows for a tentative diagnosis of CO poisoning on the basis of an
anatomical MRI study.
Functional SPECT, PET, and MRS studies are frequently abnormal, but not dia-
gnostic of CO poisoning. In other words, the abnormalities seen are compatible with
but not diagnostic of CO-induced impairment.
Heuser and Mena in 19982 studied more than 70 patients after neurotoxic
(pesticides, fumes, perfumes, etc.) exposure and found hypoperfusion in frontal, tem-
poral, and parietal lobes using SPECT. The abnormalities were often asymmetrical,
but showed no specific signature after exposure to a given neurotoxin. In other words,
8417: “8417_c020” — 2007/9/19 — 13:23 — page 450 — #2

Neuroimaging after Carbon Monoxide Exposure 451
SY CK-BRAIN 03/27/06
+4
Baseline data versus adult normals I
Right lateral view Anterior view Superior view
+3
+2
Left lateral view Posterior view Interactive view-no cerebelhm
-2
-3
-4
-5
FIGURE 20.1 (See Color insert following page 422) A SPECT scan of cortical function
after carbon monoxide poisoning. The color scale (left side) displays normal perfusion in gray,
subnormal perfusion in green and blue, and hyperperfusion in red. In other cases hypoperfusion
is found in the frontal areas. Thus, the abnormalities found vary from one patient to another.
(Credit to J. Michael Uszler)
unrelated neurotoxins resulted in similar, if not identical abnormalities on SPECT.

This is why we cannot show a typical, that is, diagnostic SPECT or PET brain scan
after CO exposure.
Figure 20.1 shows a SPECT brain scan of a patient who was chronically exposed
to CO and became symptomatic on a long-term basis. This figure is meant to illustrate
the advance in technology which now allows us to display SPECT brain scans in an
understandable and almost three-dimensional fashion. The technology and display of
SPECT was developed by Ismael Mena at UCLA (now in Santiago, Chile).
20.4 REVIEW OF THE RECENT LITERATURE

More than 100 publications on neuroimaging after CO exposure were reviewed for
this chapter and the results of significant contributions are summarized below.
8417: “8417_c020” — 2007/9/19 — 13:23 — page 451 — #3

Hopkins and Woon provide an excellent review in 20065 of structural and func-
tional neuroimaging and neuropsychological evaluation after CO poisoning. They
also discuss in some detail the mechanism of CO-induced injury. Globus pallidus
lesions are described more frequently than all others.6−22 White matter lesions are
described by many authors.9,13,15,18,21−26
Lesions can occur in many other locations such as centrum semiovale,
hippocampus, fornix, frontal, temporal, parietal, occipital lobes, cerebellum, and
of course, the basal ganglia. These locations are discussed in some detail by Hopkins
and Woon.5 While lesions can occur in any part of the brain, they are predominant in
the globus pallidus and also in the white matter.
Wu et al.27 studied ten patients within 2–5 h after CO exposure and found SPECT
to be more sensitive than CT scanning. SPECT scans showed hypoperfusion in seven
patients.
Denays et al.28 studied 12 patients on admission day and found abnormal SPECT
scans in the temporo-parietal-occipital areas in eight patients. Abnormalities were
unilateral in some patients, bilateral in others. These findings confirm my earlier
comment that lesions after neurotoxic exposure are not always symmetrical.
Parkinson et al.29 studied 73 patients during the acute exposure phase and then
followed these patients and retested them 6 months later. White matter lesions could
not be correlated to the clinical severity. Centrum semiovale lesions were correlated
with cognitive impairment—periventricular lesions were not.
Devine et al. in 200230 compared MRI and neuropsychological evaluations in
one patient who had multiple bilateral lesions in the basal ganglia 15 months after
exposure. Porter et al.31 studied MRIs and neuropsychological evaluations on the
day of exposure and 6 months thereafter in 62 patients. They found corpus callosum
atrophy and described correlation in detail. The neuropsychological test results did
not correlate with the level of corpus callosum atrophy.
Kim et al.32 studied the diffusivity of white matter lesions in five patients, describ-
ing periventricular and centrum semiovale demyelinating hyperintense bilateral
lesions.
Ku et al. in 20064 describes a case of mania and discusses the possible disinhibition
of the frontal lobes as a possible mechanism after CO poisoning (this case had frontal
white matter lesions on MRI). This discussion supports my earlier comments on
disinhibition from frontal lobe lesions.
MRI and/or CT scans were done in conjunction with neuropsychological eva-
luation by many authors: n = 16,7 n = 5,13 n = 156,14 n = 9,16 n = 21,33 n = 69.34
Sequential studies were performed and discussed: n = 5,13 n = 69,34 n = 2135 and
others.
Reports from various studies show patients being followed for 80 days,7 3
months,37 6 months,38 3 years,36 4 years,22 6 years,39 9 years,24 10 years,40 and
33 years.23 In one study, MRI and CT scans were performed in 107 patients.18
Parkinsonism was described, but does not necessarily correlate with the loca-
tion of the anatomical or functional lesions.14,16,21,41 Cerebellar lesions were also
described.7,15,23,24,30
Gale et al.33 used MRI, quantitative MRI (QMRI), SPECT, and neuropsycholo-
gical evaluation in 21 patients and found SPECT and QMRI to be the most sensitive
neuroimaging tools for the evaluation of CO poisoning.
8417: “8417_c020” — 2007/9/19 — 13:23 — page 452 — #4

Durak et al. in 20057 describe cortical atrophy, cerebellar atrophy, vermian

atrophy, corpus callosum atrophy, having examined eight patients. Neuroimaging
changes may be present even if patients are asymptomatic. Kesler et al.34 showed
that cognitive impairment can occur in the absence of abnormal imaging studies.
However, they specifically describe fornix atrophy associated with verbal memory
deficits.
Uchino et al.22 examined thirteen patients, 25 years after a coal mine explosion.
Globus pallidus and white matter lesions were found and the parieto-occipital region
was the most affected cortical area. Seven patients had definite asymmetrical cortical
and subcortical lesions.
Dunham, et al.42 showed that almost identical exposure in some of their patients,
did not lead to the same neuropsychological impairment.
Silver et al.18 studied 107 patients. Eight had CT scans. Seven of those had
typical globus pallidus and white matter lesions. Bianco and Floris in 19966 described
hemorrhagic infarction of the globus pallidus in two patients after CO exposure.
Silverman et al.19 also described hemorrhage in the globus pallidus in a patient three
years after exposure.
Pinkston et al. in a remarkable study in 2000,38 describe abnormal test studies in
two adults 3 years after chronic CO exposure and correlate these abnormalities with
the neuropsychological evaluation of these same two adults. Neuroimaging studies
showed decreased functionality in the pre-frontal cortex and the temporal lobe.
Tom et al.20 studied eighteen patients and found globus pallidus lesions in seven
patients and white matter lesions in five of those patients. Pach et al.43 studied SPECT,
MRS and neuropsychological test results in 10 patients. They found only partial
correlation in acute and 6 months postexposure studies. Hurley et al.44 showed an
increase in choline and a decrease in N-acetyl-aspartate using MRS in one of their
patients.
Prockop in a 2005 study16 described abnormal MRS studies in nine patients,
some of whom showed decreased N-acetyl-aspartate, especially in the basal ganglia.
In some patients the changes were asymmetrical, if not unilateral. MRI and
neuropsychological studies were also performed in these patients.
Sohn et al.45 studied a husband and wife pair like Pinkston et al.,38 with appar-
ently identical exposure to CO, but different clinical outcomes—only the husband
developed parkinsonism. Scanning revealed more severe white matter damage in the
husband. Only the wife had bilateral pallidal necrosis. This study illustrates the fact
that the same exposure can have different results in different people.
No review of this type would be complete without mentioning the important con-
tributions of Lapresle and Fardeau,46 who published anatomical–pathological studies
of 22 cases after severe CO poisoning.
20.5 CONCLUSIONS
A review of the literature and my own experience show that no single neuroima-
ging or other test can be used as the one and only diagnostic or prognostic indicator.
In other words, a good clinical or otherwise pertinent history, a review of med-
ical records, a physical examination, neuropsychological testing, and neuroimaging
8417: “8417_c020” — 2007/9/19 — 13:23 — page 453 — #5

studies (structural and functional) are essential for a correct diagnosis, appropriate
treatment, and for thorough follow-up evaluation.
ACKNOWLEDGMENTS
Dr Luke Curtis and Sylvia Heuser contributed valuable ideas to the subject discussed
in this manuscript. Virginia Salisbury accomplished the difficult and meticulous task
of typing the manuscript. Dr. J.M. Uszler of Santa Monica Imaging contributed the
SPECT scan in Figure 20.1.
References
1. Choi, I.S.S., Use of scanning techniques in the diagnosis of damage from carbon
monoxide, In: Carbon Monoxide Toxicity, David G. Penney, Ed., CRC Press, 2000,
pp. 363–380.
2. Heuser, G., and Mena, I. Neurospect in neurotoxic chemical exposure. Demonstration
of long-term functional abnormalities, Toxicol. Industr. Health, 14, 813–827, 1998.
3. Heuser, G., and Wu, J.C. Deep subcortical (including limbic) hypermetabolism in
patients with chemical intolerance: Human PET studies, In: The role of neural plas-
ticity in chemical intolerance, B.A. Song and I.R. Bell, Ed., Ann. NY Acad. Sci., 933,
319–323, 2001.
4. Ku, B.D., Shin, H.Y., Kim, E.J., Park, K.C., Seo, S.W., and Na, D.L. Secondary mania
in a patient with delayed anoxic encephalopathy after carbon monoxide intoxication,
J. Clin. Neuroscience, 13, 860–862, 2006 (Epub, 2006, Aug. 28).
5. Hopkins, R.O., and Woon, F.L.M. Neuroimaging, cognitive, and neurobehavioral
outcomes following carbon monoxide poisoning, Behav. Cogn. Neurosci. Rev., 5,
141–155, 2006.
6. Bianco, F., and Floris, R. MRI appearances consistent with hemorrhagic infarction as
an early manifestation of carbon monoxide poisoning, Neuroradiology, 38 (Suppl. 1),
870–872, 1996.
7. Durak, A.C., Coskun, A., Yikilmaz, A., Erdogan, F., Mavili, E., and Guven, M.
Magnetic resonance imaging findings in chronic carbon monoxide intoxication, Acta
Radiol. 46, 322–327, 2005.
8. Fine, R.D., and Parker, G.D. Disturbance of central vision after carbon monoxide
poisoning, Aust. NZ. J. Ophthalmol., 24, 137–141, 1996.
9. Gotoh, M., Kuyama, H., Asari, S., Ohmoto, T., Akioka, T., and Lai, M.Y. Sequential
changes in MR images of the brain in acute carbon monoxide poisoning, Comput.
Med. Imaging Graph, 17, 55–59, 1993.
10. Horowitz, A.L., Kaplan, R., and Sarpel, G. Carbon monoxide toxicity: MR imaging
in the brain, Radiology, 162, 787–788, 1987.
11. Kanaya, N., Imaizumi, H., Nakayama, M., Nagai, H., Yamaya, K., and Namiki, A.
The utility of MRI in acute stage of carbon monoxide poisoning, Intensive Care Med.,
18, 371–372, 1992.
12. Kleinert, A., Sinczuk-Walczak, H., and Goraj, B. Acute poisoning by carbon
monoxide affecting the extrapyramidal system, Med. Pr., 49, 573–577, 1998.
13. Klostermann, W., Vieregge, P., and Bruckmann, H. Carbon monoxide poisoning; the
importance of computed and magnetic resonance tomographic cranial findings for the
clinical picture and follow-up, ROFO, 159, 361–367, 1993.
8417: “8417_c020” — 2007/9/19 — 13:23 — page 454 — #6

14. Mimura, K., Harada, M., Sumiyoshi, S., Tohya, G., Takagi, M., Fujita, E., Takata, A.,
and Tatetsu, S. Longterm follow-up study on sequelae of carbon monoxide poisoning,
Seishin Shinkeigaku Zasshi, 101, 592–618, 1999.
15. O’Donnell, P., Buxton, P.J., Pitkin, A., and Jarvis, L.J. The magnetic resonance
imaging appearances of the brain in acute carbon monoxide poisoning, Clin. Radiol.,
55, 273–280, 2000.
16. Prockop, L.D. Carbon monoxide brain toxicity: clinical, magnetic resonance ima-
ging, magnetic resonance spectroscopy, and neuropsychological effects in 9 people,
J. Neuroimaging, 15, 144–149, 2005.
17. Sawa, G.M., Watson, C.P., Terbrugge, K., and Chiu, M. Delayed encephalo-
pathy following carbon monoxide intoxication, Can. J. Neurol. Sci., 8, 77–79,
1981.
18. Silver, D.A., Cross, M., Fox, B., and Paxton, R.M. Computed tomography
of the brain in acute carbon monoxide poisoning, Clin. Radiol., 51, 480–483,
1996.
19. Silverman, C.S., Brenner, J., and Murtagh, F.R. Hemorrhagic necrosis and vascular
injury in carbon monoxide poisoning; MRI demonstration, Przegl. Lek., 52, 267–270,
1995.
20. Tom, T., Abedon, S., Clark, R.I., and Wong, W. Neuroimaging characteristics in
carbon monoxide toxicity, J. Neuroimaging, 6, 161–166, 1996.
21. Tvedt, B., Krogstad, J.M., and Berstad, J. Hypoxic brain damage after carbon
monoxide poisoning, Tidsskr Nore Laegeforen, 116, 3005–3008, 1996.
22. Uchino, A., Hasuo, K., Shida, K., Matsumoto, S., Yasumori, K., and Masuda, K. MRI
of the brain in chronic carbon monoxide poisoning, Neuroradiology, 36, 399–401,
1994.
23. Coskun, A., Yikilmaz, A., Guven, M., and Erdogan, F. Cranial MR imaging findings of
carbon monoxide poisoning in asymptomatic patients, the chronic stage, Tani Girisim
Radvol 9, 146–151, 2003.
24. Mascalchi, M., Petruzzi, P., and Zampa, V. MRI of cerebellar white matter damage due
to carbon monoxide poisoning: case report, Neuroradiology, 38 (Suppl. 1), 873–874,
1996.
25. Murata, T., Itoh, S., Koshino, Y., Sakamoto, K., Nishio, M., Maeda, M., Yamada, H.,
Ishil, Y., and Isaki, F. Serial cerebral MRI with FLAIR sequences in acute carbon
monoxide poisoning, Comput. Assist. Tomogr., 19, 631–634, 1995.
26. Prockop, L.D., and Naidu, K.A., Brain CT and MRI findings after carbon monoxide
toxicity, J. Neuroimaging, 9, 175–181, 1999.
27. Wu, C.I., Changlai, S.P., Huang, W.S., Tsai, C., Lee, C.C., and Kao, C.H., Usefulness
of 99mTc ethyl cysteinate dimer brain SPECT to detect abnormal regional cerebral
blood flow in patients with acute carbon monoxide poisoning, Nuclear Med. Comm.,
24, 1185–1188, 2003.
28. Denays, R., Makhoul, E., Dachy, B., Tondeur, M., Noel, P., Ham, H.R., and
Mois, P. Electroencephalographic mapping and 99mTc HMPAO single photon emis-
sion computed tomography in carbon monoxide poisoning, Ann. Emerg. Med., 24,
947–952, 1994.
29. Parkinson, R.B., Hopkins, R.O., Cleavinger, H.B., Weaver, L.K., Victoroff, J., Foley,
J.F., and Bigler, E.D. White matter hyperintensities and neuropsychological outcome
following carbon monoxide poisoning, Neurology, 58, 1525–1532, 2002.
30. Devine, S.A., Kirkley, S.M., Palumbo, C.L., and White, R.F. MRI and neuropsy-
chological correlates of carbon monoxide exposure: A case report, Environ. Health
Persp., 110, 1051–1055, 2002.
8417: “8417_c020” — 2007/9/19 — 13:23 — page 455 — #7

31. Porter, S.S., Hopkins, R.O., Weaver, L.K., Bigler, E.D., and Blatter, D.O. Corpus
callosum atrophy and neuropsychological outcome following carbon monoxide
poisoning, Arch. Clin. Neuropsychology, 17, 195–204, 2002.
32. Kim, J. Chang, K., Song, I.C., Kim, K.H., Kwon, J., Kim, H.C., Kim, J.H., and Han,
M.H. Delayed encephalopathy of acute carbon monoxide intoxication: diffusivity of
cerebral white matter lesions, Am. J. Neuroradiol., 24, 1592–1597, 2003.
MRI, quantitative MRI, SPECT, and neuropsychological findings following carbon
monoxide poisoning, Brain Inj. 13, 229–243, 1999.
34. Kesler, S.R., Hopkins, R.O., Blatter, D.D., Edge-Booth, H., and Bigler, E.D. Verbal
memory deficits associated with fornix atrophy in carbon monoxide poisoning, J. Int.
Neuropsychol. Soc., 7, 640–646, 2001.
35. Hayashi, R., Hayashi, K., Inouse, K., and Yanagisawa, N. A serial computerized
tomographic study of the interval form of CO poisoning, Eur. Neurol., 33, 27–29,
1993.
36. Quattrocolo, G., Leotta, D., Appendino, L., Tarenzi, L., and Duca, S. A case of cortical
blindness due to carbon monoxide poisoning, Ital. J. Neurol. Sci., 8, 57–58, 1987.
37. Sawada, Y., Takahashi, M., Ohashi, N., Fusamoto, H, Maemura, K., Kobayashi, H.,
and Yoshioka, T. Computerised tomography as an indication of long-term outcome
after acute carbon monoxide poisoning, Lancet, 1, 783–784, 1980.
38. Pinkston, J.B., Wu, J.C., Gouvier, W.D., and Varney, N.R. Quantitative PET scan
findings in carbon monoxide poisoning: deficits seen in a matched pair, Arch. Clin.
Neuropsychol., 15, 545–553, 2000.
39. Bruno, A., Wagner, W., and Orrison, W.W. Clinical outcome and brain MRI four years
after carbon monoxide intoxication, Acta. Neurol. Scand., 87, 205–209, 1993.
40. Pasquier, F., DePoorter, M.D., Jacquemotte, N. Adnet-Bonte, C. and Petit, H. Cerebel-
lar syndrome after carbon monoxide poisoning, Magnetic Resonance imaging single
photon emission tomography, Rev. Neurol. (Paris), 149, 805–806, 1993.
41. Ikeda, K., Sasaki, S., Ichijo, S., Matsuoka, Y., and Irimajiri, S. Atlas of cranial
and spinal MRI--magnetic resonance imaging in carbon monoxide poisoning and
Parkinsonian syndrome, No To Shinkei, 50, 86–87, 1998.
42. Dunham, M.D., and Johnstone, B. Variability of neuropsychological deficits asso-
ciated with carbon monoxide poisoning, four case reports, Brain Inj., 13, 917–925,
1999.
43. Pach, D., Urbanik, A., Szczepanska, L., Hubalewska, A., Huszno, B., Groszek, B., and
Jenner, B. (99m)Tc-HmPAO single photon emission tomography, magnetic resonance
proton spectroscopy and neuropsychological testing in evaluation of carbon monoxide
neurotoxicity, Przegl. Lek., 62, 441–445, 2005.
44. Hurley, R.A., Hayman, L.A., and Taber, K.H. Applications of functional imaging to
carbon monoxide poisoning, J. Neuropsychiatry Clin. Neurosci., 13, 157–160, 2001.
45. Sohn, Y.H., Jeong, Y., Kim, H.S., Im, J.H., and Kim, J.S. The brain lesion responsible
for parkinsonism after carbon monoxide poisoning, Arch. Neurol., 57, 1214–1218,
2000.
46. Lapresle, J., and Fardeau, M. The central nervous system and carbon monoxide
poisoning. II. Anatomical study lesions following intoxication with carbon monoxide
(22 cases), Prog. Brain Res., 24, 31–74, 1967.
8417: “8417_c020” — 2007/9/19 — 13:23 — page 456 — #8

21 Recent Advances in
Brain SPECT Imaging
after Carbon Monoxide
Poisoning
S. Gregory Hipskind
CONTENTS
21.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457

21.2 Neuroimaging Modalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458
21.3 Anatomical Imaging Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458
21.3.1 Computerized Tomography (CT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458
21.3.2 Magnetic Resonance Imaging (MRI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459
21.4 Functional Imaging Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460
21.4.1 Positron Emission Tomography (PET) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460
21.4.2 Functional Magnetic Resonance Imaging (FMRI) . . . . . . . . . . . . . . . . 461
21.4.3 Magnetic Resonance Spectroscopy (MRS) . . . . . . . . . . . . . . . . . . . . . . . 461
21.4.4 SPECT Findings in Carbon Monoxide Poisoning . . . . . . . . . . . . . . . . 462
21.5 Neuroimaging Findings in Chronic, Lower-Level CO Poisoning . . . . . . . . 464
21.6 High Resolution SPECT Findings in Ten Cases of Delayed Carbon
Monoxide-Induced Encephalopathy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464
21.7 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471
21.8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 474
21.1 INTRODUCTION
It has been established in this book and earlier books in this series, that carbon
monoxide (CO) poisoning is an insidious cause of death and disability in the United
States and throughout the world.1 Tissue anoxia is most commonly implicated as
the underlying pathophysiologic mechanism of toxic CO exposure as a result of
its displacement of oxygen from hemoglobin forming carboxyhemoglobin (COHb).
In addition, CO has been shown to have a direct effect on several key biochemical
457
8417: “8417_c021” — 2007/9/19 — 13:23 — page 457 — #1

processes by binding directly to hydroxyperoxidase, cytochrome oxidase, and

cytochrome p 450 and other key enzymes in the oxidative metabolic process.2,3 Fur-
thermore, it has been shown that CO binds directly to the heme iron in the globus
pallidus in substantia nigra.4 Among the various types of neurotoxins, CO produces
a unique clinical syndrome in which, after survival of an acute intoxication, a lucid
period of variable duration can ensue followed by the onset of delayed neurological
symptoms (DNS) in sequelae or a fraction of patients.4 The mechanism for the delayed
a fraction effects of CO poisoning is not well understood, but involves elements of
reperfusion injury, vascular oxidative stress by generation of reactive oxygen species,
lipid peroxidation,5 neuronal exitotoxicity, and apoptosis.6,7 When these molecular
changes affect large enough areas of neural tissue, they may be visualized by various
neuroimaging techniques. We will see later how the pattern, distribution, and nature
of these induced neuropathalogic changes may provide characteristic neuroimaging
clues associated with CO poisoning. The role of specific neuroimaging modalities
during the various stages of CO poisoning will be reviewed.
21.2 NEUROIMAGING MODALITIES

Various neuroimaging modalities have been employed over the years to evaluate
the neuropathological changes associated with both acute and chronic CO poison-
ing. Also see Chapter 20 in this book. Typically, these modalities are divided into
two general categories—anatomical or structural studies and functional modalities.
Each modality, anatomic and functional, has made its own unique contribution to
our understanding of the neuropathological changes associated with CO poison-
ing. For purposes of this discussion, functional brain imaging will refer to those
imaging modalities that assess the level of regional differences in the metabolic
activity of brain tissue. Current prominent examples of functional brain imaging
include Single Photon Emission Computed Tomography (SPECT), Positron Emission
Tomography (PET), Functional Magnetic Resonance Imaging (fMRI), and Magnetic
Resonance (MR) Spectroscopy. The primary anatomical modalities include Compu-
terized Tomography (CT) and Magnetic Resonance Imaging (MRI) with its various
perturbations—Diffusion Weighted Imaging (DWI) and Fluid Attenuation Inversion
Recovery (FLAIR). A summary of the recent functional and anatomical imaging
findings in CO poisoning will be presented.
21.3 ANATOMICAL IMAGING FINDINGS

21.3.1 COMPUTERIZED TOMOGRAPHY (CT)
The most common finding in acute CO poisoning on CT imaging is symmetrical,
diffuse frontal lobe white matter damage (low density findings). In two relatively
large series, one by Muira (n = 60) and one by Choi (n = 129), the sensitiv-
ity of CT imaging in acute CO poisoning in detecting these changes was shown to
be 38.5% and 48.0%, respectively.8,9 The second most common finding was that
of low density areas seen in the globus pallidus, occurring in 30% of patients in
8417: “8417_c021” — 2007/9/19 — 13:23 — page 458 — #2

the Muira study8 and in 25.6% of the patients in the Choi study.9 It was sugges-
ted that the low density lesions in the cerebral white matter and globus pallidus
were related to necrosis. Although patients with white matter lesions seemed to
have a worse prognosis than those with lesions in the globus pallidus, no consistent
or reliable prediction of outcome could be established with this imaging modal-
ity because of its relatively low sensitivity, particularly for cortical lesions. These
suggested neuropathological changes are similar in location and pathology to those
found in the pathology studies of Lapresle and Fardeau who performed autopsies on
22 patients who died of acute CO poisoning.10 Decreased CT sensitivity was con-
firmed in a 2003 study by Wu et al.,11 using 99 m Tc -ECD SPECT in ten patients
with acute CO poisoning. All ten patients had negative CT scans. However, brain
SPECT imaging showed areas of abnormal hypoperfusion in the basal ganglia of five
patients and in the cortical areas of seven patients. The ongoing role and utility of
the use of standard CT imaging in the evaluation of CO poisoning continues to be
an issue.
21.3.2 MAGNETIC RESONANCE IMAGING (MRI)

Although anatomical MRI findings have a slightly superior sensitivity relative to CT,
the findings are essentially the same as CT with the exception that, in the 1992 study
by Chang,7 additional abnormalities were seen in the subcortical gray matter of the
thalamus, putamen, and caudate nucleus. In addition, 9 of 15 patients demonstrated
abnormal findings in the globus pallidus bilaterally. Abnormal cortical findings were
minimal. Subsequent studies by Gale and Bigler, et al. in 1999,12 using a 1.5 Tesla
magnet found cerebral abnormalities in only 38% of 21 patients by MRI who had
sustained moderate to severe acute CO poisoning. The majority of the abnormalities
detected were subcortical. By contrast, they identified significant abnormalities in
67% of these same patients by using either SPECT or Quantitative Magnetic Res-
onance Imaging (QMRI). The low sensitivity of MRI for the detection of cortical
lesions in CO poisoning somewhat limits it’s utility as either a diagnostic aide or
in guiding interventional strategies. However, in a study by Murata et al. in 199513
using serial MRI with FLAIR pulse imaging to follow a single patient with acute
CO poisoning, abnormal findings in cerebral white matter were noted which sug-
gested progressive demyelination in spite of normal SPECT and neuropsychological
testing. This anecdotal observation seemed to suggest a possible prognostic role for
MRI with FLAIR for detecting early neuropathological changes which might require
more aggressive interventional strategies. However, in 2005 Durak et al.14 studied
16 patients between 1 and 10 years after their acute episode of CO poisoning using
MRI with FLAIR enhancement. All patients initially presented in an unconscious
state in the acute setting. At the time of follow-up, eight patients were asymptomatic
and eight continued to experience chronic neuropsychiatric sequelae from their acute
episode. MRI with FLAIR detected white matter abnormalities in all 16 patients, with
lesions noted predominantly in the centrum semiovale area. It thus appears that early
evaluation of acute CO poisoning using MRI with FLAIR enhancement carries little
prognostic value.
8417: “8417_c021” — 2007/9/19 — 13:23 — page 459 — #3

21.4 FUNCTIONAL IMAGING FINDINGS

21.4.1 POSITRON EMISSION TOMOGRAPHY (PET)
PET gets its name from positron emitting metabolic substrate analogs, typically
[18 F]—Fluorodeoxyglucose (FDG) or [15 O]H2 O. These isotopes are typically short-
acting and are generated in cyclotrons. With their capability to directly measure
either glucose metabolism or oxygen consumption, regional differences in cerebral
metabolic activity can be expressed quantitatively in units of mL/min/100 g brain
tissue. Traditionally, PET imaging has been considered superior to SPECT in that it
has been shown to achieve intrinsic spatial resolutions of 4–5 mm versus 6–10 mm
for SPECT. In addition, PET’s ability to quantify regional differences in cerebral
metabolism directly has been considered a more accurate measurement of cerebral
function in comparison to SPECT’s relative measurements of regional cerebral blood
flow (rCBF). However, because of PET’s increased cost, decreased availability, and
the need for cyclotron-generated pharmaceuticals, it is a less practical neuroimaging
modality. In addition, with the advent of cheaper, more readily accessible SPECT
cameras which can now achieve intrinsic spatial resolutions of 2.3 mm and less
(NeuroQuad, NC Systems), PET’s advantage of superior spatial resolution no longer
exists. Furthermore, it has been shown that a very close relationship exists between
oxygen consumption, glucose metabolism, and CBF.
As seen in Figure 21.1, a very tight correlation between CBF and metabolism
is seen when inhaled Xenon 133, which measures CBF quantitatively, is compared
with 99mTc-HMPAO, the most common radiopharmaceutical used in SPECT studies.
Therefore, the differences between these modalities now become somewhat academic.
In a 1993 PET study, DeReuck et al.15 demonstrated a global decrease in cereb-
ral metabolic activity, localized primarily in the frontal and temporal cortices. A
PET study by Yoshii et al.16 of a patient two months following acute CO poisoning
demonstrated findings primarily in the basal ganglia, but with some involvement of
the caudate and amputamen. Pinkston et al.17 studied two patients, man and wife,
3 years following their acute exposure. Using statistical parametric mapping to a
derived rCBF (ml/min/100g)
120 120
ECD HMPAO
100 100
80 80
60 60
40 40
m = 0.85 m = 0.83
b=6 b = 11
20 r = 0.86 20 r = 0.93
993TC
0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120
133Xe rCBF (ml/min/100g) 133Xe rCBF (ml/min/100g)
FIGURE 21.1 Comparison of the “derived” regional cerebral blood flow (rCBF) for
99m TcECD (left) or 99m Tc HMPAO (right) scanning techniques relative to rCBF (With
permission of Ismael Mena).
8417: “8417_c021” — 2007/9/19 — 13:23 — page 460 — #4

control group they demonstrated decreased metabolic activity in the bilateral orbit
frontal, dorsolateral prefrontal, and temporal cortices.
Subsequently, Tengvar et al.18 evaluated a patient five months following acute
CO poisoning with PET and noted decreased metabolic activity in the frontal lobes,
anterior singular gyri, and sub-cortical white matter. In a study of three patients with
acute CO poisoning, Ellis-Hon et al.19 noted significant decreases in basal ganglia
metabolism in two of three patients using PET imaging.
As discussed earlier, it is apparent from the review of the four prior articles
which evaluated a total of seven patients only, that PET is not that widely used
as a neuroimaging modality for the evaluation of CO poisoning. It remains to be
determined whether or not this modality will continue to be utilized in the ongoing
neuroimaging evaluations of CO poisoning patients.
21.4.2 FUNCTIONAL MAGNETIC RESONANCE IMAGING (FMRI)

FMRI, when employed with blood oxygen level determination (BOLD) enhancement,
is able to quantitatively assess regional differences in cerebral metabolic activity.
Essentially, by measuring the oxygen extraction between the arterial and venous
circulation in various brain regions, with the increased deoxyhemoglobin levels in
the venous circulation being detected on T2 images, quantitative measurements of
oxygen consumption can be expressed in terms of mL/min/100 g brain tissue. A
study by Kondo in 200620 on one patient poisoned with CO, revealed abnormal
findings in the deep cortex, hippocampus, and globus pallidus. Hantson21 found
symmetrical distribution of deep gray and basal ganglia abnormalities in a study of
five patients with delayed effects of CO poisoning. Again, the limited availability of
this technology, along with the lack of studies with adequate sample size limit fMRI’s
ability to appropriately study the neuropathological changes that occur following
recovery from acute CO poisoning.
21.4.3 MAGNETIC RESONANCE SPECTROSCOPY (MRS)

This relatively new functional neuroimaging modality uses the response of metabolic
protons to an electromagnetic field to evaluate areas of abnormal cerebral activity. It
has been shown that the metabolic byproducts of such neuropathological processes as
demyelination, neuronal degeneration or ongoing anaerobic glycolysis can result in
regional differences in NAA (N-acetyl-aspartate). MRS detected decreases in NAA
are associated with neuronal loss or degeneration. Increased choline concentration
has been associated with active demyelination. In 1995, Murata et al.13 performed
serial MRS, MRI, and SPECT on a patient after the onset of symptoms associated
with a delayed CO encephalopathic process. Initially, increased choline was detected
which suggested axonal demyelination. The patient was followed with serial MR
spectroscopy scans which later revealed increased levels of lactate and NAA which
suggested neuronal death. These findings were detected before changes were noted
on either MRI or SPECT.
In 1998, Sakamoto et al.22 studied three patients with the chronic enceph-
alopathic form of CO poisoning using MRI, electroencephalogram (EEG),
8417: “8417_c021” — 2007/9/19 — 13:23 — page 461 — #5

N-isopropyl-p-[1231]iodoarmphetamine (IMP) SPECT, and HMRS. A white matter

lesion detected by MRI persisted long after the improvement of the patient’s clinical
symptoms. SPECT findings did not always correlate with clinical symptomatology.
MRS studies in the frontal lobe area revealed increases in choline-containing com-
pounds and reductions of NAA in all cases. There was close correlation between
clinical symptomatology and levels of lactate suggesting a possible role for MRS as
a predictor of clinical outcome in patients with the “interval form” of CO poisoning.
In 2005, Pach et al.23 studied ten patients following recovery from acute CO
poisoning. Neuropsychological testing and 99m Tc –HMPAO-SPECT was performed
at 6 months following the acute injury and MRS was performed at 8 months following
the acute episode. SPECT revealed significant abnormalities in the frontal cortex,
basal ganglia and parietal cortex. All ten patients evaluated by MRS at 8 months
post acute CO exposure exhibited elevations in either mobile lipid and/or lactate
concentration in the frontal lobes and basal ganglia.
21.4.4 SPECT FINDINGS IN CARBON MONOXIDE POISONING

A growing body of scientific information has evolved in the area of the SPECT
evaluation of acute CO poisoning. In a study of 12 patients with acute CO poisoning
in 1994, Denays et al.24 demonstrated abnormal CBF in the temporal, partial, and
occipital lobes in eight of them. In a study of ten acutely poisoned CO patients in 1998,
Kao and colleagues25 identified abnormal cortical findings in seven out of ten patients
and abnormal basal ganglia findings in six out of ten patients. In 2003, Wu11 found
abnormalities in the cortex of seven acutely poisoned patients and abnormalities in
the basal ganglia of five out of ten patients. See the chapter on scanning techniques
by Choi in Penney, 2000.26
In a 2004 study of 20 acutely poisoned CO patients by Pach et al.,27 brain SPECT
imaging with 99 mTc-HMPAO revealed abnormalities in 17 of the 20 patients (85%).
Predominant findings in this study included decreases in perfusion in the frontal and
parietal cortices and basal ganglia. Their conclusion was that the use of 99 mTc-
HMPAO brain imaging seemed to be useful in the demonstration of early central
nervous system (CNS) dysfunction. A follow-up study by the same investigators of
ten patients revealed the same findings. Another study in 2004 of five patients with
the chronic effects of CO poisoning by Kon Chu et al.28 showed decreased bilateral
white matter, globus pallidus, and frontal cortex. A summary of the SPECT findings
noted above is given in Table 21.1.
In 1995, Choi29 studied 13 patients with delayed neurological sequelae of CO
poisoning and found a diffuse, patchy, decreased pattern which was more promin-
ent in the frontal lobes. Seven of these patients were seen in follow-up six months
later at which time their SPECT scans showed increased frontal blood flow in six
out of seven patients which was also associated with improved neuropsychological
performance. In 1999, Gale and Bigler et al.12 evaluated 21 patients with delayed
sequelae of CO poisoning with the multiple modalities of MRI, QMRI, SPECT, and
comprehensive neuropsychological testing. In their study, 14/21 patients were noted
to have abnormalities on both SPECT and QMRI imaging, but only 38% of patients
were noted to have abnormal findings on routine MRI imaging. In the patients with
8417: “8417_c021” — 2007/9/19 — 13:23 — page 462 — #6

TABLE 21.1
SPECT Areas of Decreased Perfusion Seen in Studies of
Year Author Brain Area Perfusion
1994 Denays24 Temporal lobe Decreased
Parietal lobe Decreased
Occipital lobe Decreased
1995 Choi26 Diffuse patchy cortical Decreased
Bilateral frontal Decreased
1998 Kao25 Cortex Decreased
Basal ganglia Decreased
1999 Gale12 Frontal lobes Decreased
Parietal lobes Decreased
2002 Watanabe30 Bilateral frontal lobes Decreased
Bilateral insula Decreased
Right temporal lobe Decreased
2003 Wu11 Cortex Decreased
2004 Pach23 Parietal cortex Decreased
Frontal lobes Decreased
2004 Chu28 Frontal cortex Decreased
Globus pallidus Decreased
abnormal SPECT findings, the perfusion abnormalities were noted in the frontal lobes
(71%), parietal lobes (57%), and temporal lobes (36%). The QMRI findings demon-
strated decreased hippocampal size, increased ventricle to brain ratio (VBR) and
evidence of cortical atrophy.
In 2002, Watanabe et al.30 studied eight patients with delayed neurosequelae of
acute CO poisoning and ten patients with no neuropsychiatric sequelae. Using statist-
ical parametric mapping and comparing them to a control group of 44 normal patients,
patients with delayed neuropsychiatric sequelae had significantly reduced perfusion
in the bilateral frontal lobes, bilateral insula, and right temporal lobe. Interestingly,
patients with no neuropsychiatric sequelae had significantly reduced bilateral frontal
perfusion as well, worse on the left, compared to controls. This study suggests that
relative preservation of left frontal lobe function following acute CO poisoning may
have some prognostic value and that right frontal lobe dysfunction may result in more
significant neurosequelae. This observation deserves further investigation.
In the 2004 study by Chu et al.,28 five patients who manifested delayed hypoxic
encephalopathy following acute CO poisoning showed the typical T2 -weighted MRI
findings of abnormalities in bilateral periventricular white matter and white mat-
ter in the frontal and parieto-occipital areas. In addition, symmetrical abnormalities
were found in the globus pallidus. Simultaneous brain SPECT imaging using 99 mTc-
HMPAO revealed decreased profusion in the bilateral frontal cortex, globus pallidus,
and bilateral white matter.
8417: “8417_c021” — 2007/9/19 — 13:23 — page 463 — #7

21.5 NEUROIMAGING FINDINGS IN CHRONIC,

LOWER-LEVEL CO POISONING
Beck’s 1936 axiom is that “No noxious gas so potent when inhaled in atmospheric
dilutions of 1% or even less as to cause almost instantaneous death, can be incapable
of producing symptoms if inhaled at a lesser concentration for a longer period of
time”.31
Population studies have shown that even small increases in the ambient CO con-
centration can lead to significant increases in mortality,32 low birth weight,33 and
psychiatric hospitalizations.34 In a 1998 study by Bayer,35 it was reported that chronic
COHb levels between 0.4% and 5.8% were associated with persistent neurological
symptoms (PNS). Furthermore, in a 2006 study of chronic CO exposure Tellez and
colleagues36 described lack of memory, attention-concentration, and Parkinson type
movements as the main neuropsychological and neurological symptoms occurring
most frequently. If one assumes that the basic neuropathologic damage of PNS is the
same in chronic, lower-level, “occult” CO poisoning as it is in the “interval form”
of CO poisoning, it might be anticipated that the neuroimaging findings in these two
entities would also be similar.
A review of the literature reveals very few neuroimaging studies performed for
this entity. A single case of chronic CO exposure in a 2 12 month old infant in 1990
revealed bilateral basal ganglia hypodensities on MRI.37
21.6 HIGH RESOLUTION SPECT FINDINGS IN TEN

CASES OF DELAYED CARBON
MONOXIDE-INDUCED ENCEPHALOPATHY
As much as it is useful to describe the imaging findings of the neuropathological
changes which occur in the acute CO poisoning setting from a public health stand-
point, the ability to identify those cases of chronic lower-level CO exposure would
seem relevant. As has been shown, chronic lower-level CO poisoning often presents
clinically as a confusing constellation of seemingly unrelated symptoms which have
been variously described as neurasthenia, chronic fatigue syndrome, fibromyalgia,
hypochondriasis, and other psychiatric conditions. As such, it often goes undiagnosed
and therefore is improperly treated. To the extent that brain function imaging might
reveal a characteristic perfusion pattern, it might prove a useful diagnostic tool when
the clinical presentation is unclear.
This author has had the opportunity to perform high resolution brain SPECT
imaging on a series of ten patients over the last 18 months, who were experiencing
the long-term effects of acute CO poisoning. A brain dedicated four headed gamma
camera (NeuroQuad SPECT, NC Systems Inc.) with an intrinsic spatial resolution
of two mm was used. In addition, Mirage software from Segami Corporation which
used a normative database allowed comparisons of individual patients on a Brodman
area by Brodman area basis for the cortical structures and the subcortical structures
of the thalamus, lentiform nucleus, and caudate. A representation of the type of data
obtained are seen in Table 21.2.
8417: “8417_c021” — 2007/9/19 — 13:23 — page 464 — #8

TABLE 21.2
Baseline Data versus Adult Normals
Standard
Volume Maximum Minimum Mean deviation
ROI # ROI Label # ELTS (ml) (sd) (sd) (sd) (sd)
0 Area 1,2,3-Both 5756 21 1.3 −5.0 −2.6 1.9
1 Area 1,2,3-Left 2878 10 1.3 −5.0 −3.1 1.8
2 Area 1,2,3-Right 2878 10 1.3 −5.0 −2.1 1.8
3 Area 4-Both 9944 37 3.0 −5.0 −2.6 1.7
4 Area 4-Left 4972 18 3.0 −5.0 −2.7 1.7
5 Area 4-Right 4972 18 3.0 −5.0 −2.5 1.7
6 Area 5-Both 2610 9 0.3 −5.0 −2.7 1.5
7 Area 5-Left 1305 4 0.3 −5.0 −2.6 1.8
8 Area 5-Right 1305 4 0.5 −5.0 −2.8 1.1
9 Area 6-Both 15932 59 1.8 −5.0 −2.4 1.6
10 Area 6-Left 7966 29 1.8 −5.0 −2.7 1.6
11 Area 6-Right 7966 29 1.4 −5.0 −2.2 1.6
12 Area 7-Both 15052 56 1.9 −5.0 −1.3 1.7
13 Area 7-Left 7526 28 1.9 −5.0 −1.3 1.7
14 Area 7-Right 7526 28 1.7 −5.0 −1.1 1.6
15 Area 8-Both 9226 34 2.3 −5.0 −2.2 2.1
16 Area 8-Left 4613 17 2.2 −5.0 −2.3 2.1
17 Area 8-Right 4613 17 2.3 −5.0 −2.1 2.1
18 Area 9-Both 8432 31 3.6 −5.0 −2.4 2.3
19 Area 9-Left 4216 15 3.6 −5.0 −2.4 2.2
20 Area 9-Right 4216 15 3.3 −5.0 −2.3 2.3
21 Area 10-Both 11378 42 5.0 −5.0 −1.7 2.4
22 Area 10-Left 5689 21 5.0 −5.0 −1.7 2.3
22 Area 10-Right 5689 21 5.0 −5.0 −1.7 2.5
Four adults and six children underwent high-resolution brain SPECT imaging
with 99mTc-HMPAO. The average age of the adults was 34.8 years (range 31.3–44.6
years) and the average age of the children was 10.3 years (range 7.4–12.8 years). In
nine of the patients the average time from acute exposure to imaging was 2.7 years.
These nine individuals, six children and three adults, were all apparently exposed to
the same level of CO and all were found in an unconscious state. One adult was seen
4.1 years following acute exposure in which unconsciousness was also a factor in
the acute setting. All patients were experiencing various delayed neuropsychological
sequelae from their acute poisoning and most had been found to have significant
abnormalities on neuropsychological testing which primarily involved problems with
executive function and memory, as well as such affective disturbances as depression
and anxiety. Tables 21.3 and 21.4 represent a summary of their SPECT findings.
The following data (Table 21.5) are a summary of the number of statistically
abnormal findings in the Brodman areas of the nine patients experiencing the same
level of acute CO poisoning.
8417: “8417_c021” — 2007/9/19 — 13:23 — page 465 — #9

466
TABLE 21.3
Summary of SPECT Findings in 10 individuals designated as A–J
Patient Broadman Areas (S.D.) Cortical Findings Sub-Cortical Findings
(Area), (Left value), (Right value)
A 17, L—1.9, R—2.0 Bilateral frontal L > R R Globus pallidus
18, L—1.9 Bilateral temporal L > R
Bilateral occipital L > R
Patchy cortical
B 24, L—1.8, R—2.6 Bilateral medial temporal R Globus pallidus
32, R—2.3 Patchy cortical Bilateral posterior thalamic R > L
C 21, L—1.7 Bilateral frontal Left caudate body
23, L—1.9, R—2.1 Bilateral temporal L > R
24, L—2.6, R—3.0 Left medial temporal
28, L—2.3, R—2.3 Patchy cortical
32, L—2.5, R—2.3
D 1, L—2.5 21, L—2.2, R—2.4 Bilateral frontal Bilateral lentiform
2, L—2.5 24, L—3.5, R—3.2 Bilateral temporal Right posterior thalamic
3, L—2.5 25, L—4.5, R—4.1 Bilateral parietal Bilateral caudate
4, L—2.6, R—2.6 28, L—2.7, R—2.4 Patchy cortical
5, L—2.0, R—2.3 32, R—2.1
6, L—2.2 38, L—2.5, R—2.6
11, L—4.6, R—4.5 47, L—2.9, R—3.1
E 25, L—2.1 Bilateral frontal Left globus pallidus
28, L—2.14, R—2.1 Bilateral occipital Right posterior thalamus
8417: “8417_c021” — 2007/9/19 — 13:23 — page 466 — #10

31, R—1.9 Patchy cortical Left caudate
F 17, L—2.2, R—1.9 Bilateral orbital frontal Bilateral lentiform
21, L—1.9 Bilateral temporal Left globus pallidus
24, R—2.0 Bilateral occipital L > R
25, R—2.1
28, L—2.4, R—2.1
38, L—2.5, R—2.0
G 20, L—2.3 36, L—2.6 Bilateral Frontal Bilateral globus pallidus
21, L—2.4 38, R—2.3 Bilateral Temporal L>R Bilateral posterior thalamus
24, L—3.8, R—3.2 44, L—2.9 R—2.3 Bilateral Parietal Left caudate
25, L—2.6, R—2.6 45, L—2.1 R—2.5 Left Occipital
28, L—3.4, R—2.4 Patchy Cortical
32, L—3.2, R—2.7
H 20, L—2.1 R—2.0 32, L—2.8 Patchy cortical —
21, L—2.1 R—0 36, L—2.7, Bilateral frontal
R—2.9
24, L—3.9 R—2.4 38, L—2.4, Bilateral temporal
R—2.9
25, L—3.1 R—3.5 47, R—2.1 Left occipital
28, L—3.6 R—3.7
31, L—2.0
I 17, L—1.9 Right cerebellar Bilateral putamen
24, R—1.9 Patchy cortical Right thalamus
Bilateral frontal Bilateral globus pallidus
Left medial temporal
Right occipital
J 5, L—2.1 Patchy cortical Bilateral globus pallidus
17, L—2.6, R—2.5 Bilateral frontal Right putamen
Recent Advances in Brain SPECT Imaging after Carbon Monoxide Poisoning
28, L—2.0 Bilateral cerebellar Left posterior thalamus

31, L—2.0, R—2.0 Bilateral posterior Left caudate
8417: “8417_c021” — 2007/9/19 — 13:23 — page 467 — #11

temporal
Bilateral occipital
467
TABLE 21.4
Summary of Regional Abnormal Areas in Nine Patients Using High
Resolution SPECT
Cortical
Diffuse, decreased patchiness Cerebellar Frontal Temporal Parietal Occipital
8/9 2/9 8/9 8/9 1/9 5/9
Subcortical
Globus pallidus Caudate Putamen Thalamus
7/9 4 4/9 5/9
TABLE 21.5
Abnormal Cortical Findings by Age after Carbon Monoxide
Exposure
Number of Abnormal
Patient Gender Age (years) Brodman Areas
C Male 31 11
H Female 31 17
D Male 33 15
Adult average 31.7 14.3
A Male 12 3
B Male 11 1
E Male 12 4
F Male 11 9
I Male 7 2
J Male 9 6
Child average 10.3 4.2
As can be seen, given the same level of acute CO poisoning, the general cortical
and subcortical pattern of involvement was similar in both children and adults, while
the number of statistically abnormal Brodman areas in the children were less than
those seen in the adults. This finding suggests that children sustain milder neuropatho-
logical effects from the same level of acute CO poisoning. Various explanations have
been offered for this previously observed phenomenon which are consistent with the
findings of other observers that children are more resistant to the acute effects of CO
poisoning.38 In addition, the sole adult female in the group experienced a smaller
number of statistically abnormal areas than the two adult males. This, too, is consist-
ent with the epidemiological evidence suggesting that women are more resistant than
men to the effects of acute CO poisoning.39,40 (see Figures 21.2–21.6).
8417: “8417_c021” — 2007/9/19 — 13:23 — page 468 — #12

Finding No. 1(cont.):

Toxic injury from carbon monoxide poisoning
"CO Normal
poisoned"
FIGURE 21.2 (See color insert following page 422). Example of diffuse neuronal injury
two years after acute carbon monoxide poisoning.
Mild Medium Severe
FIGURE 21.3 (See color insert following page 422). SPECT scans of three patients with
mild, medium and severe cognitive defects two years after acute carbon monoxide poisoning.
8417: “8417_c021” — 2007/9/19 — 13:23 — page 469 — #13

Adult male Adult female
FIGURE 21.4 (See color insert following page 422). SPECT scans of male and female
patients two years following acute carbon monoxide poisoning with identical carboxyhemo-
globin levels (34.5% vs. 34.9%)
LC IG TM SP
RL BAW BRW
FIGURE 21.5 (See color insert following page 422). Superior, transverse views of SPECT
perfusion findings in isolated lentiform nuclei of seven patients, two years following acute
carbon monoxide poisoning. Yellow areas in the color plates are areas of abnormally decreased
perfusion.
8417: “8417_c021” — 2007/9/19 — 13:23 — page 470 — #14

ROI label # Elts Volume Maximum Minimum Mean Standard

deviation
Caudate nucleus—Left 975 0.3 % 71.1 % 26.9 % 42.6 % 9.3 %
Caudate nucleus—Right 975 0.3 % 74.1 % 37.9 % 52.9 % 7.7 %
Left lateral view Posterior view Interactive view
FIGURE 21.6 (See color insert following page 422). Six SPECT isolation views of caudate
nuclei of a patient two years following acute carbon monoxide poisoning. All areas other than
red in the color plates represent areas of abnormally decreased perfusion.
21.7 DISCUSSION
CO poisoning, for the purposes of discussion, can be divided into three distinct
categories:
1. Acute CO poisoning. This may involve immediate deficits, that is, PNS.
2. Acute CO poisoning which displays delayed effects, that is, DNS (the
so-called “interval form” of CO poisoning).
3. Chronic, lower-level or “occult” CO poisoning, which can also result in
long-term health effects.
In acute exposures, multiple organ systems can be affected with life-threatening

results. Suicide by voluntary CO inhalation remains the number one cause of toxic
death in the United States.1 Severe acute exposures can result in the development
of cardiac arrhythmias, pulmonary edema, renal failure, and metabolic acidosis
which must be managed aggressively. High concentrations of CO can cause neuro-
pathologic changes that may manifest themselves immediately or at some interval
following the acute, possibly life-threatening episode. Unfortunately, it has been
shown that the degree of neuropathologic damage caused by CO does not always cor-
relate with longterm prognosis.41 Although other organ systems can be permanently
affected, it is the neurological, neuropsychological, and neuropsychiatric sequelae
of CO poisoning that accounts for the bulk of the chronic morbidity often seen in
this disorder.42 To the extent that the neuropathological changes that occur in the
acute setting are significant and sufficiently widespread, they create “pathological
8417: “8417_c021” — 2007/9/19 — 13:23 — page 471 — #15

footprints” that can be identified by various neuroimaging techniques. In the acute

setting, from a neuroimaging standpoint, it is important to establish, if possible, the
location and extent of the brain injury. This not only aids in the prediction of clinical
outcomes, but can assist in the determination of the need for ongoing interventional
measures.
Perhaps the greatest potential contribution that neuroimaging might make is the
ability to aid in the identification of cases of lower-level, chronic CO poisoning. As
has been pointed out, chronic CO poisoning has been associated with various vague
neuropsychiatric conditions such as neurasthenia, chronic fatigue syndrome, fibromy-
algia, and other nonspecific, yet debilitating illnesses which often go undiagnosed
and improperly treated.43,44
To the extent that neuroimaging might be capable of identifying certain character-
istic perfusion patterns associated with the chronic effects of CO poisoning, improved
diagnosis, and better treatment might be possible for these vague syndromes with their
protean manifestations. The focus of this chapter, therefore, is the review of current
neuroimaging findings in acute CO poisoning, delayed neurosequelae of acute CO
poisoning and chronic CO poisoning. In addition, an attempt has been made to assess
its current role in the evaluation of the various forms of this insidious illness.
As we have seen, CO poisoning, either acute and life threatening, with delayed
neuropsychological sequelae or of the lower-level, insidious chronic type, can cause
serious neurological damage to various areas of the brain. The damage, from frank
necrosis of gray matter to demyelination of white matter, to impaired mitochondrial,
and aerobic metabolic processes, results in impaired neuronal activity. This often
results in significant neuropsychological disability and morbidity and represents a
significant disease burden for our society. Evidence from the anatomic neuroimaging
studies, CT and MRI, suggest a regional predilection for CO to damage frontal lobe
white matter including corpus callosum and the centrum semiovale as well as subcor-
tical gray areas, particularly the globus pallidus. These changes are often seen earlier
in the course of CO poisoning relative to the changes seen with PET or SPECT. CO’s
affinity for the high iron content contained within the globus pallidus and substan-
tia nigra is the proposed mechanism whereby the globus pallidus and the substantia
nigra are selectively damaged. Subcortical damage has also been seen in the caudate,
putamen and thalamus.
Regarding CO’s affect on cortical brain matter, numerous studies have shown PET
and SPECT’s superiority to CT and MRI. Since PET and SPECT primarily measure
rCBF in the more metabolically active gray matter, white matter abnormalities are
only occasionally detected. The preponderance of the functional neuroimaging data
suggest a predilection of CO to damage the frontal and temporal lobes, often the more
medial aspects of the temporal lobes. However, almost every other cortical lobe has
been implicated as well including the parietal, occipital, and cerebellar lobes of the
brain. MR Spectroscopy may serve as an alarm modality because of its apparent ability
to detect active pathologic changes sooner after the onset of the delayed sequelae of
CO poisoning as compared to MRI, PET, or SPECT. Further studies are needed to
confirm this initial impression. The apparent earlier ability of MRS to detect the
onset of neuropsychological symptoms related to CO poisoning may allow for earlier
intervention with modalities such as hyperbaric oxygen (HBO). HBO therapy has
8417: “8417_c021” — 2007/9/19 — 13:23 — page 472 — #16

been shown to be effective in treating many aspects of brain injury, including CO

poisoning (see chapter 17 in this book).
The data from the various anatomic and functional neuroimaging modalities sug-
gest that a prototypical pattern of CO’s affects on the brain is evolving. This picture
appears to be one which demonstrates abnormalities of function in the globus pal-
lidus, bilateral frontal, and bilateral temporal cortices with a tendency toward medial
versus lateral temporal involvement. In addition, a patchy cortical pattern of decreased
activity as seen in various toxic-anoxic insults to the brain is also observed often but
is fairly nonspecific. The proper neuroimaging evaluation of these findings would
suggest that a combination of CT and MRI with SPECT or PET is required on the
basis of the relative sensitivities and intrinsic spatial resolution anatomical studies
for the globus pallidus and the superior sensitivity of functional studies for cortical
structures. However, newer generation, high resolution SPECT cameras with intrinsic
spatial resolutions of 2.0 mm (NeuroQuad SPECT, NC Systems) may prove useful as
a single imaging modality. When one considers this from a radiation exposure stand-
point, the ability to do a functional neuroimaging scan with 0.26 rems of radiation
exposure versus a CT involving 6.0 rems of radiation exposure, the choice seems
obvious. Although there is zero radiation exposure associated with MRI, the abil-
ity to avoid two separate procedures and the associated increased cost would also
seem preferable. The series of nine patients imaged with the brain-dedicated Neur-
oQuad SPECT system at Brain Matters, Inc. in Denver, CO confirms its ability to not
only assess cortical neuroactivity, but also functional activity in smaller subcortical
structures, particularly the globus pallidus.
From a public health perspective, the ability of neuroimaging studies to detect CO
poisoning of the lower-level chronic type would seem very important. As has been
discussed, this insidious malady with its many manifestations often goes undiagnosed
and inappropriately treated. Increased awareness and education, particularly in the
primary care setting will be important in assisting clinicians in detecting this occult
illness in many of their patients who may present with a myriad of seemingly unrelated
symptoms. In this setting, clinicians should consider a neuroimaging evaluation to
include either a combination of CT or MRI with standard resolution SPECT, or where
available, newer generations of high resolution, brain-dedicated SPECT cameras.
With the improved diagnostic capabilities of the newer neuroimaging modalities and
increased knowledge of CO’s evolving “fingerprint,” earlier intervention and better
outcomes after CO poisoning can be achieved.
21.8 CONCLUSIONS
1. Anatomical neuroimaging studies such as CT and MRI are capable of
detecting pathological changes in frontal lobe white matter and subcortical
gray matter, particularly the globus pallidus, following the development of
the neuropsychological sequelae of CO poisoning.
2. Functional neuroanatomical studies such as PET and SPECT appear more
sensitive to changes in the regional metabolic activity and rCBF in cortical
grey matter, particularly in the frontal and temporal lobes.
8417: “8417_c021” — 2007/9/19 — 13:23 — page 473 — #17

3. A prototypical, but not necessarily pathognomonic neuroimaging profile

for CO poisoning is evolving which suggests preferential involvement
of the globus pallidus subcortically and the frontal and temporal lobes
cortically.
4. MR Spectroscopy may serve as an early indicator for evaluating neuro-
pathological changes which develop following the onset of the neuropsy-
chological sequelae of CO poisoning.
5. Neuroimaging studies, either CT or MRI coupled with SPECT or PET, or
alternatively high resolution SPECT should be considered as part of the
evaluation process for patients presenting subacutely in the primary care
setting with various vague, ill-defined symptoms which are often confused
with chronic fatigue syndrome, fibromyalgia, flu, Lyme’s disease, multiple
sclerosis, psychiatric condition, and so forth.
References
1. Cobb, N., and Etzel, RA. Unintentional carbon monoxide-related deaths in the United
States, 1979 through 1988, J. Am. Med. Assoc., 266, 659, 1995.
2. Brown, S., and Piantadosi, C. In vivo binding of CO to cytochrome oxidase in rat
brain, J. Appl. Physiol., 68, 604,1990.
3. Goldbaum, L.R., et al. Mechanism of the toxic action of carbon monoxide, Ann. Clin.
Lab. Sci. 1976; 6: 372–376.
4. Auer, R.N., Benveniste H. Hypoxia and related conditions. In: Graham DI, Langtos
PL (eds): Greenfield’s Neuropathology, London, Arnold, 1997, pp. 275–276.
5. Thom, S.R. Dehydrogenase conversion to oxidase and lipid peroxidation in brain after
carbon monoxide poisoning. J. Appl. Physiol., 73: 1584–1589, 1992.
6. Thom, S.R. Carbon monoxide-mediated brain lipid peroxidation in the rat. J. Appl.
Physiol., 68, 997, 1990.
7. Chang, K.H. Delayed encephalopathy after acute carbon monoxide intoxication: MR
imaging features and distribution of cerebral white matter lesions, Radiology. 1992;
184: 117–122.
8. Miura, T., Mitomo, M., Kawai, R., and Harada, K. CT of the brain in acute carbon
monoxide intoxication: characteristic features and prognosis, AJNR, 6, 739, 1985.
9. Choi, I.S., Kim, S.K., Choi, Y.D., Lee, S.S., and Lee, M.S. Evaluation of outcome
after acute carbon monoxide poisoning by brain CT, J. Kor. Neurol. Assoc., 8, 78,
1993.
10. Lapresle, J., and Fardeau, M. The central nervous system and carbon monoxide
poisoning. II. Anatomical study of brain lesions following intoxication with carbon
monixide (22 cases), Prog. Brain Res. 1967; 24: 31–74.
11. Wu, C.I., et al. Usefulness of 99mTc ethyl cysteinate dimer brain SPECT to detect
abnormal regional cerebral blood flow in patients with acute carbon monoxide
poisoning. Nucl. Med. Commun. 2003; 24: 1185–1188.
12. Gale, S.D., Bigler, E.D., et al. MRI, quantitative MRI, SPECT, and neuropsy-
chological findings following carbon monoxide poisoning, Brain Inj. 1999; 13:
229–243.
13. Murata, T., et al. Serial proton magnetic resonance spectroscopy in a patient with the
interval form of carbon monoxide poisoning, J. Neurol. Neurosurg. Psychiatry. 1995;
58: 100–103.
8417: “8417_c021” — 2007/9/19 — 13:23 — page 474 — #18

14. Durak, A.C., et al. Magnetic resonance imaging findings in chronic carbon monoxide
intoxication, Acta. Radiol. 2005; 46: 322–327.
15. De Reuck, J., Decoo, D., Lemanhieu, I., Stijckmans, K., Boon, P., Man Maele, G.,
Buylaert, W., Leys, D., and Petit. H. A positron emission tomography study of patients
with acute carbon monoxide poisoning treated by hyperbaric oxygen, J. Neurol., 240,
430, 1993.
16. Yoshii, R., Kozuma, R., Takahashi, W., Haida, M., Takagi, S., and Shinohara, Y.
Magnetic resonance imagin and 11c-n-methylspiperone/positron emission tomo-
graphy studies in a patient with the interval form of carbon monoxide poisoning.
Neurol. Sci. 160, 1, 1998.
17. Pinkston, J.B., Wu J.C., and Gouvier W.D. Varney NR. Quantitative PET scan find-
ings in carbon monoxide poisoning: deficits seen in a matched pair. Arch. Clin.
Neuropsychol. 2000; 15: 545–553.
18. Tengvar, C., Johansson, B., and Sorensen, J. Frontal lobe and cingulate cortical
metabolic dysfunction in acquired akinetic mutism: a PET study of the interval form
of carbon monoxide poisoning. Brain Inj. 2004; 18: 615–625.
19. Ellis-Hon, K.L., Yeung W.L., Ho C.H., Leung W.K., Li AM., Chiu-Wing Chu W.,
and Chan Y.L. Neurologic and radiologic manifestations of three girls surviving acute
carbon monoxide poisoning. J Child Neurol. 2006; 21: 737–741.
20. Kondo, A., Saito, Y., Seki, A., Sugiura, C., Maegaki, Y., Nakyama, Y., Yagi, K., and
Ohno, K. Delayed neuropsychiatric syndrome in a child following carbon monoxide
poisoning. Brain Dev. 2007; 29(3): 174–177.
21. Hanston, P., and Duprez, T. The value of morphological neuroimaging after acute
exposure to toxic substances. Toxicol. Rev. 25, 2, 1993.
22. Sakamoto, K., et al. Clinical studies on three cases of the interval form of carbon
monoxide poisoning: serial proton magnetic resonance spectroscopy as a prognostic
predictor, Psychiatry Res. 1998; 83: 179–192.
23. Pach, D., et al. Evaluation of regional cerebral perfusion using 99mTc-HmPAO
single photon emission computed tomography (SPECT) in carbon monoxide acutely
poisoned patients, Przegl Lek. 62, 6, 2005.
24. Denays, R., et al. Electroencephalographic mapping and 99mTc HMPAO single-
photon emission computed tomography in carbon monoxide poisoning, Ann. Emerg.
Med. 1994; 24: 947–952.
25. Kao, C.H., et al. HMPAO brain SPECT in acute carbon monoxide poisoning, J. Nucl.
Med. 1998; 39: 769–772.
26. Choi, I.S.S. Use of scanning techniques in the diagnosis of damage from carbon
monoxide. In: Carbon Monoxide Toxicity, D.G. Penney, Ed., CRC Press, 2000, Chapt.
16, pp. 363–380.
27. Pach, D., et al. Evaluation of regional cerebral perfusion using 99mTc-HmPAO
single photon emission computed tomography (SPECT) in carbon monoxide acutely
poisoned patients, Przegl Lek. 61, 4, 2004.
28. Chu, K., et al. Diffusion-weighted MRI and 99mTc-HMPAO SPECT in delayed
relapsing type of carbon monoxide poisoning: evidence of delayed cytotoxic edema,
Eur. Neurol. 2004; 51: 98–103. Epub Jan. 28, 2004.
29. Choi, IS., et al. Evaluation of outcome of delayed neurologic sequelae after carbon
monoxide poisoning by technetium-99m hexamethylpropylene amine oxime brain
single photon emission computed tomography, Eur. Neurol. 1995; 35: 137–142.
30. Watanabe, N., et al. Statistical parametric mapping in brain single photon computed
emission tomography after carbon monoxide intoxication, Nucl. Med. Commun. 2002;
23: 355–366.
8417: “8417_c021” — 2007/9/19 — 13:23 — page 475 — #19

31. Beck, H.G. Carbon monoxide asphyxiation: a neglected clinical problem, J. Am. Med.
Assoc., 107, 1025–1029, 1936.
32. Hexter, A.D., and Goldsmith, J.R. Carbon monoxide: association of community air
pollution with mortality, Science, 172, 265–276, 1971.
33. Ritz, B., and Yu, F. The effect of ambient carbon monoxide on low birth weight
among children born in southern California between 1989 and 1993, Environ. Health
Perspect., 107, 17–25, 1999.
34. Strathilevitz, M., Strahilevitz, A., and Miller, J.E. Air pollutants and the admission
rate of psychiatric patients, Am. J. Psychiatr., 136, 205–207, 1979.
35. Bayer, M.J., Orlando, J., McCormick, M.A., Weiner, A., and Deckel, A.W. Persistent
neurological sequelae following chronic exposure to carbon monoxide, in Carbon
Monoxide: The Unnoticed Poison of the 21st Century, Satellite Meeting, IUTOX
VIIIth International Congress of Toxicology, Dijon, France, July 3–4, 1998.
36. Tellez, J., Rodriguez, A., and Fajardo A. Carbon monoxide contamination; an
environmental health problem, Rev. Salud Publica, 8, 1, 108–117, 2006.
37. Piatt, J.P., Kaplan, A.M. Bond, G.R., and Berg, R.A. Occult carbon monoxide
poisoning in an infant, Pediatr. Emerg. Care, 6, 21, 1990.
38. White, S. Pediatric carbon monoxide poisoning. In: Carbon Monoxide Toxicity,
Penney, D.G., Ed., 2000, Chapt. 21, 463–491.
39. Howe, S., Hopkins, R.O., and Weaver, L.K. A retrospective demographic analysis of
carbon monoxide poisoned patients [abstract]. Undersea Hyperb. Med., 23 (Suppl.),
84, 1996.
40. Dodson, W.W., Santamaria, J.P., Etzel, R.A., Desautels, D.A., and Bushnell, J.D. Epi-
demiologic study of carbon monoxide poisoning cases receiving hyperbaric oxygen
treatment [abstract], Undersea Hyperb. Med., 24 (Suppl.), 38, 1997.
41. Vieregga, P., Klostermann, W., Blumm, R.G., and Borgis, K.J. Carbon monoxide pois-
oning: clinical, neurophysiological, and brain imaging observations in acute disease
and follow up, J. Neurol. , 236, 478, 1989.
42. Choi, I.S. Peripheral neuropathy following acute carbon monoxide poisoning. Muscle
Nerve, 9, 965, 1986.
43. Roy, B. Crawford, R. Pitfalls in diagnosis and management of carbon monoxide
poisoning, J. Accid. Emerg. Med. 1996; 13: 62–63.
44. Grace, T.W., and Platt, F.W. Subacute carbon monoxide poisoning; another great
imitator, J. Am. Med. Assoc., 246, 1698–1700, 1981.
8417: “8417_c021” — 2007/9/19 — 13:23 — page 476 — #20

22 Neurocognitive and
Affective Sequelae of
Carbon Monoxide
Poisoning
Ramona O. Hopkins
CONTENTS
22.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477

22.2 Mechanisms of Brain Injury. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
22.3 Cognitive Sequelae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
22.3.1 Persistent and Delayed Neuropsychological Sequelae. . . . . . . . . . . . 481
22.3.2 Cognitive Impairments in Lower Level Carbon Monoxide
Poisoning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
22.3.3 Markers of Carbon Monoxide Poisoning Severity and Outcome 484
22.3.4 Effect of HBO on Cognitive Impairments. . . . . . . . . . . . . . . . . . . . . . . . . 484
22.3.5 Relationship between Cognitive Sequelae and Neuroimaging
Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 485
22.3.6 Functional Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 486
22.4 Affective Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487
22.4.1 Depression and Anxiety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487
22.4.2 Obsessive Compulsive Disorder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488
22.4.3 Kluver–Bucy Syndrome. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488
22.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489
22.1 INTRODUCTION
Carbon monoxide (CO) is a colorless, odorless, tasteless, and nonirritating gas
produced as a by-product of combustion of carbon-containing compounds. CO is
the leading cause of poisoning injury and death worldwide,1 and the most common
cause of accidental and intentional poisoning in the United States. CO results in
approximately 40,000 emergency department visits2 and 470 unintentional deaths
per year in the United States.3 The brain and heart are particularly vulnerable
477
8417: “8417_c022” — 2007/9/11 — 12:14 — page 477 — #1

to the pathological effects of CO.1 Neurologic morbidity following CO poisoning

includes neurologic sequelae,4 abnormalities on brain imaging,4 affective changes,5
and cognitive impairments.4
22.2 MECHANISMS OF BRAIN INJURY

Exposure to CO may damage multiple organ systems with high oxygen utiliza-
tion, especially the cardiovascular and central nervous systems. For more detail
on this topic, see the Raub et al. review published in 2000.1 The mechanisms of
CO-induced neural damage are complex and multifactorial. Although the neuropath-
ological injury associated with CO poisoning are related to CO-induced hypoxia
(CO binds to hemoglobin),6 other biochemical mechanisms appear to be involved.
Mechanisms of brain injury following CO poisoning include: binding to intracellu-
lar proteins cytochrome c-oxidase, myoglobin, or cytochrome c (P-450) reductase,
leading to mitochondrial dysfunction and disruption of cellular metabolism.7 Other
mechanisms include hypoxia,8 release of excitatory amino acids (e.g., glutamate)
resulting in calcium influx and cell damage or death,9 interference with intracellu-
lar enzyme function,10 lipid peroxidation leading to oxidative injury,11 deposition of
peroxynitrate and subsequent blood vessel endothelium damage,12 oxidative stress
from intracellular iron deposition,13 and apoptosis.14 CO-mediated oxidative stress
alters myelin basic protein, resulting in an immune response and inflammation in the
central nervous system.15 Consistent with the multifactorial neuropathologic mechan-
isms, the resultant neuropathology, cognitive, affective, and neurobehavioral sequelae
are heterogeneous.
22.3 COGNITIVE SEQUELAE

Cognitive impairments frequently occur following CO poisoning in healthy
individuals.16 It is estimated that between 15% and 49% of individuals with dia-
gnosed CO poisoning will develop cognitive sequelae.17 Some CO-poisoned people
will develop persistent neuropsychological sequelae (i.e., initial impairments that
persist over time);16 however, others who have intact initial cognitive performance
may present with delayed neurologic or cognitive sequelae (i.e., normal initial cog-
nitive scores with impairment developing from 2 to 40 days post-CO poisoning; see
discussion below).
CO-related cognitive sequelae are heterogeneous regarding onset, severity, and
cognitive domain affected.4 A list of some CO-related cognitive impairments are
shown in Table 22.1. Common CO-related cognitive sequelae include impaired
memory,18 executive function,19 slow mental-processing speed, decreased intel-
lectual function,4 apraxia, aphasia, and agnosia.20 Cognitive sequelae lasting
1 month21,22 or more16,23 occurs in 25–50% of participants with loss of conscious-
ness (LOC) or carboxyhemoglobin (COHb) levels greater than 25%.22,24 Even people
with less severe CO poisoning may develop cognitive impairments.25 Recent stud-
ies utilizing comprehensive standardized neuropsychological tests find significant
cognitive impairments both immediately and several years after recovery from the
8417: “8417_c022” — 2007/9/11 — 12:14 — page 478 — #2

Neurocognitive and Affective Sequelae of Carbon Monoxide Poisoning 479
TABLE 22.1
Carbon Monoxide-Related Cognitive Impairments
Shown by Cognitive Domain
Cognitive Domain Impairments
Arithmetic Acalculia
Attention Distractibility
Divided attention
Preservative errors
Sustained attention
Executive function Decision making
Disorganization
Impulsivity
Planning
Working memory
Intelligence Verbal intelligence
Performance intelligence
Memory Anterograde memory
Delayed memory
Recall
Recognition memory
Retrograde memory
Short term or working memory
Motor Apraxia
Athetosis
Ballism
Bradykenesia
Chorea
Dyskenesia
Dystonia
Incoordination
Myoclonus
Parkinsonism
Rigidity
Tremor
Processing speed Mental-processing speed
Spatial Visuoconstruction
Visuoperception
Visuospatial
Verbal Aphasia
Dysarthria
Hypophonia
Mutism
Visual Achromotopsia
Apperceptive agnosia
Cortical blindness
Homonymous hemianopsia
Prosopagnosia
Scotomas
Visual form agnosia
8417: “8417_c022” — 2007/9/11 — 12:14 — page 479 — #3

initial CO-poisoning symptoms.4,26 The cognitive impairments range in severity from

mild, to moderate, and severe. While impairments in memory, attention, and execut-
ive function occur most frequently, a consistent pattern of neuropsychological deficits,
or a CO “syndrome,” has not been observed.
A recent review of the literature from 1995 to 2005 assessed cognitive impair-
ments following CO poisoning identified by standardized neuropsychological tests.27
Hopkins and Woon27 identified 18 group studies that included 979 CO-poisoned
patients and 16 case studies that included 35 CO-poisoned patients, for a total of
1014 patients. The mean age of the patients was 38.1 years and 40.2 years for the
group and case studies, respectively. While some studies included elderly individuals
(age >65 years), the majority of studies contained predominately young to middle
age adults. CO-poisoning severity was moderate to severe, with a mean COHb level
of 23% (normal ≤2.0%) for both the group and case studies. LOC occurred in 30.1%
(295/979) of patients in the group studies and 42.8% (15/35) of patients in the case
studies, with a range of 10.8–100% for the group studies.
All of the studies reviewed demonstrated the adverse effect of CO poisoning
on cognition. The cognitive impairments occurred in multiple cognitive domains
and were heterogeneous regarding onset, severity, and cognitive domain affected.
Of the group studies, impaired memory was the most frequently reported impair-
ment, followed by impaired attention, motor impairments, executive dysfunction,
slow mental-processing speed, and impaired visual spatial abilities. Similarly, in the
case studies, memory impairments occurred most frequently, followed by executive
dysfunction, impaired attention, motor impairments, visual spatial deficits, and slow
mental-processing speed (Figure 22.1).
These data indicate that memory is the most common cognitive domain affected
by CO poisoning.27 Memory impairments following CO poisoning can be mild,
moderate, or severe. For example, a 48-year-old male with LOC and a COHb of 9.1%
Processing speed Case studies

Group studies
Visual spatial
Cognitive impairments
Executive function
Motor
Attention
Memory
0 20 40 60 80 100
Percent of studies reporting
FIGURE 22.1 Percent of group and case studies reporting cognitive impairments following
carbon monoxide poisoning.
8417: “8417_c022” — 2007/9/11 — 12:14 — page 480 — #4

had severe memory impairments associated with hippocampal atrophy (structure in

the temporal lobe associated with the formation and recall of memories) 6 months post-
CO poisoning.28 Memory impairments following CO poisoning include impaired
short-term memory, anterograde memory, retrograde memory, recall, and recognition
memory.28−31 While memory impairments usually include the inability to learn or
remember new information, global amnesia (both anterograde and retrograde memory
impairments) are reported.31 There was the case of a patient with CO-induced amnesia
and bilateral hippocampal atrophy who recovered recognition memory, but remained
impaired for verbal and spatial memory over time.32 This suggests that some types
of memory may improve, whereas other memory impairments may not improve over
time.
Other CO-induced cognitive impairments included impaired attention, execut-
ive function, motor, visual spatial, and slowed mental-processing speed. Significant
deficits in visual tracking, visuomotor skills, visuospatial planning, and abstract
thinking occur following CO poisoning.25 Impaired executive function, attention
and concentration, visual–perceptual abilities, and information processing speed
are also reported.33 Although the common wisdom has been that most people who
survive initial CO poisoning will recover, recent studies suggest that the cognitive
impairments they incur may last years and sometimes become permanent. The identi-
fication of CO-related neurocognitive deficits may be instrumental for future studies
in determining if they are lessened through therapy, such as cognitive rehabilitation
or medications.
In addition to the more common cognitive impairments such as impaired memory
and executive function, CO poisoning can produce other impairments such as visual
agnosia.34 As an example, a 34-year-old female (patient DF) who sustained severe
CO poisoning developed visual form agnosia (e.g., inability to visually recognize
objects), owing to bilateral diffuse damage to the ventral portion of the lateral occipital
regions, including the ventral visual pathway.34 DF was unable to recognize objects,
especially line drawings of common objects. She could not discriminate between
vertical and horizontal line gratings or between simple geometric shapes.35 Her color
vision was intact and she was able to draw objects from memory, but could not
recognize the objects she drew. Despite the problem with object recognition, she had
no problem adjusting her finger-thumb grip to the width of objects.36 She was adept
at interacting manually with objects and she used the structural features of objects to
control visually guided grasping movements, showing the capacity to act upon visual
information that she was unable to report at a conscious level.36 Further, she was able
to negotiate obstacles when walking though a room.35 The ability to interact with
objects or avoid obstacles when walking was due to preservation of the dorsal visual
pathway. Thus, she had impaired ability to visually recognize objects but, preserved
visual processing for object orientation and location of objects in a room, even though
she denied “seeing” the objects.
22.3.1 PERSISTENT AND DELAYED NEUROPSYCHOLOGICAL

SEQUELAE
Neurological or cognitive sequelae can occur immediately and persist over time
(“persistent neurocognitive sequelae,” PNS), or the onset can be delayed in its
8417: “8417_c022” — 2007/9/11 — 12:14 — page 481 — #5

onset (“delayed neurocognitive sequelae” or the so-called interval form, DNS).16,22,23

Delayed neurologic or neurocognitive sequelae presentation includes a lucid period
of between 2 and 40 days after CO poisoning onset.37 A number of studies have
assessed CO-associated DNS,17,22,23,38 however few studies recognize or assess
PNS.16 PNS and DNS are common following acute CO poisoning. Delayed neuro-
psychological sequelae occur in 0.06–40% of CO poisonings.17,38 Symptoms of
DNS include mental deterioration, urinary and/or fecal incontinence, gait disturb-
ance and other neurologic problems such as cognitive impairments.39 There was
the case of a 50-year-old male, who sustained LOC owing to CO poisoning and
developed DNS 1 month later with behavioral changes, disorientation, impaired
memory, masked face, hypophonia, muscle rigidity, and bradykinesia manifested
by slow movements and gait disturbance.40 At 4 months all symptoms had resolved.
Other studies show that DNS persists over time.23 There are a paucity of data regard-
ing risk factors for DNS, however, COHb levels are not associated with the severity
of the symptoms.41 The etiology(s) of PNS and DNS are unknown, although it
has been hypothesized that DNS is due to extensive myelin and neuronal loss.42
Other theories include immuopathological damage due to activation of polymorpho-
nuclear leukocytes leading to demyelination and dopaminergic and serotonergic
disturbances.43
One method that appears to be sensitive to the underlying pathophysiology of
DNS is magnetic resonance spectroscopy (MRS). In proton MRS of the brain, signals
are present from N-acetylaspartate (located primarily in neurons, a marker for neur-
ons and axons), choline (principally phosphotidyl choline, a membrane constituent),
and creatine (used as an internal standard because its level is usually stable). For
comparison purposes, N-acetylaspartate and choline are expressed relative to creat-
ine. MRS appears to be sensitive to CO-poisoning related changes in white matter.44
Data show that N-acetylaspartate/creatine were below normal and choline/creatine
were elevated at the time of DNS symptom onset, yet both structural magnetic res-
onance (MR) and cerebral blood flow measures were normal.45 Increased severity
of the spectral changes at the onset of DNS was related to more profound clinical
symptoms. Clinical recovery correlated with normalization of the spectral changes.44
Thus, proton MRS may provide a much-needed marker of the development of DNS
following CO poisoning.
Weaver and colleagues46 assessed both PNS and DNS in a prospective outcome
study. Participants (N = 238) with acute CO poisoning were followed prospect-
ively. Persistent neuropsychological sequelae were defined as cognitive dysfunction
initially, which persisted at 2 weeks and 6 weeks after CO poisoning. Delayed neuro-
psychological sequelae are defined as a decline of at least one standard deviation
on a neuropsychological subtest score from a prior score, and meeting the definition
for cognitive sequelae at 6 weeks.46 Thirty-seven percent of participants had cognit-
ive sequelae at 6 weeks, of which 59% had PNS and 28% had DNS, a ratio of 2:1.
Hyperbaric oxygen (HBO) reduced the incidence of PNS but not DNS.46 One possible
explanation for the effect of HBO on PNS, but not on DNS is that HBO favorably
modulates mechanisms of early brain injury but does not influence mechanisms that
may be involved in the development of DNS.15
8417: “8417_c022” — 2007/9/11 — 12:14 — page 482 — #6

22.3.2 COGNITIVE IMPAIRMENTS IN LOWER LEVEL CARBON

MONOXIDE POISONING
Many clinicians believe that only patients who experience LOC37 or have moderate to
severe CO poisoning (COHb > 25%)47 will develop sequelae. A survey of hyperbaric
medical centers in North America indicated that nearly all treating facilities would
use HBO for a CO-poisoned patient with a COHb level of 40% and a COHb level of
25% was identified most often as an indication for HBO therapy.48 While most studies
of cognitive impairments following CO poisoning include patients with moderate to
severe poisoning, information is accumulating regarding cognitive impairments in
patients with lower-level or less severe CO poisoning. Less severe CO poisoning
has been defined using a variety of criteria. It has been defined as COHb level of
≤10%,49 a COHb level of 10%,50 a COHb of 5–15%,47 or COHb of 0.01–11%.25
Most studies agree that less severe CO poisoning occurs at COHb levels of <15% with
no LOC.
The effects of less severe CO poisoning can be difficult to detect as the symptoms
are often mild, therefore such individuals may not seek medical care. Symptoms
of less severe CO poisoning are similar to acute poisoning and include headache,
fatigue, and dizziness. Syncope can occur with COHb levels as low as 0.1–0.3%.25
Less severe CO poisoning can impair visual thresholds, driving skills, and auditory
discrimination at COHb levels of 0.2–0.5%.25 Impaired vigilance to incoming sensory
stimuli51 and decreased auditory-evoked potentials, impaired memory, and altered
mood17 have been reported following less severe CO poisoning. People with COHb
levels as low as 4.4% have reduced work capacity.52 Cave painters who were exposed
to lower-level CO developed hallucinations.53
Studies that assess the cognitive effects of less severe CO poisoning are limited.
Current data suggest that impaired memory and slow mental-processing speed occurs
in individuals with less severe CO poisoning.17 Amitai and colleagues25 investigated
the cognitive effects of acute lower-level CO poisoning in healthy students who were
majoring in science, law, psychology, medicine, international relations, and social
work. The healthy students were exposed to CO from kerosene stoves used to heat
the space and compared to matched controls. The students exposed to the CO showed
significantly worse performance on measures of memory, learning, visuomotor skills,
abstract thinking, and visuospatial planning compare to the students not exposed
to CO.25 Thus, even less severe CO poisoning is associated with development of
cognitive impairments.
Prospective studies of the cognitive outcomes of less severe CO poisoning are
lacking. Chambers et al.54 compared the cognitive outcomes of CO poisoned patients
with less severe poisoning to patients with more severe CO poisoning at 6 weeks,
and 6 and 12 months following their poisoning. Less severe CO-poisoning was
defined as no LOC and a COHb level ≤15%, while more severe CO-poisoning was
defined as LOC or a COHb level >15%. About 55 patients had less severe and 201
had more severe CO poisoning. Cognitive sequelae occurred in 35% versus 39%
for patients with less severe versus more severe CO poisoning, respectively. There
was no difference in the prevalence of cognitive sequelae (P = 0.91) in patients
8417: “8417_c022” — 2007/9/11 — 12:14 — page 483 — #7

with less severe compared to the more severe CO-poisoning at 6 weeks, 6 and
12 months. Regardless of less severe or more severe CO poisoning, CO-poisoned
patients had significant cognitive sequelae.54 These data suggest that CO-related
cognitive outcomes may be independent of poisoning severity.
22.3.3 MARKERS OF CARBON MONOXIDE POISONING SEVERITY

AND OUTCOME
One common belief is that markers of CO-poisoning severity, such as LOC and
COHb levels are good predictors of patient outcome. One study found the length
of LOC was related to outcome55 and development of DNS.37 Alternatively, COHb
level is not a reliable predictor of CO-poisoning severity, symptoms, or neurologic
outcome.55−57 Similarly, COHb levels do not correlate with severity of poisoning
or cognitive outcomes.58−60 For example, the rate of cognitive sequelae in patients
with severe CO poisoning with mean COHb levels of 25.2 ± 9.2% did not dif-
fer from that of patients with less severe CO poisoning with mean COHb levels of
6.8 ± 4.7%.54 Other studies have shown that COHb levels are not associated with
cognitive deficits.61 Alternatively markers of poisoning severity such as LOC, dur-
ation of coma, elevated COHb, or duration of exposure do not predict cognitive
sequelae.23
Several studies have assessed the relationship between markers of CO poisoning
severity (LOC and COHb levels) with brain imaging findings and cognitive impair-
ments. Hopkins et al.62 found that LOC was not required for the development of
cognitive sequelae following CO poisoning. Another study found no association
between corpus callosum atrophy and COHb and or LOC. Even with significantly
elevated COHb levels and LOC in approximately 50% of the patients, these markers
(e.g., COHb level and LOC) did not correlate with the presence of corpus cal-
losum atrophy or development of cognitive impairments.62 Similarly, neither fornix
atrophy nor verbal memory impairments correlate with COHb or LOC following
CO poisoning.61 Findings in primates indicate that neither severity nor duration of
CO-exposure is related to the severity of white matter damage.63 In summary, neither
symptoms of poisoning, cognitive impairment4,23 white matter hyperintensities,64
fornix atrophy,61 or corpus callosum atrophy65 are related to COHb levels or LOC.
The lack of association between COHb levels and cognitive and neuropatholo-
gical outcomes raises the question as to why this may be the case. One possible
explanation is that the lack of association between COHb and outcome measures is
due part in to the variability in measured COHb levels in CO-poisoned individuals.
The variability in measured COHb levels is directly related to the delay in removal
from the CO environment to medical treatment and the amount and duration of sup-
plemental oxygen given prior to COHb measurement.66,67 Thus, the time to COHb
measurement and supplemental oxygen impacts the result in decreased COHb levels.
22.3.4 EFFECT OF HBO ON COGNITIVE IMPAIRMENTS

The usual treatment for acute CO poisoning is 100% normobaric oxygen, com-
monly delivered by a reservoir nonrebreathing face-mask, or by HBO.41,68 HBO
8417: “8417_c022” — 2007/9/11 — 12:14 — page 484 — #8

therapy is recommended for patients with acute CO poisoning, especially if they

have experienced LOC, or have severe poisoning.2,41,69,70
Comparisons of the several randomized clinical trials in the treatment of acute
CO poisoning is difficult owing to methodological differences.22,23,71,72 A recent pro-
spective double-blind, randomized treatment trial found that HBO therapy reduced
cognitive sequelae by 46% at 6 weeks compared to normobaric oxygen.16 Both
groups of participants improved over time, but the difference in cognitive sequelae
was maintained at 12 months. Those patients with cognitive sequelae had moder-
ate to severe cognitive impairments, falling below the 16th percentile of the normal
distribution.29 They communicate and perform activities of daily living normally, but
find activities that require executive function, memory, and attention/concentration
skills as challenging or impossible.16 Similar findings of benefit for HBO have been
reported.23
Alternatively, Scheinkestel et al.71 reported that HBO might worsen outcome in
CO-poisoned patients. The study by Weaver et al.16 differed substantially from that of
Scheinkestel et al. regarding the number of intubated patients, CO exposure duration,
time from the end of the CO exposure to HBO therapy, randomization methods
(equal proportions versus cluster), follow-up rate (97% versus 46%), suicide rate
(31% versus 69%), statistical analyses, and oxygen treatment protocols.71
22.3.5 RELATIONSHIP BETWEEN COGNITIVE SEQUELAE AND

NEUROIMAGING FINDINGS
CO poisoning may result in focal and generalized neuroanatomical abnormalities
observed on MR and Computed Tomography (CT) imaging. Brain lesions following
CO poisoning occur in cortex,73 cerebellum,74 thalamus,75 and substantia nigra.76
Subcortical lesions are found in the white matter77 and basal ganglia including the
globus pallidus,78 caudate and putamen.79,80
White matter hyperintensities are common in the periventricular and centrum
semiovale or deep white matter regions.64 Generalized atrophy of white matter
structures like the corpus callosum65 and white matter degeneration in the tem-
poral, parietal and occipital regions are reported post-CO poisoning.73 A prospective
study in consecutive CO-poisoned participants demonstrated that white matter lesions
occur more frequently than basal ganglia lesions.64 While some studies indicate that
white matter lesions are the most common lesion following CO poisoning,77,81 others
indicate basal ganglia lesions are the most common.82,83 Alternative evidence sup-
porting basal ganglia lesions as the most frequent CO-related lesion comes from a
recent review that found globus pallidus lesions occurred in 32–86% of patients.84
Early (within 24 h post-CO poisoning) brain imaging can be normal with basal
ganglia lesions observed on subsequent scans.64 Alternatively, basal ganglia lesions
can occur within the first day following CO poisoning.85 Basal ganglia lesions have
been reported at 1 month,86 6 months,31 1-year,87 2 years,86 4 years,74 and 5 years88
post-CO poisoning. It appears that COHb levels are not associated with the devel-
opment of basal ganglia lesions, as the lesions occurred with COHb levels as low of
9.1%28 and as high as 54%.89
8417: “8417_c022” — 2007/9/11 — 12:14 — page 485 — #9

In addition to neural lesions, CO poisoning may cause neuronal cell loss

and concomitant structural atrophy. Atrophy has been reported in the fornix,61
hippocampus,4 and corpus callosum.65 Generalized atrophy can occur with brain
volume reduction manifested by reduced gyral volume, increased sulcal space,
and passive ventricular enlargement.4 A recent study found CO-poisoned patients
had atrophy in the putamen, caudate, and globus pallidus in the absence of basal
ganglia lesions.90 The relationship between neuropathologic findings and cognit-
ive impairments has only recently been assessed. Slow mental-processing speed
and impaired memory were associated with smaller putamen and globus pallidus
volumes in CO-poisoned patients. Impaired verbal memory was associated with
fornix atrophy,61 while slow mental-processing speed was associated with white
matter hyperintensities64 following CO poisoning. Thus, basal ganglia atrophy,
fornix atrophy, and white matter hyperintensities likely all contribute to the observed
cognitive impairments in CO victims.
There are significant correlations between neuropsychological impairments and
abnormalities in cerebral perfusion, clinical MR, and/or brain volumetric measures in
CO poisoned patients.4 A study of consecutively CO-poisoned patients using quant-
itative MR (brain volumetric measures) compared neural volumes from the initial
MR scan (e.g., day of CO poisoning) with MR 6 months post CO exposure. The
patients had atrophic changes of the fornix, corpus callosum, and basal ganglia 6
months postexposure compared to their initial MR scans that correlated with cognit-
ive impairments.61,65 Alternatively, MR and CT structural imaging carried out at the
onset of the symptoms of DNS found no association between the symptoms and the
imaging abnormalities. Pavese et al.91 found 50% of patients (11/22) had abnormal-
ities on MR imaging, whereas 27% of the patients had adverse symptoms 1 month
after the CO poisoning. Other authors report many patients with delayed symptoms
(30–42%) have normal neuroimaging examinations.92,93
22.3.6 FUNCTIONAL IMAGING

Measures of cerebral blood flow may be more sensitive to CO poisoning-related neural
changes than standard structural imaging. Decreased glucose metabolism on positron
emission tomography (PET),94 hypoperfusion on single photon emission computed
tomography (SPECT),4,95 and abnormal electroencephalography (EEG)20,21 par-
allel the focal and diffuse changes observed with structural imaging following
CO poisoning. Regional cerebral blood flow (using PET) abnormalities may also
be present in the absence of abnormalities on CT or MR.96 Neuropsychological
impairments are associated with abnormalities in cerebral perfusion, clinical MR,
and/or brain volumetric measures.4 A SPECT study found patchy hypoperfusion in
patients that developed delayed neurological sequelae.95 Parkinsonian symptoms are
associated with decreased perfusion of the basal ganglia and cognitive deficits are
associated with decreased cerebral blood flow in cortical areas.96 CO poisoning-
related cerebral blood flow abnormalities predict poor outcome (death, remote
memory impairment).97 In contrast, cerebral blood flow abnormalities did not predict
outcome 3–5 days after CO poisoning.98
8417: “8417_c022” — 2007/9/11 — 12:14 — page 486 — #10

TABLE 22.2
Carbon Monoxide-Related Affective and
Neurobehavioral Changes
Affective Sequelae
Depression
Anxiety
Apathy
Irritability
Mood swing
Elated mood
Neurobehavioral Changes
Obsessive and compulsive behaviors
Delusions
Hallucinations
Violent outbursts
Fear
Disinhibition
Anger
22.4 AFFECTIVE DISORDERS

22.4.1 DEPRESSION AND ANXIETY
Some affective and behavioral changes associated with CO poisoning are shown
in Table 22.2. Affective and personality changes following CO poisoning appear
to be heterogeneous regarding time of onset, severity, and duration.26,89,99,100
While depression is frequently reported following CO poisoning, fewer studies have
assessed anxiety.4,26,92,101 Other psychological and personality changes following
CO poisoning include obsessive and compulsive behavior,26,92,102,103 delusions and
hallucinations,20,21,92 violent outbursts,20 fear,26 and elated mood.21
Jasper and colleagues found depression and anxiety in 45% of CO-poisoned
patients at 6 weeks, 44% at 6 months, and 43% at 12 months following CO poisoning.5
The prevalence of depression and anxiety following CO poisoning varied by study
from a low of 33% to a high of 100%.4,99,104 Differences in patient populations,
patient selection, affective measures, and length of follow-up may account for the
between study differences. The consistent between study findings include the high
rate of depression and anxiety in CO-poisoned people. Similar prevalence rates of
depression and anxiety occur in other pulmonary disorders with depression reported
in 25–28% of patients with cardiac and pulmonary disorders105,106 and anxiety in
10–40% of patients with pulmonary disorders.107,108
Accidentally CO-poisoned patients are as likely as those with intentional CO
poisoning to have depression and anxiety at 6 and 12 months.5 Other studies have
reported significant depression and anxiety in participants who attempt suicide
with CO.99,104 Hay et al.99 found that depression in CO-poisoned participants was
8417: “8417_c022” — 2007/9/11 — 12:14 — page 487 — #11

similar to control patients who had a psychiatric disorder. Smith and Brandon109
found a higher prevalence of depression and personality changes in patients with
intentional CO poisoning (35%) compared with accidentally poisoned patients (21%)
and psychiatric controls (9%). Thus both intentional and accidentally CO-poisoned
patients are at risk to develop depression and anxiety.
CO-poisoned patients with cognitive impairments may develop CO morbid
depression and anxiety.4,74 Alternatively, depression following CO poisoning can
occur in the absence of cognitive impairments.78,109 Mori et al.78 described a patient
with no prior history of psychiatric disorders who developed dramatic personality
changes in the absence of cognitive deficits following accidental CO poisoning. Smith
and Brandon109 found that 33% of CO-poisoned patients developed personality and
affective morbidity, but only 11% developed cognitive impairments. CO poisoning
appears to result in a high rate of depression and anxiety.
22.4.2 OBSESSIVE COMPULSIVE DISORDER

Symptoms of obsessive-compulsive disorder include behavior mannerisms associated
with Tourettes syndrome and obsessive thoughts. Compulsive stereotypic routines
may develop in patients with basal ganglia lesions.102 A case of obsessive-compulsive
disorder secondary to CO poisoning was reported in the 34-year-old male.102 Shortly
after hospital discharge, the patient reported an irresistible urge to spit and pull
the hair on his legs. He developed obsessive thoughts of harming his friends and
family, fear of germ contamination, and sexual fantasies. His compulsive beha-
viors included repetitive hand-washing, checking behaviors, and counting rituals.
Neuroimaging showed bilateral globus pallidus lesions. Thus, CO poisoning can lead
to acquired obsessive-compulsive disorder with concomitant bilateral globus pallidus
lesions.102
22.4.3 KLUVER–BUCY SYNDROME

Muller110 reported the case of an 18-year-old female with LOC due to CO poisoning.
Her COHb was 14%. During rehabilitation she had oral tendencies (i.e., put everything
into her mouth), decreased social distance, and object agnosia. She had flattened affect
and appeared placid and indifferent toward people and events. She was diagnosed
with a Kluver–Bucy-like syndrome. Neuropsychological testing showed cognitive
deficits similar to dementia. Lesions on brain imagining were located in the lateral
temporal lobes sparing the hippocampus. The Kluver-Bucy like symptoms resolved
6 months later, but the cognitive deficits including impaired attention, distractibility,
and amnesia persisted.110
22.5 CONCLUSION
CO poisoning is common, often goes unrecognized and may result in signific-
ant morbidity. Morbidity following CO poisoning includes neurologic sequelae,
neuropathologic abnormalities on brain imaging, affective sequelae, and cognitive
impairments. Morbidity appears to be independent of poisoning severity as measured
8417: “8417_c022” — 2007/9/11 — 12:14 — page 488 — #12

by COHb level and LOC. Most CO poisoning is preventable, therefore the associated
morbidity is also preventable. Given the high rate of brain related morbidity and the
fact that the majority of CO poisoning is avoidable, awareness and prevention of
CO exposure is warranted. With increased awareness of the dangers and causes of
CO exposure, and with the availability of CO alarms, CO poisoning and its adverse
effects may be significantly reduced.
References
1. Raub, J.A. et al. Carbon monoxide poisoning: a public health perspective, Toxicology,
145, 1, 2000.
2. Hampson, N.B., Ed. Hyperbaric Oxygen Therapy: 1999 Committee Report Revised,
Undersea and Hyperbaric Medical Society, Kensington, MD, 1999.
3. Piantadosi, C.A. Carbon monoxide poisoning, N. Engl. J. Med., 347, 1054, 2002.
4. Gale, S.D. et al. MRI, quantitative MRI, SPECT, and neuropsychological findings
following carbon monoxide poisoning, Brain Inj., 13, 229, 1999.
5. Jasper, B.W. et al. Affective outcome following carbon monoxide poisoning: A
prospective longitudinal study, Cogn. Behav. Neurol., 18, 127, 2005.
6. Okeda, R. et al. Comparative study on pathogenesis of selective cerebral lesions in
carbon monoxide poisoning and nitrogen hypoxia in cats, Acta. Neuropathol. (Berl).
56, 265, 1982.
7. Piantadosi, C.A. Carbon monoxide, oxygen transport, and oxygen metabolism,
Undersea Hyperb. Med., 2, 27, 1987.
8. Caine, D. and Watson, J.D. Neuropsychological and neuropathological sequelae of
cerebral anoxia: A critical review, J. Int. Neuropsychol. Soc., 6, 86, 2000.
9. Jarrard, L.E. and Meldrum, B.S. Selective excitotoxic pathology in the rat hippocam-
pus, Neuropathol. Appl. Neurobiol., 19, 381, 1993.
10. Coburn, R.F. Mechanisms of carbon monoxide toxicity, Prev. Med., 8, 310, 1979.
11. Thom, S.R. Antagonism of carbon monoxide-mediated brain lipid peroxidation by
hyperbaric oxygen, Toxicol. Appl. Pharmacol., 105, 340, 1990.
12. Thom, S.R. et al. Vascular nitrosative stress from carbon monoxide (CO) exposure,
Undersea Hyperb. Med., 25 (Suppl.), 47, 1998.
13. Piantadosi, C.A. Carbon monoxide poisoning, Undersea Hyperb. Med., 31, 167,
2004.
14. Piantadosi, C.A. et al. Apoptosis and delayed neuronal damage after carbon monoxide
poisoning in the rat, Exp. Neurol., 147, 103, 1997.
15. Thom, S.R. et al. Delayed neuropathology after carbon monoxide poisoning is
immune-mediated, Proc. Natl. Acad. Sci. U. S. A., 101, 13660, 2004.
16. Weaver, L.K. et al. Hyperbaric oxygen for acute carbon monoxide poisoning, N. Engl.
J. Med., 347, 1057, 2002.
17. Myers, R.A., DeFazio, A. and Kelly, M.P. Chronic carbon monoxide exposure: A
clinical syndrome detected by neuropsychological tests, J. Clin. Psychol., 54, 555,
1998.
18. Hopkins, R.O., Weaver, L.K. and Kesner, R.P. Long-term memory impairments and
hippocampal magnetic resonance imaging in carbon monoxide poisoned subjects.,
Undersea Hyperb. Med., 20, 15, 1993.
19. Gale, S.D. and Hopkins, R.O. Effects of hypoxia on the brain: Neuroimaging and
neuropsychological findings following carbon monoxide poisoning and obstructive
sleep apnea, J. Int. Neuropsychol. Soc., 10, 60, 2004.
8417: “8417_c022” — 2007/9/11 — 12:14 — page 489 — #13

20. Garland, H. and Pearce, J. Neurological complications of carbon monoxide poisoning,

Q. J. Med., 36, 445, 1967.
21. Min, S.K. A brain syndrome associated with delayed neuropsychiatric sequelae
following acute carbon monoxide intoxication, Acta. Psychiatr. Scand., 73, 80, 1986.
22. Raphael, J.C. et al. Trial of normobaric and hyperbaric oxygen for acute carbon
monoxide intoxication, Lancet, 2 (8660), 414, 1989.
23. Thom, S.R. et al. Delayed neuropsychologic sequelae after carbon monoxide pois-
oning: Prevention by treatment with hyperbaric oxygen, Ann. Emerg. Med., 25, 474,
1995.
24. Gorman, D.F. et al. A longitudinal study of 100 consecutive admissions for carbon
monoxide poisoning to the Royal Adelaide Hospital, Anaesth. Intensive Care, 20,
311, 1992.
25. Amitai, Y. et al. Neuropsychological impairment from acute low-level exposure to
carbon monoxide, Arch. Neurol., 55, 845, 1998.
26. Dunham, M.D. and Johnstone, B. Variability of neuropsychological deficits associated
with carbon monoxide poisoning: Four case reports, Brain Inj., 13, 917, 1999.
27. Hopkins, R.O. and Woon, F.L. Neuroimaging, cognitive, and neurobehavioral out-
comes following carbon monoxide poisoning, Behav. Cogn. Neurosci. Rev., 5, 141,
2006.
28. Gottfried, J.A. and Chatterjee, A. Carbon monoxide-mediated hippocampal injury,
Neurology, 57, 17, 2001.
29. Bastin, C. et al. Dissociation between recall and recognition memory performance in
an amnesic patient with hippocampal damage following carbon monoxide poisoning,
Neurocase, 10, 330, 2004.
30. Bourgeois, J.A. Amnesia after carbon monoxide poisoning, Am. J. Psychiatry, 157,
1884, 2000.
31. Carlesimo, G.A. et al. Selective sparing of face learning in a global amnesic patient,
J. Neurol. Neurosurg. Psychiatr., 71, 340, 2001.
32. Henke, K. et al. Memory lost and regained following bilateral hippocampal damage,
J. Cogn. Neurosci., 11, 682, 1999.
33. Helffenstein, D. Neuropsychological evaluation of the carbon monoxide-poisoned
patient, In: Carbon Monoxide Toxicity, Penney, D.G., Ed., CRC Press, Boca Raton,
FL, 2000, Chapt. 20, pp. 439–461.
34. Farah, M. Visual Agnosia, MIT Press, Cambridge, MA, 1990.
35. Patla, A.E. and Goodale, M.A. Obstacle avoidance during locomotion is unaffected
in a patient with visual form agnosia, Neuroreport, 8, 165, 1996.
36. Goodale, M.A. et al. A neurological dissociation between perceiving objects and
grasping them, Nature, 349 (6305), 154, 1991.
37. Choi, I.S. Delayed neurologic sequelae in carbon monoxide intoxication, Arch.
Neurol., 40, 433, 1983.
38. Shillito, F., Drinker, C. and Chaughness, T. The problem of nervous and mental
sequelae in carbon monoxide poisoning, JAMA, 106, 669, 1936.
39. Choi, I.S. Carbon monoxide poisoning: systemic manifestations and complications,
J. Korean Med. Sci., 16, 253, 2001.
40. Yoshii, F. et al. Magnetic resonance imaging and C-N-methylspiperone/positron emis-
sion tomography studies in a patient with the interval form of carbon monoxide
poisoning, J. Neurol. Sci., 160, 87, 1998.
41. Ernst, A. and Zibrak, J.D. Carbon monoxide poisoning, N. Engl. J. Med., 339, 1603,
1998.
42. Yant, W. et al. Studies in asphyxia, US Public Health Bull., 211, 1, 1934.
8417: “8417_c022” — 2007/9/11 — 12:14 — page 490 — #14

43. Hurley, R.A. et al. Applications of functional imaging to carbon monoxide poisoning,
J. Neuropsychiat. Clin. Neurosci., 13, 157, 2001.
44. Sohn, Y.H. et al. The brain lesion responsible for parkinsonism after carbon monoxide
poisoning, Arch. Neurol., 57, 1214, 2000.
45. Kamada, K. et al. Cerebral metabolic changes in delayed carbon monoxide sequelae
studied by proton MR spectroscopy, Neuroradiology, 36, 104, 1994.
46. Weaver, L.K. et al. Persistent and delayed carbon monoxide cognitive sequelae,
Undersea Hyperb. Med., Vol. 32(4), No. 144, 289, 2005.
47. Bleecker, M.L. and Lindgren, K.N. The mere presence of low levels of carboxyhemo-
globin is not causal proof for altered neuropsychological performance, Arch. Neurol.,
56, 1299, 1999.
48. Hampson, N.B. et al. Selection criteria utilized for hyperbaric oxygen treatment of
carbon monoxide poisoning, J. Emerg. Med., 13, 227, 1995.
49. Sadovnikoff, N., Varon, J. and Sternbach, G.L. Carbon monoxide poisoning. An occult
epidemic, Postgrad. Med., 92, 86, 1992.
50. Crawford, R., Campbell, D.G. and Ross, J. Carbon monoxide poisoning in the home:
Recognition and treatment, Br. Med. J., 301 (6758), 977, 1990.
51. Horvath, S.M., Dahms, T.E. and O’Hanlon, J.F. Carbon monoxide and human vigil-
ance. A deleterious effect of present urban concentrations, Arch. Environ. Health, 23,
343, 1971.
52. Wright, J. Chronic and occult carbon monoxide poisoning: We don’t know what we’re
missing, Emerg. Med. J., 19, 386, 2002.
53. Guarnieri, M. Cave painting hazard?, Science, 283 (5410), 2019, 1999.
54. Chambers, C., Hopkins, R.O. and Weaver, L.K. Cognitive and affective out-
comes compared dichotomously in patients with acute carbon monoxide poisoning,
Undersea Hyperb. Med., Vol. 33(5), 338–339, 2006.
55. Jain, K.K. Carbon Monoxide Poisoning, Warren H. Green, Inc., St. Louis, 1990.
56. Levy, D.E. et al. Predicting outcome from hypoxic-ischemic coma, JAMA, 253, 1420,
1985.
57. Martindale, L.G. Carbon monoxide poisoning: The rest of the story, J. Emerg. Med.,
15 (2 Pt 1), 101, 1989.
58. Camporesi, E.M. Hyperbaric Oxygen Therapy: A Committee Report, Undersea and
Hyperbaric Medical Society, Bethesda, MD, 1996.
59. Weaver, L.K. et al. Double-blind, controlled, prospective randomized clinical trial
(RCT) in patients with acute carbon monoxide poisoning: Outcome of patients treated
with normobaric oxygen or hyperbaric oxygen (HBO2) - an interim report., Undersea
Hyperb. Med., 22 (Suppl.), 14, 1995.
60. Winter, P.M. and Miller, J.N. Carbon monoxide poisoning, JAMA, 236, 1502,
1976.
61. Kesler, S.R. et al. Verbal memory deficits associated with fornix atrophy in carbon
monoxide poisoning, J. Int. Neuropsychol. Soc., 7, 640, 2001.
62. Hopkins, R.O. et al. Severe anoxia with and without concomitant brain atrophy and
neuropsychological impairments, J. Int. Neuropsychol. Soc., 1, 501, 1995.
63. Ginsberg, M.D., Myers, R.E. and McDonagh, B.F. Experimental carbon monoxide
encephalopathy in the primate: II. Clinical aspects, neuropathology, and physiologic
correlation, Arch. Neurol., 30, 209, 1974.
64. Parkinson, R.B. et al. White matter hyperintensities and neuropsychological outcome
following carbon monoxide poisoning, Neurology, 58, 1525, 2002.
65. Porter, S.S. et al. Corpus callosum atrophy and neuropsychological outcome following
carbon monoxide poisoning, Arch. Clin. Neuropsychol., 17, 195, 2002.
8417: “8417_c022” — 2007/9/11 — 12:14 — page 491 — #15

66. Sokal, J.A. and Kralkowska, E. The relationship between exposure duration, carboxy-
hemoglobin, blood glucose, pyruvate and lactate and the severity of intoxication
in 39 cases of acute carbon monoxide poisoning in man, Arch. Toxicol., 57, 196,
1985.
67. Sokal, J.A. The effect of exposure duration on the blood level of glucose, pyruvate
and lactate in acute carbon monoxide intoxication in man, J. Appl. Toxicol., 5, 395,
1985.
68. Weaver, L.K. Carbon monoxide poisoning, Crit. Care Clin., 15, 297, 1999.
69. Tibbles, P.M. and Edelsberg, J.S. Hyperbaric-oxygen therapy, N. Engl. J. Med., 334,
1642, 1996.
70. Piantadosi, C.A. Diagnosis and treatment of carbon monoxide poisoning, Respir. Care
Clin. N. Am., 5, 183, 1999.
71. Scheinkestel, C.D. et al. Hyperbaric or normobaric oxygen for acute carbon monoxide
poisoning: A randomized controlled clinical trial, Med. J. Aust., 170, 203, 1999.
72. Ducasse, J.L., Celsis, P. and Marc-Vergnes, J.P. Non-comatose patients with acute car-
bon monoxide poisoning: Hyperbaric or normobaric oxygenation?, Undersea Hyperb.
Med., 22, 9, 1995.
73. Uchino, A. et al. MRI of the brain in chronic carbon monoxide poisoning,
Neuroradiology, 36, 399, 1994.
74. Bruno, A., Wagner, W. and Orrison, W.W. Clinical outcome and brain MRI four years
after carbon monoxide intoxication, Acta Neurol. Scand., 87, 205, 1993.
75. Chang, K.H. et al. Delayed encephalopathy after acute carbon monoxide intoxication:
MR imaging features and distribution of cerebral white matter lesions, Radiology, 184,
117, 1992.
76. Kawanami, T. et al. The pallidoreticular pattern of brain damage on MRI in a
patient with carbon monoxide poisoning, J. Neurol. Neurosurg. Psychiatr., 64, 282,
1998.
77. Watanabe, N. et al. Statistical parametric mapping in brain single photon computed
emission tomography after carbon monoxide intoxication, Nucl. Med. Commun., 23,
355, 2002.
78. Mori, E. et al. Isolated athymhormia following hypoxic bilateral pallidal lesions,
Behav. Neurol., 9, 17, 1996.
79. Hsiao, C.L., Kuo, H.C. and Huang, C.C. Delayed encephalopathy after carbon monox-
ide intoxication—long-term prognosis and correlation of clinical manifestations and
neuroimages, Acta Neurol. Taiwan, 13, 64, 2004.
80. Martinez Bermejo, A. et al. [Bilateral hypodensity of the basal ganglia. Clinico-
evolutionary correlation in children], Rev. Neurol., 33, 101, 2001.
81. Smallwood, P. and Murray, G.B. Neuropsychiatric aspects of carbon monoxide pois-
oning: A review and single case report suggesting a role for amphetamines, Ann. Clin.
Psychiatry, 11, 21, 1999.
82. Pracyk, J.B. et al. Brain computerized tomography after hyperbaric oxygen therapy
for carbon monoxide poisoning., Undersea Hyperb. Med., 22, 1, 1995.
83. Roohi, F., Kula, R.W. and Mehta, N. Twenty-nine years after carbon monoxide
intoxication, Clin. Neurol. Neurosurg., 103, 92, 2001.
84. Hopkins, R.O. et al. Basal ganglia lesions following carbon monoxide poisoning,
Brain Inj., 20, 273, 2006.
85. Gottfried, A.W. Intellectual consequences of perinatal anoxia, Psychol. Bull., 80, 231,
1973.
86. Stuppaeck, C.H. et al. Akathisia induced by necrosis of the basal ganglia after carbon
monoxide intoxication, Mov. Disord., 10, 229, 1995.
8417: “8417_c022” — 2007/9/11 — 12:14 — page 492 — #16

87. Shimosegawa, E. et al. Cerebral blood flow and glucose metabolism measurements
in a patient surviving one year after carbon monoxide intoxication, J. Nucl. Med., 33,
1696, 1992.
88. Vion-Dury, J. et al. Sequelae of carbon monoxide poisoning: An MRI study of two
cases, J. Neuroradiol., 14, 60, 1987.
89. Jaeckle, R.S. and Nasrallah, H.A. Major depression and carbon monoxide-induced
parkinsonism: Diagnosis, computerized axial tomography, and response to L-dopa,
J. Nerv. Ment. Dis., 173, 503, 1985.
90. Pulsipher, D.T., Hopkins, R.O. and Weaver, L.K. Basal ganglia volumes following
carbon monoxide poisoning: A prospective longitudinal study, Undersea Hyperb.
Med., 33, 245, 2006.
91. Pavese, N. et al. Clinical outcome and magnetic resonance imaging of carbon
monoxide intoxication. A long-term follow-up study, Ital. J. Neurol. Sci., 20, 171,
1999.
92. Lee, M.S. and Marsden, C.D. Neurological sequelae following carbon monoxide pois-
oning clinical course and outcome according to the clinical types and brain computed
tomography scan findings, Mov. Disord., 9, 550, 1994.
93. Choi, I.S. and Cheon, H.Y. Delayed movement disorders after carbon monoxide
poisoning, Eur. Neurol., 42, 141, 1999.
94. de Reuck, J. et al. A positron emission tomography study of patients with acute carbon
monoxide poisoning treated by hyperbaric oxygen, J. Neurol., 240, 430, 1993.
95. Choi, I.S. et al. Evaluation of outcome of delayed neurologic sequelae after carbon
monoxide poisoning by technetium-99m hexamethylprophylene amine oxime brain
single photon emission computed tomography, Eur. Neurol., 35, 137, 1995.
96. Kao, C.H. et al. HMPAO brain SPECT in acute carbon monoxide poisoning, J. Nucl.
Med., 39, 769, 1998.
97. Turner, M. and Kemp, P.M. Isotope brain scanning with Tc-HMPAO: A predictor of
outcome in carbon monoxide poisoning?, J. Accid. Emerg. Med., 14, 139, 1997.
98. Sesay, M. et al. Regional cerebral blood flow measurements with Xenon-CT in
the prediction of delayed encephalopathy after carbon monoxide intoxication, Acta
Neurol. Scand. Suppl., 166, 22, 1996.
99. Hay, P.J. et al. The neuropsychiatry of carbon monoxide poisoning in attempted
suicide: A prospective controlled study, J. Psychosom. Res., 53, 699, 2002.
100. Vieregge, P. et al. Carbon monoxide poisoning: Clinical, neurophysiological, and
brain imaging observations in acute disease and follow-up, J. Neurol., 236, 478,
1989.
101. Jefferson, J. Subtle neuropsychiatric sequelae of carbon monoxide intoxication: Two
case reports, Am. J. Psychiatry, 133, 961, 1976.
102. Escalona, P.R. et al. Obsessive-compulsive disorder following bilateral globus
pallidus infarction, Biol. Psychiatry, 42, 410, 1997.
103. Lugaresi, A. et al. "Psychic akinesia" following carbon monoxide poisoning, Eur.
Neurol., 30, 167, 1990.
104. Skopek, M.A. and Perkins, R. Deliberate exposure to motor vehicle exhaust gas: The
psychosocial profile of attempted suicide, Aust. N. Z. J. Psychiatry, 32, 830, 1998.
105. Silverstone, P.H. Prevalence of psychiatric disorders in medical inpatients, J. Nerv.
Ment. Dis., 184, 43, 1996.
106. Silverstone, P.H. et al. The prevalence of major depressive disorder and low self-
esteem in medical inpatients, Can. J. Psychiatry., 41, 67, 1996.
107. Karajgi, B. et al., The prevalence of anxiety disorders in patients with chronic
obstructive pulmonary disease, Am. J. Psychiatry, 147, 200, 1990.
8417: “8417_c022” — 2007/9/11 — 12:14 — page 493 — #17

108. Pollack, M.H. et al., Prevalence of panic in patients referred for pulmonary function
testing at a major medical center, Am. J. Psychiatry, 153, 110, 1996.
109. Smith, J.S. and Brandon, S. Morbidity from acute carbon monoxide poisoning at
three-year follow-up, Br. Med. J., 1, 318, 1973.
110. Muller, N.G. and Gruber, O. High-resolution magnetic resonance imaging reveals
symmetric bitemporal cortical necrosis after carbon monoxide intoxication, J.
Neuroimaging, 11, 322, 2001.
8417: “8417_c022” — 2007/9/11 — 12:14 — page 494 — #18

23 Neurocognitive and
Neurobehavioral
Sequelae of Chronic
Carbon Monoxide
Poisoning: A
Retrospective Study and
Case Presentation
CONTENTS
23.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 496

23.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 496
23.1.2 Incidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497
23.1.3 Physical, Cognitive and Emotional/Affective Sequelae of
Chronic Carbon Monoxide Poisoning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498
23.1.4 Published Chronic Carbon Monoxide Studies Utilizing
Neuropsychological Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 502
23.2 Helffenstein Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 506
23.2.1 Sample and Exposure Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 506
23.2.1.1 Admission Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 506
23.2.1.2 Sample Demographics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507
23.2.1.3 Source and Location of Exposure . . . . . . . . . . . . . . . . . . . . . . . 507
23.2.1.4 Level of Carbon Monoxide Exposure . . . . . . . . . . . . . . . . . . . 507
23.2.1.5 Duration and Frequency of Exposure . . . . . . . . . . . . . . . . . . . 507
23.2.1.6 Time Since Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 508
23.2.2 Battery Administered . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 508
23.2.3 Norms Utilized . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509
23.2.4 Symptoms Experienced by Participants During Carbon
Monoxide Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509
23.2.5 Persisting Symptoms Reported by Participants . . . . . . . . . . . . . . . . . . . 511
495
8417: “8417_c023” — 2007/9/11 — 12:14 — page 495 — #1

23.2.5.1 Physical Symptoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513

23.2.5.2 Visual Symptoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513
23.2.5.3 Cognitive Symptoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515
23.2.5.4 Psychological / Behavioral Symptoms . . . . . . . . . . . . . . . . 515
23.2.6 Neuropsychological Testing Outcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516
23.2.6.1 Index Scores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516
23.2.6.2 IQ Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 518
23.2.6.3 Halstead–Reitan Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522
23.2.6.4 Memory Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522
23.2.6.5 Academic Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 523
23.2.6.6 Visual–Visual Perceptual Testing . . . . . . . . . . . . . . . . . . . . . . 525
23.2.6.7 Speed of Information Processing . . . . . . . . . . . . . . . . . . . . . . 527
23.2.6.8 Other Motor Skills . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528
23.2.6.9 Miscellaneous Tests of Executive Function . . . . . . . . . . . 529
23.2.6.10 Language Comprehension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 530
23.2.7 Minnesota Multiphasic Personality Inventory-2 . . . . . . . . . . . . . . . . . . 531
23.2.8 Vocational Outcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 536
23.2.9 Summary of Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538
23.3 Case Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540
23.3.1 Patient Demographics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540
23.3.2 Exposure Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540
23.3.3 Educational History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540
23.3.4 Persisting Symptoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 541
23.3.5 Results from Initial Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 541
23.3.6 Neuropsychological Re-evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542
23.3.6.1 Circumstances of Re-evaulation . . . . . . . . . . . . . . . . . . . . . . . 542
23.3.6.2 Self-reported Symptoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542
23.3.6.3 Comparison of Test Scores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543
23.3.6.4 Results from Re-evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543
23.3.7 Summary of Case Study. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545
23.3.8 Takeaway Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545
23.4 Addendum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 546
Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 547
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 548
23.1 BACKGROUND
23.1.1 Introduction
Carbon monoxide (CO) is the most common cause of poisoning in the United States
and may result in neuropathologic changes which in turn result in a wide range
of cognitive, visual, affective, and neurologic sequelae. The effects of moderate
to severe acute CO poisoning have been well documented in the literature. For
example, Gale et al.1 utilized neuropsychological testing in addition to brain ima-
ging techniques [i.e., Single Photon Emission Computed Tomography (SPECT),
8417: “8417_c023” — 2007/9/11 — 12:14 — page 496 — #2

Neurocognitive and Neurobehavioral Sequelae of Chronic Carbon Monoxide 497
Magnetic Resonance Imaging (MRI), and quantitative Magnetic Resonance Imaging

(qMRI)] to identify the patient’s residual neuropsychological and neuropathological
impairments. Ninety-three percent of the sample group showed cognitive impair-
ments on neuropsychological testing, including difficulties with attention, memory,
executive function, and mental processing speed. Ninety-five percent of the patients
were experiencing ongoing problems with depression and anxiety. The results of the
imaging studies revealed that 38% of the patients had abnormal MRI scans, 67% had
abnormal SPECT scans, and 67% had abnormal qMRI findings. The qMRI technique
identified hippocampal atrophy and/or diffuse cortical atrophy. The SPECT identified
cerebral profusion deficits, most notably in the frontal and temporal lobes. Significant
correlation was identified between the neuroimaging techniques and deficits noted on
neuropsychological testing.
The effects of chronic CO poisoning are less well researched and documented.
Most CO poisoning studies to date have focused on short-term effects of a one time,
lower-level exposure to CO in experimental settings or on the long-term effects of
accidental acute CO poisoning. As with any toxin, there are three components in
determining the severity of exposure. These include the level or amount of toxin the
individual is exposed to, the frequency of exposure, and the duration of each exposure.
One difficulty faced in evaluating the effects of chronic CO poisoning is that it is often
difficult to calculate the above variables with any degree of certainty. However, the
health risks of exposure to lower levels of CO repeatedly or for an extended duration
should not be minimized. Wright2 states, “There is a strong possibility that low level
exposure to CO is responsible for widespread and significant morbidity. However, the
clinical syndrome produced is often overlooked because of a range of presentation,
obscure symptoms, and a lack of awareness of the problem” (p. 387).
There is now a growing body of evidence, which clearly shows that chronic
exposure to CO can, and often does, result in permanent neurological, cognitive,
and visual dysfunction. Regarding this issue, Penney3 states, “Clearly, prolonged
exposure to this poison even at what were previously thought to be ultra-low levels
is capable of producing many and varied residual health effects. Furthermore, the
incidence of such unpleasant and often debilitating effects is far higher than was
previously believed by the medical and public health community and can continue
for a very long period of time” (pp. 414–415). Indeed, there is some evidence to
suggest that because of the repeated exposures, which typically occur in cases of
chronic CO poisoning, the pathophysiological changes and damage to the brain may
actually be more significant than in cases of acute poisoning.4
23.1.2 INCIDENCE
CO exposure is the most common cause of death by poisoning in the US and results in
an estimated 40,000 emergency room (ER) visits per year.5 Typically, these are acute
poisoning incidents which have resulted in severe medical problems requiring emer-
gency intervention. Townsend and Maynard6 note, “The Institute for Environment
and Health commented in their 1998 publication on CO that, ‘It is likely that many
more sub-acute CO intoxications occur than are brought to the attention of med-
ical practitioners.’ Because of the difficulty in recognizing the effects of exposure
8417: “8417_c023” — 2007/9/11 — 12:14 — page 497 — #3

to low concentrations of CO, there is currently limited knowledge on the size of the
problem” (p. 708). These authors go on to note that there are two key questions,
which must be addressed. One, the number of people potentially affected by low
levels of CO in their homes, workplaces or other settings, and, two, the likelihood
of long-term negative health effects. With regard to the second question, there is
some evidence to suggest that a significant proportion (43%) of individuals who
experience a chronic exposure to CO have permanent neurological sequelae at 3-year
follow-up.7 The number of individuals potentially affected by low levels of CO is
more difficult to estimate, but is certainly more common than acute poisoning events.
Hampson8 points out that the signs and symptoms of CO poisoning are nonspe-
cific and that under-diagnosis in ERs is well described in the literature. He goes on
to point out that not all patients chronically exposed to CO are treated in ERs and
are often seen in medical offices or clinics and, therefore, their statistics would not
appear in the national database. It is also noted that these patients will often attrib-
ute their nonspecific symptoms to alternate causes (e.g., viral illness) and not seek
medical attention. This latter event is typically referred to as “occult CO poisoning.”2
The problem with accurately diagnosing chronic CO poisoning has been well docu-
mented by other authors. Regarding this issue, Penney3 states, “For every single case
of chronic CO poisoning reported/successfully diagnosed, there are ten cases that go
unreported/undiscovered/undiagnosed” (pp. 396–397).
The answer to the question regarding the frequency of chronic CO poisoning
remains elusive. Heckerling et al.9 estimated that 3–5% of individuals who present to
urban ERs for headaches and dizziness might well be victims of chronic CO poisoning.
Whatever the number of chronic occult CO poisonings, it is clearly a significant health
risk in the United States. Regarding this issue, Halpern10 states, “Chronic occult CO
poisoning is a diagnosis that is not frequently recognized in patients seen initially
in an ER or by a primary care provider. It is not readily recognized because of a
limited history, vague and variable clinical presentation and a failure of emergency
care providers to suspect the cause of symptoms. It is a serious and potentially lethal
condition and should be suspected whenever a person is seen at a health care facility
for multiple ‘flu-like’ complaints, especially during the winter months when homes
are heated, or in the late fall when furnaces are started” (p. 107).
23.1.3 PHYSICAL, COGNITIVE AND EMOTIONAL/AFFECTIVE

SEQUELAE OF CHRONIC CARBON MONOXIDE POISONING
To date, there has been several retrospective survey studies conducted concerning
the residual or persisting symptoms associated with chronic CO poisoning. These
include two studies conducted by Penney3 and a comprehensive questionnaire study
of individuals chronically exposed to CO in the United Kingdom conducted by CO
Support in 1996, a registered charity, headed by Ms. Debbie Davis. The CO Support
Study was originally published as a technical paper in October 1997,11 and a detailed
summary of that study was later published by Hay et al.12
Penney3 conducted two retrospective studies of chronic CO poisoning. Study A
included data from 66 individuals who had sustained chronic CO poisoning, defined
as CO exposure lasting more than 24 h. Data were obtained through the Internet.
8417: “8417_c023” — 2007/9/11 — 12:14 — page 498 — #4

Duration of exposure, as reported by 29 participants ranged from 0.18 to 120 months.

The mean duration of exposure was 30.7 months plus or minus 6.0 months (SEM).
Air CO concentrations reported in 15 instances were 427.8 ppm, plus or minus 115.2
ppm. Carboxyhemoglobin (COHb) level, as reported in 11 instances was 9.65%,
plus or minus 2.46%. Study A assessed symptoms that respondents indicated they
experienced during their chronic CO exposure, as well as their residual or persisting
symptoms. Respondents reported 103 different symptoms in total during their expos-
ure. The typical respondent reported multiple symptoms in multiple systems. A partial
list of these symptoms appears in Table 23.1. Penney notes that the misdiagnosis rate
in cases of chronic CO poisoning is very high. He believes chronic CO poisoning is
not better recognized because “It almost invariably presents with too many dispar-
ate, seemingly unrelated and for the most part, nonspecific symptoms. This tends to
confuse physicians who act mainly on pattern recognition of one or a few symptoms
to come up with a probable diagnosis, or at least a ‘short list.’ The result of being
presented with 5, 10, or 15 or more symptoms is likely to yield a diagnosis of hypo-
chondriasis (faking), psychiatric condition, or both” (p. 395). When an ER or family
physician is presented with a history of multiple symptoms in multiple systems, it
is easy to understand how they may misdiagnose chronic CO poisoning. Halpern10
emphasizes that, “A high index of suspicion is necessary to recognize this condition.”
In addition, accurate and detailed history taking is important in identifying chronic
CO poisoning.
TABLE 23.1
Symptoms Reported During Chronic Carbon Monoxide
Poisoning Exposure—Penney Study A, 20003
Physical Headaches Weakness
Nausea Tinnitus
Vomiting Syncope, partial/total
Fatigue/lethargy Sleep disturbance
Dizziness/vertigo Vision problems
Shortness of breath Heart palpitations
Muscle/joint aches Paresthesias
Balance problems Muscle cramps
Hearing problems Chest pain/tightness
Tremors
Cognitive Short-term memory Spelling
Mental confusion Speech
Attention/concentration Disorientation
Word finding
Emotional/affective Depression Personality changes
Anxiety Mood swings
Irritability Apathy
a Summarized and reprinted with permission of author.
8417: “8417_c023” — 2007/9/11 — 12:14 — page 499 — #5

TABLE 23.2
Persisting Symptoms Following Chronic Carbon Monoxide Exposure—
Penney Study A, 20003
Physical Ataxia Headaches
Balance Hearing problems
Muscle/joint pain Multiple chemical sensitivity
Temperature deregulation Nausea
Chest pain/tightness Paresthesias
Choking Peripheral neuropathy
Motor incoordination Heart palpitation
Muscle cramps Motor tremors
Dizziness Photophobia
Dysarthria Phonophobia
Fatigue Sleep disturbance
Shortness of breath Vision problems
Tinnitus
G. I. problems
Cognitive Attention/concentration Reading
Short-term memory Speech
Mental confusion Spelling
Disorientation Writing
Executive dysfunction Vocabulary (reduced)
Slow speed of information processing
Math
Paraphasias (literal and verbal)
Verbal fluency (word finding)
Emotional/affective Depression Aggression
Anxiety Libido (reduced)
Panic attacks Motivation (reduced)
Irritability
Personality change
a Summarized and re-printed with permission of author.
Study A also assessed residual or persisting symptoms experienced by the 66

respondents. Ninety-five different persisting symptoms were identified. A select
sample of those symptoms is presented in Table 23.2.
Penney’s Study B3 involved the analysis of questionnaires completed by 82 indi-
viduals who claimed to have suffered chronic CO poisoning. The mean duration
of exposure was 28.4 months, plus or minus 4.4 months, with a range of 3 weeks
to 120 months. The mean period after termination of the CO exposure, to the time
their responses were given was 21.4 months, plus or minus 2.2 months, or nearly
2 years after the exposure stopped. This study reviewed only residual or permanent
symptoms that the respondents were experiencing at the time they completed their
questionnaires. Regarding physical symptoms, 100% reported persisting problems
with fatigue. Greater than 80% reported residual problems with headaches, muscle
8417: “8417_c023” — 2007/9/11 — 12:14 — page 500 — #6

and joint pain, dizziness/balance problems, and sleep disturbance. Greater than 50%
reported ongoing difficulties with chest pain, tingling and numbness, and vision prob-
lems. More than 45% reported ongoing change in their perception of or sensitivity to
smell or taste.
From a cognitive standpoint, greater than 70% reported ongoing problems with
decision-making, following directions, short-term memory, and attention and con-
centration. More than 40% reported ongoing difficulties with spatial disorientation
and organization.
With regard to emotional and affective symptoms, more than 80% of the sample
reported ongoing problems with mood change/swings, temper problems/irritability,
and personality changes. More than 40% reported ongoing social and family prob-
lems. Thirty percent reported school problems. Regarding these results, Penney3
states, “This study suggests that a multitude of physical, cognitive, and emotional
symptoms persist for very long periods of time following chronic exposure to CO.
The CO exposure need not produce altered consciousness at any time for this to occur.
In fact, the CO concentrations and COHb saturations are quite low and in the range
previously thought incapable of producing lasting health harm in humans” (p. 413).
The CO Support study11 involved the analysis of questionnaires completed by
65 individuals who were chronically exposed to CO. None of those individuals
lost consciousness (LOC) as part of their exposure. The results of 12 question-
naires completed by individuals who did experience LOC were also reviewed. Ten
of those individuals were involved in chronic CO poisonings and two experienced
acute poisonings. This study is unique in that it used controls matched for gender,
age, and income. For the chronically exposed patients, data regarding symptoms
experienced during exposure were summarized as well as persisting symptoms. A
detailed summary of this study will not be presented as part of this chapter. For a
detailed review of the results of this study, the reader is referred to Hay et al.12 It
is important to note, however, that as in the Penney studies; respondents experi-
enced multiple symptoms in multiple systems during their exposure. It is important
to note that results of this study also suggest that, while some symptoms do abate
to some degree once the exposure stops, in every case the symptoms may persist
long-term.
The CO Support study also gathered data regarding employment outcome.
Although this was not discussed in detail in the Hay et al.12 summaries, a review
of the original technical paper provide more detail regarding employment outcome.
Those results are presented in Table 23.3. Of note, 32% of the patients chronically
exposed to CO were unable to return to work following the exposure. When LOC
occurred because of the CO poisoning, 75% were unable to return to competitive
employment. This would suggest that when LOC occurs in conjunction with CO
poisoning, the possibility of total and permanent vocational disability increases dra-
matically. Both of these figures stand in contrast to the control group where none of
those individuals was disabled from employment.
Hay and his colleagues12 conclude their summary of this study by stating, “The
results of this survey indicate that there is a continuing and unrecognized prob-
lem associated with chronic exposure to CO. Most physicians do not recognize the
symptoms of CO poisoning, and, therefore, do not diagnose it. Many individuals
8417: “8417_c023” — 2007/9/11 — 12:14 — page 501 — #7

TABLE 23.3
Employment Outcome—CO Support Study, 199711
CO Poisoning Matched Nondisabled
Chronic CO Poisoning with LOC* Control Group
(N = 65) (N = 12) (N = 65)
Able to work full time 35% 17%
Able to work part time 8% 8%
Unable to work 32% 75%
Not applicable 22% 0%
Working full time 44.6%
Working part time 24.6%
Unable to work 0%
Not applicable 30.8%
∗ This group included ten individuals chronically exposed to CO and two individuals who experienced
acute exposures.
suffer for many years because of their exposure to this gas, and as the survey indic-
ates, many people continue to suffer symptoms years after the exposure has stopped.
Respondents of the questionnaire indicate that they have experienced a wide range
of symptoms up to 2 years after the exposure ended” (p. 434).
23.1.4 PUBLISHED CHRONIC CARBON MONOXIDE STUDIES

UTILIZING NEUROPSYCHOLOGICAL ASSESSMENT
In 1990, Ryan13 presented the case study of a 48-year-old, right-handed, married
woman who reported a 3-year history of constant headaches, lethargy and memory
problems. She indicated that she did not have any difficulty recalling events from
the distant past but had difficulty recalling information that is more recent. She was
reporting ongoing episodes of mental confusion, periods of depression, and anxiety.
She reported that on one occasion she nearly LOC in her basement and, therefore, had
the gas company check her furnace. The furnace was found to be releasing 180 ppm
CO. The patient was running a typing service out of her basement and the possibility
existed that the exposure may have persisted for up to 3 years. No COHb level was
obtained and it is noted that the patient never LOC. While the woman’s headaches
stopped once her furnace was replaced, her memory problems persisted. Her history
was negative for alcohol or drug abuse, head trauma, and psychiatric problems. She
had never previously been exposed to any other toxic substances.
The results of her neuropsychological testing revealed deficits in the area of incid-
ental memory, as measured by the Digit Symbol Incidental Memory Test, as well as
clear deficits in her ability to both learn and recall verbal and visual information.
Dr. Ryan summarizes the results by stating, “There is no doubt that this patient has
developed a clinically significant memory disorder. Prior to her exposure, she worked
in positions that placed demands on concentration and memory skills; following
8417: “8417_c023” — 2007/9/11 — 12:14 — page 502 — #8

this exposure, she was unable to keep track of verbal information that was presen-
ted seconds earlier, and had difficulty accurately retrieving both verbal and visual
information that she had learned within the past 30 min” (p. 64). Based on his over-
all assessment of the case, Dr. Ryan attributed her neuropsychological disturbance,
affective disorder, and somatic complaints to her 3-year history of low level CO
exposure. Dr. Ryan further notes that memory disturbances are one of the most fre-
quently reported cognitive problems following CO poisoning. He also notes that this
case supports the conclusion that a LOC is not necessary for the development of
neuropsychiatric symptoms following a period of CO exposure.
Roy Myers et al.14 state that “Chronic exposure to CO produces a clinical syn-
drome that is often overlooked because of obscure symptomatology, a range of
presentations, and a lack of awareness of the problems” (p. 555). They go on to
note that neurological exams will often not identify subtle changes in functioning and
that neuropsychological testing is often more sensitive to the neurotoxic effects of
chronic CO poisoning. Seven patients were included in this study. Each individual had
been exposed to CO intermittently or constantly over periods ranging from 3 weeks
to 3 years. Once the exposure was identified, each was sent for hyperbaric oxygen
(HBO) treatment. Each of the individuals was exposed to a minimum of 200 ppm CO.
Each individual was administered the CO Neuropsychometric Screening Battery, as
well as other specific neuropsychological tests including the WAIS-R, Trails A and B,
Finger Tapping Test, Logical Reasoning and Visual Reproduction from the Wechsler
Memory Scale, and Minnesota Multiphasic Personality Inventory (MMPI) or MMPI-
2. Six of the seven individuals received HBO therapy, ranging from 5 to 59 treatments.
The individual who received only five treatments was intolerant to the chamber and
that individual’s treatment was considered incomplete. One individual was included
in the study who did not receive any HBO treatment. Individuals who received HBO
treatment were tested every 2 weeks “until the psychometric tests reached a plateau
or returned to normal” (p. 557). The individual who did not receive HBO treatment
was tested at two months after the exposure stopped, and then after 10 months of
rehabilitation, was tested again at 1 year after the exposure stopped.
In addition to neuropsychological testing, a questionnaire was also completed
by each participant regarding his or her symptoms. The most common symptoms
experienced during the exposure (acknowledged by 50% or more of the group)
included problems with headaches, dizziness, motor tremors, difficulties with short-
term memory, sleep disturbance, cognitive set loss, anxiety, reading comprehension,
vision, gait/balance, muscle tremors, paresthesias, altered sense of smell, body aches,
tinnitus, and spatial disorientation. One participant who received ten HBO treatment
sessions reported significant resolution of their symptoms following treatment. One
individual who received 50 HBO treatments reported moderate resolution of their
symptoms. Three of the participants who received 19, 59, and 29 HBO treatments,
respectively, reported minimal or no functional improvement of their symptoms. The
individual who received five incomplete treatments reported only minimal improve-
ments in his symptoms. The individual, who received no HBO treatment, but ten
months of rehabilitation, reported that his condition improved significantly, but
that patient continued to report ongoing and multiple symptoms. In summary, five
of the seven (71% of the sample) reported minimal or no resolution of their symptoms
8417: “8417_c023” — 2007/9/11 — 12:14 — page 503 — #9

after a period of recovery from the exposure, HBO treatment, and/or rehabilitation
services.
With regard to the neuropsychological testing, the authors found that the CO
Neuropsycho-metric Screening Battery was of little value in identifying residual
cognitive deficits and recommended neuropsychological evaluation that is more
detailed. By the completion of testing, four of the seven participants continued to
demonstrate a significant split in their Verbal and Performance IQs (PIQs). Three of
those individuals had Verbal IQs (VIQs) significantly greater than PIQs. One had PIQ
significantly greater than VIQ. Three of the seven continued to demonstrate impair-
ments on the Trails B Test, a test of alternating attention and logical sequencing. Three
of the seven also demonstrated residual deficits in fine motor speed. One weakness
of this study is that there is no discussion of the impact of practice effects. It appears
that most individuals in this study were tested every 2 weeks and repeated exposure
to these tests can result in significant gains due to practice. The authors note that
ongoing problems with emotional lability, irritability, depression, and anxiety are
common sequelae of chronic CO poisoning.
Pinkston et al.15 conducted Positron Emission Tomography (PET) scans and
neuropsychological testing of two adult patients 3 years following a chronic CO
poisoning. The patients were both right-handed, middle-aged individuals who had
been married for many years. Both worked in professional occupations and they had
no history of prior psychiatric or neurologic conditions. They suffered exposure to
CO for a 3-year period due to faulty furnace exhaust/ducting.
Neuropsychological testing was conducted on both subjects four times over a
3-year period. The results of the testing indicated a significant anterior frontal lobe
syndrome. In addition, both individuals demonstrated frontal symptoms in their
activities of daily living, such as indecisiveness, mental passivity, and disorganiza-
tion. Both individuals experienced a significant vocational disability because of their
persistent and ongoing symptoms. Indeed, both individuals were rendered vocation-
ally disabled because of their residual deficits. In addition, both experienced losses in
their level of independence and both experienced difficulties with various activities
of daily living.
Both subjects demonstrated a similar pattern of hypometabolism on PET imaging.
Substantially reduced metabolism was evident in the orbital frontal and dorsal lateral
prefrontal cortex, as well as areas of the temporal lobe for both individuals. It was
determined that the individual scans were consistent with the patients’ presenting
symptoms and reduced level of functioning. While hypometabolism was evident in
various regions of the temporal lobe, both individuals were performing within normal
limits on memory tests by the time of their final evaluation. However, both individuals
were reporting and experiencing significant memory dysfunction in their day-to-day
activities and activities of daily living.
It is noted that one of the subjects received a more significant exposure, being in
the home for significant more prolonged periods of time. That subject demonstrated
more problems on neuropsychological testing and his PET scan demonstrated more
areas of significant hypometabolism. That individual also demonstrated more severe
behavioral/affective problems. Both patients ultimately developed epilepsy and were
begun on Depakote. The authors note that the development of a seizure disorder was
8417: “8417_c023” — 2007/9/11 — 12:14 — page 504 — #10

not surprising given the temporal and hippocampal dysfunction noted on the PET
imaging.
Hartman,16 presents a case study of a 65-year-old woman chronically exposed
to CO over a six-to-seven-year period. The exposure occurred because of improper
installation of a water heater. The woman had 2 years of college and a professional
degree. Her symptoms began within a year of the new water heater being installed. As
is typical for cases of chronic CO poisoning, her symptoms gradually emerged and
worsened over time. Physical symptoms that she developed during the exposure were
fatigue, muscle spasms, loss of muscle tone in her face, paresthesias, muscle fatigue,
migrating neuritis (sharp pains), problems with balance, her fingernails, and toenails
turned black, sleep disturbance, and one near-blackout episode. Other symptoms that
she experienced included depression, panic attacks, sleep disturbance, problems with
short-term memory, spatial disorientation, and difficulties with vision. By the end
of the exposure, she had severe body pain, even at rest. Also during the exposure
period, she experienced frequent urinary tract infections, other chronic infections,
and developed severe allergies, suggesting a possible compromise of her immune
system.
It was also noted during the exposure that house plants died and silverware turned
black quickly. It was noted that her symptoms did abate to some degree when she
would leave the house for several days at a time, but worsened upon returning home.
Once the exposure stopped, some symptoms resolved but some persisted, most not-
ably difficulty with her vision, short-term memory, allergies, and sleep disturbance. It
was also noted that she had become sensitive to a variety of chemicals and substances
(e.g., perfume).
The patient underwent serial neuropsychological testing. By the time she was
evaluated several years after the exposure stopped, she was functionally reporting
ongoing problems with language comprehension, verbal short-term memory, and
occasional episodes of spatial disorientation, mild emotional lability, and intermittent
sleep disturbance. She was continuing to note ongoing sensitivity to various chemicals
and substances (e.g., pesticides). The patient had also developed Crohn’s disease.
On neuropsychological testing, she was demonstrating persistent sensory/motor and
spatial integration deficits. There was a bilateral loss of her ability to discriminate
one versus two-point touch on her fingertips and some errors in fingertip localization.
Fine motor coordination on the Grooved Pegboard Test was severely impaired. The
patient’s ultimate DSM-III-R diagnosis was 294.80 Organic Mental Disorder, not
otherwise specified/probable CO exposure etiology.
Divine et al.17 present a case study of a 45-year-old woman chronically exposed
to CO for approximately 1 year. The exposure occurred because of a faulty furnace
at her place of employment where she worked as a cook. For approximately 1 year,
she experienced the following symptoms: “severe flu,” inability to walk in a straight
line, bumping into things, problems with balance, severe headache, fatigue, verbal
fluency, hearing problems, paresthesias, irritability, and facial pain. Her condition
was misdiagnosed as a sinus infection. She had been off work for a period of 5 days
and upon returning to work immediately became ill and contacted the gas company.
“Extremely high” levels of CO were identified in her work area, at which point she
left the premises. The exact CO concentration in her work space was not given.
8417: “8417_c023” — 2007/9/11 — 12:14 — page 505 — #11

Seventeen months following cessation of the exposure, many of her symptoms had
resolved. However, she was reporting ongoing problems with reading, writing, speak-
ing, verbal fluency, and dysarticulation. An MRI performed at 15-months after the
exposure ended was read as abnormal. The scan revealed multiple small lesions bilat-
erally in the basal ganglia. The lesions were more severe in the globitus palidus than in
the putamen. The radiologist concluded that the lesions were consistent with chronic
CO poisoning. The patient had a Bachelor’s degree and no prior neuropsychological
history. Neuropsychological testing conducted at 17-months after the end of exposure
revealed deficits in attention and concentration, learning and memory retrieval. Test-
ing also suggested problems with depression at that time. Behaviorally, during the
course of the evaluation, lapses in attention, perseverations, sequencing problems,
slight concreteness, and verbal fluency difficulties were noted. The authors concluded
that the testing was consistent with subtle frontal lobe dysfunction. Retesting was per-
formed 12-months later and similar deficits were found, suggesting persistent frontal
lobe dysfunction.
23.2 HELFFENSTEIN STUDY

23.2.1 SAMPLE AND EXPOSURE DATA
23.2.1.1 Admission Criteria
Participants in this study were 21 consecutively evaluated patients who had been
chronically exposed to CO. There were a number of criteria established for admission
to the study:
1. The full WAIS-III was administered as part of the evaluation

2. No prior psychological or psychiatric history requiring treatment
3. No prior neuropsychological history (e.g., no prior head injuries, toxic
exposures, or neurological illness)
4. No history of substance abuse
As all participants were involved in some type of active litigation, no patient

was admitted to the study if there was any indication of symptom magnification
or exaggeration. All 21 participants of this study performed satisfactorily on three
symptom validity tests. Eighty-one percent of the participants were administered
and passed the Tombaugh Test of Memory Malingering, Hiscock Digit Recognition
Test, and Rey II 15 Item Memory Malingering Test. The remaining 19% of the cohort
were administered and passed the Computerized Assessment of Response Bias, Word
Memory Test, and Tombaugh Test of Memory Malingering. In addition, the MMPI-2
profiles of all 21 participants were devoid of any indications of symptom magnification
or exaggeration. In addition, all 21 participants demonstrated behavioral indicators
of good effort during the neuropsychological testing process.
Participants were admitted to the study only if there was some documented evid-
ence of dangerous levels of CO in their living or work environments. In 14 cases,
measurements of CO in ppm had been made in the living or work space. For two of the
8417: “8417_c023” — 2007/9/11 — 12:14 — page 506 — #12

participants, COHb levels were obtained shortly after the exposures were identified.
For five of the participants high levels of CO were identified as being emitted from
a furnace or other appliance and there was a reasonable mechanism by which CO
entered the living or workspace of the individual.
23.2.1.2 Sample Demographics

The mean age of participants in this study was 39.6 years (range 16–78). The mean
level of education was 13.8 years (range 9–20 years of formal education). All parti-
cipants in this study were Caucasian. Eighteen were female and three were male. All
participants were right-handed. Nineteen of the participants were working full time
at the time of the exposure and two were full-time students.
23.2.1.3 Source and Location of Exposure

Fifteen of the 21 participants experienced chronic CO poisoning in their homes. Six of
the participants experienced CO exposure at work. Eleven of the participants exper-
ienced CO exposure because of a faulty furnace or boiler. Five were exposed as a
result of a faulty furnace and water heater combination. One experienced CO expos-
ure because of a faulty gas range and water heater combination. Two participants
experienced CO exposure as a result of faulty gas fireplace installation. One parti-
cipant experienced CO exposure because of a faulty water heater installation. One
participant experienced CO exposure because of auto exhaust fumes from a parking
garage entering his apartment.
23.2.1.4 Level of Carbon Monoxide Exposure

As noted above, in 14 of the cases included in this study, actual measurements of
CO in parts per million were made in the living or work space of the individual.
The average concentration of CO was 123 ppm, with a range of 13–467 ppm. It is
important to note that 100-ppm CO exposure over many hours will result in a COHb
saturation of 14%. Two of the participants had their COHb levels measured shortly
after the presence of CO was discovered. Regression (i.e., back) calculations were
performed to determine the COHb levels at the time the participant left the toxic
environment. On average, these two individuals had COHb saturations of 14.5% at
that time. For four of the participants in the study, a furnace and water heater at their
place of employment were housed in a confined space, which limited combustion
air. This created a situation where both the furnace and water heater had a tendency
to back draft. The furnace was producing 200+ ppm CO and the water heater was
producing 2000+ ppm CO. These readings were taken at the exhaust vents. When
back drafting occurred, CO was entering the work space of these individuals.
23.2.1.5 Duration and Frequency of Exposure

For each of the 21 participants, it was possible to estimate with some certainty the
duration of exposure. The mean duration was 28.9 months (2.4 years) with a range of
8417: “8417_c023” — 2007/9/11 — 12:14 — page 507 — #13

0.5–120 months. It was also possible to estimate with some certainty the frequency
of exposure. The average frequency of exposure was 6.3 days per week with a range
of 3–7 days per week.
23.2.1.6 Time Since Exposure

Each of the 21 participants was evaluated because they were experiencing persisting
symptoms, which were great enough to negatively impact their day-to-day, academic,
and/or work activities. Clearly, this sample represents the sub-set of chronically
exposed CO patients who do not make a full and complete recovery. Testing was
conducted an average of 46.8 months (3.9 years) postexposure. The range was 16–111
months postexposure (1.3–9.25 years).
23.2.2 BATTERY ADMINISTERED

Each participant was administered a comprehensive battery of neuropsychological
tests. Each individual received the Expanded Halstead–Reitan Neuropsychological
Test Battery, as well as most of the tests renormed for age, gender and education
by Heaton, Grant and Matthews.19 Each individual was administered the complete
WAIS-III, as well as the Peabody Individual Achievement Test (PIAT). In addition to
the above, the following tests were administered:
1. Complex ideation subtest of the Boston Diagnostic Aphasia Examination

2. Nonverbal agility and verbal agility subtest of the Boston Diagnostic
Aphasia Examination
3. Ruff Figural Fluency Test (Ruff FFT)
4. Hooper visual organization test
5. Padula visual midline screening test20
6. Line bisection test
7. Behavioral dyscontrol scale
8. Buschke verbal selective reminding test
9. Rey-Osterreith complex figure test
10. Paced Auditory Serial Addition Test (PASAT) (Levin version)
In addition to the above, each of the participants completed a MMPI-2 or Min-

nesota Multiphasic Personality Inventory-Adolescent (MMPI-A). For each of the
participants, the following index scores were generated:
1. Halstead Impairment Index (HII)

2. Average Impairment Rating (AIR)
3. Global Deficit Score (GDS)
4. General Neuropsychological Deficit Scale (GNDS)
For an earlier discussion of neuropsychological testing in a case study of

CO-poisoning by this author, see Helffenstein, 2000.18
8417: “8417_c023” — 2007/9/11 — 12:14 — page 508 — #14

23.2.3 NORMS UTILIZED

For the purpose of this study, the Heaton et al.19 demographic norms were used.
These norms are demographically corrected for age, gender, and education. In addi-
tion, for the purpose of this study, the performance levels proposed by Heaton et
al. were utilized. Table 23.4 shows the ranges of performance proposed by Heaton
et al. as well as the accompanying T-scores and percentile scores. Where nondemo-
graphically corrected scores (i.e., non-Heaton norms) are utilized, those norms will be
individually identified by level of performance and will be discussed using the Heaton
guidelines.
23.2.4 SYMPTOMS EXPERIENCED BY PARTICIPANTS DURING

CARBON MONOXIDE EXPOSURE
Table 23.5 is a summary of the most common symptoms reported by the participants
of this study during chronic CO poisoning. Consistent with prior studies of chronic
CO poisoning, people experienced multiple symptoms in multiple systems. Table 23.5
clearly shows that the symptoms involve a wide variety of physical, sensory-
motor, visual, cognitive, and affective/mood conditions. Penney3 discussed in some
detail misdiagnosis of chronic CO poisoning, and notes that common misdiagnoses
include chronic fatigue syndrome, viral/bacterial/pulmonary or gastrointestinal
infection, “rundown condition,” endocrine problem, immune deficiency disorders,
psychiatric/psychosomatic problems, allergies, and food poisoning.
As noted earlier in this chapter, prior studies have found that during chronic
CO poisoning, individuals typically experience multiple symptoms in multiple organ
systems. On average, the participants in the current study experienced 25 symptoms
during the chronic CO poisoning event, and, as noted by Dr. Penney, most of these
individuals received a variety of misdiagnoses during their exposure. Often, when
TABLE 23.4
Heaton Range of Performance on Neuropsychological Testing
(Heaton, et al.19 )
Level of Performance T-Scores Percentile Scores
Above average 55+ 68+
Average 45–54 30–67
Below average* 40–44 12–29
Mild impairment 35–39 6–13
Mild/Moderate impairment 30–34 3–5
Moderate impairment 25–29 1–2
Moderate/severe impairment 20–24 <1
Severe impairment 1–19 <1
∗ Some authors (e.g., Jarvis and Barth21 ) also refer to this range of scores as the
“Borderline” range.
8417: “8417_c023” — 2007/9/11 — 12:14 — page 509 — #15

TABLE 23.5
Symptoms Reported During Chronic Carbon
Monoxide Exposure, Helffenstein Study (N = 21)
Symptom % of Participants Reporting
Headaches 100
Fatigue 90
Dizziness 90
Mental confusion 90
Attention and concentration 90
Sleepiness 86
Irritability 86
Nausea 81
Muscle/joint aches 81
Short-Term memory 81
Shortness of breath 76
Sleep disturbance 76
Chest pain (tightness) 71
Word finding 71
Slow mental processing 71
Heart palpitations 67
Motor incoordination 67
Paresthesias 67
Cough 57
Balance problems 57
Anxiety 57
Light sensitivity 57
Depression 52
Motor weakness 52
Muscle spasms/tremors 52
Noise sensitivity 48
Diarrhea 48
Tinnitus 43
Vomiting 38
Sensitivity to chemicals 38
Blurry vision 38
Emotional lability 33
they did not respond to various treatments, their medical providers concluded that they
were likely suffering from a psychological or psychiatric disorder. Dr. Penney notes
that, because of a lack of training with regard to chronic CO poisoning, and human
CO poisoning generally, most health care providers have a low index of suspicion for
this condition, which results in “shockingly high rates of misdiagnosis.”
In addition, to the symptoms listed in Table 23.5, some percentage of the cohort
also experienced the following symptoms: constipation, blackouts, double vision,
decreased fine motor skills, spatial disorientation, problems with multitasking,
8417: “8417_c023” — 2007/9/11 — 12:14 — page 510 — #16

paraphasic errors in their speech, cognitive set loss, difficulties with planning and
organization, metallic taste in their mouth, hair loss, decreased hearing, bruising
easily, change in skin color, decreased night vision, abdominal pain, incontinence
of feces, incontinence of urine, hypersensitive sense of smell, and visual inattention
resulting in them bumping into things frequently.
23.2.5 PERSISTING SYMPTOMS REPORTED BY PARTICIPANTS

Each of the 21 participants in the study underwent a comprehensive clinical interview
with this author concerning their residual or persisting symptoms. Again, consistent
with prior studies, these individuals typically experience a wide range of persisting
symptoms. For the purposes of this study, the symptoms were clustered into per-
sisting physical symptoms (Table 23.6), persisting visual symptoms (Table 23.7),
persisting cognitive symptoms (Table 23.8), and persisting psychological/behavioral
symptoms (Table 23.9). Again, the norm was for each participant in the study to
experience multiple symptoms in all four areas of functioning.
As noted above, each of the participants of this study underwent a comprehensive
structured clinical interview. A list of possible symptoms was generated based on a
TABLE 23.6
Persisting Physical Symptoms, Helffenstein Study (N = 21)
Chemical sensitivity 100
Fatigue 100
Physical pain (e.g., joints, muscles) 95
Headaches 95
Muscle cramps/spasms 81
Motor incoordination 81
Phonophobia (noise sensitivity) 81
Impaired auditory gating (filtering background noise) 81
Temperature deregulation 81
Tinnitus 76
Shortness of breath 67
Paresthesias 67
Gait/balance problems 67
Dizziness 62
Motor Tremors 62
Altered sense of smell 52
Nausea/vomiting 43
Altered sense of taste 43
High blood pressure 29
Metallic taste 29
Compromised immune system 29
Vertigo 19
8417: “8417_c023” — 2007/9/11 — 12:14 — page 511 — #17

TABLE 23.7
Persisting Visual Symptoms, Helffenstein Study (N = 21)
Photophobia 90
Visual scanning deficit 90
Accommodation 86
Depth perception 86
Bumps into things 86
Perception of movement in peripheral vision 86
Eye fatigue 76
Veers off center 76
Blurry vision 71
Oscillopsia 24
Double vision 14
Decreased visual acuity at night 5
Note: Bumping into things more frequently and veering off center when walking or
driving in a straight line are common manifestations of hemispatial inattention.
TABLE 23.8
Persisting Cognitive Symptoms, Helffenstein Study
(N = 21)
Attention and concentration 100
Cognitive set loss 100
Short-term memory 100
Verbal fluency (word finding) 100
Slow speed of processing 100
Multitasking 90
Problem solving/decision making 90
Paraphasic errors in speech 86
Reading comprehension 86
Planning and organization 86
Math 81
Language comprehension 76
Spelling 71
Transposition errors 71
Spatial disorientation 71
Accessing remote memories 57
Writing skills 48
Memory confabulation 29
Memory contamination 19
Initiation 14
8417: “8417_c023” — 2007/9/11 — 12:14 — page 512 — #18

TABLE 23.9
Persisting Psychological/Behavioral Symptoms, Helffenstein
Study (N = 21)
Irritability 90
Depression 81
Sleep disturbance 76
Decreased libido 63
Emotional lability 52
PTSD/specific phobia 48
Appetite (inconsistent or decreased) 48
Generalized anxiety 24
Social isolation 33
Anhedonia 19
Decreased motivation 14
review of prior chronic CO poisoning outcome studies. Given the multiple possible
symptoms in multiple organ systems, this type of interview was highly preferable to
a more “open-ended” style interview.
23.2.5.1 Physical Symptoms

A review of the symptoms in Table 23.6 reveals that patients chronically exposed
to CO often have residual physical symptoms in a wide variety of organ systems
including auditory, gustatory, motor, sensory, vestibular, pulmonary, gastrointestinal,
and auto-immune. As discussed above, such a wide variety of symptoms can create
a significant diagnostic challenge for primary healthcare providers, especially if the
chronic CO poisoning has gone unidentified.
It is important to note that 100% of the sample reported persistent problems with
fatigue. In most instances participants in the study found that they became more tired
more quickly whether performing physical or cognitive activities. All were reporting
reduced stamina and endurance to varying degrees. In some instances participants
in the study were already utilizing a psychostimulant medication (e.g., Concerta,
Adderall, Provigil, or Ritalin) at the time of their evaluation. In some instances a
prescription for a psychostimulant was recommended as part of the individual’s treat-
ment plan. It was this author’s impression that the fatigue experienced by the majority
of the sample was organically based and did not relate to the patient’s psychological
status.
23.2.5.2 Visual Symptoms

A variety of visual problems have been well documented following moderate to severe
acute CO poisoning. The Penney studies A and B,3 as well as the CO Support study,11
document that “vision problems” in general are common following chronic CO
8417: “8417_c023” — 2007/9/11 — 12:14 — page 513 — #19

poisoning. However, neither study describes in detail what types of vision problems
are likely to occur. The current study helps to define in some detail the constellation
of visual problems, which are common following chronic CO poisoning.
We now understand that acquired brain injuries of all types can result in a wide
variety of vision problems, which as a group has come to be referred to as a Post-
Traumatic Vision Syndrome (PTVS). Indeed, PTVS is extremely common following
any neurotrauma, including toxic encephalopathy. Many of the functional vision
problems identified in Table 23.7 relate to organically based problems, such as:
Esotropia—one eye turns inward

Exotropia—one eye turns outward
Convergence insufficiency—ability to aim the eyes accurately at a target and
then track the target as it moves closer or further away
Accommodation—ability to accurately focus on a target, sustain focus, and
change focus when focal length changes
Binocular vision—eyes working/tracking together as a team so that accom-
modation, convergence, and eye alignment are all integrated
Hemispatial inattention—inconsistent ability of the brain to attend to one visual
field
Nystagmus—shaking of the eyes, which the individual cannot control
For a more detailed understanding of these types of visual deficits, the reader is
referred to Politzer.21
A brief description of some of the vision problems listed in Table 23.7 follows:
Photophobia: Sensitivity to bright sunlight and at times flourescent lighting.

Visual scanning deficit: Patients will often report that their eyes do not track
effectively. They will, for example, lose their place when they are reading
or miss things they are looking for.
Accommodation: Patients will often report that their eyes are slow to refocus
when shifting their gaze from near to far, or vice versa.
Depth perception: Patients will often report that they have difficulty accurately
judging distance. This problem often presents itself in driving situations.
Hemispatial inattention: Reduced attention to one visual field. Two of the more
frequently reported functional problems related to hemispatial inattention is
bumping into things more frequently and veering off center when walking
or driving in a straight line.
Unilateral saccades: Refers to a unilateral eye-tracking motion. One eye
will scan to the side independently, at which time the individual will often
perceive motion in their peripheral vision when there is nothing there.
Oscillopsia: Refers to the perception that stationary objects are moving.
8417: “8417_c023” — 2007/9/11 — 12:14 — page 514 — #20

When a patient who has been chronically exposed to CO reports a constellation of

visual problems consistent with a PTVS, referral for further evaluation and treatment
is indicated. Referral to a neuro-optometrist or neuro-ophthalmologist familiar with
these types of brain-related vision problems can often be beneficial. In this author’s
experience, neuro-optometric rehabilitation (i.e., vision therapy) and/or use of yoked
prisms can both be helpful in ameliorating many of the above vision problems.
23.2.5.3 Cognitive Symptoms

Table 23.8 presents a list of the persisting cognitive problems most commonly repor-
ted by the participants of this study. As a group, the participants in this study felt that
their residual cognitive deficits constituted some of the most disabling and debilitat-
ing aspects of their injury. For the most part, these symptoms can be placed into four
general categories: executive functions, memory skills, academic abilities, and lan-
guage skills. The ultimate findings of this study suggest that executive and memory
functions are cognitive abilities most impacted by chronic CO poisoning. In addition,
the findings of this study suggest that left hemisphere functions are typically more sus-
ceptible than right to the effects of chronic CO poisoning. Thus, there is a high degree
of consistency between the participants’ self-reported persisting cognitive symptoms
and the ultimate findings of this study.
23.2.5.4 Psychological / Behavioral Symptoms

Table 23.9 presents a summary of the most common persisting psychological and
behavioral symptoms reported by the participants in this study. Organically based
mood disorders and personality change are extremely common following moder-
ate to severe acute CO poisoning. The CO Support study11 found that 38.5% of
their sample reported persisting problems with depression, 49.3% reported persisting
problems with personality/emotional problems, 29.2% reported persisting problems
with panic attacks, and 47.7% reported ongoing problems with sleep disturbance. The
Penney A and B studies also found that a significant percentage of the sample reported
persisting problems with depression and mood disturbance. Again, the current study
has elaborated on these findings.
Ninety percent of the sample reported ongoing problems with irritability.
Specifically, the participants often reported that they would become easily upset and
angry over small things that would not have bothered them previously. As a group, they
tended to report reduced coping and stress management abilities. In many instances,
this problem was so significant that it would negatively impact personal, family, and
work relationships.
Persisting problems with depression was reported by 81% of the sample. By far,
this was the most common affective or mood problem reported by the participants in
this study. Commonly reported clinical symptoms of depression included feelings of
sadness, reduced energy, emotional lability, decreased motivation, anhedonia, social
isolation, reduced appetite, and decreased libido. However, it is important for clini-
cians to understand and take into account that many of these classic indicators of
depression may occur independently of depression because of organic brain injury.
8417: “8417_c023” — 2007/9/11 — 12:14 — page 515 — #21

As a group, the participants in this study responded well to use of an antidepress-

ant medication in the Select Serotonin Re-uptake Inhibitor (SSRI) class. In some
instances, the participants were already utilizing an SSRI-type antidepressant at the
time of their evaluation or began use of such an antidepressant following the evalu-
ation. In each case, the participants reported that use of an SSRI medication typically
resulted in improvements in their mood, a reduction in their irritability, improvement
in their stress tolerance, improved stamina, and endurance, and reduced emotional
lability.
Sleep disturbance was also a commonly reported persisting symptom by the par-
ticipants in this study. Seventy-six percent were experiencing some disruption of
their sleep pattern. Most of the sample had difficulty getting to sleep and an even
greater percentage had difficulty staying asleep. Many of the participants in the study
reported they would frequently wake up during the night and had difficulty getting
back to sleep. Many also reported a sense that their sleep was now “restless.” In most
cases, it was evident that the sleep disturbance was compounding the individual’s
organically based fatigue. For most of the participants in the study, they were either
already utilizing a sleep aid at the time of their evaluation or one was recommended
following their assessment.
Anxiety was another commonly reported symptom by the participants in this
study. Forty-eight percent of the participants were diagnosed with post-traumatic
stress disorder (PTSD) or a specific phobia directly related to the CO poisoning
event. Another 24% were experiencing more generalized anxiety. In many cases, this
related to performance anxiety. The individuals were well aware that they were not
functioning as well as they had prior to the exposure and were thus experiencing some
anxiety associated with their reduced performance.
23.2.6 NEUROPSYCHOLOGICAL TESTING OUTCOME

23.2.6.1 Index Scores
As noted above, four index scores were generated as part of each participant’s test
battery.22 These included the HII, AIR, GDS, which is generated from the Heaton
battery of tests19 and the GNDS. A summary of performance on three of these index
scores is presented in Table 23.10. The HII, which incorporates seven of the individual
test scores, was found to be the least sensitive of the index scores when demograph-
ically corrected norms were utilized. Sixty percent of the participants scored within
normal limits on this index. Fifteen percent of the sample had a HII score in the below
average/borderline range. Only 25% had HII scores in the impaired range. The HII
score ranges from 0.0 to 1.0, with a higher score reflecting more significant impair-
ment. Dr. Ralph Reitan, who originally developed the HII, has historically reported
that a score of 0.5 or greater on this index is reflective of an acquired brain injury.
In this study the mean HII score was 0.395 (range 0.1–0.9). The modal HII score
was 0.3. Only 25% of the sample had HII scores of 0.5 or greater. Therefore, whether
using the demographically corrected norms or the traditional cut-off score of 0.5, only
one-fourth of the participants had impaired scores on this index.
8417: “8417_c023” — 2007/9/11 — 12:14 — page 516 — #22

TABLE 23.10
Index Score Performance: Percentage of Participants Scoring in Each Range
of Performance
Mild/ Moderate/
Above BA/ Mild Moderate Moderate Severe
Average Average Bdl. Impairment Impairment Impairment Impairment
H.I.I. 15 45 15 10 10 5 0
(7 Scores)
A.I.R. 23.8 33.3 9.5 14.3 14.3 4.8 0
(12 Scores)
G.D.S 5.5 11 28 44.5 11 0 0
(20 Scores)
Note: Demographically corrected norms have never been generated for the GNDS.
B.A = Below Average; Bdl. = Borderline
The AIR, which takes into account 12 of the individual test scores, was somewhat
more sensitive to the effects of chronic CO poisoning. Scores on this index range
from 0.0 to 5.0 with higher scores suggesting more impairment. Thirty-three point
four percent of the sample performed in the impaired range, 9.5% performed in the
below average/borderline range, and 57.1% performed within normal limits when
demographically corrected norms were utilized. The mean AIR score was 1.14 with
a range of 0.5–3.0.
The GDS, which in this study took into account 20 of the demographically cor-
rected scores, was even more sensitive to the effects of chronic CO poisoning. The
Boston Naming Test was not administered as part of the test battery and, therefore,
only 20 of the 21 tests that comprise the GDS were utilized. In this sample, 55.5%
of the participants had GDS scores in the impaired range, 28% had scores in the
borderline/below average range, and only 16.5% had scores that were within normal
limits. The mean GDS score was 0.634 with a range of 0.0–1.40.
The GNDS takes into account 42 of the individual test scores. The GNDS produces
a score ranging from 0 to 116 with higher scores representing more significant cerebral
dysfunction. When Drs. Reitan and Wolfson23 first developed this index, they found
that a cut-off score of 26 best differentiated a group of known moderately to severely
brain-injured individuals from a group of nonbrain injured individuals. In this study,
the mean GNDS score was 26.67 with a range of 16–47. Eleven (or 52%) of the
participants had GNDS scores of 26 or greater. Therefore, 48% of the sample had
GNDS scores that were below the recommended cut-off score of 26.
In summary, it appears that the GDS is the index score most sensitive to the effects
of chronic CO poisoning. It is important to note, however, that three of the participants
had scores within normal limits on all of the index scores. Two of the participants
had scores within normal limits on three of the indexes, and only one below average/
borderline score. Therefore, as is true for other neurological conditions, these index
scores appear to be sensitive to generalized moderate to severe cerebral dysfunction,
8417: “8417_c023” — 2007/9/11 — 12:14 — page 517 — #23

but may be insensitive to more subtle or localized cerebral dysfunction. For example,
when the GNDS was originally developed, a cut-off score of 26 was found to be
the most effective in differentiating a group of nonbrain-injured individuals from a
group of moderate to severely impaired brain-injured individuals. Indeed, when a
cut-off score of 26 was utilized, this cut-off score only misidentified 8% of moderate
to severely brain-injured individuals as nonbrain-injured (sensitivity 92%) and only
misidentified 10% of nonbrain-injured individuals as brain injured (specificity 90%).
In the current study, 52% (11/21) of the participants had GNDS scores of 26 or greater.
Forty-eight percent (10/21) of the participants had scores of 25 or less. As with other
neurological populations (e.g., mild traumatic brain injury) where the individual’s
cognitive deficits may be more subtle or localized, the traditional GNDS score of
26 is not recommended for patients who have been chronically exposed to CO. As
a result, it would be more important to consider the individual’s GNDS score in
comparison to the nonbrain-injured control group and the results of the current study.
It is also important to note that the Global Deficit Scale is the only index score that
incorporates performance on tests of short-term memory. The GDS takes into account
performance on tests of both verbal and visual short-term memory. As memory deficits
are by far the most common following chronic CO poisoning, this index score would
be especially important to consider.
23.2.6.2 IQ Tests
WAIS-III Full-Scale, VIQ and PIQ results are presented in Table 23.11. Also presented
are the Verbal Comprehension Index (VCI), Perceptual Organization Index (POI),
Working Memory Index (WMI), and Processing Speed Index (PSI) score results. In
summary, 19.1% of the sample had impaired Full Scale IQs; 14.3% had Full Scale
IQs in the below average/borderline range; and 66.6% had Full Scale IQs within
normal limits. Therefore, approximately one-third of the sample obtained Full Scale
IQs, which were below expectation or clearly impaired.
TABLE 23.11
WAIS-III I Q Testing: Percentage of Participants Scoring in Each Range of
Performance
Mild/ Moderate/
Above Mild Moderate Moderate Severe
Average Average BA/Bdl. Impairment Impairment Impairment Impairment
FSIQ 2.4 42.8 14.3 9.5 4.8 4.8 0
VIQ 23.8 23.8 28.6 14.0 9.5 0 0
PIQ 38.1 33.3 19 0 0 9.5 0
VCI 38.1 28.6 9.5 19 0 4.8 0
POI 42.9 47.6 0 0 4.8 4.8 0
WMI 4.8 19 33.3 14.3 19 4.8 0
PSI 38.1 23.8 9.5 9.5 4.8 4.8 9.5
8417: “8417_c023” — 2007/9/11 — 12:14 — page 518 — #24

With regard to VIQ, 23.5% of the sample had impaired VIQs; 28.6% had VIQs
in the below average/borderline range; and 47.6% had VIQs within normal limits.
VIQ subtest score performance is presented in Table 23.12. With regard to PIQ,
9.5% of the sample had PIQs in the impaired range; 19% had PIQs in the below
average/borderline range; and 71.4% had PIQs that were within normal limits. PIQ
subtest score performance is presented in Table 23.13. Contrary to studies with other
neurologically impaired populations, VIQ appeared to be more significantly impacted
by chronic CO poisoning than PIQ. It is important to note that one-third or more of
the participants obtained borderline or impaired scores on five of the seven verbal
TABLE 23.12
WAIS-III Verbal Subtest Scores: Percentage of Participants Scoring in Each
Range of Performance
Mild/ Moderate/
Mild Moderate Moderate Severe
Above BA/ Impair- Impair- Impair- Impair-
Average Average Bdl. ment ment ment ment
Vocabulary 33.3 42.9 14.3 4.8 4.8 0 0
Letter-Number 4.8 28.6 19 19 23.8 0 0
Sequencing
Arithmetic 19 38.1 9.5 19 4.8 0 0
Digit Span 0 33.3 33.3 14.3 14.3 4.8 0
Information 19 2.9 9.5 19 4.8 0 0
Similarities 23.8 47.6 14.3 14.3 0 0 0
Comprehension 19 47.6 19 14.3 0 0 0
TABLE 23.13
WAIS-III Nonverbal Subtest Scores: Percentage of Participants Scoring in
Each Range of Performance
Mild/ Moderate/
Mild Moderate Moderate Severe
Above BA/ Impair- Impair- Impair- Impair-
Average Average Bdl. ment ment ment ment
Digit Symbol 23.8 33.3 9.5 19 9.5 0 4.8
Symbol Search 52.4 14.3 9.5 9.5 9.5 0 4.8
Picture 28.6 33.3 19 9.5 4.8 4.8 0
Arrangement
Picture 33.3 33.3 19 4.8 4.8 0 0
Completion
Block Design 47.6 42.9 0 4.8 0 4.8 0
Matrix 47.6 38.1 9.5 4.0 0 0 0
Reasoning
8417: “8417_c023” — 2007/9/11 — 12:14 — page 519 — #25

subtests. These included the Comprehension, Information, Digit Span, Arithmetic,

and Letter-Number Sequencing subtests.
Difficulties with the Digit Span and Letter-Number Sequencing subtests most
likely relate to problems with sustained auditory attention and concentration.
Problems with the Arithmetic subtest most likely relate to a working memory deficit.
Difficulties with the Information subtest most likely relate to a reduced ability to
access previously learned information (i.e., memory retrieval) and difficulties with
the Comprehension subtest most likely relate to problems with verbal reasoning and
problem solving. To some degree, each of these problems can be related to frontal
lobe dysfunction, which is found to be common in cases of chronic CO poisoning in
other studies.
Within the PIQ subtests there were three subtests where one-third or more of
the sample performed in the below average to impaired ranges. This included the
Picture Completion subtest, Digit Symbol Coding, and Picture Arrangement. Picture
Completion assesses attention to salient visual detail. Digit Symbol Coding assesses
sustained visual attention and concentration as well as eye-hand coordination. Picture
Arrangement assesses logical planning and organization.
More than one-third of the sample performed in the below average to impaired
range on the Symbol Search subtest as well. This test evaluates sustained visual atten-
tion and accurate visual discrimination. Difficulties with planning and organization
on the Picture Arrangement subtest may relate to aspects of sequencing and organiz-
ation as mediated by the frontal lobes. Exposure to CO is known to negatively impact
a variety of visual and visual perceptual functions that may account for the sensitivity
of the Picture Completion and Symbol Search subtests to this mechanism of injury.
Difficulties with the Digit Symbol-Coding task most likely relate to difficulties with
sustained visual attention and concentration, which again may be frontally mediated.
Table 23.14 lists in descending order the percent of the sample who scored in the
below average to impaired range on each of the WAIS-III subtests. It is important to
note that three of the four subtests that appear to be most sensitive to the effects of
chronic CO poisoning are verbal subtests. This may account for why VIQ appears to
be more significantly affected by chronic CO poisoning than PIQ. It is also important
to note that all 13 of the WAIS-III subtests, which were administered as part of this
study can be negatively affected by chronic CO poisoning.
As part of the current investigation, VIQ/PIQ difference or “split” was also studied.
For the purposes of this study, a significant VIQ/PIQ split was considered to be 15
points or one standard deviation. Such a difference occurs in the nonbrain injured
population less than 10% of the time. In the current study 9 of the 21 participants or
42.9% had VIQ/PIQ splits of 15 points or greater (range 15–22 points). For three of
the participants (14.3% of the sample) their VIQ was greater than the PIQ. For six
of the participants (28.6% of the sample) their PIQ was greater than their VIQ. Thus,
twice as many participants had a PIQ that was significantly greater than their VIQ.
This is yet another indication to suggest that VIQ appears to be more susceptible to
the effects of chronic CO poisoning.
There is some evidence to suggest that blood flow is slightly greater in the left
cerebral hemisphere for individuals who are right-hand dominant. This may be due
to greater oxygen requirements of the dominant hemisphere. Therefore, the hypoxia
8417: “8417_c023” — 2007/9/11 — 12:14 — page 520 — #26

TABLE 23.14
WAIS-III Subtests: Percentage of
Participants who Scored in the
Below Average to Impaired Range
Digit Span 66.7

Letter-Number Sequencing 66.6
Digit Symbol-Coding 42.8
Arithmetic 42.8
Picture Arrangement 38.1
Information 38.1
Picture Completion 33.4
Symbol Search 33.3
Comprehension 33.3
Similarities 28.6
Vocabulary 23.8
Matrix Reasoning 14.3
Block Design 9.6
created by chronic CO poisoning may have a greater impact on the left hemisphere,
as its oxygen requirements are greater. In addition, since blood flow is greater in the
left hemisphere, more CO is likely to be “downloaded” into this hemisphere of the
brain resulting in greater injury. Both of these theories could explain why VIQ was
more impacted than PIQ in this study.
As a group, the sample demonstrates greater subtest variability than would
typically be seen in nonbrain-injured individuals. Utilizing the Heaton et al.18
demographically corrected norms, the average subtest variability was 2.79 stand-
ard deviations with a range of 1.6–3.9 SDs. The average Scaled Score variability was
7.2 with a range of 5–10.
In addition, scores from the WAIS-III that were analyzed included the Digit
Symbol-Coding, incidental learning, free recall, and pairing tasks. Thirty-six per-
cent of this sample had below average to impaired scores on the free recall task;
25.5% had borderline to impaired scores on the pairing task.
Table 23.15 presents the percentage of participants who scored in the Below
Average to Impaired Range on each of the summary IQ Scores and summary Index
Scores. The results of this study support the notion that WMI Score and the VIQ
score are most susceptible to the effects of chronic CO poisoning. More than half
of the sample scored in the below average to impaired range on these two summary
scores. It is important to note that the WMI score is based on scores from the subtests
that measure working memory with verbal material. This is yet another finding to
suggest that left hemisphere function is especially vulnerable to the effect of chronic
CO poisoning.
8417: “8417_c023” — 2007/9/11 — 12:14 — page 521 — #27

TABLE 23.15
WAIS-III Summary Scores: Percentage of
Participants who Scored in the Below
Average to Impaired Range
Working Memory Index (WMI) 71.4

Verbal IQ (VIQ) 52.1
Processing Speed Index (PSI) 38.1
Full Scale IQ (FSIQ) 33.4
Verbal Comprehension Index(VCI) 33.3
Performance IQ (PIQ) 28.5
Perceptual Organization Index (POI) 9.6
23.2.6.3 Halstead–Reitan Tests

Table 23.16 presents the percentage of the cohort who scored within each level of
performance on the scores that can be generated from the standard Halstead–Reitan
Neuropsychological Test Battery. Table 23.17 identifies the percentage of the sample
who scored in the below average to impaired range on each of the standard Halstead–
Reitan tests. Of these tests it appears that the Category Test, Tactual Performance
Test (TPT) Memory, Rhythm Test, TPT-Localization, Trails B, and Speech-Sounds
Perception Test are the most sensitive to the effects of chronic CO poisoning. The
Category Test is a measure of the patient’s ability to generate solutions to a new
and ambiguous problem. The TPT Memory and Localization tests are both measures
of incidental memory. The Rhythm Test and Speech-Sounds Perception Test are
both sensitive to sustained auditory attention. The Trails B test measures alternating
attention and logical sequencing abilities. Consistent with prior studies, it appears that
frontal functions such as sustained attention and concentration and problem solving
abilities as well as memory (as measured by the TPT Memory and Localization tasks)
are most sensitive to the effects of chronic CO poisoning.
23.2.6.4 Memory Testing

Each participant was administered two tests of verbal learning and retention and
two tests of nonverbal (visual) learning and retention. Tests used were the Story
Memory Test and Figure Memory Test as renormed and administered via the Heaton
et al.18 guidelines. In addition, the Buschke Verbal Selective Learning Test utilizing the
Larabee norms and the Rey–Osterreith Complex Figure Test utilizing the Meyers and
Meyer’s norms were used. Table 23.18 summarizes the percentage of the sample that
scored in each level of performance for each of the above memory tests. Table 23.19
summarizes the percentage of the sample scoring in the below average to impaired
range for each of the above memory tests. It appears well established in the literature
that CO poisoning typically has a profound effect on short-term memory abilities. The
literature is also clear that the temporal lobes and hippocampus are brain structures
that are highly vulnerable to the effects of CO poisoning. These are brain structures
8417: “8417_c023” — 2007/9/11 — 12:14 — page 522 — #28

TABLE 23.16
Halstead–Reitan Tests: Percentage of Participants who Scored in each Range
of Performance
Mild/ Mod./
Category Test 30 10 25 20 10 5 0
TPT-Dominant 50 30 10 5 5 0 0
TPT- 35 5 10 0 0 0 0
Nondominant
TPT-Both 35 30 20 10 0 0 5
TPT-Memory 5 35 20 25 10 0 5
TPT-Location 15 30 20 20 15 0 0
Rhythm Test 9.5 3.3 23.8 19 14.3 0 0
Speech-Sounds 4.3 38.1 28.6 9.5 9.5 0 0
Perception
Tapping- 28.6 47.6 9.5 0 4.8 4.8 4.8
Dominant
Tapping- 45 49 5 10 5 0 0
Nondominant
Trails A 28.6 38.1 9.5 19 0 0 4.8
Trails B 23.8 28.6 28.6 4.8 9.5 0 4.8
Grooved
Pegboard
Dominant 33.3 38.1 4.8 9.5 9.5 0 4.8
Nondominant 25 55 10 10 0 0 0
that play a major role in short-term memory. The results of this study are consistent
with prior studies, suggesting that memory skills are highly vulnerable to the effects
of chronic CO poisoning. The results of this study clearly indicate that verbal and
nonverbal learning and retention are all extremely vulnerable.
23.2.6.5 Academic Testing

Each participant of this study was administered the original PIAT of reading recog-
nition, reading comprehension, and spelling subtests, as this was the version of the
PIAT which Dr. Heaton used in his norming process. The math section from the
PIAT-R utilizing those published norms was also administered. Table 23.20 presents
the percent of the sample that scored within each level of performance. Table 23.21
summarizes the percentage of the sample scoring in the below average to impaired
range. Results suggest that reading, math, and spelling skills are all vulnerable to the
effects of chronic CO poisoning, with spelling being the most sensitive.
8417: “8417_c023” — 2007/9/11 — 12:14 — page 523 — #29

TABLE 23.17
Halstead–Reitan Tests:Percentage
of Participants who Scored in the
Category Test 60
TPT-Memory 60
Rhythm Test 57.1
TPT-Localization 55
Trails B 47.7
Speech-Sounds Perception 47.6
TPT-Nondominant 35
TPT-Both 35
Trails A 33.3
Pegs-Dominant 28.6
Tapping-Dominant 23.9
Tapping-Nondominant 20
Peg-Nondominant 20
TPT-Dominant 20
TABLE 23.18
Memory Testing: Percentage of Participants that Scored in Each Range of
Performance
Mild/ Moderate
Mild/ Moderate Moderate/ Severe Severe
Above BA/ Impair- Impair- Impair- Impair- Impair-
Average Average Bdl. ment ment ment ment ment
Story memory
Learning 14.3 19 4.8 9.5 33.3 9.5 0 9.5
Retention 14.3 14.3 0 4.8 33.3 9.5 9.5 14.3
Figure Memory
Learning 19 14.3 23.8 23.8 14.3 4.8 0 0
Retention 14.3 23.8 28.6 0 9.5 19 4.8 0
Buschke VSR
Learning (CLTR) 5 15 25 25 0 30 0 0
30-min. Recall 10 20 10 10 15 0 35 0
Rey-osterreith
Complex Figure
Initial recall 10 20 10 10 15 15 20
30-min. Recall 10 25 5 15 29 9 25 9
8417: “8417_c023” — 2007/9/11 — 12:14 — page 524 — #30

TABLE 23.19
Memory Tests: Percentage of Participants
Scoring in the Below Average to Impaired
Range
Buschke (CLTR) learning 95
Story Memory 4-h delay recall 71.4
Buschke 30-min delay recall 70
Rey-Osterreith IR (Learning) 70
Figure Memory-learning 66.7
Story Memory-learning 66.6
Rey-Osterreith-30-min delay recall 65
Figure Memory-4-h delay recall 61.9
TABLE 23.20
Academic Testing: Percentage of Participants that Scored in Each Range of
Performance
Mild/ Moderate/
Math 15 45 15 15 10 0 0
Reading 40 35 20 5 0 0 0
Recognition
Reading 30 25 5 15 15 5 5
Comprehension
Spelling 15 25 20 30 5 0 5
TABLE 23.21
Academic Testing: Percentage of Par-
ticipants Scoring in the Below Aver-
age to Impaired Range
Spelling 60
Reading Comprehension 45
Math 40
Reading Recognition 30
23.2.6.6 Visual–Visual Perceptual Testing

The literature clearly shows that a wide variety of visual and visual perceptual
functions are vulnerable to the effects of CO poisoning. As part of the battery admin-
istered to each of the participants, a number of tests were administered to evaluate
8417: “8417_c023” — 2007/9/11 — 12:14 — page 525 — #31

constructional praxis, including copy of a key figure, the Spatial Relations score from
the Heaton battery, and an analysis of the patient’s copy of the Rey-Osterreith Complex
Figure. For the key drawing which was administered as part of the Reitan-Indiana
Aphasia Screening Test, scoring guidelines outlined in Reitan and Wolfson24 were
utilized. About 27.6% of the sample had key constructions that were considered to
be within normal limits. About 28.6% produced keys that were considered to be
borderline and for 23.3% of the cohort, their key drawings were considered to be
clearly impaired. For the Rey–Osterreith Complex Figure test scoring guidelines
presented in the manual25 were utilized. Sixty percent of the sample had Rey draw-
ings that were clearly impaired. Ten percent produced Rey drawings that were in
the below average to borderline range. Thirty percent of the sample-produced draw-
ings were considered to be within normal limits. It is important to note that 45%
of the sample produced Rey copies were poorly planned and organized, suggesting
a frontal quality to the construction. Fifty-seven percent of the sample had Spatial
Relations scores that were in the below average to impaired range. Forty-two point
eight percent had Spatial Relations scores that were considered to be within normal
limits. In summary, constructional praxis appears to be a cognitive ability, which is
vulnerable to the effects of chronic CO poisoning, and this problem can be observed
in a number of visual-constructional tasks.
As part of the test battery, each participant was administered the Line Bisection
Test,26 a measure of hemispatial inattention. There was a suggestion of a possible
subtle hemispatial inattention evident in 15% of the participants. The results were
suggestive of possible left hemispatial inattention for two of the participants and a
right hemispatial inattention for one participant. Amuch more pronounced hemispatial
inattention was evident in the results of two of the participants, one suggesting a clear
right hemispatial inattention, and one suggesting a clear left hemispatial inattention.
The Padula Visual Midline Screening Test20 was also sensitive to identifying
hemispatial inattention. This task found that 14% of the sample was experiencing a
left visual inattention, 19% were experiencing a right visual inattention, and that one
participant tended to miss visual stimuli in both visual fields (i.e., bifixation). When
bifixation occurs, the individual will tend to miss important stimuli in one visual field
if there is competing stimuli in the opposite visual field. Reitan and Wolfson24 present
a case of a woman who suffered severe CO poisoning and loss of consciousness for
5 h following an attempted suicide by running her car engine in a closed garage.
The individual demonstrated a right homonymous hemianopsia, suggesting that CO
poisoning can result in lateralized visual defects. The results of this study suggest that
chronic carbon CO has the potential to cause less severe visual field problems in the
form of hemispatial inattention.
All participants were administered the Digit Vigilance Test, a test of visual scan-
ning speed and accuracy. Thirty percent of the participants obtained scores in the
below average to impaired range on the speed component of the Digit Vigilance
Test, whereas 70% had scores that were within normal limits. Half of the sample
had visual scanning accuracy scores in the below average to impaired range. Fifty
percent had scores that were within normal limits. These results suggest that both
visual scanning speed and accuracy are vulnerable to the effects of chronic CO
poisoning.
8417: “8417_c023” — 2007/9/11 — 12:14 — page 526 — #32

TABLE 23.22
Visual/Visual Perceptual Tests: Percentage of Participants Scoring in Each
Range of Performance
Mild/
Above BA/ Mild Moderate Moderate
Average Average Bdl. Impairment Impairment Impairment
Digit Vigilance Test
Speed 20 50 15 5 5 5
Accuracy 30 20 20 25 0 5
Spatial Relations 38.1 4.8 23.8 9.5 14.3 9.5
Hooper VOT 15 60 20 5 0 0
TABLE 23.23
Visual/Visual Perceptual Tests: Percentage
of Participants who Scored in the Below
Spatial Relations Score 57
Digit Vigilance Test
Accuracy score 50
Speed score 30
Hooper Visual Organization Test 25
The Hooper Visual Organization Test27 was also administered to 20 of the 21 par-
ticipants. One participant performed in the mildly impaired range and four performed
in the below average to borderline range on this test of complex visual organization.
Seventy-five percent of the sample tested performed within normal limits. Table 23.22
summarizes the percentage of participants who scored in each range of performance
for the Digit Vigilance Test, Hooper Visual Organization Test, and Spatial Relations
Score. Table 23.23 summarizes the percentage of participants who scored in the below
average to impaired range on these measures.
23.2.6.7 Speed of Information Processing

Twenty of the 21 participants were administered the PASAT, a test of complex attention
and speed of auditory information processing. Table 23.24 summarizes the percentage
of the sample score at each level of performance. In summary, 10% of the participants
were not able to proceed past the practice trial of this task. For 30% of the sample,
the task was discontinued after Trial 2 as a result of poor performance and high levels
of frustration. Twenty percent went on to obtain below average or impaired scores
on trial three. Forty percent of the sample (8 out of 20 participants) were able to
perform within normal limits on all four trials of this task. Sixty percent had some
8417: “8417_c023” — 2007/9/11 — 12:14 — page 527 — #33

TABLE 23.24
Paced Auditory Serial Addition Test: Percentage of Participants Scoring in
Each Range of Performance
Mild/ Moderate/
Above BA/ Mild Moderate. Moderate Severe
PASAT
T1 5.5 38.9 5.5 5.5 16.7 16.7 11.1
(N = 18)
T2 11.1 38.9 5.5 11.1 11.1 5.5 16.7
(N = 18)
T3 8.3 58.3 8.3 16.6 0 8.3 0
(N = 12)
T4 0 66.7 25 0 8.3 0 0
(N = 12)
Note: The PASAT was not attempted with one subject. Two subjects were unable to proceed past the
practice trials. The test was stopped after Trial 2 for six subjects as a result of high levels of frustration
and poor performance.
level of difficulty. This would suggest that speed of auditory information processing
is a cognitive ability commonly affected by chronic CO poisoning.
The best measure of visual information processing speed contained in the battery
was the PSI score from the WAIS-III. Thirty-eight percent of the sample performed
in the below average to impaired range. Sixty-two percent performed within normal
limits.
In summary, 60% of the cohort had some level of difficulty with auditory
information processing speed, whereas only 38% demonstrated problems with visual
information processing speed. This would be another indicator that left hemisphere
functions are more susceptible to the effects of chronic CO poisoning.
23.2.6.8 Other Motor Skills

Fine motor speed and dexterity, as measured by the Finger Tapping Test and Grooved
Pegboard, were discussed under the heading of Halstead-Reitan Tests. Several addi-
tional tests of motor skills were also administered as part of the battery. Grip strength
was measured by use of the Hand Dynamometer. Grip strength with the dominant
hand was in the borderline to impaired range for 33.4% of the sample. Sixty-six point
six percent had scores that were within normal limits. Grip strength with the nondom-
inant hand was in the borderline to impaired range for 15% of the sample. Eight-five
percent had grip strength within normal limits for the non-dominant hand.
Executive motor skills were evaluated by the Behavioral Dyscontrol Scale. This
is a set of tasks designed to measure motor programming and sequencing abilities.
Forty percent of the sample performed within normal limits on this task. Twenty-five
8417: “8417_c023” — 2007/9/11 — 12:14 — page 528 — #34

percent of the sample obtained scores that were below average/borderline, 20%
demonstrated mild impairment, 10% demonstrated mild/moderate impairment; 5%
demonstrated moderate impairment. Of those who demonstrated scores in the
below average/borderline and impaired ranges, 60% demonstrated motor per-
severation type errors. Fifty-five percent demonstrated motor disinhibition type
errors. Sixty-five percent demonstrated right-left confusion errors. Fifty-five per-
cent demonstrated echopraxic errors and 85% demonstrated motor sequencing
errors. In summary, executive motor dysfunction, as measured by motor program-
ming and sequencing abilities, appears to be quite common following chronic CO
poisoning.
Executive motor functioning was also evaluated by the nonverbal agility and
verbal agility subtest of the Boston Diagnostic Aphasia Examination. Ten percent of
the sample demonstrated impairment on the Verbal Agility subtest. A striking 90%
demonstrated impairment on the nonverbal agility subtest, indicating that oral-motor
dyspraxia is a common sequelae of chronic CO poisoning. Table 23.25 summarizes
the percentage of participants who scored in the below average to impaired range on
the supplemental motor tasks.
23.2.6.9 Miscellaneous Tests of Executive Function

As part of the battery, we attempted to administer the Wisconsin Card Sorting Test to
each participant. Four of the subjects (19% of the sample) were unable to complete
the task. Of the 17 subjects who were able to complete the Wisconsin Card Sorting
Test, 23.6% had scores in the below average to impaired ranges. Therefore, 38%
of the total sample had some degree of difficulty with the Wisconsin Card Sorting
Test, a test of mental flexibility and problem solving. Twenty-one percent of the
sample demonstrated from one to three losses of cognitive set on the Wisconsin Card
Sorting Test.
Verbal fluency was measured by the Thurstone Word Fluency Test. Thirty percent
of the sample performed in the below average to impaired range. Seventy percent
performed within normal limits.
TABLE 23.25
Other Motor Functions: Percentage of
Participants who Scored in the Below
Hand Dynamometer
Dominant hand 33.4
Nondominant hand 15
Behavioral Dyscontrol Scale 60
Boston Diagnostic Aphasia Exam
Verbal Agility subtest 10
Nonverbal Agility subtest 90
8417: “8417_c023” — 2007/9/11 — 12:14 — page 529 — #35

TABLE 23.26
Miscellaneous Tests of Executive Function: Percentage of Participants Scoring
in each Range of Performance
Mild/ Moderate/
WCST 65 11.8 11.8 0 11.8 0 0
(Note: Four [4] participants were unable to successfully complete this task.)
Thurstone VFT 40 30 10 5 5 0 10
Ruff FFT
• Unique design
Score 40 30 20 5 5 0 0
Error Ratio
Score 35 35 15 15 0 0 0
• Perseverations 25 60 0 10 5 0 0
Stroop
Interference
Score 25 50 20 0 5 0 0
Behavioral
Dyscontrol
Scale 0 40 25 20 10 5 0
Figural fluency was measured by the Ruff FFT. Thirty percent of the sample
had below average to impaired scores on the Unique Design score, a measure of
figural fluency. Thirty percent had below average to impaired Error Ratio scores, a
measure of planning efficiency. Fifteen percent demonstrated a significant number of
perseverations on this task.
Response inhibition was measured by the interference score of the Stroop. Twenty-
five percent of the sample scored in the below average to impaired range on this
task. For each of these miscellaneous tests of executive function, the percentage
of participants scoring in each range of performance is presented in Table 23.26.
Table 23.27 presents the percentage of participants who scored in the below average
to impaired range on each of the measures of executive function. Clearly the PASAT
(speed of auditory information processing and complex attention) and the BDS (motor
programming and sequencing) were the most sensitive to the effects of chronic CO
poisoning.
23.2.6.10 Language Comprehension

Language comprehension was evaluated utilizing the Complex Ideation subtest of the
Boston DiagnosticAphasia Examination. Forty-three percent of the sample performed
in the below average to impaired range. Fifty-seven percent performed within normal
limits. The percentage of participants scoring in each range of performance is
presented in Table 23.28.
8417: “8417_c023” — 2007/9/11 — 12:14 — page 530 — #36

TABLE 23.27
Executive Function Tests: Percentage
of Participants who Scored in the
PASAT 62
Ruff FFT Unique Design score 30
Ruff FFT Error Ratio score 30
Stroop Interference 25
WCST 23.6
Thurstone Word Fluency Test 20
TABLE 23.28
Language Comprehension: Percentage of Participants Scoring in Each Range of
Performance
Mild/ Moderate/
Complex 33.3 23.8 14.3 14.3 9.5 0 4.8
Ideation
subtest
of
BDAI
Note: 43 percent of the sample scored in the Below Average to the Impaired Range.
23.2.7 MINNESOTA MULTIPHASIC PERSONALITY INVENTORY-2

The MMPI-2 was completed by 19 of the subjects in this study. The other two par-
ticipants were adolescents and completed the MMPI-A. The MMPI-2 and MMPI-A
are the most commonly used of all paper-and-pencil personality tests and are widely
utilized by neuropsychologists as part of a comprehensive neuropsychological test
battery. As part of the MMPI-2, the patient reports current or longstanding symptoms
or personality traits by responding true or false to 567 statements. The inventory
produces 14 Clinical Scales, 9 Restructured Clinical Scales, 15 Content Scales, and a
variety of alternate subscales are also available. Four validity scales provide inform-
ation about the patient’s approach to this inventory, as well as the validity of the
profile. The validity scales of the MMPI-2 also provide information about symptom
magnification/exaggeration, as well as symptom denial. The profile produced by the
ten Clinical Scales is compared to those of normal control subjects as well as psychi-
atric patients with a wide variety of diagnoses. In addition to providing information
8417: “8417_c023” — 2007/9/11 — 12:14 — page 531 — #37

about the patient’s personality traits and characteristics, the MMPI-2 also provides
information about the patient’s current emotional and psychological status.
The MMPI-2 is an objective test of personality and emotional status and should
be viewed as a self-report symptom checklist. The MMPI-2 is a “here and now”
questionnaire and the patient is asked to report their current symptoms or problems. As
a result, the effects of actual impairment may be manifested in the patient’s response to
items associated with physical, cognitive, emotional, or behavioral changes, which
may relate directly to their injuries and resulting impairment. Indeed, 111 of the
MMPI-2 items relate directly to symptoms commonly associated with a wide range of
neurological disorders. With regard to this issue, Lezak et al.28 state, “Since so many
MMPI items describe symptoms common to a variety of neurological disorders, self-
aware and honest patients with these symptoms may produce MMPI profiles which
could be misinterpreted as evidence of psychiatric disturbance even when they do
not have a psychiatric or behavioral disorder” (p. 749). Lezak notes that the most
commonly elevated MMPI-2 scales for individuals who have sustained some type
of neurological injury are Scale 1, Hypochondriasis; Scale 2, Depression; Scale 3,
Hysteria; Scale 7, Psychesthenia; and Scale 8, Schizophrenia. Cripe29 also notes that
neurologic patients as a group tend to elevate on Scales 1, 2, 3, 7 and 8. With regard
to this issue, Dr. Cripe states, “The basic reason neurologic patients elevate on Scales
1, 2, 3, 7, and 8 is that the inventory is loaded with many items that can be endorsed
by a neurologic patient because of their neurologic disorders and the resulting real
world problems rather than necessarily due to psychiatric disorders, emotional factors
and maladjustment” (p. 296). Dr. Cripe goes on to note, “The safest and most logical
assumption to make if a medical or neurologic patient elevates on Scales 1, 2, 3,
7, or 8 is that the patient has some awareness of his/her problems and is reporting
the problems within the limitations and constrictions of the MMPI item pool. The
patients simply see themselves as having difficulties related to their medical problems
and are trying to communicate this awareness. The fact that they are aware of the
problems does not necessarily indicate that they are distraught about the problems or
emotionally maladjusted” (p. 300).
A similar pattern of elevations has been identified in patient populations exposed
to a variety of neurotoxins. Morrow et al.30 identified elevations on Scales 1, 2, 3,
and 8 in workers seen in an occupational health clinic with complaints of cognitive
problems following exposure to a variety of neurotoxins. Bowler et al.31 identified a
1, 2, 3, 8, 7 MMPI profile among women workers exposed to organic solvents.
As noted above, 19 of the 21 participants in the current study were administered
the MMPI-2. The mean T-scores for all of the validity and clinical scales, as well
as the T-score range of these 19 profiles, are presented in Table 23.29. Consistent
with prior studies, as a group patient’s chronically exposed to CO demonstrated
elevations on Scales 3, 1, 2, 7, and 8 in order of descending T-score elevation.
In each case the elevation was above the clinical cut-off level of T = 65. A group
MMPI-2 profile of these 19 individuals is presented in Figure 23.1. The elevations on
Scales 1 and 3 reflect the group’s on-going report of physical concerns and symptoms.
Their elevation on Scale 2 likely reflects feelings or other symptoms of depression.
However, this scale also contains a number of items related to cognitive dysfunction,
which may well relate to the chronic CO poisoning. The elevation on Scale 7 most
8417: “8417_c023” — 2007/9/11 — 12:14 — page 532 — #38

TABLE 23.29
MMPI-2 Validity and Clinical Scales
Summary (N = 19)
Scale T-Score Mean T-Score Range
VRIN 47.26 38–66
TRIN 57.63 50–73
F 60.37 41–96
F(b) 61.52 42–108
F(p) 46.37 41–73
L 53.79 38–76
K 52.16 35–65
S 52.21 33–65
1. HS 79.11 49–101
2. D 76.05 46–105
3. Hy 87.63 49–120
4. Pd 59.47 37–76
5. Mf 48.32 30–74
6. Pa 62.53 45–85
7. Pt 74.68 51–103
8. Sc 72.06 44–114
9. Ma 54.31 37–72
10. Si 52.53 34–80
100
87.63
80 79.11
74.68
T-scores
76.05
72.06
65T 62.53
60.37
60 59.47
54.31
53.79 52.16
52.23
48.32
40
L
Hs
Hy
Pd
Mf
Pa
Pt
Sc
Ma
Si
FIGURE 23.1 Group MMPI-2 validity and clinical scales profile of 19 patients chronically
exposed to carbon monoxide—Helffenstein Study.
likely reflects ongoing feelings of anxiety. Scale 8 contains many items associated
with organic brain injury.
Table 23.30 presents the mean T-scores and T-score ranges for the Content Scales.
As a group, the patients chronically exposed to CO were elevated on the Anxiety
8417: “8417_c023” — 2007/9/11 — 12:14 — page 533 — #39

TABLE 23.30
MMPI-2 Content Scales Sum-
mary (N = 19)
ANX 66.58 43–89
FRS 49.79 35–72
OBS 58.37 41–75
DEP 58.31 39–83
HEA 73.32 49–96
BIZ 52.42 39–70
AMG 51.31 31–72
CUM 44.21 38–64
ASP 42.05 33–56
TPA 50.10 36–64
LSE 57.79 44–92
SOD 54.42 32–82
FAM 47.73 36–65
WRK 62.37 50–82
TRT 56.63 39–95
TABLE 23.31
MMPI-2 Health Concerns Subscales Summary
(N = 19)
Subscale T-Score Mean T-Score Range
Gastrointestinal symptoms 61.3 43–86
Neurological symptoms 75.5 45–114
General health concerns 72.1 48–89
and the Health Concerns Content Scales. This would be consistent with the clinical
interviews conducted with each of these patients in which they discussed their anxiety
and concern regarding their current and future health status. Table 23.31 contains the
mean T-scores and T-score ranges for the three Health Concerns subscales. As a group,
the patients chronically exposed to CO demonstrate an elevation on the General Health
Concerns subscale and the Neurological Symptoms subscale. Thus, as a group, these
individuals appear to be anxious and concerned about their health status, particularly
their neurological symptoms.
Table 23.32 presents the mean T-scores and T-score ranges for the MMPI-2
Restructured Clinical Scales. Figure 23.2 presents a group MMPI-2 Restructured
Clinical Scales profile for the 16 participants in this study for which the Restructured
Clinical Scales were calculated. As a group, they were elevated on RC1, RC2, and
RCd, in descending order of profile elevation. The elevation on RCd (Demoraliza-
tion) is interpreted as an indication of overall emotional discomfort that the individual
8417: “8417_c023” — 2007/9/11 — 12:14 — page 534 — #40

TABLE 23.32
MMPI-2 Restructured Clinical
Scales Summary (N = 16)
RCd 65.06 50–55
RC 1 75.0 55–99
RC 2 67.31 42–100
RC 3 42.0 33–66
RC 4 47.5 37–60
RC 6 47.5 41–70
RC 7 53.38 32–84
RC 8 56.44 39–82
RC 9 44.75 34–61
100
80
75
T-scores
67.31
65.06
60
56.44
47.5 53.38
42 47.5
44.75
40
RCd dem
RC1 som
RC2 lpe
RC3 cyn
RC4 asb
RC6 per
RC7 dne
RC8 abx
RC9 hpm
dem = Demoralization cyn = Cynicism dne = Dysfunctional Negative Emotion

som = Somatic Complaints asb = Antisocial Behavior abx = Aberrant Experiences
lpe = Low Positive Emotion per = Ideas of Persecution hpm = Hypomanic Activation
FIGURE 23.2 Group MMPI-2 Restructured Clinical (RC) scales profile of 16 patients
chronically exposed to CO-Helffenstein Study
is experiencing.32 Individuals with similar elevations often describe themselves as

discouraged, generally demoralized, insecure, and pessimistic. In addition, they will
also frequently report poor self-esteem. As a group, this population also appears
to elevate most notably on RC1 (Somatic Complaints). Individuals with elevations
on RC1 are typically experiencing a variety of physical and other health problems.
An elevation on RC1 does not necessarily indicate that the individual is excessively
preoccupied with bodily concerns but may simply be accurately reporting their ongo-
ing physical symptoms associated with an actual injury. This scale contains many
8417: “8417_c023” — 2007/9/11 — 12:14 — page 535 — #41

TABLE 23.33
DSM-IV GAF Scores, Helffenstein Study
(N = 21)
Level of Functioning % of Sample
Mild symptoms (GAF 61 → 70) 19
Moderate symptoms (GAF 51 → 60) 76
Serious symptoms (GAF 41 → 50) 5
items related to fatigue, motor weakness, and chronic pain, all of which are common
sequelae of chronic CO poisoning. Also as a group, this sample elevated on RC2 (Low
Positive Emotions). Individuals who elevate on this scale typically are reporting a
lack of positive emotions in their lives and may also relate to feelings of depression,
unhappiness, and a sense of failure or helplessness.
Table 23.33 presents a summary of the DSM-IV GlobalAssessment of Functioning
(GAF) scores for the participants in this study. Clearly, the majority met the criteria
for “Moderate Symptoms” or moderate difficulty in social, occupational, and school
functioning.
23.2.8 VOCATIONAL OUTCOME

Nineteen of the 21 participants of this study were working full time at the time that
they were exposed to CO. Two of the participants were students. The results of the
study clearly documents the residual cognitive deficits, inconsistencies, or relative
weaknesses associated with chronic CO poisoning. The study also documents that
fatigue is a sequelae of CO poisoning that was experienced by all of the subjects. To
some degree, all reported that their stamina and endurance for physical and cognitive
activities had been compromised. To some degree, all subjects reported that they
became more tired more quickly whether performing physical or cognitive activities.
For the majority of the subjects of this study, it appeared that to a large degree the
fatigue that they were experiencing was organically based. In some cases, sleep
disturbance, chronic pain, and depression were serving to exacerbate further their
fatigue.
Table 23.34 presents vocational outcomes for the 19 participants of the study
who were working at the time of exposure. Only one of the participants was able
to return to his prior job on a full-time basis. However, this individual owned and
personally managed two film processing plants on opposite sides of the city where
he lived. Following the exposure, he found he was unable to manage successfully
both plants. To some degree, this related to a cognitive inability to manage effectively
all aspects of both plants. To some degree, it also related to the residual fatigue that
he experienced postexposure. As a result, he was forced to close one of his plants,
which led to a significant reduction in his earning potential. At the time of follow-up,
it was learned that this patient had made the decision to sell his remaining plant and
to seek alternate employment. The patient made this decision because he had lost so
8417: “8417_c023” — 2007/9/11 — 12:14 — page 536 — #42

TABLE 23.34
Vocational Outcome (N = 19 Working At
Time of Exposure)
Return to prior job full time 5% (1/19)*
Return to prior job part time 0% (0/19)
Return to lower level job full time 21% (4/19)
Return to lower level job part time 37% (7/19)
Opted to retire 11% (2/19)**
Total and permanent vocational disability 26% (5/19)
∗ This individual owned his own business but in order to
effectively manage his business postexposure, he had to close

one of his two plants which significantly reduced his earning
potential.
∗∗ These two individuals were close to retirement but would
have worked longer had it not been for their CO positioning
many commercial contracts postexposure that the business was no longer financially
viable. Loss of the contracts was attributed by this author directly to the sequelae of
his CO poisoning.
None of the participants who were working at the time of the exposure returned
to their prior employment on a part-time basis. However, this author suspects this is a
fairly common vocational outcome for this population. If an individual is working in
a job that is over-learned, they may well be able to continue performing that job from
a cognitive standpoint postexposure. If that individual experiences ongoing problems
with fatigue, then return to his prior job on a part-time basis would certainly be a
viable possibility.
Four of the 19 participants who were working fulltime at the time of the exposure
were able to return to work in a cognitively less challenging (i.e., lower level) job on
a full-time basis. For these individuals, while they were experiencing some ongoing
problems with fatigue, they were able to continue to work full-time, but from a
cognitive standpoint were unable to perform successfully the job that they were doing
prior to the exposure.
One of the more striking findings of the study is that 7 of the 19 individuals who
were working full-time at the time of the exposure (37%) were able to return to work,
only in less cognitively complex jobs on a part-time basis. For this subset of the study,
the individual’s residual cognitive deficits and fatigue combined to impact negatively
their vocational functioning in two separate ways.
Two of the participants of the study who were working full-time at the time of their
exposure made the choice to retire postexposure. Each of these individuals was close
to retirement but would have worked longer had it not been for the CO poisoning. In
each case, they made the decision to retire because of the combined effect that their
cognitive deficits and fatigue was having on their vocational functioning.
Another striking finding of the study is that five of the 19 individuals who were
working full-time at the time of the exposure (26%) were totally and permanently
8417: “8417_c023” — 2007/9/11 — 12:14 — page 537 — #43

vocationally disabled because of the exposure. In each case, the individual’s residual
cognitive dysfunction, fatigue and other deficits combined to render them totally and
permanently disabled from a vocational standpoint. None of these individuals was
able to maintain successfully substantial gainful work activity postexposure. This
figure is consistent with the results of the CO Support Study11 in which 32% of their
cohort were totally and permanently disabled. In summary, all 19 of the individuals
who were working full-time at the time of their chronic CO poisoning experienced
some degree of vocational disability postexposure.
23.2.9 SUMMARY OF FINDINGS

Table 23.35 summarizes the 32 measures from the neuropsychological test battery that
were most sensitive to the effects of chronic CO poisoning. That is, at least 38% (or
slightly more than one-third) of the participants obtained scores in the below average to
the impaired range on these measures. A review of these measures finds that they tend
to cluster in a meaningful fashion. Twelve of the 32 measures (37.5%) relate to exec-
utive functions. These functions include sustained attention and concentration, speed
of information processing, alternating attention, and logical sequencing, generating
solutions to new and ambiguous problems, planning and organization, and executive
motor skills, such as motor programming and sequencing, and oral motor dyspraxia.
Eleven of the 32 measures (34.4%) relate to some aspect of short-term memory func-
tioning. This includes various measures of verbal and non-verbal (visual) learning
and retention, as well as incidental memory and working memory. Analyzing the data
in this manner clearly suggests that executive functions and memory functions are
most susceptible to the effects of chronic CO poisoning.
It is also important to note that 7 of the 32 measures (21.9%) relate to some aspect
of language or academic skills. Two of the measures (6.3%) relate to some aspect of
vision or visual perceptual functions.
As noted in the introduction to this chapter, Gale et al.1 identified hypoperfusion
in the frontal and temporal lobes on SPECT studies of patients that had experienced
acute CO poisoning. This frontal-bilateral temporal pattern appears to be fairly well
established in the literature associated with acute CO poisoning. Results of the current
study would suggest that these same regions of the brain are most vulnerable to the
effects of chronic CO poisoning. Pinkston et al.15 identified hypometabolism in the
frontal regions of the brain in two individuals chronically exposed to CO. Their
neuropsychological testing identified a variety of deficits in executive functioning.
Executive dysfunction is frequently associated with frontal lobe involvement. In that
same study, the participants were reporting significant ongoing problems with short-
term memory. That study also identified hypometabolism in the temporal regions
of the brain, which would be consistent with the types of memory problems the
patients were reporting. This finding is consistent with several known aspects of brain
physiology. The blood brain barrier is more permeable in the frontal and temporal
regions, which would result in more uptake of CO into these regions. In addition, the
frontal and temporal regions of the brain are less able to compensate for a hypoxic
event. Thus, we would expect that these regions would be more vulnerable to the
effects of chronic CO poisoning, which is supported by the current study.
8417: “8417_c023” — 2007/9/11 — 12:14 — page 538 — #44

TABLE 23.35
Combined Battery: Percentage of Par-
ticipants Scoring in the Below Aver-
age to Impaired Range
Buschke (CLTR) learning 95
Oral Motor Dyspraxia 90
Story Memory 4-h delay recall 71.4
Working Memory Index (WAIS-III) 71.4
Buschke 30-min delay recall 70
Rey-Osterreith IR (Learning) 70
Digit Span (WAIS-III) 66.7
Figure Memory—learning 66.7
Story Memory—learning 66.6
Letter-Number Sequencing (WAIS-III) 66.6
Rey-Osterreith 30-min delay recall 65
PASAT 62
Figure Memory—4-h delay recall 61.9
Category test 60
TPT—Memory 60
Spelling (PIAT) 60
Rhythm Test 57.1
Spatial Relations score 57
TPT—Localization 55
Verbal IQ 52.1
DVT—Accuracy 50
Trails B test 47.7
Speech Sounds Perception test 47.6
Reading Comprehension (PIAT) 45
Language Comprehension 43
Digit Symbol—Coding (WAIS-III) 42.8
Arithmetic (WAIS-III) 42.8
Math (PIAT) 40
Processing Speed Index (WAIS-III) 38.1
Picture Arrangement (WAIS-III) 38.1
Information (WAIS-III) 38.1
A review of Table 23.35 also suggests that left hemisphere functions are more
susceptible to the effects of chronic CO poisoning. As noted earlier in the chapter, there
is literature to suggest that, for right-hand dominant individuals, blood flow is slightly
greater in the left hemisphere. As noted earlier in this chapter, all 21 participants of
this study were right-hand dominant. Again, the theory is that because of greater blood
flow in the left hemisphere this resulted in greater uptake of CO into that hemisphere
which resulted in greater injury to left hemisphere functions. In addition, the left
hemisphere would be expected to be more susceptible to the effects of a hypoxic event
given its greater demand for oxygen. Heuser and Mena33 conducted SPECT studies
8417: “8417_c023” — 2007/9/11 — 12:14 — page 539 — #45

on 72 right-handed adults who had been exposed to a wide variety of neurotoxic

chemicals including solvents, pesticides, CO, chlorine gas, polychlorinated biphenyls
(PCBs), formaldehyde, herbicides, and heavy metals. Each participant of the study
was experiencing persisting cognitive dysfunction including problems with short-
term memory. The SPECT studies were abnormal for all participants when their
scans were compared to normal controls. Heuser and Mena33 state, “Bilateral, often
asymmetrical, impairment of perfusion was found, mostly in the frontal, temporal and
parietal lobes. This hypoperfusion was predominantly left-sided in young patients and
predominantly right-sided in the elderly” (p. 813). The findings of this study suggest
that the left cerebral hemisphere is more vulnerable to the effects of a wide range
of neurotoxins in right-hand dominant, nonelderly adults. Clearly, this represents a
subset of the population totally consistent with the population of the current chronic
CO poisoning study.
23.3 CASE STUDY

23.3.1 PATIENT DEMOGRAPHICS
The patient (H.H.) was one of the two adolescents included in the above study sample.
At the time of her initial neuropsychological evaluation, she was 16 years old. She is
Caucasian and is right-handed. At the time of her original evaluation, she was in the
tenth grade in high school.
23.3.2 EXPOSURE INFORMATION

H.H. was chronically exposed to CO in her home. A new furnace had been installed
in the family home and fitted with a natural gas orifice. The home had liquid propane
(LP) gas, which required a different orifice. This error led to over-firing of the burner
and incomplete combustion of the gas, producing large amounts of CO (1500 ppm
CO found in the furnace exhaust). Inspection of the furnace revealed pinhole cracks
in the heat exchanger, which allowed CO to enter the living space around the furnace.
In addition, it was determined that there was passive infiltration of CO into the home
because of CO filtering back through the water heater exhaust vent. On testing, 30 ppm
CO was found in the furnace room and 33 ppm CO was found being emitted from the
registers in the home. It is likely that the CO exposure occurred over a 7-year period
and began when H.H. was 7 months old, continuing until she was approximately 7
years old.
23.3.3 EDUCATIONAL HISTORY

H.H. and her mother reported that she had not had any major difficulty learning and
had been an A/B student all of her school career. However, they are aware that each
year school was becoming more difficult and that she was having to study harder to
maintain good grades. She had never had any behavioral or social problems. At the
time of her initial evaluation, H.H. reported that her goal was to complete a college
degree and then pursue some type of higher education in the healthcare field. H.H. had
8417: “8417_c023” — 2007/9/11 — 12:14 — page 540 — #46

never been diagnosed with any type of learning disability nor had she ever received
special education or remedial services.
23.3.4 PERSISTING SYMPTOMS

During the course of her clinical interview, H.H. reported a variety of persisting
physical symptoms and problems. This included occasional right leg spasms, daily
headaches, tinnitus, photophobia, difficulty filtering out background noise, temper-
ature deregulation, shortness of breath, cognitive and physical fatigue, and chemical
sensitivity. Cognitively, she was aware of ongoing problems with attention and con-
centration, cognitive set loss, short-term memory, verbal fluency, reduced speed of
processing, difficulties with problem solving, spatial disorientation, and occasional
problems with initiation. Visually, she was aware of difficulties with visual scanning,
depth perception and photophobia with headaches. At the time of her evaluation, she
was utilizing Celexa, 20 mg per day, as an antidepressant medication. She denied any
feelings of depression at the time of testing. She was acknowledging some inconsistent
motivation and appetite.
23.3.5 RESULTS FROM INITIAL TESTING

On her original neuropsychological evaluation, H.H. demonstrated a pattern of deficits
suggestive of executive dysfunction. On the testing, she demonstrated inconsistent
sustained attention and concentration abilities, which ranged from the 8th to the
95th percentiles. Figural fluency was mildly to moderately impaired at the 3rd per-
centile. Verbal fluency was in the borderline range at the 21st percentile. Speed of
auditory information processing was inconsistent, ranging from the 12th to the 48th
percentile. Cognitive flexibility was in the borderline range at the 18th percentile.
She demonstrated a severe oral motor dyspraxia, as well as other problems with
motor programming and sequencing. Left temporal/hippocampal involvement was
suggested by verbal learning, ranging from the 2nd to the 21st percentile. Some right
temporal/hippocampal involvement was suggested by a mild impairment of her abil-
ity to learn new complex visual information at the 12th percentile. Therefore, her
initial test results did suggest a frontal/bilateral temporal pattern of dysfunction. In
addition, on her original testing, she demonstrated a mild constructional dyspraxia
on two separate tasks and she also showed deficits in the areas of math and incidental
memory.
With regard to the pattern of deficits identified on testing, the following observa-
tion was made in her narrative neuropsychological report, “Her pattern of weaknesses
and deficits is subtle but nevertheless consistent with chronic CO poisoning. From her
report and a review of school records, it appears that she is compensating well, but she
reports it is becoming more difficult each year to maintain the same grades. The areas
particularly affected by CO are the frontal and bilateral temporal lobes of the brain,
which are areas affected in H.H. Further, these areas of the brain do not fully develop
until a person is at least 21 years of age. Therefore, the full extent of H.H.’s difficulties
may not be currently evident. In addition, it is common that, when someone is in school
and they can devote extensive amounts of time to homework, they can compensate for
8417: “8417_c023” — 2007/9/11 — 12:14 — page 541 — #47

their residual cognitive deficits. Unfortunately, when cognitive demands increase at

the college level and in the work world and it is no longer possible to spend extended
periods of time to compensate for residual cognitive deficits, her subtle problems on
testing may become more evident at a later date.” Although her cognitive deficits,
inconsistencies, and relative weaknesses were rather subtle at age 16, our concern
was that over time executive and memory function deficits would become more not-
able when compared to her demographically similar peer group. Because of injury
to the frontal and temporal regions of the brain, our concern was that these functions
would not develop normally. As a result, a neuropsychological re-evaluation was
recommended at a later date.
23.3.6 NEUROPSYCHOLOGICAL RE-EVALUATION

23.3.6.1 Circumstances of Re-evaulation
H.H. underwent a neuropsychological re-evaluation when she was 19 years old.
She had graduated from high school with a cumulative grade point average of 3.7
(A-minus average). At the time of her re-evaluation, she had successfully completed
1 year of college (32 semester hours) and her grade point average was 2.8 on a 4.0
scale. Her major was elementary education. H.H. was clear in reporting that, since
the time of her last evaluation, she has experienced greater and greater difficulty
cognitively and academically. At the time of her re-evaluation, she was extremely
concerned about her ability to successfully complete her college degree and function
successfully in the world of work.
23.3.6.2 Self-reported Symptoms

H.H. was continuing to experience a wide variety of persistent physical symptoms,
including ongoing problems with headaches, restless leg syndrome, tinnitus, prob-
lems with auditory gating, increasing problems with fatigue (most likely due to the
increased cognitive demands of college), shortness of breath on minor exertion, and
left-sided weakness. She continued to be sensitive to a wide variety of chemicals and
substances. Visually, she was noting ongoing problems with double vision, blurry
vision, visual scanning, depth perception, accommodation, photophobia, and eye
fatigue. From a cognitive standpoint, she was reporting increasing problems with
attention and concentration, cognitive set loss, multitasking, short-term memory, new
learning, and cumulative memory. She was continuing to note a variety of language-
based problems, including difficulties with verbal fluency, reading comprehension,
paraphasic errors in her speech, and language comprehension. As her math courses
became more difficult, she has noted increasing problems with math. In addition,
as speed demands and time pressures increased in college, she was becoming more
aware of slowed speed of information processing. At the time of her re-evaluation,
she was acknowledging minimal feelings of depression, most significantly related to
the cognitive problems she encountered in college and concern about her long-term
academic and vocational future. Since the time of her last evaluation, she developed
panic attacks and was experiencing increased problems with emotional lability. Since
8417: “8417_c023” — 2007/9/11 — 12:14 — page 542 — #48

the time of her last evaluation, she began participating in individual psychotherapy
and was being followed by a psychiatrist. She was utilizing Celexa as an antidepress-
ant medication and Clonazepam on an as-needed basis for anxiety. Her psychiatrist
attributed her affective and mood disturbance directly to the CO exposure.
23.3.6.3 Comparison of Test Scores

A detailed comparison of H.H.’s re-evaluation test scores was made to her test scores
obtained as part of her original evaluation. A clinically significant change was con-
sidered to be any increase or decrease of 0.7–0.9 standard deviations. A statistically
significant change was considered to be any increase or decrease of 1.0 or more
standard deviations. Her performance remained essentially the same on 41 of the
individual neuropsychological test scores. She demonstrated three clinically signific-
ant gains and ten statistically significant improvements on testing. She demonstrated
11 clinically significant declines and 18 statistically significant declines when her re-
evaluation test scores were compared to her prior test scores. Thus, she demonstrated
13 gains versus 29 significant declines. As part of her re-evaluation, it was noted that,
while specific scores did not fall into the impaired range or even the borderline range
on the re-evaluation, the fact that she demonstrated a significant decline in cognitive
functioning over time was clinically significant. It was felt that this finding suggested
that these were cognitive abilities whose development had not “kept pace” with her
demographically similar peers. The relative declines, which she demonstrated from
her prior testing, clustered almost exclusively into three areas: (1) Executive/frontal
functions; (2) Memory functions; and (3) Academic/Language functions. There were
also several declines in motor/sensory functions and visual/visual-perceptual func-
tions. Therefore, the areas of cognitive functioning where H.H. demonstrated her
most significant relative decline were areas of functioning found to be most vulner-
able to the effects of chronic CO poisoning in the current study. It is also important
to note that H.H. demonstrated a notable decline in all seven of the summary scores
generated as part of the WAIS-III testing when compared to her demographically
similar peers.
23.3.6.4 Results from Re-evaluation

As noted in the body of this chapter, the index scores generated as part of the test
battery are not always sensitive to the more subtle or localized effects of chronic
CO poisoning. Indeed, H.H.’s scores on all four of the indexes used as a measure
of generalized cognitive functioning were within normal limits. However, her per-
formance on the WAIS-III was suggestive of some residual and more subtle cognitive
dysfunction. Her VIQ was in the borderline range at the 27th percentile. Her VCI
score was also in the borderline range at the 27th percentile. Her WMI score was at the
low end of the average range at the 31st percentile. When utilizing demographically
corrected scores, subtest variability ranged from the 14th to the 92nd percentile. This
represented a 2.5 standard deviation variability.
H.H. demonstrated a clear pattern of executive dysfunction on the re-evaluation.
Sustained attention and concentration abilities were inconsistent, ranging from the
8417: “8417_c023” — 2007/9/11 — 12:14 — page 543 — #49

16th to the 82nd percentile. Abstract verbal concept formation, as measured by the
Similarities subtest of the WAIS-III, was in the borderline range at the 18th percentile.
Reasoning and judgment abilities, as measured by the comprehension subtest of the
WAIS-III, were in the borderline range at the 24th percentile. Planning, organizing
and logical sequencing abilities, as measured by the Picture Arrangement subtest of
the WAIS-III, was in the borderline range at the 24th percentile. Verbal fluency was
mildly to moderately impaired at the 4th percentile. Speed of auditory information
processing, as measured by the PASAT, was inconsistent and far below expectation,
ranging from the 3rd to the 28th percentile. Visual information processing speed
was well within normal limits at the 62nd percentile. Severe oral motor dyspraxia
remained evident on the non-verbal agility subtest of the Boston Diagnostic Aphasia
Examination.
Left temporal/hippocampal dysfunction was suggested by a mild impairment of
her ability to learn new narrative verbal information presented in paragraph form
(Story Memory Test-Learning Component, 11th percentile). Retention of this inform-
ation following a 4-h-delay period was in the borderline range at the 21st percentile.
H.H. was unable to recall 14.3% of the information that she had previously learned.
Cueing was of some help to her memory. Learning of new rote verbal information,
as measured by the Buschke Verbal Selective Reminding Test (CLTR score), was
severely impaired at <1st percentile. Retention of this information following a 30-
min-delay period was mildly to moderately impaired at the 4th percentile. Language
comprehension was in the borderline range at the 21st percentile.
Right temporal/hippocampal dysfunction was suggested by a mild impairment of
her ability to recall newly learned complex visual information following a 30-min
delay period (Modified Taylor Complex Figure Test, 30-min delay, 10th percentile).
H.H. was unable to recall 17% of the information that she had previously learned
following a 30-min delay.
H.H. also demonstrated a variety of academic problems on her re-evaluation.
Reading Recognition was in the borderline range at the 16th percentile
(10.3 grade level). Reading Comprehension was mildly impaired at the 10th per-
centile (10.7 grade level). These were essentially the same scores that she obtained
when she was 16 years old. Therefore, her re-evaluation suggested that she had made
no appreciable improvement in her reading skills or abilities since that time. Indeed,
she was reporting increased reading problems over time. Although there had not been
a decline in her actual reading skills and abilities, there was obviously an increase
in the reading demands, which would have appeared to H.H. as a decline in her
reading skills and abilities. It was felt that this represented an excellent example of
how H.H.’s cognitive abilities had not developed at a pace commensurate with her
demographic peer group. Over time, the lack of development had led to an increased
functional cognitive problem. The same type of relative decline, when compared to
her demographic peer group, was evident in her vocabulary and arithmetic skills. Her
general fund of information, as measured by the Information subtest of the WAIS-III,
was in the mildly impaired range at the 14th percentile on retesting. It was con-
cluded that her general fund of information of the type generally acquired in school
and through various life experiences had not kept pace with her demographic peer
group.
8417: “8417_c023” — 2007/9/11 — 12:14 — page 544 — #50

On her re-evaluation, H.H. demonstrated a moderate to severe right visual inatten-

tion on the Padula Visual Midline Screening Test. She also continued to demonstrate
deficits or relative weaknesses in the area of incidental memory.
There are multiple indicators in H.H.’s re-evaluation to suggest that left hemi-
sphere functions had been more significantly affected than right. Indicators of more
lateralized cerebral dysfunction included the following: (1) PIQ, well within normal
limits and VIQ being in the borderline/low average range; (2) POI being 14 points
greater than VCI; (3) Verbal short-term memory being notably more impaired than
visual short-term memory; (4) Moderate to severe right visual inattention; (5) Tact-
ile sensitivity on the right being in the borderline range, whereas tactile sensitivity
on the left was well within normal limits; (6) Visual processing speed being in the
high average range versus verbal processing speed being in the borderline to mildly
to moderately impaired range; and (7) Verbal fluency being mildly to moderately
impaired at the 4th percentile, whereas figural fluency was within normal limits.
23.3.7 SUMMARY OF CASE STUDY

This case study was chosen because it demonstrates several key points and because it
correlates well with the overall results of the study. This study demonstrates that the
effects of chronic CO poisoning on neuropsychological testing may at times be quite
subtle, as evidenced by H.H.’s original neuropsychological evaluation. The results
of her re-evaluation also emphasize that the index scores are not always sensitive to
the effects of chronic CO poisoning and in many instances it is more important to
look at the overall pattern of individual deficits, relative weaknesses, and inconsist-
encies. This case study emphasizes the importance of conducting re-evaluations in
late adolescence or early adulthood in cases of chronic CO poisoning. Clearly, if one
is evaluating a child or adolescent following chronic CO poisoning, the full extent
of their long-term cognitive deficits may not be evident until their early 20s. The
pattern of dysfunction identified in H.H.’s initial neuropsychological evaluation was
consistent with the pattern identified in the study. It was also confirming of the overall
results of the study that on re-evaluation H.H. demonstrated her most notable declines
in the areas of executive/frontal functions, memory functions, and academic/language
functions. Those are areas of cognitive functioning found to be most sensitive to the
effects of chronic CO poisoning. It was also confirming that on re-evaluation she
was demonstrating more significant left hemisphere dysfunction, which was again
consistent with the findings of this study.
23.3.8 TAKEAWAY MESSAGES

1. Chronic CO poisoning can and often will result in permanent neurocognit-
ive and neurobehavioral dysfunction.
2. Loss of consciousness during exposure is not required for the exposure to
result in permanent impairment.
3. During the exposure, expect that the individual will have experienced
multiple symptoms in multiple systems.
4. Expect to see frequent misdiagnoses during the period of exposure.
8417: “8417_c023” — 2007/9/11 — 12:14 — page 545 — #51

5. If the patient does not make a full and complete recovery, again expect
to see multiple residual deficits and problems in multiple systems. Expect
to find persisting physical, fatigue, visual, cognitive, affective, mood, and
behavioral changes.
6. This study clearly documents that a wide range of cognitive functions can
be affected by chronic CO poisoning. However, this study also suggests
that executive/frontal, memory, and language/academic functions appear
to be highly susceptible to the effects of chronic CO poisoning.
7. This study also clearly suggests that left hemisphere functions tend to be
more susceptible to the effects of chronic CO poisoning than right hemi-
sphere functions. However, as expected, right hemisphere functions are
not immune to the effects of chronic CO poisoning.
8. When evaluating a child or adolescent following chronic CO poisoning,
it is important to realize that, because the human brain does not fully
develop until the early 20s, neuropsychological re-evaluation is recommen-
ded. Over time, the individual may well experience increasing cognitive
problems and deficits when compared to their demographically similar
peers. Essentially, brain development does not keep pace with their demo-
graphically similar peers and, therefore, over time they will functionally
experience greater and greater cognitive difficulty. In addition, mood and
affective problems can develop long after the CO poisoning stops.
9. When individuals experience permanent residual neurocognitive and neur-
obehavioral deficits associated with chronic CO poisoning, this will almost
always have a negative impact on their vocational functioning. Indeed, total
and permanent vocational disability is quite common following chronic CO
poisoning.
23.4 ADDENDUM
Just prior to the publication of this chapter, Pearson Assessments, who have the
scoring rights to the MMPI-2, made the decision to provide the Lees-Haley Fake Bad
Scale34 score as part of their Extended Score Protocol. It is important for clinicians
to understand that the construct validity of the Lees-Haley Fake Bad Scale (FBS)
has been criticized in the literature (e.g., Butcher, et al.35 and Arbisi and Butcher36 ).
The primary concern regarding the construct validity of this scale is that it contains
many symptoms common to a wide variety of medical and neurological disorders.
Twenty-four of the 43 items contained in this scale relate to possible neurological
symptoms including problems with attention and concentration, physical pain, head-
aches, fatigue, reduced stress tolerance, tinnitus, vision problems, sleep disturbance,
temperature deregulation, alteration in sense of taste, dizziness, and decreased libido.
Therefore, if a neurologically impaired individual is honestly reporting their symp-
toms on the MMPI-2 profile, then they are likely to demonstrate artificially elevated
scores on the FBS. Regarding this issue, Butcher35 states, “The results indicate that the
FBS is more likely to measure general maladjustment and somatic complaints rather
than malingering. The rate of false-positives produced by this scale is unacceptably
8417: “8417_c023” — 2007/9/11 — 12:14 — page 546 — #52

high, especially in psychiatric settings. This scale is likely to classify an unacceptably

large number of individuals who are experiencing genuine psychological distress as
malingerers. It is recommended that the FBS not be used in clinical settings nor should
it be used during disability evaluations to determine malingering” (pp. 473–474). The
potential for a false-positive error is recognized by Pearson Assessments and their
website contains a cautionary statement regarding use of the FBS. The PearsonAssess-
ment website states, “Scores on the FBS should be considered in the context of scores
on the other validity scales, the circumstances of the assessment and any conditions
such as a significant physical injury or disease that could artificially elevate scores on
the FBS.”
As has been noted earlier, the current study clearly supports that individuals
who have sustained chronic exposure to CO will often experience multiple persisting
symptoms in multiple systems. As a result, this population would be at risk for obtain-
ing elevated FBS scores simply because they are honestly reporting their persisting
symptoms on the MMPI-2. Therefore, prior to publication of this chapter, the FBS
score was calculated for each of the 19 adults who completed the MMPI-2 as part
of this study. The Pearson Assessment website suggests that raw scores above 23 on
the FBS should “raise concerns” about the validity of self-reported symptoms and
that raw scores above 28 should raise “very significant concerns” about the validity
of self-reported symptoms. Of the 19 participants in this study, only seven had FBS
scores of 23 or less. Three had FBS scores ranging from 23 to 27 and 9 had FBS scores
of 28 or greater. The average FBS score was 26.2 (range 17–38). As noted earlier
in the chapter, no individual was admitted to this study if there was any indication
of symptom magnification, exaggeration, or malingering. None of the 19 adults in
this study demonstrated any behaviors suggestive of symptom magnification. They
all passed three formal symptom validity tests and the standard and well accepted
validity indicators of the MMPI-2 did not suggest any symptom over reporting.
A critical item analysis of each participant’s FBS score was conducted. It was
determined that 50% or more of the subjects acknowledged persisting problems with
attention and concentration, physical pain, headaches, fatigue, reduced coping and
stress management, vision problems, sleep disturbance, gastro intestinal (GI) distress,
temperature deregulation, and dizziness on the FBS items. The current study supports
that these are common residual symptoms experienced by many individuals who have
been chronically exposed to CO. It was this author’s opinion that the fact that they
honestly reported these symptoms on the MMPI-2 artificially elevated their FBS
score. It is also this author’s opinion that the FBS should not be used, or at least be
used with extreme caution, as a validity indicator or an indication of symptom over
reporting for individuals who have been chronically exposed to CO.
ACKNOWLEDGEMENT
I would like to thank our psychometricians, Doug Wise, B.A., Amber Wolffrum,
B.A., and Vickie Novak, M.A. who administered the test batteries and also assisted
with organizing the volumes of data generated by this study. My wife, Diana, was,
as always, a wonderful support to me throughout this project. Technical support was
8417: “8417_c023” — 2007/9/11 — 12:14 — page 547 — #53

provided by C.N.A.’s Director of Operations, Brian Reusink, M.Ed. My thanks to

Robert Sokol, PhD who “helped to polish” the final product.
References
MRI, Quantitative MRI, SPECT and neuropsychological findings following carbon
monoxide Poisoning, Brain Inj., 13, 22–243, 1999.
2. Wright, J. Chronic and occult carbon monoxide poisoning: we don’t know what we’re
missing, Emerg. Med. J.; 19: 386–390, 2002.
3. Penney, D.G. Chronic carbon monoxide poisoning, In: Carbon Monoxide Toxicity,
Penney, D.G., Ed., CRC Press, Boca Raton, FL, 2000.
4. Hopkins, R.O. Neuropsychological and neuroimaging effects of carbon monoxide,
National Academy of Neuropsychology, 22nd Annual Conference, Miami Beach, FL,
October 11, 2002.
5. Weaver, L.K. Carbon monoxide poisoning, Critical Care Clinics, 15, 297–317, 1999.
6. Townsend, C.L., and Maynard, R.L. Effects on health of prolonged exposure to low
concentrations of carbon monoxide, Occup. Environ. Med.; 59: 708–711, 2002.
7. Broome, J.R., Pearson, R.R., and Skrine, H. Carbon Monoxide Poisoning: forgotten
not gone!, Br. J. Hosp. Med., 39, 298–305, 1988.
8. Hampson, N.B. Carbon Monoxide Poisoning and Its Management in the United States,
In: Carbon Monoxide Toxicity, Penney, D.G., Ed., CRC Press, Boca Raton, FL,
2000.
9. Heckerling, P.L., Leikin, J.M., Maturen, A., and Perkins, J. Predictors of occult carbon
monoxide poisoning in patients with headache and dizziness, Ann. Intern. Med., 107,
174–176, 1987.
10. Halpern, J.S. Clinical Notebook: Chronic occult carbon monoxide poisoning, J.
Emerg. Nursing, 15, (Part 1), 1989.
11. Carbon Monoxide Support The Effects of Chronic Exposure to CO: A Research Study
Conducted by CO Support, Technical Paper, 1–47 pp., appendices, October, 1997.
12. Hay, A.W.M., Jaffer, S., and Davis, D. Chronic Carbon Monoxide Exposure: The CO
Support Study. In: Carbon Monoxide Toxicity, Penney, D.G., Ed., CRC Press, Boca
Raton, FL, 2000.
13. Ryan, C.M. Memory Disturbance Following Chronic, Low-Level Carbon Monoxide
Exposure, Arch. Clin. Neuropsychology; 5, 59–67, 1990.
14. Myers, R.A.M., DeFazio, A., and Kelly, M.P. Chronic carbon monoxide exposure:
A clinical syndrome detected by neuropsychological tests, J. Clin. Psychology, 54,
555–567, 1998.
findings in carbon monoxide poisoning: Deficits seen in a matched pair, Arch. Clin.
Neuropsychology, 15, 545–553, 2000.
16. Hartman, D.E. Neuropsychological Toxicology Identification and Assessment of
Human Neurotoxic Syndromes (2nd ed.), Plenum Press, New York and London, 1995.
17. Devine, S.A., Kirkley, S.M., Palombo, C.L., and White, R.F. MRI and neuropsy-
Perspect., 101, October, 2002.
18. Helffenstein, D.A. Neuropsychological evaluation of the carbon monoxide-poisoned
patient. In: Penney, D.G., Ed., Carbon Monoxide Toxicity, CRC Press, Boca Raton,
FL, 2000, Chapt. 20, pp. 439–461.
8417: “8417_c023” — 2007/9/11 — 12:14 — page 548 — #54

19. Heaton, R.K., Grant, I., and Matthews, C.G. Comprehensive Norms for an Expanded
Halstead-Reitan Battery, Psychological Assessment Resources, Odessa, FL, 1991.
20. Padula, W.V., and Argyris, S. Post-trauma vision syndrome and visual mid-line shift
syndrome, Neuro Rehab., 6, 165–171, 1996.
21. Politzer, T. Vision function, examination and rehabilitation in patients suffering from
traumatic brain injury. In: Minor Traumatic Brain Injury Handbook: Diagnosis and
Treatment, Jay, G.W., Ed., CRC Press, 2000.
22. Jarvis, P.E., and Barth, J.T. The Halstead-Reitan Neuropsychological Test Bat-
tery: A Guide to Interpretation and Clinical Application, Psychological Assessment
Resources, Odessa, FL, 1994.
23. Reitan, R.M., and Wolfson, D. Traumatic Brain Injury, Vol. II, Recovery and
Rehabilitation, Neuropsychology Press, Tucson, AZ, 1988.
24. Reitan, R.M., and Wolfson, D. The Halstead-Reitan Neuropsychological Test Battery:
Theory and Clinical Interpretation, 2nd ed, Neuropsychology Press, Tucson, AZ,
1993.
25. Meyers, J.E., and Meyers, K.R. Complex Figure Test and Recognition Trial:
Professional Manual, Psychological Assessment Resources, Inc., Odessa, FL, 1995.
26. Schenkenberg, T., Bradford, D.C., andAjax, E.T. Line Bisection and unilateral neglect
in patients with neurologic impairment, Neurology, 30, 509–517, 1980.
27. Hooper, H.E. Hooper Visual Organization Test. LA: Western Psychological Service,
1983.
28. Lezak, M.D., Howieson, D.B., and Loring, D.W. Neuropsychological Assessment,
4th ed, Oxford University Press, NY, 2004.
29. Cripe, L.I. Use of the MMPI with mild closed head injury. In: The Evaluation and
Treatment of Mild Traumatic Brain Injury, Varney, N.R. and Roberts, R.J., Eds.,
Lawrence Erlbaum Associates, Mahwah, NJ, 1999.
30. Morrow, L.A., Ryan, C.M., Hodgson, M.J., and Robin, N. Alterations in cognitive
and psychological functioning after organic solvent exposure, J. Occup. Med., 32,
444–449, 1990.
31. Bowler, R.M., Mergler, D., Rauch, S.S., et al. Affective and personality disturbances
among female former microelectronics workers, J. Clin. Psychol., 47, 41–52, 1991.
32. Tellegen, A., Ben-Porath, Y.S., McNulty, J.L., Arbisi, P.A., Grahm, J.R., and Kaem-
mer, B. The MMPI-2 Restructured Clinical (RC) Scales - Development, Validation,
and Interpretation, University of Minnesota Press, Minneapolis, MN, 2003.
33. Heuser, G., and Mena, I. NeuroSPECT in neurotoxic chemical exposure demonstration
of long-term functional abnormalities, Toxicol. Industr. Health, 14, 813–827, 1998.
34. Lees-Haley, P.R., English, L.T., and Glenn, W.J. A Fake Bad Scale on the MMPI-2
for Personal Injury Claimants, Psychol. Rep., 68, 203–210, 1991.
35. Butcher, J.N., and Arbisi, J.L. The Construct Validity of the Lees-Haley Fake Bad
Scale – Does this scale measure somatic malingering and feigned emotional distress?
Arch. Clin. Neuropsychol., 18, 473–485, 2003.
36. Arbisi, P.A., and Butcher, J.N. Failure of the FBS to predict malingering of somatic
symptoms: response to critiques by Greve and Bianchini and Lees-Haley and Fox,
Arch. Clin. Neuropsychol., 19, 341–345, 2004.
8417: “8417_c023” — 2007/9/11 — 12:14 — page 549 — #55

8417: “8417_c023” — 2007/9/11 — 12:14 — page 550 — #56
24 Chronic Carbon
Monoxide Poisoning: A
Case Series
David G. Penney
CONTENTS
24.1 Study Background and Demographics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551

24.2 Symptomatic Evaluation Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 553
24.3 Use of Self-Report Questionnaires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 558
24.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 564
24.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 565
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 565
24.1 STUDY BACKGROUND AND DEMOGRAPHICS

This is a retrospective chart-review study. Patients data files were enrolled on two
bases, (1) they were exposed to carbon monoxide (CO) for more than 24 h (i.e.,
chronic), and (2) they were adult, that is, 18 years of age and above. The cohort was
subdivided into those patients who were active smokers (18) and those who were not
smokers (43). Diagnostic questionnaires on file had been completed by the patients,
some completing two or even three questionnaires over the course of their evaluation.
In all instances, the most recently completed questionnaire (i.e., longest time period
since CO exposure) was chosen for use. The questionnaires used were dated from
2001 to 2006.
Patients resided throughout the United States. The most highly represented states
were Michigan, Texas, and Washington, four each; Nevada, New York, West Virginia,
California, Florida, Kansas, and Missouri, three each; New Mexico, Pennsylvania,
Utah, Virginia, Georgia, Idaho, Indiana, and Maine, two each, and the rest with
one each—Alaska, Alabama, Connecticut, Iowa, Illinois, Montana, North Carolina,
Nebraska, Rhode Island, South Carolina, and Vermont.
Table 24.1 lists the sources of CO identified by each of the patients, as they knew
it. Notice that “furnace” contributed 38.8% to the total of sources. Related to that
was “heater” at 8.9%, fireplace at 14.9%, “boiler” at 4.5%, and “heating system” at
1.5%. Therefore, space heating equipment contributed over two-thirds of the sources
of CO identified. Water heaters made up 10.4%. Motor vehicles in total contributed
551
8417: “8417_c024” — 2007/9/11 — 12:14 — page 551 — #1

TABLE 24.1
Sources of Carbon Monoxide in Study C
Nonsmoker % Smoker % Total %
Furnace 17 36.1 9 45.0 26 38.8
Fireplace 10 21.3 0 0 10 14.9
Heater 4 8.5 2 10.0 6 8.9
Water heater 4 8.5 3 15.0 7 10.4
Automobile 3 6.4 1 5.0 4 6.0
Boiler 2 4.3 1 5.0 3 4.5
Stove 1 2.1 1 5.0 2 3.0
Truck 1 2.1 1 5.0 2 3.0
Heating system 1 2.1 0 0 1 1.5
Oven 1 2.1 0 0 1 1.5
Race car 1 2.1 0 0 1 1.5
Track hoe 1 2.1 0 0 1 1.5
Lift truck 1 2.1 0 0 1 1.5
Diesel engine 0 0 1 5.0 1 1.5
Rocket fuel 0 0 1 5.0 1 1.5
Total 47 20 67
TABLE 24.2
Demographics of Patients Enrolled in Study C
Exposure Time After CO
Male/ Duration poisoning
Female Age (years) CO (ppm) COHb (%) (months) (months)
Nonsmokers 12/31 41.9 ± 2.2 150.5 ± 24.2 8.1 ± 1.5 24.9 ± 4.1 23.4 ± 3.2
n= 43 20 7 30 39
Smokers 8/10 48.9 ± 3.3 143.3 ± 32.0 10.5 ± 1.7 33.9 ± 8.9 22.3 ± 3.7
n= 18 3 5 18 17
Combined 20/41 43.9 ± 2.3 149.6 ± 23.4 9.2 ± 1.3 27.8 ± 4.7 23.0 ± 3.0
n= 61 23 12 56 56
less than 10% as the source of chronic CO poisoning. It is not clear why none of the
smokers identified “fireplace” as the CO source. It may be related to the fact that this
group was quite small compared to the nonsmoker group.
The demographics of the two subgroups and of the total enrollees in the study
are shown in Table 24.2. The nonsmoker group of 43 individuals contained 12 men
and 31 women, with an average age of 41.9 ± 2.2 years (mean, standard error of the
mean). The smoker group of 18 individuals contained 8 men and 10 women, with an
average age of 48.9 ± 3.3 years. This difference verges on significance at the p > .05
level. Mean Air CO concentration on the nonsmoker group was 150.5 ± 24.2 ppm,
8417: “8417_c024” — 2007/9/11 — 12:14 — page 552 — #2

Chronic Carbon Monoxide Poisoning: A Case Series 553
and 143.3 ± 32.0 ppm in the smoker group. Combined mean CO concentration was
149.6 ± 23.4 ppm. All values of CO are those at the time of measurement, and may
or may not be representative of CO concentration at other times during exposure.
Carboxyhemoglobin (COHb) was measured in only twelve of the enrollees, seven
nonsmokers, and five smokers. The latter had COHb of 8.1% and the former, 10.5%.
Overall, cohort COHb was 9.2 ± 1.3%. No “back-calculation” to time zero (i.e., time
of leaving site of CO poisoning) was attempted. Mean overall exposure duration was
27.8 months, with the smokers being nonsignificantly longer by 9 months. The time
after the CO poisoning at which the diagnostic questionnaire was completed was
about 23 months in both groups, that is, approximately 2 years.
24.2 SYMPTOMATIC EVALUATION DATA

The initial/immediate symptoms experienced by the enrollees are shown in Table 24.3.
The symptoms are listed in decreasing order of frequency of complaint, down to a
total of two. As in my earlier studies of chronic CO poisoning,1 a wide variety of
symptoms were involved. The data show that headache was the most frequently
reported symptom in both the nonsmoker and smoker groups. This was followed by
nausea, dizziness, fatigue, confusion, forgetfulness, and so forth.
The immediate symptoms reported only once are as follows: allergies; anger;
apathy; appetite poor; arm pain; back pain; back spasm; backache; blacking out;
body pain; brain buzzing, burning, sizzling; bruising, bleeding; congestion; coordin-
ation, loss; couldn’t get out of bed; couldn’t identify people; couldn’t move; cramps;
déjà vu; digestive problems; dreams, horrible; dropping things; dry skin; ear aches;
ear pain; eyes burning; eyes, spots; face numb; fainting; fever; frustration; hair falling
out; head rushes; hearing loss; heart painful when active; heart squeezed; heartbeat
irregular; hoarseness; ill; incontinence; intestinal problems; irritable bowel; itchy;
light-headedness; lost in familiar place; miscarriage; motivation lacking; palpita-
tions; paralysis; poor coordination; pressure in head; projects, trouble completing;
respiratory problems; restless leg syndrome; runny nose; skin red hot; sore throat;
standing up, fell down; stomach, with gas in it; “strobing,” stumbling; throat burning;
transient ischemic attack (TIA); trembling; urine output reduced; vertigo; walking all
night; and wheezing.
Table 24.4 lists the symptoms or conditions that existed when the question-
naire was completed many months after termination of the CO exposure. Of the
persistent physical symptoms (Table 24.4A), “fatigue” was said to be present by
93.4% of enrollees. Next in line at decreasing frequency were “sleep problems,”
“headache,” “muscle pain,” “weakness,” and so forth. Note that headache, reported
as the most frequent immediate symptom is less frequently reported as a long-
term, persistent symptom. The most frequently reported sensory-motor symptom
is “eye/vision problems,” and the most frequently reported gross neurologic condi-
tion is reduced “physical strength.” Of the persistent cognitive-memory symptoms
(Table 24.4B), “memory” problems were said to be present by 98.4% of enrollees.
Next in line at decreasing frequency were “attention-concentration,” “more distract-
ible,” “multitasking,” and so forth. The affective-emotional symptoms seen with
8417: “8417_c024” — 2007/9/11 — 12:14 — page 553 — #3

TABLE 24.3
Immediate Symptoms during the Carbon Monoxide Exposure in Study C
Symptom/Condition Nonsmokers Smokers Combined
Headache 32 13 45
Nausea 24 7 31
Dizziness 21 6 27
Fatigue 13 9 22
Confusion 9 8 17
Forgetfulness 8 3 11
Sleepy 6 4 10
Disorientation 6 3 9
Dyspnea/S.O.B. 6 2 8
Flu-like symptoms 3 4 7
Vomiting 4 3 7
Weakness 3 4 7
Numbness 4 3 7
Sleep problems 5 2 7
Irritability 7 0 7
Tiredness 7 0 7
Chest pain 3 3 6
Muscle pain 5 1 6
Blurred vision 4 2 6
Memory problems 6 0 6
Depression 6 0 6
Pain 3 2 5
Vision problems 2 3 5
Diarrhea 3 2 5
Tingling 3 2 5
Loss of balance 3 2 5
Tinnitus 4 1 5
Attention-concentration 5 0 5
Decreased mental capacity 3 1 4
Cough 4 0 4
Leg pain 1 2 3
Sweating 2 1 3
Cold in day 2 1 3
Joint pain 3 0 3
Sleeping longer 3 0 3
Thinking difficulty 3 0 3
Tachycardia 1 2 3
Stomach problems 1 1 2
Double vision 1 1 2
Panic attacks 1 1 2
Nervousness 1 1 2
Moodiness 2 0 2
Anxiety 2 0 2
Photosensitivity 2 0 2
Abdominal pain 2 0 2
Hallucinations 2 0 2
Felt like rubbed raw inside with sandpaper 2 0 2
Sleepiness 2 0 2
Neck pain 2 0 2
Consciousness altered 2 0 2
Crying 2 0 2
Blood pressure increased 2 0 2
8417: “8417_c024” — 2007/9/11 — 12:14 — page 554 — #4

TABLE 24.4A
Incidence of Persistent Symptoms (Physical, Sensory-Motor and Gross
Neurologic) after Carbon Monoxide Exposure in Study C
Combined
Symptom/Condition Nonsmokers Smokers Combined Percent
Fatigue 41/43 16/18 57/61 93.4
Sleep problems 38/43 17/18 55/61 90.2
Headache 37/43 16/18 53/61 86.9
Muscle pain 35/43 17/18 52/61 85.3
Weakness 36/43 15/17 51/60 85.0
Balance 32/40 15/16 47/56 83.9
Physical strength 34/41 15/18 49/59 83.1
Joint pain 32/43 17/18 49/61 80.3
Numbness 35/43 13/17 48/60 80.0
Eye vision problems 31/43 16/17 47/60 78.3
Tingling 34/43 13/18 47/61 77.0
Fine motor problems 30/40 11/16 41/56 73.2
Walking problems 29/42 13/18 42/60 70.0
Dizziness 28/43 14/18 42/61 68.9
Taste/smell changes 27/43 14/18 41/61 67.2
Shortness of breath 33/43 7/17 40/60 66.7
Hearing problems 28/43 11/18 39/61 63.9
Handwriting changed 26/42 12/18 38/60 63.3
Thermoregulatory dysfunction 25/41 12/18 37/59 62.7
Chest pain 24/43 9/18 33/61 54.1
Tremor 21/43 12/18 33/61 54.1
Other gross motor problems 26/41 7/17 33/58 56.9
Nausea 21/43 7/18 28/61 45.9
Vertigo 20/39 4/17 24/56 42.9
Urinary incontinence 17/43 7/17 24/60 40.0
Diarrhea 16/43 7/18 23/61 37.7
Paresthesia 5/14 2/4 7/18 38.9
Vomiting 8/43 4/18 12/61 19.7
greatest frequency was “more anxiety/fear” at 98.2%. This was followed by “irrit-
ability,” “mood changes,” “anger/temper,” “depression,” “more apathetic,” and so
forth. The belief that others perceived them as “different than before” the CO pois-
oning was endorsed by 94.6% of enrollees. One-fifth of enrollees indicated they had
had thoughts related to suicide after the CO poisoning.
Table 24.5 presents the severity of each of the persistent physical symptoms as
reported by the enrollees, for the nonsmoker, smoker, and combined group. Overall,
“fatigue” was reported to have the greatest severity, 3.5 ± 0.2 (mean, ± standard error
of the mean). The intensity scale ranged from 0 to 5. The absence of a symptom is
equivalent to a zero. Above zero, 1 = slight, 2 = mild, 3 = moderate, 4 = severe, and
5 = extremely severe. Thus, 3.5 is midway between moderate and severe. Because the
value for fatigue is a mean carrying with it some variance [as expressed by standard
8417: “8417_c024” — 2007/9/11 — 12:14 — page 555 — #5

TABLE 24.4B
Incidence of Persistent Symptoms (Cognitive-Memory, Affective-Emotional)
after Carbon Monoxide Exposure in Study C
Combined
Symptom/Condition Nonsmokers Smokers Combined Percent
Memory 42/43 18/18 60/61 98.4
Attention-concentration 42/43 18/18 60/61 98.4
More anxiety/fear 39/40 16/16 55/56 98.2
Irritability 40/41 16/16 56/57 98.2
More distractible 26/27 12/12 38/39 97.4
Mood changes 42/43 17/18 59/61 96.7
Anger/temper 35/37 15/15 50/52 96.2
Depression 33/34 16/17 49/51 96.1
More apathetic 37/36 12/13 47/49 95.9
Loss of interest in life 23/25 14/14 37/39 94.9
Seen as different than before 38/40 15/16 53/56 94.6
Loss of motivation 22/24 13/13 35/37 94.6
Multitasking 40/43 16/17 56/60 93.3
Word-finding 40/43 16/17 56/60 93.3
Panic attacks 28/31 13/13 41/44 93.2
Lose track in actions 38/41 16/18 54/59 91.5
Mental confusion 39/43 16/18 55/61 90.2
Stays in room/home 38/42 15/17 53/59 89.8
Weep/cry 16/18 9/10 25/28 89.3
Slow mental processing 27/31 13/14 40/45 88.9
Makes more mistakes 25/29 14/15 39/44 88.6
Lower self-esteem 24/27 12/14 36/41 87.8
Understanding others 26/29 13/15 39/44 86.6
Avoids social situation 37/42 14/17 51/59 86.4
Afraid of CO poisoning again 36/41 15/18 51/59 86.4
Decision difficulty 36/43 16/18 52/61 85.3
Initiation of actions 26/32 14/15 40/47 85.1
Following directions 32/39 15/18 47/57 82.5
Less pleasure taken 27/32 10/13 37/45 82.2
Diminished libido 31/39 14/16 45/55 81.8
Sorting/organizing 36/43 13/17 49/60 81.7
Speaking problems 36/43 11/17 47/60 78.3
Mental/written math 24/30 11/15 35/45 77.8
Reading problems 22/29 10/14 32/43 74.4
More suspicious 20/26 9/14 29/40 72.5
Disorientation 29/42 11/16 40/58 69.0
Finding familiar places 26/41 14/18 40/59 67.8
Can’t balance checkbook 27/39 12/17 39/56 69.6
More compulsive 18/25 5/13 23/38 60.5
Suicide thoughts 3/25 4/9 7/34 20.6
8417: “8417_c024” — 2007/9/11 — 12:14 — page 556 — #6

TABLE 24.5
Physical Symptoms in Study C
Muscular Joint Chest Therm Sleep
H.A. Pain Pain S.O.B Diarrhea Pain Fatigue Weakness Dz. Nz. Vertigo Vomiting Numbness Tingling Reg. Dys Problems
Nonsmokers 2.9 ± 0.2 2.7 ± 0.3 2.4 ± 0.3 2.3 ± 0.3 0.9 ± 0.2 1.4 ± 0.2 3.6 ± 0.2 2.8 ± 0.3 1.7 ± 0.2 1.3 ± 0.3 1.1 ± 0.2 0.4 ± 0.1 2.4 ± 0.3 2.1 ± 0.2 1.6 ± 0.3 3.4 ± 0.2
n= 43 42 43 42 43 43 43 43 43 40 36 43 41 42 38 41
Smokers 2.7 ± 0.3 2.9 ± 0.2 2.5 ± 0.3 1.1 ± 0.4 0.6 ± 0.2 0.8 ± 0.3 3.3 ± 0.3 2.9 ± 0.4 2.1 ± 0.3 0.9 ± 0.3 1.2 ± 0.3 0.4 ± 0.2 1.9 ± 0.4 1.8 ± 0.4 1.8 ± 0.4 3.0 ± 0.3
n= 17 17 17 17 17 17 17 16 17 17 17 18 16 17 16 17
Chronic Carbon Monoxide Poisoning: A Case Series
Combined 2.8 ± 0.2 2.7 ± 0.2 2.4 ± 0.2 1.9 ± 0.2 0.8 ± 0.2 1.2 ± 0.2 3.5 ± 0.2 2.8 ± 0.2 1.8 ± 0.2 1.2 ± 0.2 1.1 ± 0.2 0.4 ± 0.1 2.3 ± 0.2 2.0 ± 0.2 1.7 ± 0.2 3.3 ± 0.2
n= 60 59 60 59 60 60 60 59 60 57 53 61 57 59 54 58
H.A. = Headache, SOB = Shortness of breath (dyspnea), Dz = Dizziness, Nz = Nausea.
8417: “8417_c024” — 2007/9/11 — 12:14 — page 557 — #7

557
error of the mean (SEM)], some enrollees reported less fatigue and some reported
more than the mean. Recall that nearly all enrollees reported fatigue of some degree
(i.e., 93.4%). A symptom of slightly lesser severity was “sleep problems,” at 3.3. This
was followed by “headache” and “weakness” at 2.8, and then “muscle pain” (2.7),
“joint pain” (2.4), “numbness” (2.3) and “tingling” (2.0). There appeared to be no
significant differences between smokers and nonsmokers. The symptoms reported at
the lowest level of severity were “diarrhea” (0.8) and “vomiting” (0.4).
Table 24.6 presents the severity of each of the persistent sensory-motor and gross
neurologic symptoms as reported by the enrollees, for the nonsmoker, smoker, and
combined group. Overall, problems with “physical strength” were rated most highly,
at 2.5. Following this was “eye/vision” problems (2.2), “balance” problems (2.1), and
“hearing” problems, “fine motor” problems, and “walking” problems, all reported at
the same mean level of severity (1.8). There appeared to be no significant differences
between smokers and nonsmokers. The symptoms reported at the lowest level of
severity were “paraesthesia” and “urinary incontinence” (1.0).
Owing to the amount of data on cognitive and memory symptoms, Table 24.7
was divided into A and B sections. Each section contains ten aspects each of
cognitive-memory data, as reported by the enrollees for the nonsmoker, smoker
and combined group. Overall, problems involving “memory” (Table 24.7A) and
“attention-concentration” (Table 24.7B) were reported to persist with the highest
level of intensity after CO poisoning, 3.5 ± 0.2 for both. Following these were
“multi-tasking” (3.3), “more distractible” (3.1), “word-finding” problems (3.0), and
“decisions difficult” and “makes more mistakes”, both at 2.9. There appeared to
be no significant differences between smokers and nonsmokers. The symptoms
reported at the lowest levels of severity were “finding familiar places” (1.9) and
“disorientation” (1.5).
Owing to the amount of data on affective-emotional symptoms, Table 24.8 was
divided into A and B sections. Each section contains ten aspects each of the affective-
emotional data, as reported by the enrollees for the nonsmoker, smoker, and combined
group. Overall, problems involving “mood changes” (3.6) and “increased anxiety”,
“depression” and “irritability” (all 3.5) (Table 24.8A) were reported to persist with
the highest level of intensity after CO poisoning. Other affective-emotional aspects
were reported to persist at somewhat lower levels of intensity: “afraid of CO pois-
oning again” and “anger/temper” at 3.2, “different per others” at 3.1, and “loss of
motivation” at 3.0. There appeared to be no significant differences between smokers
and nonsmokers. The symptoms reported at the lowest level of severity was “suicide
thoughts” (0.3).
24.3 USE OF SELF-REPORT QUESTIONNAIRES

The use of questionnaires is basic to clinical psychology, the health sciences, and
other fields in attempting to understand a person’s background, thoughts, beliefs,
health status, and so forth. For this reason a number of special purpose questionnaires
were devised to obtain such data relative to the study of chronic CO poisoning. As
such, the questionnaires used are information gathering tools, not testing tools.
8417: “8417_c024” — 2007/9/11 — 12:14 — page 558 — #8

TABLE 24.6
Sensory–Motor and Gross Neurologic Symptoms in Study C
Other
Gross Fine
∗ Eye- Taste- Motor Motor Hand Walking Physical Urinary
Vision Hearing Smell Paresthesia Balance Tremor Problems Problems Writing Problem Strength Incontinence
Nonsmokers 2.1 ± 0.3 1.9 ± 0.3 1.8 ± 0.3 1.0 ± 0.4 2.0 ± 0.2 1.3 ± 0.3 1.9 ± 0.3 1.9 ± 0.3 1.2 ± 0.4 1.9 ± 0.3 2.6 ± 0.3 1.1 ± 0.2
n= 38 39 42 14 40 42 38 37 26 40 41 43
Smokers 2.5 ± 0.4 1.8 ± 0.4 1.4 ± 0.3 1.0 ± 1.0 2.2 ± 0.3 1.2 ± 0.4 0.9 ± 0.3 1.6 ± 0.4 1.1 ± 0.6 1.7 ± 0.4 2.4 ± 0.4 0.8 ± 0.3
n= 15 17 16 3 15 17 16 16 9 18 18 17
Combined 2.2 ± 0.2 1.8 ± 0.2 1.7 ± 0.2 1.0 ± 0.4 2.1 ± 0.2 1.3 ± 0.2 1.6 ± 0.2 1.8 ± 0.2 1.2 ± 0.3 1.8 ± 0.2 2.5 ± 0.2 1.0 ± 0.2
n= 53 56 58 17 55 59 54 53 35 58 59 60
∗ Eye-vision, hearing, and so forth abnormalities.
8417: “8417_c024” — 2007/9/11 — 12:14 — page 559 — #9

559
560
TABLE 24.7A
Cognitive–Memory Symptoms in Study C
Finding
Speaking Word-Finding Memory Decisions Understanding Following Familiar
Problems Problems Problems Difficult Others Reading Directions Places Confusion Disorientation
Nonsmokers 2.4 ± 0.2 3.1 ± 0.2 3.5 ± 0.2 2.9 ± 0.3 2.5 ± 0.2 2.3 ± 0.3 2.6 ± 0.3 1.8 ± 0.3 2.7 ± 0.2 1.5 ± 0.2
n= 39 42 42 43 29 29 39 41 41 42
Smokers 1.9 ± 0.4 2.9 ± 0.3 3.5 ± 0.3 2.9 ± 0.4 2.5 ± 0.3 2.4 ± 0.3 2.4 ± 0.3 2.3 ± 0.4 2.7 ± 0.3 1.6 ± 0.3
n= 17 17 17 18 15 14 18 18 17 16
Combined 2.3 ± 0.2 3.0 ± 0.2 3.5 ± 0.2 2.9 ± 0.2 2.5 ± 0.2 2.3 ± 0.3 2.6 ± 0.2 1.9 ± 0.2 2.7 ± 0.2 1.5 ± 0.2
n= 56 59 59 61 44 43 57 59 58 58
8417: “8417_c024” — 2007/9/11 — 12:14 — page 560 — #10

TABLE 24.7B
Cognitive–Memory Symptoms in Study C (cont.)
Attention Slower Make Lose Balance Mental/
Concentration Starting More Mental More Track Sorting/ Check Written
Problems Projects Distractible Processing Mistakes of Self Organizing Multitasking Book Math
Nonsmokers 3.6 ± 0.2 2.9 ± 0.3 3.1 ± 0.2 2.6 ± 0.3 2.8 ± 0.3 2.9 ± 0.2 2.8 ± 0.3 3.4 ± 0.2 2.5 ± 0.3 2.4 ± 0.3
n= 43 32 27 31 29 41 43 41 39 27
Smokers
3.4 ± 0.3 2.4 ± 0.3 3.1 ± 0.4 3.3 ± 0.3 3.1 ± 0.4 2.8 ± 0.4 2.5 ± 0.4 3.2 ± 0.3 2.2 ± 0.5 2.4 ± 0.5
n= 18 15 11 14 15 18 17 17 16 14
Combined 3.5 ± 0.2 2.7 ± 0.2 3.1 ± 0.2 2.8 ± 0.2 2.9 ± 0.2 2.9 ± 0.2 2.7 ± 0.2 3.3 ± 0.2 2.4 ± 0.3 2.4 ± 0.3
n= 61 47 38 45 44 59 60 58 55 41
8417: “8417_c024” — 2007/9/11 — 12:14 — page 561 — #11

561
562
TABLE 24.8A
Affective–Emotional Symptoms in Study C
Afraid of More More Decreased Mood Increased Panic More
CO Again Suspicious Compulsive Self-Esteem Changes Anxiety Attacks Apathy Depression Irritability
Nonsmokers 3.3 ± 0.3 2.7 ± 0.4 2.6 ± 0.4 2.7 ± 0.3 3.7 ± 0.2 3.4 ± 0.2 2.8 ± 0.3 3.1 ± 0.2 3.5 ± 0.2 3.6 ± 0.2
n= 39 24 23 25 38 40 31 36 34 41
Smokers 3.1 ± 0.4 1.9 ± 0.5 1.5 ± 0.6 2.9 ± 0.5 3.3 ± 0.4 3.7 ± 0.3 3.2 ± 0.3 2.4 ± 0.5 3.3 ± 0.4 3.3 ± 0.3
n= 17 13 13 13 14 16 13 13 17 16
Combined 3.2 ± 0.2 2.4 ± 0.3 2.2 ± 0.3 2.8 ± 0.2 3.6 ± 0.2 3.5 ± 0.2 2.9 ± 0.2 2.9 ± 0.2 3.5 ± 0.2 3.5 ± 0.2
n= 56 37 36 38 52 56 44 49 51 57
8417: “8417_c024” — 2007/9/11 — 12:14 — page 562 — #12

TABLE 24.8B
Affective–Emotional Symptoms in Study C (cont.)
Loss of Avoid
Weeping/ Anger Interest Loss of Suicide Depressed Social Stay in Gets Less Different
Crying Temper in Life Motivation Thoughts Libido Situations Room Pleasure Per Others
Nonsmokers 2.4 ± 0.3 3.4 ± 0.3 2.7 ± 0.3 2.7 ± 0.3 0.1 ± 0.1 2.6 ± 0.3 2.8 ± 0.3 3.3 ± 0.3 2.8 ± 0.34 3.1 ± 0.4
n= 18 37 25 24 25 31 36 37 31 17
Smokers
2.9 ± 0.5 2.9 ± 0.3 3.4 ± 0.4 3.4 ± 0.4 0.8 ± 0.4 2.9 ± 0.5 2.9 ± 0.5 3.4 ± 0.4 2.8 ± 0.5 3.0 ± 0.8
n= 10 15 14 13 9 14 16 17 13 5
Combined 2.6 ± 0.3 3.2 ± 0.2 2.9 ± 0.2 3.0 ± 0.2 0.3 ± 0.1 2.7 ± 0.3 2.9 ± 0.2 3.3 ± 0.2 2.8 ± 0.3 3.1 ± 0.3
n= 28 52 39 37 34 45 52 54 44 22
8417: “8417_c024” — 2007/9/11 — 12:14 — page 563 — #13

563
A history questionnaire (not used here) was designed to gather historical informa-
tion regarding medical, social, and psychological events, as well as basic demographic
data about the patient. It is six pages in length and requires 30–60 min to complete.
The diagnostic questionnaire used here asks respondents about their health,
broadly defined, at the time of the toxic event, as well as at the present time, termed
“now.” The responses elicited are both qualitative and quantitative—does the symp-
tom or condition apply to the respondent and what is the intensity of the symptom
or condition as they sense it. The intensity scale runs from 0 to 5, that is, 0, 1, 2,
3, 4, and 5. An answer of “no” is equivalent to a zero. Zero indicates the symptom
or condition is not present, 1 = slight, 2 = mild, 3 = moderate, 4 = severe, and
5 = extremely severe.
The collateral A questionnaire (not used here) is designed for a third party who
knows the patient extremely well (i.e., intimately) to answer. This person is usually a
“significant other.” This questionnaire is also both qualitative and quantitative like the
diagnostic questionnaire. Respondents are asked to complete the form from his/her
own knowledge of the subject, not by quering the subject.
Routinely, during analysis of questionnaire data, the threshold of significance is
set at “mild” (= or >2). That means that responses that fail to reach a level of 2.0 are
considered not to be significant.
The collateral B questionnaire is another kind of third party questionniare. It is
designed for a person to answer who knows the subject well, but may not know him
or her intimately. It is qualitative only—did the symptom or sign preexist the CO
poisoning incident, or did it occur immediately afterward.
The credibility and validity of a patient’s responses can be evaluated in a number
of ways. Patients can be asked to complete questionnaires several times over, sep-
arated by several month intervals. When this is done, patients are asked to attest to
the fact that they have not retained photocopies of previously completed question-
naires. Responses on these questionnaires are examined for consistency over time. A
second approach to the problem of credibility and validity of responses has been to
embed questions in the questionnaire whose function is to reveal those who would
blindly indicate that they have every symptom/sign presented (i.e., “control items”).
A third approach is to perform cluster analysis of symptoms. This looks at multiple
questions concerned with a related symptom or condition. Again, information about a
patient’s consistency can be determined in this way. Finally, the use of questionnaires
completed by third parties (i.e., collaterals), indicate whether significant others, close
friends, family, and so forth see the same symptoms as the patient reports, the same
intensity of those symptoms, and whether the symptoms are new after the toxic insult
or predated the insult. Some patients unconsciously under-report their symptoms,
while others over-report. Information regarding such tendencies can also be gained
though use of collateral questionnaires.
24.4 DISCUSSION
This study describes extensive and severe symptoms residual after lower-level chronic
CO poisoning where the mean reported COHb was 9.2%. COHb in Study A was
9.65% and 9.0% in Study B, corrected in the later instance to 14.0% after allowance
8417: “8417_c024” — 2007/9/11 — 12:14 — page 564 — #14

for CO washout prior to the drawing of blood. Similar extensive and severe health
damages are shown in the study of chronic CO poisoning by Dr. Helffenstein in this
book2 (Chapter 23), which parallels and also greatly extends this one to areas of
neuropsychological damage.
Dr. Hopkins (Chapter 22) discusses how less-severe CO poisoning has been
defined in terms of COHb by various investigators as ≤10%, 5–15%, and 0.01–11%.3
She described the study by Chambers et al.4 who compared the cognitive outcomes of
CO victims with less-severe acute poisoning to those with more severe CO poisoning
at 6 weeks, and 6 and 12 months after their poisoning. Less-severe CO poisoning
was defined as no loss of consciousness (LOC) and COHb saturation ≤15%. More
severe CO-poisoning was defined as involving LOC or a COHb saturation of >15%.
Two-hundred and one patients were included in the more severe CO-poisoning group
and 55 in the less severe CO poisoning group. Cognitive sequelae occurred in 39%
of the first group and in 35% of the second group. The difference was not statistically
significant at any time. Thus, both the more and the less severe groups had signi-
ficant cognitive sequelae. Therefore, CO-induced cognitive outcomes appear to be
unrelated to measures of apparent CO poisoning severity.
One could ask why it has taken so long for the effects of less severe, acute, and
chronic CO poisoning to be recognized. With respect to chronic CO poisoning, this
has come about primarily through the work of neuropsychologists such as Bronstein
et al., Ryan, Hartman, Pinkston, Devine, Helffenstein, Hopkins, and others,2,3,5−9
toxicologists such as Hay and Penney1,10,11 and epidemiologists such as Ritz, Morris,
and others,12,13 not largely through the work of physicians. This is probably owing
to the fact that most of the symptoms produced by chronic CO poisoning are not
considered seriously by those in mainline medicine and the practice of neuropsycho-
logical evaluation is a discipline that has grown up outside of internal medicine. In
my case as a specialized toxicologist, I may have recognized them because they were
thrust into my face by victims of chronic CO poisoning, and moreover since I entered
the field from pure science and had no preconceived ideas about what to expect.
24.5 CONCLUSION
This study of a series of 61 patients demonstrates that chronic CO poisoning is not
without long-term health consequences. The study results show not only the frequency
of the reporting of symptoms known to be part of the Carbon Monoxide Poisoning
Syndrome1 but now also the intensity with which each of these symptoms is reported.
Again, we see a multiplicity of symptoms and signs in the five arenas, many more than
most of the other usual or more common diseases seen by health care professionals.
And again, these symptoms and signs appear to involve a number of organ systems, but
all with the common connection of the central nervous system and the well recognized
damage that CO can do to this organ system.
References
1. Penney, D.G. Chronic carbon monoxide poisoning. In: Carbon Monoxide Toxicity,
D.G. Penney, Ed.,CRC Press, NY, 2000, Chapt. 18, pp. 393–418.
8417: “8417_c024” — 2007/9/11 — 12:14 — page 565 — #15

2. Helffenstein, D.A. Cognitive and behavioral sequelae of chronic CO poisoning: Data

on large case series. In: Carbon Monoxide Poisoning, D.G. Penney, Ed., CRC-Taylor
and Francis Press, NY, 2007, Chapt. 23.
3. Hopkins, R.O. Neurocognitive and affective sequelae of carbon monoxide poisoning
In: Carbon Monoxide Poisoning, D.G. Penney, Ed., CRC-Taylor and Francis Press,
NY, 2007, Chapt. 23.
4. Chambers, C., Hopkins, R.O., and Weaver, L.K. Cognitive and affective out-
comes compared dichotomously in patients with acute carbon monoxide poisoning.
Undersea Hyperbar. Med., 48, 2006.
5. Bronstein, A.C., Kadushin, F.S., and Teitelbaum, D.T. Neurobehavioral findings in
two cases of chronic low level carbon monoxide poisoning Veter. Human Toxicol.,
29, 479, 1987 (abstract only).
6. Ryan, C.M. Memory disturbances following chronic, low-level carbon monoxide
exposure. Arch. Clin. Neuropsychol., 5, 59–67, 1990.
7. Hartman, D.E. Neuropsychological Toxicology Identification and Assessment of
Human Neurotoxic Syndromes (2nd ed.), Plenum Press, New York and London, 1995.
findings in carbon monoxide poisoning: Deficits seen in a matched pair. Arch. Clin.
Neuropsychol., 15, 545–553, 2000.
9. Devine, S.A., Kirkley, S.M., Palumbo, C.L., and White, R.F. MRI and neuropsy-
Perspect., 2002; 110: 1051–1055.
10. Hay, A.W.H., Jaffer, S., and Davis, D. Carbon Monoxide Support. Effects of chronic
exposure to CO:Aresearch study conducted by CO Support. Technical Paper. October,
1997. 47 pp, appendices.
11. Hay, A.W.H., Jaffer, S., and Davis, D. Chronic carbon monoxide exposure: The CO
Support study. In: Carbon Monoxide Toxicity, D.G. Penney, Ed., CRC Press, NY,
2000, Chapt. 19, pp. 419–437.
12. Ritz, B., and Yu, F. The effect of ambient carbon monoxide on low birth weight
among children born in Southern California between 1989 and 1993. Environ. Health
Perspect., 107, 17–25, 1999.
13. Morris, R.D. Low-level carbon monoxide and human health. In: Carbon Monoxide
Toxicity, D.G. Penney, Ed., CRC Press, NY, 2000, Chapt. 17, pp. 381–391.
Editor’s note: On March 7, 2007 I presented a 50 minute talk in a room at the House
of Lords, London, UK, along with others. Several lords and members of parliament
were in attendance at various times. The meeting was organized by “CO Awareness”,
an English “charity”, whose mission is the reduction of injuries and deaths from unin-
tentional CO poisoning in that country. I spoke about the dangers of CO poisoning, its
proper diagnosis, this book, and particularly about the patient (case) series described
in this chapter.
The impression I gained at the meeting and in talking to a number of parti-
cipants, live and via E-mail, was that some British are suspicious of statistics released
by the government on the numbers of deaths from CO poisoning. Many believe
the numbers are far higher than official values. They are also frustrated by what
they perceive as inaction on the part of governmental agencies that could address
the problems of CO exposure. For more on this, see the comments by Rob Aiers
8417: “8417_c024” — 2007/9/11 — 12:14 — page 566 — #16

in Chapter 11, who manages the number one website for CO poisoning matters,
“carbonmonoxidekills.com”.
Another impression I gained at the House of Lords meeting is that in Britain
there are inadequate numbers of people with special training and experience in CO
poisoning. Victims see, for the most part, only physicians, who are principally general
practitioners, and who are unschooled and inexperienced in toxicology and in the
outcomes of CO poisoning. The same can be said for the neurologists, internists,
cardiologists, etc. that they see. As discussed elsewhere (Chapters 14 and 19) these
are not generally the health professionals who can be most helpful in the diagnosis,
testing and management of people with CO poisoning. In some respects this is similar
to the situation in the USA.
Finally, it was my impression that high quality neuropsychological evaluation
for CO victims such as we have in the USA (see Chapters 22, 23 and 25) may not
be encouraged by the medical establishment and/or is unavailable to most victims
of CO poisoning in Britain, even though it is often needed. Since the majority of
the lasting health effects of CO poisoning generally occur in the cognitive-memory
and affective-emotional arenas, and neuropsychological evaluation is considered the
gold-standard for assessing CO-induced brain damage, this problem is extremely
serious with regard to the proper testing and long-term management of CO poisoned
patients.
8417: “8417_c024” — 2007/9/11 — 12:14 — page 567 — #17

8417: “8417_c024” — 2007/9/11 — 12:14 — page 568 — #18
25 Functional and
Developmental Effects of
Carbon Monoxide
Toxicity in Children
Carol L. Armstrong and Jacqueline Cunningham
CONTENTS
25.1 Prevalence and Symptoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569

25.2 Mechanisms of Toxicity Effects on the Brain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 570
25.3 Evidence of Structural Brain Injury . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 571
25.4 Neurobehavioral Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 571
25.5 Carbon Monoxide Toxicity as a Model of White Matter Injury . . . . . . . . . . 572
25.6 Developmental Effects of Carbon Monoxide Toxicity . . . . . . . . . . . . . . . . . . . . 573
25.6.1 Pediatric Vulnerability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 573
25.6.2 Age as a Critical Variable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 574
25.6.3 Developmental Neuropsychological Approach to Assessing
Toxicity Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 575
25.7 Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 577
25.7.1 Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 577
25.7.2 Case Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585
25.8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 586
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
25.1 PREVALENCE AND SYMPTOMS

Carbon monoxide (CO) poisoning is a relatively common environmental toxic
exposure risk that can masquerade as other illnesses. Attribution to CO toxicity
can be made based on clinical findings and knowledge of a malfunctioning heater,
without evidence of elevated carboxyhemoglobin (COHb) levels or acute mental
status changes. In the United States, CO toxicity is implicated in more than 40,000
emergency department visits made annually.1,2 CO is found wherever there is incom-
plete combustion of carbonaceous material, and, after carbon dioxide, it is the most
569
8417: “8417_c025” — 2007/9/11 — 12:14 — page 569 — #1

abundant air pollutant.3 Common sources of CO are fires, faulty combustion heating
systems, exhaust from internal combustion engines, and heating gases other than
natural gas. Auto exhaust and exhaust gases from oil heat were most commonly asso-
ciated with elevation of COHb in a study of serially admitted children to an urban
emergency department who had elevated COHb levels.2 Normal COHb saturation
is 0.4–0.7% at rest, while the ambient level is 0.5–1.5% in the general population
due to added environmental exposure. Tobacco smokers have COHb levels ranging
from 4% to 20%; the mean for one-pack/day smokers is 5–6%. COHb levels may be
higher during pregnancy. Infants of mothers who smoke may have COHb elevations
up to 4.3%.
The presenting symptoms of a chronic CO exposure are occult, different and less
specific than those that are more readily (though not easily) diagnosed in an acute/peak
exposure. These differences in presenting symptoms, neuroimaging, COHb measure-
ments, and other clinical characteristics often lead to misdiagnosis.4 Symptoms of
chronic exposure may be mistaken as flu-like symptoms or other etiology, especially
if there was no loss of consciousness (LOC).
Early symptoms of an acute or a peak CO exposure are more obvious. These
may include headache (experienced at COHb levels of ≥ 10%), fatigue and leth-
argy, dizziness, paresthesias, chest pain, palpitations, and nausea. Severe exposures
result in obtunded consciousness, reducing the victim’s ability to recognize danger.
Peak exposure may be followed by an acute encephalopathy, abnormal hyperin-
tensity signal changes in the brain on magnetic resonance imaging (MRI), and
long-lasting neurobehavioral and cognitive changes. Delayed sequelae, or enceph-
alopathy, follows a period of apparent recovery after acute encephalopathy. Problems
in functioning can have a delayed onset, and appear suddenly after days or weeks
of apparent normal functioning. Symptoms may progress to severe neurocognit-
ive impairments and neuropsychiatric (personality) changes, indicating evidence of
hypoxic/anoxic brain damage.
25.2 MECHANISMS OF TOXICITY EFFECTS ON THE

BRAIN
The harm created by CO exposure is mainly due to reduction in oxygen transport
rather than to any significant poisoning effect on respiratory enzymes.5 The major
determining factors of toxicity are CO concentration, duration of exposure, and alve-
olar ventilation. The disruption to cerebral blood flow created by reduction in oxygen
transport produces brain injury through hypoxia. The basic mechanisms of behavioral
changes are hypoxic brain damage resulting from: (1) the displacement of O2 when
CO binds with hemoglobin and interferes with O2 delivery to tissues, (2) the inhibi-
tion of mitochondrial oxidative respiration and cardiomyopathy due to displacement
of O2 from myoglobin by CO, with associated hypotension and systemic acidosis,
and (3) the increased permeability of the vascular endothelium caused by CO. Neural
inflammatory processes may also be involved in delayed neurobehavioral problems
because of brain iron extravasion into neural tissues and release of myelin basic
protein following CO poisoning.6
8417: “8417_c025” — 2007/9/11 — 12:14 — page 570 — #2

Functional and Developmental Effects of Carbon Monoxide Toxicity in Children 571
Minimal effects on cognitive performance are reliably observed at 7% COHb

level, and a 5% COHb level causes decrements in vigilance tests and in difficult
dual task tests. A 2–4% COHb level has been associated with reduced video game
performance.7−10
25.3 EVIDENCE OF STRUCTURAL BRAIN INJURY

Although initial COHb levels do not predict later impairment well, radiographic
neural changes are predictive of functional impairments. Peak exposures are more
likely to produce obvious structural changes on clinical MRI scans of the brain when
caused by hypoxia and by cardioembolic or hypotensive strokes related to myocar-
dial ischemia. The most common MRI finding on T2-weighted images and Fluid
Attenuation Inversion Recovery (FLAIR) is often, but not always, bilateral symmet-
ric hyperintensity of the white matter, greater in the centrum semiovale, with relatively
less involvement of the temporal lobes and anterior frontal lobes.11,12 The incidence
of white matter hyperintensities varies in natural history group studies. In one large
prospective controlled group study (N = 73), the incidence was 12%, significantly
more in the periventricular area than in the centrum semiovale.13 Cognitive impair-
ment was greater in those patients with hyperintensities in the centrum semiovale,
and signal changes persisted over the 6-month study period. Atrophy is common
in chronic CO exposure. Most reports of structural abnormalities are based on case
reports, but one group report cited cerebral cortical atrophy in 60% of adults, mild
atrophy of cerebellar hemispheres in 50%; vermian atrophy in almost 70%, with less
common atrophy found in the globus pallidus and corpus callosum.6,12 Atrophy of
the hippocampus,11 basal ganglia,14 thalamus and medial temporal lobe15 can also
occur. Structures with high-energy demands, including the hippocampus and the cere-
bellum, are particularly vulnerable to hypoxia,11 as are those high in iron content,
including the basal ganglia.16 Initial Computed Tomography (CT) or MRI often fails
to detect low-density lesions in the globus pallidus, a frequent site of injury, and has
poor clinical correlation as abnormalities seen on the scan may be found in both well-
recovering and comatose patients. At follow-up, lesions may disappear, diminish, or
remain unchanged, and resolution of white matter lesions, ventricular enlargement,
and cortical atrophy later on may be represented by an improved clinical picture with
persistent neurocognitive and neuropsychiatric impairment at a less severe level.6,17,18
However, neurocognitive impairments are likely even when neurological symptoms
resolve. Visual evoked potentials are often done, but may miss effects, if white matter
lesions result in demyelination in other parts of the brain.19 The most robust findings
may be found at long-term follow-up.
25.4 NEUROBEHAVIORAL FINDINGS

Neurobehavioral symptoms vary and correspond with damage to cortex, subcortical
nuclei, and white matter. The following symptoms and signs have been reported:
amnesia,11,20 apraxia,21 visual object agnosia,22 akinetic mutism,23,24 parkinsonism,25
choreoathetosis,26 peripheral neuropathy,27,28 cortical blindness,29,30 depression and
8417: “8417_c025” — 2007/9/11 — 12:14 — page 571 — #3

anxiety,31 and impaired attention and executive functioning.32 Varied memory pro-
cesses have been implicated in CO toxicity, including incidental memory, recognition,
paired associate learning, delayed recall, and visual short- and long-term memory,
without effects on intelligence, visuospatial functions, speed, dexterity, nor letter
fluency.33
25.5 CARBON MONOXIDE TOXICITY AS A MODEL OF

WHITE MATTER INJURY
Chang et al.34 identified three types of white matter injury that occur following CO
poisoning: (1) multiple small necrotic foci in the centrum semiovale and interhemi-
spheric commissures; (2) areas of necrosis in the deep periventricular white matter that
is associated with axonal destruction and lipid-laden macrophages; and (3) demyelin-
ation in the deep white matter.13,32,35,36 The multifocal and diffuse injuries associated
with CO toxicity link it with traumatic brain injury (TBI) in terms of its common-
ality with the latter as a white matter disease.37 Whereas the acute clinical course
of CO toxicity is differentiated from that of deceleration or rotational brain injury,
the chronic state of either condition is associated with events, including the produc-
tion of free radicals and excitatory amino acids and the disruption of normal calcium
homeostasis, which act in concert to exacerbate the hypoxic-ischemic insult that
can occur in association with diffuse axonal injury (DAI).38 Given the commonality
of neurochemical mechanisms underlying axonal injuries, TBI may be a relevant
comparison when examining the neurobehavioral effects of CO toxicity, as well as
in speculating on the responsiveness of either condition to similar approaches to
rehabilitation.39
In the cognitive domain, a pattern of relative strengths and weaknesses has
long been known to characterize DAI: the relative preservation of “crystallized”
and prior-learned knowledge, and the severe effects on “fluid” processing,40
which is resource-limited. Resource-limited processes require higher levels of pro-
cessing resources for optimal performance and are affected by inter-and intra-task
competition.41
Fluid intelligence is largely nonverbal and a culture-reduced form of mental effi-
ciency that is used when a task requires adaptation to a new situation. Crystallized
intelligence is content-related, and represents the corpus of what has been learned.
Whereas no narrowly defined cognitive pattern has been identified as characterizing
CO toxicity,33 a frequent cognitive profile obtained in the neuropsychological eval-
uation of the CO-poisoned patient is one that mirrors that found in TBI in that the
profiles demonstrate relative sparing of crystallized abilities, but serious involvement
of fluid processes.42 Most common deficits in the CO-poisoned patient are observed
in executive functioning (i.e., planning, monitoring, resisting distraction, and main-
taining flexibility in thinking), attention and concentration, memory, visual perceptual
abilities, and speed of information-processing.42 In both adults and children, deficits
in these areas are also expected to occur, singly or in combination, as sequelae to
closed head injury.43
8417: “8417_c025” — 2007/9/11 — 12:14 — page 572 — #4

25.6 DEVELOPMENTAL EFFECTS OF CARBON

MONOXIDE TOXICITY
25.6.1 PEDIATRIC VULNERABILITY
It is important to differentiate the typical CO exposures in adults as compared to
children. Adults often are exposed to more lethal doses, but children have heightened
vulnerability to CO toxicity based on constitutional factors.44 Low birth weight is a
well-known sequelae of CO exposure.45,46 Across all pediatric age groups, roughly
50% of fatalities are unintentional and of those, 50% are related to motor vehicle
exhaust inhalation.47 Parameters that are relevant to neural development are the dur-
ation of exposure (chronic or peak exposure); acute, subacute or delayed symptoms,
timing of exposure (prenatal vs postnatal), and dose (moderate levels analogous to
cigarette smoking or higher levels).
Symptoms in children appear to manifest more variably than in adults, with
even poorer clinical correlation with a wider range of COHb levels,16 and with vari-
able outcomes occurring even in related individuals with similar exposure levels.48
This variability can lead to the conclusion that no clear relationship can be pre-
dicted between severity of CO poisoning (COHb levels at admission) and residual
neuropsychological deficits.13,48 Even in cases of known elevated COHb levels,
some individuals demonstrate little chronic cognitive impairment, although a degree
of impairment remains likely. There are very few reported case outcomes for
CO-exposed children. Severe memory impairment was reported in four pediatric
cases with chronic exposure over several years; in three of the four children, expos-
ure was lower-level and without LOC.33 Physiological factors putting children at
greater risk than adults include children’s rapid uptake of CO into the bloodstream,
high metabolic needs coupled with small blood volume, reduced pulmonary trans-
port of CO while sleeping (occurring for longer periods in children), and reduced
pulmonary transport of CO due to young age.16
There is also heightened risk for children, as compared with adults, from expos-
ure to CO in utero. CO is a known potent fetotoxin and teratogen with prenatal
toxicity retarding growth and development.49 Direct effects on cognition are shown
by translational studies. Long-lasting learning and memory deficits have been seen
to result from gestational exposure of rats to levels of COHb in the maternal blood
supply equivalent to cigarette smoking in humans. Postnatal exposure to this dose
of CO from 1 to 10 days postbirth has resulted in no memory or learning deficits at
3 and 18 months of age in rats.50 However, exposure to the same level of CO from
day 0 to day 20 of gestation in rats has led to impairments in habituation, working
memory, and the ability to explore novel objects, but no alteration in spontaneous
motor activity.51 The behavioral effect of prenatal exposure may be caused in part
by changes in mesolimbic dopaminergic transmission52 and in cholinergic and cat-
echolaminergic pathways.53 In one study, guinea pigs were exposed to CO for 10 h
a day from Day 23 to 25 of gestation until term at 68 days. The fetus brains were
examined 5–7 days prior to birth. Prenatal exposure to CO affected cholinergic and
catecholaminergic pathways in the medulla, particularly in cardio-respiratory centers,
a finding which may be related to smoking and increased rate of sudden infant death
8417: “8417_c025” — 2007/9/11 — 12:14 — page 573 — #5

syndrome.53 Prenatal exposure can also cause developmental alterations in the areas
of the brainstem responsible for respiratory control, causing increased sensitivity to
CO after birth.54
Other factors mediate the impact of CO as a toxicant on child development,
some potentiating and some moderating. The level of CO poisoning, age at the
time of exposure, and the intellectual level of the child mediate long-term functional
capacity.55
25.6.2 AGE AS A CRITICAL VARIABLE

Because the growth of children is incomplete at any one point in time, the devel-
opmental stage reached at the time of a brain insult holds particular importance
in assessing the injury’s significance. The primary challenge to successful coping
imposed by pediatric brain injury, is that an acquired brain injury generates “rippling
effects” that have long-term consequences over the whole course of a child’s future
development.56 The most enduring effects are thus expected to occur at the youngest
ages when the child has not yet had the opportunity to master fundamental develop-
mental tasks. In the period from birth to 2 years of age, the child is expected to achieve
an understanding of cause-and-effect relationships. By understanding that certain
events are routinely paired together, he/she becomes capable of basic self-regulation,
an essential achievement that paves the way for the ability to integrate thinking,
emotion, and behavior—the task of the developmental period from 3 to 5 years. In
the middle-childhood years, the achievement of classroom self-organizational skills
forms the precursor for the development of the planning and goal-direction capacit-
ies necessary for movement and problem-solving through the middle school years.
Failure to develop age-appropriate skills at any stage in the developmental trajectory
thwarts the development of the judgment skills eventually necessary for successful
progress toward adolescent autonomy.56 The commonly reported finding that per-
manent deficits in mood regulation and executive functioning, in areas including
decision-making and attention-monitoring57 frequently follow CO poisoning, war-
rants looking at developmental factors which are influential in the expression of
impairment.
Disruptions in achieving major milestones in the first year of life create lock-step
constitutional changes that affect overarching consequences as the child matures.
For example, the infant who fails to understand that certain events are routinely
paired together becomes despondent when the mother does not immediately attend
to his/her cry, resulting in the infant’s difficulty in regulating the sleep-wake cycle.58
Taking a developmental perspective on pediatric CO toxicity extends the focus
from the physiological mechanisms causing dysfunction to psychosocial consid-
erations. In the absence of such a perspective, valuable information may be lost
because subtle signs, such as irritability, are not detected as markers of organi-
city. The paradoxical results obtained in a study59 that rated infants with relatively
high levels of COHb as asymptomatic have in hindsight been attributed to failure
to consider the significance of feeding difficulty and irritability in interpreting the
findings.16
8417: “8417_c025” — 2007/9/11 — 12:14 — page 574 — #6

Empirical findings of CO toxicity in children, both in utero and in early child-

hood, have shown wide-ranging degrees of neurobehavioral injury. Faulty heater
installation resulted in a family being exposed to high levels of accumulated CO in
an apartment in which they had lived for 1 year. A 4-year-old boy and a 5-year-old
girl were hospitalized with COHb levels of 20% and 13% respectively, and were
treated with oxygen through masks. Over the course of a year following exposure,
behavioral symptoms were recurrent: headaches, lethargy, sleepiness, hyperactivity,
violent tantrums, and extreme mood swings in the boy, and irritability, intermittent
fatigue, and falling asleep in school in the girl. However, testing of cognitive ability
(receptive language, vocabulary, adaptive behavior) was normal in both children 5
months after admission for treatment. The mechanism of the effect was thought to be
hypoxemia (rather than hypoxia) partly because psychometric tests were normal.60
In another study, a 22-year-old pregnant woman exposed to elevated CO levels, who
developed neurologic symptoms as well as tachycardia, tachypnea, signs of preterm
labor, and a mildly elevated COHb level, delivered her baby at full term, without
sequelae for the infant, following treatment with hyperbaric O2 .61
25.6.3 DEVELOPMENTAL NEUROPSYCHOLOGICAL APPROACH TO

ASSESSING TOXICITY EFFECTS
The particular imprint CO toxicity bears on a child’s functioning may represent the
toxicant’s direct impact on specific cell groups or brain structures. Impact on neural
structures which are vulnerable to CO toxicity and which are essential to the expres-
sion of certain functions will be represented as focal effects of injury. As examples,
injury to the globus pallidus and the basal ganglia will affect the development of
motor repertoires, and injury to structures critically involved in memory, such as the
hippocampus, will disrupt this domain of cognitive functioning.11 In contrast to the
specific brain-behavior relationships which can, at least to a significant degree, be
ascribed to injury in the well-differentiated adult brain, the assessment of injury in
the less-differentiated brain must consider the injury’s potential for interfering with
complex processes that interact with one another and that represent consequences of
both biological and psychosocial developmental forces. A multiplicity of neural and
nonneural factors interact to determine a toxin’s impact in children’s functioning.62
In the child, whose behavioral repertoire is in the process of developing, toxins do not
simply hit given neural structures—the “where”; they hit a “what” at both a “where”
and a “when”.63 Because knowledge of a toxin’s particular signature in behavior
demands a consideration of how injury to specific structures (the “where”) disrupts
a developing system over time (the “what” and the “when”), the evaluation of the
functional effects of toxicity in a child suspected of CO poisoning is necessarily
multifaceted.
A developmental neuropsychological approach to assessing effects of toxicity
proposed by Bernstein64 stipulates that two levels of data-gathering are needed to
account for the interactions possible between exposure to a toxicant and the devel-
opmental events occurring at the time of exposure. At a first level of assessment,
data are obtained in broad domains of functioning to determine whether or not
8417: “8417_c025” — 2007/9/11 — 12:14 — page 575 — #7

deficits actually exist. Domains of interest include behavioral regulation, attentional

capacities, learning skills, memory capacities, problem-solving skills, motor skills
(gross, fine, graphomotor), sensory capacities, general cognitive ability, commu-
nicative competence/language abilities, visuospatial processing, social cognition,
socio-emotional status/adjustment, and academic achievement. These domains are
tapped by the direct physical and neurological examination of the child, administration
of standardized tests, and observation of the child’s behavior in both naturalistic and
structured situations. Naturalistic observations complemented by information from
behavioral questionnaires and history-taking interviews, contributed by caretakers
including parents, teachers, and health-care professionals, form relevant components
of assessment. Their importance resides in the fact that the well-structured condition
of formal testing may not detect the derailing of self-regulatory functions in the child’s
familiar and habituated settings. Often, the injured child’s behavior is much worse at
home than in school in part because behavioral disruption may actually be leading
to a child’s inability to organize moment-to-moment behavior when the situation is
not strongly structuring environmental stimuli. That such derailing is possible as a
result of toxicity requires documenting the child’s ability to adapt to the demands of
ordinary situations as an important aspect of evaluation, one usually based on family
report.
As proposed by Bernstein,64 the second level of a developmental neuro-
psychological assessment moves from identifying deficits to determining their
linkages to a particular neurotoxicant. Knowledge of the toxicant’s mode of action
and predilection for specific brain systems and/or processes, as well as of the devel-
opmental significance of the timing, amount, and duration of exposure, determines
the ability to address the questions of specificity, sensitivity, and causation posed at
this level. Information on the where, when, and what of CO toxicity comes from
animal studies on early brain injury. However, the “where” of CO toxicity (decreased
oxygenation of blood) and the “when” (last trimester) in the association between
developmental age and CO’s power to produce injury, leave open the knowledge of
“what” is disrupted as a result of damage.
In work with pre- and postgestational rats and kittens, different functional out-
comes have been robustly associated with the precise age at injury.65 The most optimal
outcomes have been associated with embryonic stages characterized by maximal
astrocyte generation and synapse development, when the developing brain carries
the potential for recovery due to the redundancy in neural generation occurring
at that point. Conversely, the developing brain is at its worst time for compens-
ation when it is in a stage of neural migration and of the initiation of synaptic
formation. In rats, this is the first week of life, a period that is likened to the
third trimester in the human fetus.65 During a period when the fetus is most vul-
nerable to CO poisoning owing to pulmonary development, it is also at a stage
when genetically-programmed developmental processes work against its ability
to withstand the effects of neural toxicity. Thus, the timing of a CO insult can
cause cumulative biological risks66 to a developing system, which can be diverse
and lifelong in their aversive impacts).67 Four case studies are next presented to
describe the wide-ranging variability in outcomes possible in cases of pediatric
CO toxicity.
8417: “8417_c025” — 2007/9/11 — 12:14 — page 576 — #8

25.7 CASE STUDIES

The following case studies are shown to illustrate the complexity of consequences
possible in young children following CO exposures. The first three cases, CB, LB
and DB, are siblings with chronic lower-level exposures lasting from 4.5 years in
two children (ages 18 months and 3 years at the beginning of exposure), to lower
level in utero exposure plus 2.5 postnatal years in the youngest. A fourth case is
presented, HK, who had chronic exposure in utero and postnatally, with a probable
peak exposure at almost two years of age.
Table 25.1 summarizes psychometric test data as well as information obtained
clinically. The children’s neuropsychological findings demonstrate a range of outcome
morbidity. Test batteries for the children vary largely because of the applicability of
measures at different age groups. For example, there was no neuropsychological
battery for the 4-year-old at the time of the assessment, and the Wechsler Intelligence
Battery for Children68 was a primary instrument, whereas by age six, a theoretically
based neuropsychological battery 69 could be used. However, the constructs that the
different tests measured are given, and all tests are converted to a single standard (i.e.,
percentile), in order to facilitate comparisons.
25.7.1 OBSERVATIONS
The two cases exposed in utero (DB and HK) showed respiratory symptoms
(postnatal plethora, chronic respiratory infections, asthma, premature lung devel-
opment). Respiratory problems were not observed in the other two children whose
exposure onset was at 4.5 and 2.5 years of age. Skin conditions were observed in
three of the four children, all from the same household and thus the same exposure
conditions. Systemic problems that were observed in all cases were nervous habits,
difficulty concentrating, emotional lability and irritability, and attention deficit dis-
orders. Partial autistic behaviors were observed in the child with the most severe
exposure (HK), which included regression in speech at age two years, poor eye con-
tact, difficulty playing rule-guided games, lack of imaginative play, and little social
cognition. Furthermore, the longitudinal observations in this child with the most
severe exposure indicated that he changed from a pleasant-appearing but poorly con-
centrating child at age four years, to a rigid, fearful, and easily frustrated child at age
six and a half.
Motor development was delayed in two of the four children (LB and HK). Some
dyspraxia was measured in all four children. In the child with longitudinal data, his
ability to copy figures (constructional or graphomotor praxis) declined significantly
in relation to developmental expectations. By 6.5 years of age, he was significantly
impaired in all motor skills tested. Tactile discrimination (identifying two fingers
touched simultaneously) was impaired in the two oldest children in whom this could
be tested.
Developmental language impairment was often observed by the parents as poor
comprehension or impaired articulation, and slow or regressed speech development
in two children (LB and HK). The child with the most severe exposure as well as
a history of familial reading disability (HK) had receptive and expressive language
8417: “8417_c025” — 2007/9/11 — 12:14 — page 577 — #9

578
TABLE 25.1
Summary of Psychometric Test and Clinical Data for Four Case Studies
Cases
Case Initials CB LB DB HK Time 1 HK Time 2

Clinical characteristics
Age at initial CO 3 years 1.5 years From conception Chronic low-level
exposure in utero, and peak
exposure age 1 year, 11
months
Years of CO exposure 4.5 years 4.5 years 2.5 years + in utero Possible chronic exposure
in utero, one
documented peak
exposure with
unresponsiveness,
vomiting, diarrhea,
pallor, dilated, equal,
slowed pupil responses;
other less severe peak
exposures possible.
Total exposure 2 years +
in utero
Peak or chronic CO Chronic Chronic Chronic Chronic + peak
exposure
Source of CO exposure Faulty furnace emissions Faulty furnace emissions Faulty furnace emissions Faulty furnace Faulty furnace
and ventilation into and ventilation into and ventilation into
8417: “8417_c025” — 2007/9/11 — 12:14 — page 578 — #10

house house house
Age at test 8 years, 7.5 months 7 years, 0 months 4 years, 4.0 months 4 years, 1.5 months 6 years, 6.0 months
Symptoms observed by Chronic cough, fatigue, Skin bumps, flushing of Red skin at birth that Asthma onset postnatally, Symptoms persist with
family without other skin bumps, itching, cheeks, speech normalized, became premature lung exception of
known cause restlessness, difficulty problems, loss of plethoric 2 days after development, failed to hyperactivity but
concentrating, difficulty appetite, nausea, birth and required crawl, hyperactive, falls Attention Deficit
learning, sleep irritability, severe suctioning, chronic often, impaired Disorder persists, poor
disturbance, chronic constipation, incident of cough and respiratory concentration, emotional attention/concentration,
diarrhea, frequent acute flu-like symptoms problems that onset lability, bites fingernails, poor comprehension,
muscle cramps, and cyanotic hands, fine postnatally, often tired, temper tantrums, overly dependent on
nausea/vomiting, motor incoordination low energy, frequent fatigued parents, partial
irritable, hyperactive, and dressing difficulty at knee pain, impaired autistic-like behaviors
episode of lethargy and age of 6 years, problems articulation and reduced (regression in speech at
hallucination, nervous of speech articulation intelligibility of speech, age 2 years following
habits, messy dysphagia, asthma, peak exposure) with
habits/smearing of feces, borderline anemia recovery of speech,
temper tantrums lacks imaginary play,
poor eye contact,
difficulty playing
rule-guided games, little
interest in others
Hand dominance Right Right Right Right No change
Other risk factors None known None known None known Paternal reading disability Same
including family (mild); mother smoked
history 15 cigarettes/day during
pregnancy
Parental educational
level
Mother 4 years college 4 years college 4 years college 12th grade Same
Father 2 years college 2 years college 2 years college 10th grade
Depression None, though doesn’t None, happy, None, happy, outgoing, None reported, pleasant None, fearful, easily
Functional and Developmental Effects of Carbon Monoxide Toxicity in Children
8417: “8417_c025” — 2007/9/11 — 12:14 — page 579 — #11

enjoy school even-tempered verbally quiet presentation frustrated, rigid
(Continued)
579
TABLE 25.1
580
(Continued)
Cases

Early motor and speech Normal Motor development Normal Developmental and Globally impaired, with
development normal prior to CO neurological evaluations greater expressive
exposure; speech at age 2 years speech
development slowed 7–8 months revealed impairment—first
following 6 months of clumsiness, impaired sentence structure at age
CO exposure balance, lack of speech 5.5 years, communicates
or language progression primarily by gesture;
after 3 months of genetic syndrome and
therapy; delay in autism ruled out
behaviors: cognitive
8 months, gross motor
10 months, fine motor
7 months,
personal/social 9 months
Timing of onset of Slow, first noticed after Slow Lethargy and viral-like Respiratory symptoms
symptoms 2 years of exposure symptoms postnatally postnatally;
speech/motor symptoms
after peak exposure
Neuropsychological findings
Sensory and motor function
Copying hand 9th percentile 9th percentile 25th percentile 5th percentile
movements
8417: “8417_c025” — 2007/9/11 — 12:14 — page 580 — #12

Copying figures 25th percentile (NEPSY) 9th percentile (NEPSY) 63rd percentile (NEPSY) 66th percentile (Beery) 5th percentile (NEPSY)
Grooved Pegboard test 63rd percentile on right 32nd percentile on right
25th percentile on left 19th percentile on left

Visuomotor precision 75th percentile
Oral praxis 1st percentile 1st percentile
Finger localization test Impaired: age seven Impaired: age six
equivalent equivalent
Body part naming 50th percentile
Attention and information processing speed
Symbol digit 47th percentile n/a
modalities test—oral
version (coding)
Coding (digit symbol 84th percentile
subtest—written
version)
Trail making test—A 31st percentile 1st percentile
(tracking of numbers)
Trail making test—B 19th percentile 1st percentile
(tracking of letters
and numbers)
Visual cancellation 63rd percentile 63rd percentile 75th percentile 1st percentile
tests
Language
Word fluency 99th percentile 5th percentile
(phonemic)
Semantic fluency 84th percentile >80th percentile 75th percentile
Token test (linguistic 81st percentile 19th percentile 1st percentile
comprehension)
Comprehension of 84th percentile
instructions
Peabody picture 77th percentile 77th percentile 73rd percentile <1st percentile 1st percentile
8417: “8417_c025” — 2007/9/11 — 12:14 — page 581 — #13

vocabulary test-III
(Continued)
581
582
TABLE 25.1
(Continued)
Cases

Vocabulary knowledge 5th percentile
Boston naming test 77th percentile 39th percentile
Picture naming 99th percentile
Speeded naming 37th percentile 25th percentile
Spelling— 13th percentile 42nd percentile
phonological
decoding
Reading (word and 8th percentile 77th percentile
sentence)
Sentence repetition 63rd percentile 2nd percentile
Written expression Impaired graphia, High rate of letter and Severely impaired
numeral reversals, poor numeral reversals (58% graphia, unable to
allographic alpha numeral reversals) sequence alphanumeric
quality, impaired symbols, impaired
number sequencing, orthographic quality to
symbol matching letters of his name
Verbal memory
Paragraph story recall 84th percentile 37th percentile 75th percentile
Face-name associative 84th percentile 75th percentile
memory
Word list learning, 75th percentile 16th percentile 1st percentile
5 trials
8417: “8417_c025” — 2007/9/11 — 12:14 — page 582 — #14

Learning rate over 92nd percentile 23rd percentile 1st percentile
5 trials
Recall after distractor 28th percentile 15th percentile 1st percentile
list
Recall after 30 min 77th percentile 3rd percentile 1st percentile
Visuospatial perception and memory
Benton facial 59th percentile 50th percentile
recognition
Road map test 85th percentile 20th percentile
Visuospatial Average 6th percentile (mean)
performance subtests
from IQ battery
Complex figure copy 57th percentile 31st percentile
Immediate recall of 51st percentile 11th percentile
complex figure
Delayed recall of 39th percentile 6th percentile
complex figure
Memory for faces 50th percentile 25th percentile
Delayed recall of faces 37th percentile 37th percentile
Executive functions
Wisconsin card sorting 68th percentile 84th percentile
test—Categories
achieved
Wisconsin card sorting 33rd percentile 22nd percentile
test—Perseverative
errors
8417: “8417_c025” — 2007/9/11 — 12:14 — page 583 — #15

(Continued)
583
584
TABLE 25.1
(Continued)
Cases

Porteus maze test 27th percentile 19th percentile
Tower test 25th percentile
Knock-tap test (Go, 26th percentile
no-go task)
Similarities test 1st percentile
Comprehension test 2nd percentile
Wide range 19th percentile 32nd percentile 1st percentile
achievement test-III:
Arithmetic
8417: “8417_c025” — 2007/9/11 — 12:14 — page 584 — #16

impairment in areas including vocabulary knowledge, sentence repetition, and writ-

ten expression at age 6 years. Another child also demonstrated reading impairment
in both word recognition and phonological decoding (spelling). In this child (CB),
her toxic exposure from age 3 to 7 occurred during the period of development of
prereading and productive reading skills—beginning age 5—which is a possible basis
for impairment in her reading ability. Both this child and her sister (who was exposed
from age 1.5 to 5 years of age and who had no impairment in reading) demonstrated
a high rate of letter and number reversals in writing. Such reversals are thought to
represent impairment in tactile or somatosensory mapping,70 and both children had
impairments in tactile mapping. The most severely affected child developed the most
severe graphic impairment. Therefore, development of written expression appears to
be very sensitive to CO toxicity.
Memory impairment was not common among the children with lower-level expos-
ure only, but in the child with a known peak exposure (HK), despite treatment with
oxygen during hospital transfer, memory impairment was severe. The likelihood of
hippocampal involvement in ischemic effects of peak exposure may be greater in this
case. In a second child (LB), with lower-level CO exposure from age 1.5 to 5 years of
age, the only memory deficit was delayed recall (two of three measures were impaired
showing an abnormal decline in recall over time), suggesting hippocampal injury.11
Executive functions require a high level of cognitive control of resource-driven
processing, and are dissociated from crystallized knowledge. However, only the most
severely CO-exposed child demonstrated defective reasoning; the two other children
of sufficient age for such testing, had scores in the low average, but not statistically
impaired range.
25.7.2 CASE DISCUSSION

Bernstein’s recommendations for assessment of toxicity effects on children begin with
the examination of functional domains sensitive to disruption and which include beha-
vioral regulation, social cognition, emotional adjustment, adaptability, attentional
capacities, memory, and specific cognitive abilities. Radiographic demonstration of
common foci of neural injuries (globus pallidus, basal ganglia, hippocampus, and
centrum semiovale) predict residual motor and memory dysfunction. Such injur-
ies carry the potential to interfere with complex processes, resulting from rippling
effects on development. Predictions of functional effects of CO toxicity in these vari-
ous domains appear to be accurate on the basis of assessment findings in the cases
presented. Furthermore, the prediction of aversive consequences of injury during
neural migration and initiation of synaptic formation in the human third trimester,
is consistent with the finding that the child or children with significant CO toxicity
during fetal development demonstrated the strongest effects. The children with the
most severe exposure or with the longest lower-level exposure had the most severe
cognitive disorders.
The children with the lower-level, chronic CO exposures had impairments mainly
limited to motor development and praxis, tactile mapping, and written language
development, with some memory impairment (in delayed recall). The conditions of
their exposure put them at risk for milder neurodevelopmental effects. Chronic CO
8417: “8417_c025” — 2007/9/11 — 12:14 — page 585 — #17

concentrations that are from small, localized sources, in their case faulty furnaces and
ventilation systems, can raise the ambient level and thereby increase the destructive
effects of that local source.71 CO also has a greater effect during cold temperatures
because its effects are amplified by decreased indoor air exchange, increased use
of furnaces, and other indoor sources of combustion,71 all of which also raise the
ambient level that can be augmented to a very dangerous level by a small local
source, such as a faulty furnace. All the children met these risk criteria, especially the
three siblings who resided in the north at approximately 43◦ latitude. CO levels also
have significant adverse cardiological effects, putting at risk those with underlying
disease.
The children demonstrated many of the residual symptoms associated with chronic
CO exposure: tiredness/fatigue, sleep problems, gastrointestinal problems, breath-
ing difficulties, difficulty concentrating, and emotional lability. Moreover, they
demonstrated developmental problems which ran the gamut from mild to severe,
and which disrupted maturation of important adaptive skills, such as dressing,
social engagement, involvement in physical play, learning, and complex skill devel-
opment. Thus, their presentations differed from those expected in adults whose
neurobehavioral impairments following CO toxicity are primarily in memory, con-
centration, and personality change. In these children, their impairments disrupted
future development.
25.8 CONCLUSIONS
This chapter provides an overview of the functional effects of CO poisoning. It
discusses the differences in outcomes predicted for adults versus those for children.
Case studies serve to illustrate the difference between the disruption of a whole
system of development as opposed to changes from an a priori functional status, as
expected for adults. If symptoms in the latter case can be said to represent delta—
change, in children they can be said to represent omega—the future endpoint of a
system. A developmental approach to neuropsychological assessment was found to be
a useful approach to understand the complex factors governing outcomes, including
age at the time of CO exposure, severity and length of exposure, fetal exposure,
source of CO, and risk from raised ambient levels. Mitigating influences, such as the
child’s preinjury intellectual level as demonstrated by parental achievement, were
also relevant to these cases.
As a conclusion to the discussion, it is proposed that a developmental approach to
rehabilitation be also considered important with regard to knowing how to alleviate
consequences of injury. Well-publicized work in the TBI literature has advocated for
such a developmental approach and can provide a guide to approaches to moderating
the far-reaching consequences of a brain insult sustained in development. Brain injury
can be considered a rippling event in a child’s life.56 Rippling biological events are
greater in children than in adults because of the incompletely developed status of
the child’s brain, which needs time and experience to mature. Because new abilities
build on established skills over time, a point of departure driving rehabilitation goals
is the identification of the critical developmental task missed as a result of injury and
8417: “8417_c025” — 2007/9/11 — 12:14 — page 586 — #18

the provision of a program systematically designed to teach mastery of the task. The
critical task for developmental ages, birth-to-two years, is achieving understanding of
cause-effect relationships; at ages, 3–5 years, the integration of thinking, emotions,
and behaviors; at ages, 6–11, school skills; at ages 12–15, planning and organization;
and at 16–19, judgment and autonomy.
A developmental rehabilitative model has clear relevance to the treatment of the
child following a toxic insult. Moreover, given the often insidious character of CO
toxicity, a caretaker’s knowledge of developmental expectations and aberrations may
be essential for generating the rehabilitative decisions and actions that could reduce
or halt a chronic and undetected course of injury created by poisoning.
References
1. Hampson, N.B. Emergency department visits for carbon monoxide poisoning in the
Pacific Northwest. J. Emerg. Med., 16, 695, 1998.
2. Baker, M.D., Henretig, F.M. and Ludwig, S. Carboxyhemoglobin levels in children
with nonspecific flu-like symptoms. J. Pediatr., 108, 501, 1988.
3. Jaffe, F.A. Pathogenicity of carbon monoxide. Am. J. Foren. Med. Path., 18, 406,
1997.
4. Penney, D.G. Chronic carbon monoxide poisoning. In Carbon Monoxide Toxicity,
Penney, D.G., Ed., CRC Press: New York, 2000; p. 393.
5. Robkin, M.A. Carbon monoxide and the embryo. J. Dev. Biol., 41, 283, 1997.
6. Weaver, L.K. and Hopkins, R.O. Hemorrhagic infarction in white matter following
acute carbon monoxide poisoning. Neurology, 64, 1101, 2005.
7. Wright, G., Randell, P. and Shephard, R.J. Carbon monoxide and driving skills. Arch.
8. Insogna, S. and Warren, C.A. The effects of carbon monoxide on psychomotor
function. In Trends in ergonomics/human factors, Mital, I.A., Ed., Elsevier: North.
Holland, 1984.
9. Bunnell, D.E. and Horvath, S.M. Interactive effects of physical work and carbon
monoxide on cognitive task performance. Aviat. Space Environ. Md., 59, 1133, 1988.
10. Benignus, V.A., et al. Effect of low level carbon monoxide on compensatory tracking
and event monitoring. Neurotoxicol. Teratol., 9, 227, 1987.
11. Reynolds, C.R., Hopkins, R.O. and Bigler, E.D. Continuing decline of memory
skills with significant recovery of intellectual function following severe carbon
monoxide exposure: Clinical, psychometric, and neuroimaging findings. Arch. Clin.
Neuropsych., 14, 235, 1999.
12. Durak, A.C., et al. Magnetic resonance imaging findings in chronic carbon monoxide
intoxication. Acta. Radiol., 46, 322, 2005.
13. Parkinson, R.B., et al. White matter hyperintensities and neuropsychological outcome
following carbon monoxide poisoning. Neurology, 58, 1525, 2002.
14. Prockop, L.D. Carbon monoxide brain toxicity: clinical, magnetic resonance imaging,
magnetic resonance spectroscopy, and neuropsychological effects in 9 people. J.
15. Bruno, A.,Wagner, W. and Orrison, W.W. Clinical outcome and brain MRI four years
after carbon monoxide intoxication. Acta. Neurol. Scand., 87, 205, 1993.
16. White, S.R. Pediatric carbon monoxide poisoning. In Carbon Monoxide Toxicity,
Penney, D.G., Ed., CRC Press: New York, 2000; p 463.
8417: “8417_c025” — 2007/9/11 — 12:14 — page 587 — #19

17. Cocito, L., et al. Cognitive recovery after delayed carbon monoxide encephalopathy.
Clin. Neurol. Neurosurg., 107, 347, 2005.
18. Inagaki, T., et al. A long-term follow-up study of serial magnetic resonance images
in patients with delayed encephalopathy after acute carbon monoxide poisoning.
Psychiat. Clin. Neuros., 51, 421, 1997.
19. Vieregge, P., et al. Carbon monoxide poisoning: Clinical, neurophysiological, and
brain imaging observations in acute disease and follow-up. J. Neurol., 236, 478,
1989.
20. Spinnler, H., Sterzi, R. and Vallar, G. Amnesic syndrome after carbon monoxide
poisoning: A case report. Schweiz. Arch. Neurol. Neuroch. Psych., 127, 79, 1980.
21. Min, S.K. A brain syndrome associated with delayed neuropsychiatric sequelae
following acute carbon monoxide intoxication. Acta. Psych. Scand., 73, 80, 1986.
22. Garland, H. and Pearce, J. Neurological complications of carbon monoxide poisoning.
Q. J. Med., 144, 445, 1967.
23. Tengvar, C.,Johansson, B. and Sorensen, J. Frontal lobe and cingulate cortical meta-
bolic dysfunction in acquired akinetic mutism: A PET study of the interval form of
carbon monoxide poisoning. Brain Inj., 18, 615, 2004.
24. Smith, J.S., Brierley, H. and Brandon, S. Akinetic mutism with recovery after repeated
carbon monoxide poisoning. Psychol. Med., 1, 172, 1971.
25. Jaeckle, R.S. and Nasrallah, H.A. Major depression and carbon monoxide-induced
parkinsonism: Diagnosis, computerized axial tomography, and response to l-dopa. J.
Nerv. Ment. Dis., 173, 503, 1985.
26. Schwartz, A., Hennerici, M. and Wegener, O.H. Delayed choreoathetosis following
acute carbon monoxide poisoning. Neurology, 35, 98, 1985.
27. Garcia, A. and Maestro, I. Reversible motor and sensory peripheral neuropathy in a
patient following acute carbon monoxide intoxication. Electromyo. Clin. Neur., 45,
19, 2005.
28. Snyder, R.D. Carbon monoxide intoxication with peripheral neuropathy. Neurology,
20, 177, 1970.
29. Quattrocolo, G., et al. A case of cortical blindness due to carbon monoxide poisoning.
Ital. J. Neurol. Sci., 8, 57, 1987.
30. Karakurum, B., et al. Partial recovery from cortical blindness following carbon
monoxide intoxication. Int. J. Neurosci., 115, 143, 2005.
31. Jasper, B.W., Hopkins, R.O. and Weaver, L.K. Affective outcome following carbon
monoxide poisoning. Cogn. Behav. Neurol., 18, 127, 2005.
32. Porter, S.S., et al. Corpus callosum atrophy and neuropsychological outcome
following carbon monoxide poisoning. Arch. Clin. Neuropsych., 17, 195, 2002.
33. Ryan, C.M. Memory disturbances following chronic, low-level carbon monoxide
exposure. Arch. Clin. Neuropsych., 5, 59, 1990.
34. Chang, K.H., et al. Delayed encephalopathy after acute carbon monoxide intoxication:
MR imaging features and distribution of cerebral white matter lesions. Radiology, 184,
117, 1992.
35. Choi, I.S. Use of scanning techniques in the diagnosis of damage from carbon monox-
ide. In Carbon Monoxide Toxicity, Penney, D.G., Ed., CRC Press: Boca Raton, FL,
2000; p. 363.
36. Tom, T., et al. Neuroimaging characteristics in carbon monoxide toxicity. J.
37. Povlishock, J.T., Erb, D.E. and Astruc, J. Axonal response to traumatic brain
injury: Reactive axonal change, deafferentation and neuroplasticity. J. Neurotraum.,
9 (Suppl. 1), 5189, 1992.
8417: “8417_c025” — 2007/9/11 — 12:14 — page 588 — #20

38. Novack, T.A., Dillon, M.C. and Jackson, W.T. Neurochemical mechanisms in brain
injury and treatment: A review. J. Clin. Exp. Neuropsyc., 18, 685, 1996.
39. Bigler, E.D. Brain imaging and behavioral outcome in traumatic brain injury. In
Childhood traumatic brain injury: Diagnosis, assessment, and intervention, Bigler,
E.D., Clark, E. and Farmer, J. E., Eds., PRO-ED: Austin, TX, 1997; p. 7.
40. Horn, J.L. and Cattell, R.B. Age differences in fluid and crystallized intelligence.
Acta. Psychol., 26, 107, 1967.
41. Norman, D.A. and Bobrow, D.G. On data-limited and resource-limited processes.
Cognitive Psychol., 7, 44, 1975.
42. Helffenstein, D.A. Neuropsychological evaluation of the carbon monoxide-poisoned
patient. In Carbon Monoxide Toxicity., Penney, D.G., Ed., CRC Press: Boca Raton,
FL, 2000; p. 439.
43. Yeates, K.O. Closed-head injury. In Pediatric neuropsychology: Research, theory,
and practice, Yeates, K.O., Ris, M. D. and Taylor, H. G., Eds., The Guilford Press:
New York, 2000; p. 92.
44. Gemelli, F. and Cattani, R. Carbon monoxide poisoning in childhood. Br. Med. J.
Clin. Res. Ed., 291, 1197, 1985.
45. Secker-Walker, R.H., et al. Smoking in pregnancy, exhaled carbon monoxide, and
birth weight. Obstet. Gynecol., 89, 648, 1998.
46. Ritz, B. and Yu, F. The effect of ambient carbon monoxide on low birth weight among
children born in southern California between 1989 and 1993. Environ. Health Persp.,
107, 17, 1999.
47. Cobb, N. and Etzel, R.A. Unintentional carbon monoxide-related deaths in the United
States 1979 through 1988. J. Amer. Med. Assoc., 266, 659, 1995.
48. Dunham, M.D. and Johnstone, B. Variability of neuropsychological deficits associated
with carbon monoxide poisoning: Four case reports. Brain Inj., 13, 917, 1999.
49. Penney, D.G. Effects of carbon monoxide on developing animals and humans.
In Carbon Monoxide, Penney, D.G., Ed., CRC Press: Boca Raton, FL, 1996;
p. 109.
50. Tattoli, M., et al. Effects of early postnatal exposure to low concentrations of carbon
monoxide on cognitive functions in rats. Pharmacol. Res., 40, 271, 1999.
51. Giustino, A., et al. Prenatal exposure to low concentrations of carbon monoxide
alters habituation and non-spatial working memory in rat offspring. Brain Res., 844,
201, 1999.
52. Cagiano, R., et al. Effects of prenatal exposure to low concentrations of carbon
monoxide on sexual behaviour and mesolimbic dopaminergic function in rat offspring.
Brit. J. Pharmacol., 125, 909, 1998.
53. Tolcos, M., et al. Chronic prenatal exposure to carbon monoxide results in a
reduction in tyrosine hydroxylase-immunoreactivity and an increase in choline
acetyltransferase-immunoreactivity in the fetal medulla: Implications for sudden
infant death syndrome. J. Neuropath. Exp. Neur., 59, 2000.
54. McGregor, H.P., Westcott, K. and Walker, D.W. The effect of prenatal exposure to
carbon monoxide on breathing and growth of the newborn guinea pig. Pediatr. Res.,
43, 126, 1998.
55. Klees, M., Heremans, M. and Dougan, S. Psychological sequelae to carbon monoxide
intoxication in the child. Sci. Total Environ., 44, 165, 1985.
56. Dise-Lewis, J.E., Calvery, M.L. and Lewis, H.C. Brain Stars Manual. Brain Stars
Program: Denver, CO, 2002.
57. World Health Organization (WHO), Environmental Health Criteria 213. Carbon
Monoxide, 2nd edition, Geneva, Switzerland, 1999.
8417: “8417_c025” — 2007/9/11 — 12:14 — page 589 — #21

58. Hofer, M.A. The organization of sleep and wakefulness after maternal separation in
young rats. Dev. Psychobiol., 9, 189, 1976.
59. Lattere, M., Raspino, M. and Vietti, R.M. Acute carbon monoxide poisoning in
children. Pediatria Medica e Chirurgica, 16, 565, 1994.
60. Khan, K. and Sharief, N. Chronic carbon monoxide poisoning in children. Acta
Paediatr., 84, 742, 1995.
61. Silverman, R.K. and Montano, J. Hyperbaric oxygen treatment during pregnancy in
acute carbon monoxide poisoning: A case report. J. Reprod. Med., 42, 309, 1997.
62. Greenough, W.T., Black, J.E. and Wallace, C.S. Experience and brain development.
Child Dev., 58, 539, 1987.
63. Nowakowski, R.S. Basic concepts of CNS development. Child Dev., 58, 568, 1987.
64. Bernstein, J.H. Assessment of developmental toxicity: Neuropsychological batteries.
Environ. Health Persp., 102 (Suppl. 2), 141, 1994.
65. Kolb, B. and Gibb, R. Early brain injury, plasticity, and behavior. In Developmental
cognitive neuroscience, Nelson, C.A., Luciana, M., Eds., MIT Press: Cambridge,
MA, 2001; p. 175.
66. Dennis, M. Childhood medical disorders and cognitive impairment: Biological risk,
time, development, and reserve. In Pediatric neuropsychology: Research, theory, and
practice., Yeates, K.O., Ris, M. D. and Taylor, H.G., Eds., The Guildford Press: New
York, 2000; p. 3.
67. Koger, S.M., Schettler, T. and Weiss, B. Environmental toxicants and developmental
disabilities: A challenge for psychologists. Am. Psychol., 60, 243, 2005.
68. Wechsler, D. Wechsler Intelligence Battery for Children-III. The Psychological
Corporation: San Antonio, TX, 1991.
69. Korkman, M., Kirk, U. and Kemp, S., NEPSY: A developmental Neuropsychological
Assessment. The Psychological Corporation: San Antonio, TX, 1998.
70. Zeiner, H.K. and Prigatano, G.P. Information processing deficits in hydrocephalic and
letter reversal children. Neuropsychol, 20, 483, 1982.
71. Morris, R.D. Low-level carbon monoxide and human health. In Carbon Monoxide
Toxicity, Penney, D.G., Ed., CRC Press: New York, 2000; p. 381.
8417: “8417_c025” — 2007/9/11 — 12:14 — page 590 — #22

26 Issues in Rehabilitation
and Life Care Planning
for Patients with Carbon
Monoxide Poisoning
James M. Gracey
CONTENTS
26.1 Impact of Carbon Monoxide Poisoning on Cognitive, Physical and

Emotional/Behavioral Functioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592
26.1.1 Rehabilitation Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592
26.1.2 Cognitive Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
26.1.3 Physical Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
26.1.4 Emotional and Behavioral Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596
26.2 Rehabilitation Counseling as a Valued Service for Community
Reintegration of Patients with Carbon Monoxide Poisoning . . . . . . . . . . . . . 598
26.2.1 Rehabilitation Counseling Defined . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 598
26.2.2 Rehabilitation Counseling as Part of the Multidisciplinary Team 599
26.2.3 Rehabilitation Counseling Tools for Community Reintegration . 600
26.2.4 Functional Capacity Evaluation (FCE) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 600
26.2.5 Work Hardening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 601
26.2.6 Supported Employment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 601
26.2.7 Selective Job Development and Prevocational Planning . . . . . . . . . 602
26.3 Research Relating to Long-Term Community Adjustment of Patients
with Carbon Monoxide Poisoning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 603
26.4 Life Care Planning for Patients with Carbon Monoxide Poisoning . . . . . . 605
26.4.1 General Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605
26.4.2 Developing the Life Care Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606
26.5 Case Study of a Patient with Carbon Monoxide Poisoning . . . . . . . . . . . . . . . 606
26.5.1 General Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606
26.5.2 Activities of Daily Living (ADL) Issues. . . . . . . . . . . . . . . . . . . . . . . . . . . 607
26.5.3 Transferable Skills Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 607
26.5.4 Outcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608
26.5.5 Life Care Plan Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609
26.5.5.1 Medical Care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609
591
8417: “8417_c026” — 2007/9/11 — 12:14 — page 591 — #1

26.5.5.2 Therapeutic Modalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609

26.5.5.3 Medication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 610
26.5.5.4 Laboratory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 610
26.5.5.5 Diagnostic Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 610
26.5.5.6 Recreation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 610
26.5.5.7 Care Providers/Residential Care . . . . . . . . . . . . . . . . . . . . . . . . 610
26.5.5.8 Transportation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 610
26.5.5.9 Complications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 611
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 611
Appendix 1 - Sample Life Care Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 612
26.1 IMPACT OF CARBON MONOXIDE POISONING

ON COGNITIVE, PHYSICAL AND
EMOTIONAL/BEHAVIORAL FUNCTIONING
26.1.1 REHABILITATION EVALUATION
The challenge for patients with carbon monoxide (CO) poisoning is similar to patients
with other serious medical problems, to successfully reintegrate at home and work.
The process to achieve that goal begins with a rehabilitation evaluation of the impact
of CO poisoning on the patient’s cognitive, emotional, and physical functioning in the
community. The rehabilitation evaluation is a systematic, comprehensive assessment
process, often using a team approach, which is highly individualized. The goal is to
assess residual work potential and skills by means of a variety of tools, including
possible testing (or reviewing neuropsychological test data) or other means which
are individualized for the patient. The rehabilitation evaluation of a CO-poisoned
patient requires specialized knowledge of neuropsychological assessment tools, and
in virtually all cases the rehabilitation counselor should consult directly with the
neuropsychologist and physician early in the process and prior to initiating any
return-to-work plan. The reality of rehabilitation counseling practice today is that
counselors have probably had little exposure to assisting CO patients with return-
to-work issues and may not be familiar with the myriad of symptoms which are
being experienced. Thus, it is critical that counselors begin to share information and
knowledge on the subject. Regardless of counselor sophistication on CO, the bet-
ter rehabilitation plans are ones that represent a consensus of opinion among key
professionals and the patient/family members. By working toward a consensus, the
rehabilitation counselor improves the probability that the individualized return-to-
work plan is the right one for the CO poisoned patient. The four steps in the evaluation
process include information gathering, analysis, synthesis, and interpretation.1,2 Of
critical importance is a comprehensive review of available medical records regard-
ing the CO poisoning and subsequent treatment. The variables to be considered by
the rehabilitation counselor in that assessment are extensive, but most are known or
can be determined with quality interview and assessment tools. There is no cadre of
experienced rehabilitation counselors throughout the United States who specialize in
rehabilitative evaluation and treatment for persons with CO poisoning. Virtually all
8417: “8417_c026” — 2007/9/11 — 12:14 — page 592 — #2

Issues in Rehabilitation and Life Care Planning 593
states have rehabilitation counselors who specialize in working with persons with
brain injuries, but the constellation of chronic symptoms which often accompanies
CO poisoning presents unique challenges for community reintegration.
26.1.2 COGNITIVE ISSUES

Rehabilitation counselors who are evaluating patients with CO poisoning sequelae
must rely on neuropsychologists, neurologists, physiatrists, and psychiatrists for a
clear understanding of what the cognitive deficits are. Often it is the neuropsycholo-
gist, who after completing a comprehensive assessment will refer the CO patient
to a rehabilitation counselor for evaluation and treatment. It would be difficult,
if not impossible, to complete a rehabilitation evaluation without a comprehens-
ive neuropsychological assessment because the tests are essential to objectively
defining the cognitive, and emotional and behavioral deficits which have resul-
ted from the CO exposure.3 The responsibility of the rehabilitation counselor is to
understand the deficits as reported in the neuropsychological evaluation and then
determine the functional impact of those deficits on all aspects of the patient’s activ-
ities of daily living (ADLs), including work. In addition to reviewing available
medical/neuropsychological records, rehabilitation counselors should independently
document ongoing cognitive symptoms (through interview) which the patient is
reporting, including problems with memory, concentration, verbal and nonverbal
learning, reading, speed of information processing, vision, hearing, chemical sens-
itivity, executive functioning, decision-making, motor speed and coordination and
fatigue. Any or all of these symptoms/deficits may be occurring at the time of refer-
ral, and the presence of the deficits will negatively impact work and daily living.
The rehabilitation counselor’s evaluation focuses on individualized manifestations
of the cognitive symptoms. It must be noted that patient response to CO poison-
ing differs by the extent of brain damage, and the range of neuropsychological
dysfunction varies greatly. Helffenstein4 summarized the likely areas of cognitive
dysfunction, which are typically impacted following CO poisoning in Tables 26.1
through 26.4.
Regarding memory, CO-poisoned patients often have multiple areas of concern as
noted in Table 26.2. Regarding sensory-motor functioning, Table 26.3 illustrates com-
mon areas of concern which must be considered by the rehabilitation counselor in any
rehabilitation planning. Table 26.4 illustrates common vision and information pro-
cessing deficits of CO-poisoned patients. Common manifestations of the symptoms,
which are noted in Tables 26.1 through 26.4 are summarized in Table 26.5.
This short list and other manifestations are reviewed by Penney.5 Such manifesta-
tions of CO-related cognitive dysfunction will have a negative impact on the patient’s
return-to-work capacity and depending on the degree of deficit, often prevent a suc-
cessful reintegration at work. Several studies5 suggest that these cognitive problems
will continue for years, making it likely that the rehabilitation counselor will need to
leave cases open or on hold for much longer periods of time than other catastrophic
injuries.
It is important to corroborate the symptoms/problems which are reported to the
rehabilitation counselor with medical records or in conversation with key people who
know the patient, especially within a medical-legal practice. Rehabilitation counselors
8417: “8417_c026” — 2007/9/11 — 12:14 — page 593 — #3

TABLE 26.1
Likely Areas of Executive Dysfunction
• Generating solutions to problems
• Planning organizing sequencing
• Initiation
• Persistence/follow through
• Self-regulation/maintenance of focus
• Self-monitoring for errors
• Mental flexibility
• Working memory
• Multitasking
Source: Adapted from Helffenstein, located at www.

coheadquarters.com/CO1.htm, Penney, D.G., 2007
TABLE 26.2
Likely Areas of Short-Term Memory
Dysfunction
• Verbal
• Visual (nonverbal) memory
• Learning
• Retention
• Incidental memory
• Contamination/confabulation
• Cumulative memory
• Lateralized dysfunction is possible

TABLE 26.3
Likely Areas of Sensory-Motor Dysfunction
• Tactile sensitivity
• Stereognosis
• Fine motor speed
• Grip strength
• Motor programming and sequencing (Executive
motor-skills)
• Verbal and nonverbal agility (Oral motor function)
• Psychomotor problem solving

8417: “8417_c026” — 2007/9/11 — 12:14 — page 594 — #4

TABLE 26.4
Likely Areas of Vision and Information
Processing Dysfunction
• Visual scanning speed and accuracy
• Attention to visual detail
• Visual analysis, synthesis and organization (problem
solving)
• Constructional praxis
• Speed of auditory processing
• Speed of visual processing

TABLE 26.5
Common Functional Cognitive Issues
• Impaired ability to remember instructions or take medications properly
• Fatigue and loss of energy (with or without daily napping)
• Increased need for structuring the daily routine with cuing
• Reduced capacity for initiating task or problem solving completion
• Varying degrees of social disinhibition
• Impaired multi-tasking capacity
• Poor insight regarding deficits
• Compromised safety in the kitchen and around machinery
• Tendency toward concrete thinking and planning with a related reduction in generalization capacity
• Poor money management skills
• Directionality/driving problems (even to familiar places)
Source: Adapted from Helffenstein, located at www.coheadquarters.com/CO1.htm, Penney, D.G., 2007
often corroborate symptoms with others who know and observe the patient, includ-
ing family members, employers, friends, and other care providers (physicians,
psychologists, cognitive therapists, physical and occupational therapists, etc.). When
discussing the cognitive symptoms with the patient (or family members), it is import-
ant to cite examples of how that symptom is affecting their life on a daily basis. By
the conclusion of the evaluation process, the rehabilitation counselor should have a
solid and down-to-earth understanding of how the cognitive symptoms are affecting
the daily functioning of the patient with CO poisoning.
26.1.3 PHYSICAL ISSUES

Patients with CO poisoning often experience a variety of physical symptoms which
must be factored into the rehabilitation evaluation. Physical factors alone can be a
major determinant of how well the patient successfully accommodates at work and
8417: “8417_c026” — 2007/9/11 — 12:14 — page 595 — #5

home, although the cognitive problems are often the major barriers to successful
community reintegration. CO-related physical deficits include paresthesias, changes
in muscle tone, muscle spasms/tremors, loss of coordination, muscular weakness,
headaches, sleep disturbance, changes in vision or hearing, light or noise sensitivity
and pain.3,5 As with the cognitive deficits, the rehabilitation counselor reviews each
physical symptom in terms of how it is impacting the patient’s daily functioning.
Many of the physical problems have probably been reviewed with a physician or
other caregiver, but the rehabilitation counselor will focus on the practical implic-
ations (functional assessment) of the symptoms for daily living and work. This is
an important issue as the counselor determines the impact of specific physical defi-
cits on a patient’s ability to function at home and work. Without this comprehensive
review of symptoms, the rehabilitation counselor is not aware of important phys-
ical problems which can have a negative impact on the outcome of a rehabilitation
plan. Physicians or other caregivers are often not aware of every physical symptom
of the CO-poisoned patient which are reviewed during the rehabilitation evaluation
process.5 For example, vision and hearing deficits may not be known by the physi-
cian of a CO-poisoned high school student who is experiencing significant cognitive,
physical, and emotional symptoms. For that student, his primary life role is to study
and learn in class and through reading assignments. Vision symptoms in an academic
setting are problematic if the student did not previously discuss the problems and
how they were affecting reading comprehension, or how a hearing loss is impacting
the ability to hear the teacher and other students. For this reason, the rehabilita-
tion counselor should always check with the patient and teacher regarding whether
such physical problems are involved. In an active clinical practice, rehabilitation
counselors often discuss such symptoms with the patient for the first time since the
exposure. In the community-based model of rehabilitation within which most rehab-
ilitation counselors operate, patients do not always bring up important symptoms
unless they are asked. In addition, physicians may not be knowledgeable regarding
the multiple physical symptoms which are caused by CO poisoning. The rehabilit-
ation counselor should review any new symptom with the physician who can order
appropriate assessments in order to make the diagnosis. This level of multidiscip-
linary give and take is essential to proper case management since caregivers are not
always aware of the myriad of physical problems facing the patient who is struggling
to reintegrate post-CO exposure at home, work, or school.
26.1.4 EMOTIONAL AND BEHAVIORAL ISSUES

Multiple researchers3,5–9 report that emotional and behavioral sequelae post-CO
exposure are common and often require treatment intervention. Common symptoms
include a sense of personality change, depression, increased irritability and reduced
frustration tolerance, generalized anxiety, emotional lability (including frequent epis-
odes of tearfulness), intense fearfulness regarding the future, apathy, impulsivity, and
social disinhibition. Post-traumatic stress (PTSD) symptoms should also be assessed
by the patient’s neuropsychologist. Other emotional and behavioral sequelae include
problems with poor sleep, anger management, and CO-induced psychotic symptoms.
8417: “8417_c026” — 2007/9/11 — 12:14 — page 596 — #6

TABLE 26.6
Likely Areas of Emotional and Behavioral
Executive Dysfunction
• Reduced self-control/inappropriate behavior
• Impulsivity
• Emotional lability
• Irritability
• Emotional flattening
• Excitability
• Erratic carelessness
• Rigidity
• Reduced motivation

Helffenstein4 described several emotional/behavioral sequelae as directly related to

executive dysfunction as illustrated in Table 26.6.
Table 26.6 is instructive for the rehabilitation counselor because the sequelae com-
monly occur and should be noted during the evaluation process. The rehabilitation
counselor can usually review the symptoms in advance by looking at the neuropsy-
chological or neurological reports, and some rehabilitation clinics conduct preintake
screenings of the complex symptoms by phone with the patient and significant family
members. Preintake screenings are helpful in reducing the amount of time which the
patient and family members will have to be present physically in the office. Consid-
ering that many patients are significantly sleep-deprived and that they already have
multiple outpatient appointments, such techniques are helpful in accurately gathering
data, establishing rapport, and reducing in-office patient time. The goal of the rehab-
ilitation counselor is to first identify significant emotional and behavioral symptoms
that are interfering with daily functioning. The counselor then documents how each
symptom is impacting their life and work. Regarding the emotional and behavioral
sequelae of CO patients, the rehabilitation counselor should work cooperatively with
the treating neuropsychologist and psychiatrist, including a possible meeting between
the rehabilitation counselor, psychologist, psychiatrist, and patient in order to review
treatment goals and establish how the rehabilitation counselor will complement the
treatment process.10 Neuropsychologists and psychiatrists often welcome such inter-
disciplinary planning with the goal of all parties ultimately agreeing to a particular
return-to-work strategy.10,11 Most rehabilitation counselors are used to working in
a community-based model of treatment intervention and can be the ones to initiate
these rehabilitation planning sessions.
It is not difficult to imagine how such pervasive emotional and behavioral symp-
toms can interfere at home and work. Patients often report that they no longer
recognize who they are and are concerned that they are “losing their minds.” For
this reason, psychological and psychiatric follow-up is usually critical as the patient
8417: “8417_c026” — 2007/9/11 — 12:14 — page 597 — #7

struggles with issues related to community reintegration. Patients often feel a sense of
having lost control of their lives and do not know how to regain lost self-confidence.
Most patients can accurately recall how their lives were before the CO exposure and
just “want their lives back.” Such emotional turmoil is inevitable after injury, and
the rehabilitation counselor can offer concrete reassurance that the rehabilitation pro-
cess will be a positive and necessary link between their past and future. Experienced
clinicians report that virtually all patients express fearfulness that their capacity to
return-to-work is impaired. They are usually looking for a methodology from the
rehabilitation counselor to constructively explore the return-to-work process, and
that is probably the most important service provision of the rehabilitation counselor
during this treatment phase.
26.2 REHABILITATION COUNSELING AS A VALUED

SERVICE FOR COMMUNITY REINTEGRATION OF
PATIENTS WITH CARBON MONOXIDE
POISONING
26.2.1 REHABILITATION COUNSELING DEFINED
Community-based rehabilitation counseling services are readily utilized in postacute
settings by physicians, psychologists, neuropsychologists, and other disciplines for
the purpose of assisting persons with a disability to return (reintegrate) to work and
resume as normal a lifestyle as possible.12–14 Patients who have been exposed to CO
poisoning are included in that referral network. The actual role and function of the
rehabilitation counselor has been actively evolving over the past 40 years, partially
through a series of major legislative achievements in the United States. The 1973
Rehabilitation Act for the first time reflected a congressional commitment to serving
persons with severe physical, intellectual, and emotional problems/deficits, and that
commitment was reiterated in a series of amendments to the Act (1974, 1976, 1978,
1986, and 1992).12 Rehabilitation counseling as a profession has been energized by
such legislation changes, which include the removal of architectural and transport-
ation barriers, development of social security disability, individualized educational
programs (for school age children), supported employment and developmental dis-
ability services expansion and other changes. The Americans with Disabilities Act
(ADA)15,16 through its five titles, prohibits discrimination on the basis of disability
in employment, public accommodations, public services, and telecommunications.
The importance of these legislative accomplishments has been to make it possible
for persons who are experiencing significant disability to achieve better functional
outcomes at work and throughout their lives.
Rehabilitation counselors in the United States work with diverse caseloads
including persons with injuries/functional deficits to multiple areas of the body.
The issue for rehabilitation counselors throughout the world is how to successfully
reintegrate patients who are experiencing significant medical problems back to the
workplace and to a normalized home environment.14 The aforementioned legislative
changes have been foundational to the development of a community-based model
8417: “8417_c026” — 2007/9/11 — 12:14 — page 598 — #8

for vocational rehabilitation services. In essence, the changes have been helpful
in leveling the playing field between able-bodied Americans and persons with all
types of disability. In addition to understanding these changes in law and prac-
tice, rehabilitation counselors know that there are three fundamental steps to the
rehabilitation process. The most important of the three is a thorough rehabilitation
evaluation. The second step involves developing an appropriate plan of action which
takes into account the client’s strengths, deficits, goals, and values. The third step
after plan development is plan initiation. Realistically, proper implementation of this
process probably varies greatly by rehabilitation counselor, community, and availab-
ility of knowledgeable professionals regarding CO and its long-term consequences.
Community-based treatment of CO poisoning requires a blend of interdisciplinary
treatment professionals, and the rehabilitation counselor is an important member of
that team.
26.2.2 REHABILITATION COUNSELING AS PART OF THE

MULTIDISCIPLINARY TEAM
Postacute multidisciplinary treatment planning for CO patients is essential in an effort
to maximize community reintegration success.3,9 Rehabilitation counselors can play
an important role in that treatment process because their goal involves the prac-
tical implementation of the patient’s desire to return to normal functioning at home
and work. Rehabilitation counselors recognize the importance of defining transient
and permanent work limitations that the CO poisoning has caused, but also seek to
quantify residual strengths and find ways around the multiple deficits. The CO pois-
oning and residual symptoms have interrupted normal planning for the high school
student, college student, or seasoned worker and without assistance and restruc-
turing their life plans and experiences will be significantly delayed or lost forever.
Career planning post-CO poisoning is usually difficult because of the combination
of symptoms, hospitalization interruptions, need for multiple outpatient interven-
tions, and low energy for rehabilitation. A common issue for treaters is when to refer
for vocational rehabilitation services and whether early referral is beneficial.17,18
Although the research on this variable is not definitive, it is clear that successful long-
term community integration is dependent on many factors and not just the timing of
the referral.18 More important variables include counselor skill and resourcefulness
and whether the right physical, psychological, and cognitive interventions are being
implemented at the right time. Even though much is not known about the long-term
sequelae of CO poisoning, planning for return to home and work should begin soon
after stabilization. The rehabilitation counselor, in concert with other team members,
should identify which resources will be utilized, and the choices depend on iden-
tification and prioritization of deficits in daily functioning. For example, assisting
a student to return successfully to school or a worker to their preinjury employer
after stabilization from CO poisoning is a crucial beginning aspect of community
reintegration.19
Before deciding upon any particular strategy, the rehabilitation counselor will
want to summarize the key postinjury deficits which can be expected to inter-
fere with the return-to-work effort. The counselor will know those deficits from
8417: “8417_c026” — 2007/9/11 — 12:14 — page 599 — #9

the initial evaluation process (including interviews with the patient, physician,
neuropsychologist, and other caregivers plus review of medical records). Deficits
and expected problems should be prioritized and matched with the patient’s preinjury
work history and present goals. To that end the Vocational Diagnosis and Assess-
ment of Residual Employability (VDARE) process1 can be quite useful because it
translates data into objective terms, identifies transferable skills, helps to determ-
ine a vocational objective, and recommends additional services that are necessary
to raise the patient’s level of functioning sufficiently so that he/she can successfully
return to school or work. This methodology is helpful to the distraught CO poisoning
patient who is having significant doubts that anything can be done for him or her.
Decisions can then be made regarding whether the patient is ready to seek compet-
itive employment, return to school, return to a preinjury employer in the same or
modified position, do a structured volunteer program or request supported employ-
ment assistance. Successful community reintegration starts with helping the patient
identify options for services. The rehabilitation counselor should review options with
other professionals, especially the neuropsychologist and cognitive therapist. The
goal of such interdisciplinary planning is to increase the probability that the final
rehabilitation plan for the CO patient is the right one. Considering the complexity of
cognitive, physical, and emotional and behavioral sequelae, which typically occur
after CO poisoning, the rehabilitation counselor would be well advised to review any
return to work or school plans with such treaters.
26.2.3 REHABILITATION COUNSELING TOOLS FOR

COMMUNITY REINTEGRATION
The rehabilitation counselor has several important rehabilitation tools available which
can be helpful in implementing a successful return-to-work/school plan. They include
functional capacity evaluation (FCE), work hardening, and supported employment.
All three are useful for chronic CO patients who are struggling to reintegrate as
productive workers. The following sections will discuss the importance of patients
reconnecting with employers, post-CO stabilization, and will review the benefits of
utilizing prevocational plans for CO patients who are not yet ready for a competitive
work environment.
26.2.4 FUNCTIONAL CAPACITY EVALUATION (FCE)

Patients with CO poisoning are often experiencing so many symptoms that an FCE
becomes essential in documenting how the deficits are impacting the capacity to sus-
tain work activities over any part of a workday.20 In practice, physicians often prefer
such testing techniques because the physical and mental capacity can be stated in
occupational terms, such as strength, posture, and mobility tolerances. FCEs offer
baseline information which can be compared to job task requirements (determin-
ations of whether the patient can perform the essential functions of a job). FCEs
typically include recommendations for task avoidance, temporal pacing, and exer-
tional levels and observational data during testing which is invaluable in planning
return-to-work/school strategies. Clinical observation during FCEs is very helpful
8417: “8417_c026” — 2007/9/11 — 12:14 — page 600 — #10

in further documenting how the CO poisoning has impacted the patient’s cognitive,
physical, and emotional and behavioral functioning and whether he/she is ready to
return to school or work.
26.2.5 WORK HARDENING

The concept of work hardening has been variously described as work conditioning,
work readiness, or work capability training.21 Work hardening programs can be a
useful tool for CO patients because they emphasize real or simulated work activities,
and staff observe and assess patients in conjunction with physical/mental condi-
tioning tasks which are designed to improve their biomechanical, neuromuscular,
cardiovascular, and behavioral functioning.22 Such programs are readily available
and can be an important step for the CO patient because they are very goal dir-
ected, professionally staffed, and the focus is on helping the patient explore their
capacity over an extended period of time (up to several months). The functional
deficits can be readily observed over time, and remediation plans can be developed
with the patient, physician, neuropsychologist, rehabilitation counselor, and other
providers. The rehabilitation counselor is often the logical team member who will
interface with the work hardening staff. CO patients can sometimes move directly to
a return-to-work/school plan after completion of a work hardening program.
26.2.6 SUPPORTED EMPLOYMENT

The concept of supported employment involves selective placement of persons with
severe disability in a quasi-competitive work environment as opposed to a sheltered
workshop. As the number of sheltered workshops has declined in the United States,
the need for more intensive and long-term follow along services has not. Most CO
patients would not be comfortable or appropriate in a closed workshop setting, but
many are not ready for competitive interviewing and placement because of the unre-
solved symptoms. This community-based model of rehabilitation intervention has
been available for several decades and offers the multidisciplinary team another altern-
ative if the patient is not ready for competitive employment. The basic theory of such
services has been in vogue for the past 35 years. Service provision involves locating
a suitable employer and then identifying important job and social skills which can be
taught to a client, training clients to utilize such skills (through a job coach, if neces-
sary), and placing them and providing long-term follow-up support. Beginning in the
1970s this author and other providers termed such supported employment services
“work stations in industry.” A criticism of such supported work service is around
the use of a job coach.23 The job coach is the trainer and follow-up troubleshooter
for clients at the job site, and a concern has been noted that the cost of such per-
sonnel is prohibitive in many clinical settings. An additional concern is that the role
of a job coach is intensive and can stigmatize the client as different at a work site.
The reality for clinicians is that such resources are valuable and allow the rehabil-
itation counselor options for selective placement, namely to re-engage in real work
settings.
8417: “8417_c026” — 2007/9/11 — 12:14 — page 601 — #11

26.2.7 SELECTIVE JOB DEVELOPMENT AND

PREVOCATIONAL PLANNING
The reality for persons with significant disability, including CO poisoning, is that
their labor force participation rates and wages are lower than that of able-bodied
Americans, and the differences are measurable and significant.24 The vast major-
ity of all persons with disabling conditions want to work competitively,2 and
CO-poisoned patients are no exception. CO patients are facing financial stress
and typically feel that they must return to work despite their limitations. Sev-
eral researchers have stressed the difficulties which persons with serious cognitive,
physical, and emotional problems face in finding and keeping jobs.25 Because
of this problem, rehabilitation counselors should promote return-to-work as soon
as possible after stabilization. Most people will want to return to their pre-CO
exposure job or career, if possible, and to a different job (career) if not. The
rehabilitation counselor will have to determine a reasonable approach, consider-
ing the CO patient’s situation. Once an option is determined and agreed upon,
a gradual return-to-work process is usually essential, and the rehabilitation coun-
selor will probably want to have access to the employer for feedback on work
performance. Depending on symptom severity, a job tryout can last for several
months and should include structured follow-up sessions with the counselor for
troubleshooting issues. The length of follow-up will vary by severity of problems
and employer compatibility, but a range of 3 to 6 months is not unusual. It is
not uncommon to find benevolent employers (family members, close friends, etc.)
who will do virtually anything to help the person return to work. During the job
tryout, it is important to emphasize to both parties that the tryout is to be a learn-
ing experience (an evaluation) so that he/she can test out their residual strengths
in a supportive work environment where the employer is also a participant in the
evaluation and return-to-work process. Over time, problems can be identified and
worked through with practical solutions by the rehabilitation counselor or support-
ing cognitive therapist. Although rehabilitation counselors are uniquely trained to
assist patients with developing such return-to-work plans, this does not mean that
other disciplines (especially occupational therapists or cognitive therapists) can-
not enhance the effort. The dramatic increase in cognitive rehabilitation programs
has created various subspecialty professionals such as life skills trainers, beha-
vior specialists, case managers, job coaches, and others. These individuals are
often skilled in community reintegration techniques including setting up job tryouts
with preinjury or new employers and volunteer placements and then troubleshoot-
ing with an employer or volunteer agency and patient during the follow-up
phase.
In a similar manner, prevocational plans can be developed with a benevol-
ent employer or volunteer placement if the CO patient’s symptoms are stable but
still too severe for competitive employment. The important goal here is to assist
the patient to be productive again, even in a volunteer setting. Prevocational
plans should precede a formal return-to-work plan and should not become per-
manent unless the CO patient’s symptoms are so severe that he/she is no longer
employable.
8417: “8417_c026” — 2007/9/11 — 12:14 — page 602 — #12

26.3 RESEARCH RELATING TO LONG-TERM

COMMUNITY ADJUSTMENT OF PATIENTS WITH
Prediction of success or failure in the community for persons with significant phys-
ical, psychiatric, and cognitive disability is complex, multifaceted, and has been
researched extensively for several decades.26–31 A consideration for rehabilitation
counselors who work with CO patients is that studies which emphasize long-term
community adjustment themes important to them are few and far between. Weaver
et al.32 evaluated whether four patients with severe CO poisoning had better func-
tional outcomes with or without hyperbaric oxygen treatment. The study mentions
employment status of the four patients several months posthospital discharge and con-
cludes that three of the four had a favorable outcome and had successfully returned
to work. The study offers no insight regarding the researchers’ definition of “success-
ful” return-to-work. Likewise, Weaver33 summarized numerous studies regarding the
long-term consequences of CO poisoning. The report documents multiple affective,
cognitive, and neurologic sequelae including malaise, apathy, memory disturbances,
depression, anxiety, focal neurologic abnormalities, parkinson-like symptoms, and
other problems. Job-related problems are likely with such severe symptoms, but the
study did not document employment outcomes. Likewise, other researchers have
noted that CO patients experience a variety of emotional problems with depression,
anxiety, and increased lability in 1-year follow-up studies,7 but the articles do not
offer much detail as to whether CO patients are working successfully. Most clinicians
and researchers regard return to productivity as a reasonable goal of treatment,29
and experienced clinicians realize that the combination of physical, cognitive, and
emotional symptoms which result from CO exposure complicate that return-to-work
effort.
Regarding long-term employability of CO patients, the brain injury literature is
instructive since the cognitive and emotional and behavioral components are similar
between the two diagnostic categories. Crisp30 reviewed 19 studies that emphasized
outcomes for persons with brain injury and identified key factors that were consistently
related to vocational outcome. For traumatic brain injury (TBI) patients, employment
status was related to post-traumatic amnesia (PTA) and length of coma, and the cog-
nitive deficits which most strongly predicted return to productivity involved verbal
and/or visual memory and attention. Employment status was also negatively impacted
by major depressive symptoms. An earlier study by Crisp28 reviewed 29 studies (from
the 1970s and 1980s) which had identified critical factors of persons with brain injury
which were associated with vocational outcomes. In that review of the literature,
cognitive and personality deficits (problems with memory, concentration, and per-
sonality) were key variables in whether TBI patients successfully returned to work.
The next most commonly cited variable in successful return-to-work was whether
patients were experiencing postinjury psychosocial adaptive difficulties (social isola-
tion, reduced capacity for engaging in appropriate social interaction and poor family
adjustment). Of the 29 studies which Crisp surveyed, only 7 reported re-employment
statistics of 68% or better for their respective samples, and 20 studies ranged between
12% and 64% employed. Two studies did not provide employment data. The mean
8417: “8417_c026” — 2007/9/11 — 12:14 — page 603 — #13

TABLE 26.7
Return-to-Work Analysis for 208 TBI Patients at 3.5 Years Postinjury
Follow-up
Pre-TBI Employed at date of injury N = 170
Unemployed at date of injury N = 38
Post-TBI Postinjury Variable N Percentage
Working at follow-up 79 46.5%
Returned to work but no longer working 23 13.5%
No attempt at return to work 68 40%
Total not working who were employed prior to TBI 91 53.5%
Source: Adapted from Fleming et al. Brain Injury, 13, 417–431, 1999.
number of employed persons in those 20 studies was only 38%, and the mean number
of years since injury was almost four.
Fleming et al.29 studied 208 patients who were admitted to a head injury program
between 1991 and 1995. The mean time from initial injury to follow-up was approx-
imately 3.5 years. Table 26.7 presents a postinjury analysis of how many TBI patients
had returned to work after an average of 3.5 years from the date of injury. Of the 208
subjects, 170 were working before the TBI, and 38 were not. Of that working group
(N = 170), only 46.5% (N = 79) had successfully returned to work after the injury
with most of the working group going back to the same or similar jobs. Thirteen and a
half percent of the working group (N = 23) had been employed since injury but were
not employed at follow-up, and 40% (N = 68) of the preinjury working group had not
returned to work at all. Thus, a total of almost 53.5% (N = 91) were not working in
any capacity at follow-up even though this same group was employed before the TBI.
The authors discussed factors which may contribute to successful community (home
and job) reintegration. The results were consistent with previous studies in that the
variables of PTA, older age at injury, shorter mean duration of acute hospitalization,
premorbid occupational status and cognitive variables were predictive of community
integration success or failure.
The Fleming et al.29 results are echoed in the CO Support Group research34
regarding return-to-work rate and outcomes of CO-exposed patients. Of the chronic
(N = 65) and unconscious groups (N = 12), 32% and 75%, respectively, did not
return to any employment. The incapacity of CO-exposed patients was further docu-
mented in pre- and postexposure patients by income. The postinjury income drop
for the unconscious group was around 50% after exposure, and the two groups
overall had a significant and continuing reduction in household earnings. Such res-
ults are not surprising, considering the myriad physical, cognitive, and emotional
symptoms which interfere with all aspects of personal and vocational functioning
after exposure. It is understandable that patients with pain/cramps, pins-and-needles
sensation/stiffness, headaches, fatigue/weakness, poor concentration, memory loss,
dizziness, gastrointestinal problems, cardiac issues, vision problems, depression, and
other sequelae are not easily able to resume a career and re-establish earning capacity.
8417: “8417_c026” — 2007/9/11 — 12:14 — page 604 — #14

Of particular concern for community re-entry programs such as vocational rehabilit-

ation, CO patients in both groups and in large numbers reported that the symptoms
were persisting. Although there were postexposure drops in the prevalence of some
symptoms, the study noted that significant numbers of people continued to suffer
ill health effects long after exposure. For example, 60% of the chronic group con-
tinued to suffer from tiredness, pain, headaches, and problems with concentration
and memory. Such pervasive and unresolved medical problems will significantly and
negatively affect long-term employability and necessitate early vocational rehabilit-
ation intervention. Helffenstein35 has published vocational outcome data on 19 CO
poisoned patients who were all employed at the time of exposure, and the results
underscore the CO Support Group research and the TBI results. Only one of the
19 returned to his preinjury job, but that individual was self-employed and could
only complete the work transition by closing one of his two plants, thus signi-
ficantly reducing earnings. Fifty-eight percent returned to lower skill part-time or
full-time jobs and 26% were determined to be permanently and totally disabled.
Dunham and Johnstone31 further document the impact of CO poisoning on cognitive,
emotional/behavioral, and vocational functioning. Their study recommends careful
vocational planning to include postexposure routinized job duties, frequent monit-
oring, more time to learn new tasks and alternative methods of instruction than just
verbal.
Much more research is needed as there are insufficient studies which assess how
the unresolved symptoms are impacting work on a long-term basis. As noted above,
the traumatic brain injury literature on employability is instructive regarding CO
patients. It is likely that the CO exposure will have a major impact on work and earning
capacity. As a whole, this problem merits more study, rehabilitation interventions,
and awareness training for rehabilitation professionals. The vast majority of patients
are working or in school at the time of exposure, and most will want to resume
their work or schooling after returning home. This review of literature asserts what
experienced clinicians are seeing on a daily basis, namely, that chronic CO exposure
has a permanent and negative impact on earning capacity, long-term employability,
and ability to readapt to preinjury career plans and goals. The problem is particularly
troublesome since few rehabilitation counselors are even aware of these issues.
26.4 LIFE CARE PLANNING FOR PATIENTS WITH

26.4.1 GENERAL INTRODUCTION
The International Academy of Life Care Planners defines a Life Care Plan (LCP)
as a “dynamic document based upon published standards of practice, comprehensive
assessment, data analysis and research, which provides an organized, concise plan for
current and future needs with associated costs for individuals who have experienced
catastrophic injury or have chronic health care needs.”36,37 The history of life care
planning has been carefully reviewed in The Guide to Rehabilitation38 and The Life
Care Planning and Case Management Handbook39 and will not be discussed here
8417: “8417_c026” — 2007/9/11 — 12:14 — page 605 — #15

in any detail. In summary, the development of LCPs began in the mid-1980s and
has continued to accelerate to the present. The life care planning process has been
described as transdisciplinary and is usually undertaken by nationally certified and
licensed professionals within a health care or rehabilitation discipline.40 There are
specific training programs for life care planners which lead to certification (CLCP).
Most CLCPs are nurses or rehabilitation counselors, although other health care profes-
sionals who have become certified have degrees in medicine, chiropractic, psychiatry,
psychology, speech therapy, and others.41 Developing an LCP for patients who have
experienced CO poisoning is challenging since many CLCPs have little background
with CO-exposed persons. However, this will be an area of increasing need as more
patients and family members are asking what medical and rehabilitation products and
services will be required throughout their lifetime because of a CO exposure.
26.4.2 DEVELOPING THE LIFE CARE PLAN

The LCP is intended to be a realistic assessment of current and future med-
ical/rehabilitative care needs for a catastrophically injured person or someone with a
chronic health condition. The scope of the LCP should include a careful document-
ation of needs regarding medical care, evaluations, medications, durable medical
equipment, supplies, therapeutic modalities, laboratory testing, diagnostics, surgery,
transportation, adaptive aids, hospitalizations, home modifications, residential/home
care, recreational adaptations, orthotics/prosthetics, and other areas. The life care
planning process usually includes a comprehensive review of medical records, patient
and family interviews, and consultations with caregivers (physicians, neuropsycholo-
gists, therapists, and other relevant health care specialists). The CLCP, in conjunction
with other treatment professionals determines each item or service, the duration of
need (e.g., lifetime, 1 year, 5 years, etc.), frequency of usage (e.g., 1x/day, 1x/week,
etc.), purpose, cost (e.g., per month, year, etc.) and supplier/vendor (if known). Only
items/services that are determined to be probable from a medical and rehabilitation
point of view (more likely than not) are included in the LCP. Items or services that
may become necessary for someone in the future are not included as line items in the
LCP except as complications.
Life care plans for CO-poisoned patients are being developed in increasing num-
bers and often in a legal context. However, a well-prepared and documented LCP
becomes an excellent plan of treatment for patients and families because it is such
a comprehensive assessment of long-term care needs. While this can be said of any
good LCP, it is particularly important for recovering CO-poisoned patients because
of the complexity and diversity of their ongoing symptoms and care needs.
26.5 CASE STUDY OF A PATIENT WITH CARBON

MONOXIDE POISONING
26.5.1 GENERAL INFORMATION
Ms. G.B. was 39 years old at referral and was referred for a vocational rehabilitation
and LCP evaluation after experiencing a 48-h exposure to CO (caused by a faulty
8417: “8417_c026” — 2007/9/11 — 12:14 — page 606 — #16

heater) in which a roommate died. The exposure was significant and resulted in a
60-day period of initial hospitalization, with 20 hyperbaric dives in the first 20 consec-
utive days. The diagnoses included: (1) CO poisoning with hypoxic encephalopathy,
(2) Aspiration pneumonia, (3) Status postrhabdomyolysis with renal insufficiency,
(4) Deep vein thrombosis in the left calf, (5) Bilateral mild hearing loss.
She had significant impairments of mobility, cognition, ADLs, speech, and swal-
lowing. Her carboxyhemoglobin (COHb) saturation at admission was 34%. By the
end of the initial hospitalization, she was described as having made significant gains
in mobility and ADL functioning, but with slower improvements in cognition and
communication.
She was transferred to an inpatient rehabilitation facility where she had extensive
medical and neuropsychological work-ups and was followed closely by a neurologist,
neuropsychologist, cognitive therapist, physical therapist, and occupational therapist.
At the time of transfer, she was experiencing moderately severe to severe impair-
ments of auditory comprehension, expressive communication, initiation, memory,
attention, distractibility, and auditory processing. She had significant visual scanning
impairments, was perseverative on two-step commands, and had an apraxic gait. She
transitioned to outpatient status after about 45 days but required verbal or visual cuing
to remain on task. Ongoing problems with executive functioning and reasoning defi-
cits limited her ability to work. Because of the deficits, she lived with family members
and was legally assigned a guardian. She continued to experience mild dexterity prob-
lems, tremors, difficulties with going down stairs as well as multitasking, organizing
information and problem solving, cognitive flexibility, feeling insecure, being alone,
managing finances, and social judgment. She was referred for a vocational rehabilit-
ation and LCP evaluation at about 16 months postinjury. At the time of referral, she
had not been able to return successfully to work. Her preinjury vocational background
included a postgraduate degree and successful career as a financial planner.
26.5.2 ACTIVITIES OF DAILY LIVING (ADL) ISSUES

At the intake interview (16 months postinjury), she was experiencing multiple phys-
ical, cognitive and emotional and behavioral symptoms (noted above) which were
interfering with all aspects of daily functioning. She was having cognitive problems
with fatigue, attention, concentration, short-term memory, reading comprehension,
slowed speed of information processing, executive dysfunction, and lack of insight.
She was also experiencing physical and emotional problems with balance, headaches,
tremors, behavioral dysregulation, low frustration tolerance, and impatience. After
the CO exposure, family members noted problems with decision-making and prob-
lem solving, although Ms. G.B. insisted that she was doing better than the caregivers
and family members believed.
26.5.3 TRANSFERABLE SKILLS ANALYSIS

In the case of Ms. G.B., a cognitive therapist was providing routine follow-along ses-
sions with her at the time of referral for a rehabilitation evaluation. The rehabilitation
counselor and cognitive therapist agreed that they would closely monitor progress
8417: “8417_c026” — 2007/9/11 — 12:14 — page 607 — #17

once a prevocational plan was developed. A vocational plan should only be initiated
after a consensus of opinion has been determined among caregivers (including the
rehabilitation counselor), patient and family members that the CO-poisoned patient
is more likely than not going to be successful in a return-to-work plan. Although fail-
ure cannot be prevented, a systematic rehabilitation evaluation (as described above)
will reduce the probability of failure substantially because a thorough evaluation will
match the patient’s residual skills with known training or job requirements through
a transferable skills analysis (TSA). The TSA is “the foundation of any attempt to
identify similar or related jobs that are consistent with or equal to the functional skill
levels of the worker.”2
Considering the complex combination of physical, cognitive, and emotional defi-
cits that the CO patient is facing, a careful analysis of residual skills (TSA) is crucial
prior to implementing any return to productivity plan. In the case of Ms. G.B.,
her work skills were significantly diminished from before the CO exposure. As a
financial planner before the injury, she was responsible for reviewing assets, liabil-
ities, and earning power of clients and helped them develop plans to increase assets
and maximize income. Required skills include an ability to work well with people,
perform work with accuracy and analysis, make judgments and decisions, express
personal thoughts, make recommendations, and communicate clearly.42 The job also
required frequent travel to clients’ homes and long irregular hours. Once her residual
skills were matched to the list of essential job duties of a financial planner (through
the TSA), it was clear that she could not resume this career in the future. That
decision was strongly supported by family members as they were convinced that
the deficits were of such severity that she would fail any return-to-work plan as a
financial planner. After further discussion and review with her doctors it was further
determined that Ms. G.B. was probably not competitively employable. She was exper-
iencing pervasive problem solving inefficiencies, impaired executive functioning,
plus moderate problems with attention, speed of information processing, forgetful-
ness, and disorganization, and she had little to no insight regarding any of these
problems.
26.5.4 OUTCOME
After referral to the rehabilitation counselor (16 months postinjury), a consensus
among caregivers formed that Ms. G.B. could not return to her career as a financial
planner. The rehabilitation counselor met or spoke by phone with family members,
physicians, neuropsychologist, physical therapist, cognitive therapist, and with her
supervisor at the financial firm. The counselor carefully documented her pre- and
postexposure home and work functioning. The counselor noted that after the CO
exposure family members assumed a primary caretaker role with frequent input from
the neuropsychologist. All parties agreed that Ms. G.B. was not employable and that
she would require extensive medical and rehabilitation follow-up care throughout
her lifetime. The referral of Ms. G.B. to the rehabilitation counselor had the effect
of solidifying opinions that she could not sustain competitive employment in the
future. However, despite the severity of her problems/deficits, Ms. G.B. wanted to
work in some capacity. The rehabilitation counselor, in concert with other treatment
8417: “8417_c026” — 2007/9/11 — 12:14 — page 608 — #18

team members recommended a series of college courses in areas of personal interest,

and this was followed by referral to a volunteer organization, which was also of
interest to the patient. The patient began a set schedule of volunteer hours with a
local nonprofit agency, plus occasional college courses through various schools. The
rehabilitation counselor followed up for about a year until satisfied that Ms. G.B. was
stabilized at her volunteer placement and at home.
26.5.5 LIFE CARE PLAN ISSUES

The LCP for Ms. G.B. was determined through face-to-face and phone consultations
with her physicians, neuropsychologist, cognitive therapist, and physical therapist.
All caregivers were cooperative in providing recommendations for the LCP, and the
attached plan was considered to be an accurate assessment of her current and future
medical and rehabilitative care needs. Once the plan was completed in the attached
chart format, it was submitted to the caregivers for a final review of its completeness
and accuracy. The caregivers were asked to approve or change the recommendations
in writing.
26.5.5.1 Medical Care

Her doctors recommended that she would require lifetime follow-up by a physiat-
rist, otolaryngologist (ENT), and psychiatrist. The physiatrist was to provide medical
supervision of the CO-related problems with paresthesia, tremors, loss of coordin-
ation, headaches, muscle weakness and discomfort, as well as monitor the multiple
cognitive and emotional deficits. The ENT was to monitor and treat the mild bilateral
hearing loss, and the psychiatrist was to provide psychotropic medication manage-
ment with particular focus on regulation of depression and behavioral dyscontrol
issues.
26.5.5.2 Therapeutic Modalities

Her doctors recommended that she would require long-term and community-based
cognitive therapy at a rate of between two and four times per month for as long as she
continued to live in her own apartment. Her doctors stated that she would probably
not remain in her own apartment beyond a period of 15 to 20 years after which time
she would need to live in a brain injury assisted living program for the duration of
her lifetime. The physicians stated that she would probably experience early onset
dementia after that time, although proof for this assertion in the literature is lack-
ing. The community-based cognitive therapy was designed to monitor compensatory
strategies and improve/maintain her independent living skills. The neuropsycholo-
gist recommended individual therapy at the rate of one to two times per month for
life. The psychotherapy focused on crisis intervention, loss of social relationships,
behavior management, and depression. Her physicians and physical therapist recom-
mended a few sessions of physical therapy per year for life in order to treat neck and
back pain and spasms.
8417: “8417_c026” — 2007/9/11 — 12:14 — page 609 — #19

26.5.5.3 Medication
Her physicians recommended medications for depression and pain control. In addi-
tion, movement disorder medications were recommended to treat the tremors. It was
determined that she would require these or similar classes of medication throughout
her lifetime.
26.5.5.4 Laboratory
Because of the lifetime need for the aforementioned medications, she also needed
periodic comprehensive metabolic panels in order to assess the physiological effects
of the drugs.
26.5.5.5 Diagnostic Studies

The physicians recommended one Magnetic Resonance Imaging (MRI) during the
next 20 years to evaluate the expected neurologic deterioration and early onset
dementia, although as stated above, the underlying reason for this remains unclear.
26.5.5.6 Recreation
Adult recreational camps for persons with brain injury were recommended to provide
an opportunity for supervised recreation and socialization. Ms. G.B. had difficulty
sustaining relationships after the CO exposure and was not able to resume normal
preinjury recreational outings without significant family support.
26.5.5.7 Care Providers/Residential Care

Her caregivers recommended that she would need home care assistance (a companion)
as long as she remained at home. She was relatively independent in personal care
but required significant daily support with home safety, routine problem solving,
shopping, meal preparation, and daily planning. The companion will implement
treatment goals in consultation with the cognitive therapist, neuropsychologist, and
case manager. The physicians opined that she would probably experience early onset
dementia within the next 15–20 years at which point she will have to transition
into a brain-injury-assisted living program for life. The assisted living program will
provide long-term treatment support in the least restrictive environment. Ms. G.B. also
required routine follow-along support with a case manager who continually assessed
quality of care issues and level of independence in ADLs. The case manager was
available by phone and in person to troubleshoot and problem solve with Ms. G.B.
and her family members.
26.5.5.8 Transportation
The cost of mileage reimbursement for travel to medical and rehabilitative appoint-
ments is included in the LCP.
8417: “8417_c026” — 2007/9/11 — 12:14 — page 610 — #20

26.5.5.9 Complications
Ms. G.B. is at risk for a variety of medical complications, which, if they occur,
will increase the cost of her future medical and rehabilitative care. These complica-
tions include seizure disorder, visual problems, motoric deterioration, and worsening
depression. This listing is intended to educate the reader in understanding that there
are multiple factors which could influence Ms. G.B.’s future medical needs, but which
cannot now be included in this LCP. This LCP includes all of the products and ser-
vices that can now be anticipated and are considered reasonable and necessary. The
future cost of these products and services may vary according to Ms. G.B.’s actual
health requirements. Because she is at risk for the complications which are listed in
the LCP, the actual costs may be higher if her condition changes from what can now
be anticipated. Any evaluation of her future care plan requirements should consider
this list of potential complications and expected costs.
ACKNOWLEDGMENTS
Thanks to my wife, Anthea Blanas Gracey, for her assistance in editing and finalizing
the manuscript; to Ruth Zebarth, nurse and Life Care Planner for her assistance
in finalizing the LCP; and to Lynn Dalton and MarLene Nelson for their technical
assistance.
References
1. Havranek, J., Grimes, J., Field, T., and Sink, J. Vocational Assessment: Evaluating
Employment Potential, Elliott and Fitzpatrick, Athens, 1994, Chapt. 5.
2. Weed, R. and Field, T. Rehabilitation Consultant’s Handbook, Elliott and Fitzpatrick,
Athens, 2001, Chapt. 6.
3. Helffenstein, D. Neuropsychological evaluation of the carbon monoxide-poisoned
patient, In Carbon Monoxide Toxicity, Penney, D.G., ed., CRC Press, Boca Raton,
2000, Chapt. 20.
4. Penney, D.G. www.coheadquarters.com/CO1.htm, 2006.
5. Penney, D.G. Chronic carbon monoxide poisoning, In Carbon Monoxide Toxicity,
Penney, D.G., ed., CRC Press, Boca Raton, 2000, Chapt. 18.
6. Balzan, M., Agius, G., and Debono, A. Carbon monoxide poisoning: Easy to treat but
difficult to recognise, J. Postgrad. Med., 72, 470–473, 1996.
7. Hay, P. and Denson, L. The neuropsychiatric effects of carbon monoxide poisoning:
Preliminary findings from a prospective study, J. Undersea Hyperbar. Med., 25,
47, 1998.
8. Hopkins, R. and Weaver, L. Longterm outcome in subjects with carbon monoxide
poisoning, J. Undersea Hyperbar. Med., 21, 17, 1994.
9. Wilson, F., Harpur, J., Watson, T., and Morrow, J. Adult survivors of severe cerebral
hypoxia — Case series survey and comparative analysis, J. Neurol.Rehabil., 18,
291–298, 2003.
10. Barisa, M.T. and Barisa, M.W. Neuropsychological evaluation applied to vocational
rehabilitation, J. Neurol. Rehabil., 16, 289–293, 2001.
8417: “8417_c026” — 2007/9/11 — 12:14 — page 611 — #21

11. Ultmann, M., Geller, T., Chilakamairi, J., and Kaplan, S. Multidisciplinary rehabil-
itation management of depression in the carbon monoxide injured patient, Pediatric
Rehab., 2, 101–106, 1998.
12. Rubin, S. and Roessler, R. Foundations of the Vocational Rehabilitation Process,
PRO-ed., Austin, 1995, Chapt. 2.
13. Garner, W.E. “An Identification of Competencies Critical to Practicing Rehabilita-
tion Counselors: Implications of Validating the Rehabilitation Counselor Certification
Examination.” PhD diss., Southern Illinois University, Carbondale, IL, 1985.
14. Rubin, S., Matkin, R., Ashley, J., Beardsley, M., May, V., Onstott, K., and Puck-
ett, F. Roles and functions of certified rehabilitation counselors (special issue), Rehab.
Couns. Bull., 27, 199–224, 238–245, 1984.
15. The Americans with Disabilities Act of 1990, 101, 42, U.S.C., 12112.
16. Berkowitz, E. Disabled policy: A personal postscript, J. Disability Policy Studies, 3,
2–16, 1992.
17. Marnetoft, S. and Selander, J. Long-term effects of early versus delayed vocational
rehabilitation — A four year follow-up, Disabil. Rehabil., 24, 741–745, 2002.
18. Marnetoft, S., Selander, J., Bergroth, A., and Ekholm, J. Vocational rehabilitation —
early versus delayed. The effect of early vocational rehabilitation compared to delayed
vocational rehabilitation among employed and unemployed, long-term sick-listed
people, Int.. J. Rehabil. Res., 22, 161–170, 1999.
19. VanLierop, B. and Nijhuis, F. Assessment, education and placement: An integrated
approach to vocational rehabilitation, Int.. J. Rehabil. Res., 23, 261–269, 2000.
20. Wickstrom, J. Functional capacity testing, In Multidisciplinary Perspectives in Voca-
tional Assessment of Impaired Workers, Scheer, S., ed., Aspen Press, Rockville, 1990,
73–88.
21. Lett, C., McCabe, N., Tramposh, A., and Tate-Henderson, S. Work hardening, In Work
Injury: Management and Prevention, Isernhagen, S., ed., Aspen Press, Rockville,
1988, 195–229.
22. Commission on Accreditation of Rehabilitation Facilities. 1994 Standards Manual
and Interpretive Guidelines for Organizations Serving Persons with Disabilities,
Tucson, 1994.
23. Rusch, F., Conley, R., and McCaughrin, W. Benefit-cost analysis of supported
employment in IL., J. Rehabil., 59, 31–36, 1993.
24. U.S. Census Bureau. Disability – Work Experience and Mean Earnings – Work Dis-
ability Status of Civilians 16 – 74 Year Old, by Educational Attainment and Sex,
Washington, DC, 2005.
25. McMahan, B. and Shaw, L. Work Worth Doing: Advances in Brain Injury Rehabilit-
ation, PMD Press, Orlando, 1991, Chap. 7.
26. Cattelani, R., Tanzi, F., Lombardi, F., and Mazzucchi, A. Competitive re-employment
after severe traumatic brain injury: Clinical, cognitive and behavioral predictive
variables, Brain Injury, 16, 51–64, 2002.
27. Cifu, D., Keyser-Marcus, L., Lopez, E., Wehman, P., Kreutzer, J., Englander, J., and
High, W. Acute predictors of successful return to work 1 year after traumatic brain
injury: A multicenter analysis, Arch. Phys. Med. Rehab., 78, 125–131, 1997.
28. Crisp, R. Return to work after traumatic brain injury, J. Rehabil., 58, 27–33, 1992.
29. Fleming, J., Tooth, L., Hassell, M., and Chan, W. Prediction of community integra-
tion and vocational outcome 2–5 years after traumatic brain injury rehabilitation in
Australia, Brain Injury, 13, 417–431, 1999.
30. Crisp, R. Key factors related to vocational outcome: Trends for six disability groups,
J. Rehabil., 71, 30–37, 2005.
8417: “8417_c026” — 2007/9/11 — 12:14 — page 612 — #22

31. Dunham, M. and Johnstone, B. Variability of neuropsychological deficits associated

with carbon monoxide poisoning: Four case reports, Brain Injury, 13, 917–925, 1999.
32. Weaver, L., Hopkins, R., and Larson-Lohr, V. Neuropsychologic and functional recov-
ery from severe carbon monoxide poisoning without hyperbaric oxygen therapy, Ann.
Emerg. Med., 27, 736–740, 1996.
33. Weaver, L. Carbon monoxide poisoning, Environ. Emerg., 15, 297–317, 1999.
34. Carbon Monoxide Support group. Effects of chronic exposure to CO: A research study,
Technical paper, U.K., 1997.
35. Helffenstein, D. Neurocognitive and neurobehavioral sequelae of chronic car-
bon monoxide poisoning: A retrospective study and case presentation, In Carbon
Monoxide Poisoning, Penney D., ed., Chapt. 23, 2007.
36. International Academy of Life Care Planners (IALCP), Standards of Practice for Life
Care Planners, Ankeny, 2000, 1.
37. McCallum, P. Field review: Revised standards of practice for life care planners, J. Life
Care Planning, 4, 67–74, 2005.
38. Deutsch, P. and Sawyer, H. A Guide to Rehabilitation, Ahab Press, White Plains,
2005, Chapt. 5.
39. Weed, R., ed. Life Care Planning and Case Management Handbook, CRC Press,
Boca Raton, 1999/2004, Chapt. 1.
40. Reavis, S. Standards of practice, J. Life Care Planning, 1, 49–57, 2002.
41. May, R. Certification in life care planning is alive and well, J. Life Care Planning, 1,
59–61, 2002.
42. Eureka. Colorado Career Information System, (ECOCIS) [computer software],
Eureka, 2005.
8417: “8417_c026” — 2007/9/11 — 12:14 — page 613 — #23

APPENDIX 1
614
Sample Life Care Plan of a Patient with Carbon Monoxide Poisoning

Life Care Plan
Medical Care
Item or Service Duration Frequency Purpose Estimated Cost Supplier/Vendor
Physical medicine and Lifetime 5–6x/year for 2 years; Evaluate function, cognition, $130/visit Western Medical
rehabilitation then 4x/year for 3 years; mobility and neurologic status and $650–780/year for 2 years; then Associates
specialist then 1x/year treat problems related to CO $520/year for 3 years; then $130/year
exposure
Otolaryngologist Lifetime 1x/year Monitor and treat hearing loss $200/visit B. Alpha, M.D.
$200/year
Psychiatrist Lifetime 4x/year Psychotropic medication $86–120/visit A. Smith, M.D.
management $344–480/year
Life Care Plan
Therapeutic Modalities
Cognitive therapy∗ 15–20 years 24–48 sessions/year Provide cognitive assistance; monitor $185–215/h Western
compensatory strategies to improve $4,440–10,320/year Rehabilitation
executive functioning/frustration Hospital
tolerance; productivity,
socialization and recreational goals
Psychotherapy Lifetime 12–24 sessions/year Psychological support for crisis $150–225/session A. Beta, Psy.D.
intervention, loss of social $1,800–5,400/year
relationships; behavior management
and depression
Physical therapy Lifetime 4–6 sessions/year Treat neck and back pain and spasms $120–164/session R. Omega, R.P.T.
8417: “8417_c026” — 2007/9/11 — 12:14 — page 614 — #24

$480–984/year
∗ Community-based cognitive therapy will no longer be required once she transitions to a brain-injury-assisted living program in 15–20 years owing to early onset
dementia.
Life Care Plan
Medications∗
Zoloft 50 mg Lifetime 1x/day Antidepressant $2.73/day Bigstore
(Generic) $996/year Pharmacy
Ultram 50 mg Lifetime 1–2 tabs/day Analgesic for headaches $25.99–38.99/30 days Bigstore
$0.87–1.30/day Pharmacy
$318–475/year
Movement disorder Lifetime 1x/day Treat neuromotor movement disorder $1.27–5.42/day Bigstore
medication∗∗ $464–1,978/year Pharmacy
∗ Ms. G.B. will be on these or similar classes of medications throughout her lifetime.
∗∗ Alternative anti-Parkinson medications were recommended for treatment of the movement disorder. These medications include Mirapex 3 mg per day; Sinemet 25/100
once per day and Requip 5–10 mg per day. The estimated cost for these medications ranges from $1.27 to $5.42 per day. The estimated annual cost is $464–$1978 per year.
Issues in Rehabilitation and Life Care Planning
Life Care Plan

Laboratory
Comprehensive Lifetime 2x/year Evaluate physiological effects of $99–133/lab Health Partners
metabolic panel medication $198–266/year Lab
Life Care Plan
Diagnostic Studies
Brain MRI Lifetime 1x/next 20 years Evaluate neurologic deterioration, $1,425–2,755/facility fee Health Partners
early onset dementia $285–551/professional fee MRI
8417: “8417_c026” — 2007/9/11 — 12:14 — page 615 — #25

$1,710–3,306/Total
615
616
Life Care Plan

Recreation
Adult recreational Lifetime 1-2x/year Provide supervised recreation and socialization $200–400/year Various
camps
Life Care Plan
Care Providers/Residential Care∗
Companion 15–20 years 8–12 h/day Provide structure and support; evaluate environment $18–20/h ABC Home Services
for safety; assist with problem solving needs and $144–240/day
shopping; meal preparation and so forth $52,560–87,600/year
Brain injury assisted Begin in 15–20 years Daily Long-term rehabilitative care and supported living $275/day Rehab Assisted Living
living program for life $100,375/year
Case manager Lifetime 1–2 h/month Client advocate; assess quality of care/activities; $80–100/h Continuous Services
coordinate care $960–2,400/year
Heavy housecleaning 15–20 years 2x/month Provide housecleaning and laundry assistance $75–100/week Once Over Cleaning
$1,800–$2,400/year
Life Care Plan
Transportation
Mileage 15–20 years Monthly Travel to medical/rehabilitative appointments $0.445/mile Various
reimbursement $900–1,200/year
∗ Her caregivers have recommended that she will need home care assistance (companion) as long as she remains at home. She is relatively independent in personal care
but will require significant daily support with home safety, routine problem solving, shopping, meal preparation, daily planning and heavy housecleaning. The
companion will implement treatment goals in consultation with the cognitive therapist, neuropsychologist and case manager. The physicians have stated that she will
8417: “8417_c026” — 2007/9/11 — 12:14 — page 616 — #26

experience early onset dementia within the next 15–20 years at which point she will probably have to transition into a brain injury assisted living program for life. The
assisted living program will provide long-term treatment support in the least restrictive environment.
Life Care Plan
Complications
Complication Secondary To Length of Risk Surgery Hospitalization Possible Outcome
Seizure disorder Carbon monoxide poisoning Lifetime No Possible Additional brain injury, increased care needs
Visual problems Carbon monoxide poisoning Lifetime No Possible Balance disturbance, altered mobility, falls, need for
yoked prism lenses and vision therapy
Motoric deterioration Carbon monoxide poisoning Lifetime No Possible Altered mobility and independence, falls, increased care
and equipment needs
Severe depression Cognitive impairments and limitations Lifetime No Possible Withdrawal, anger, paranoia, confusion, anxiety,
sadness, hostility, reduced self-esteem and
self-confidence
Issues in Rehabilitation and Life Care Planning
8417: “8417_c026” — 2007/9/11 — 12:14 — page 617 — #27

617
8417: “8417_c026” — 2007/9/11 — 12:14 — page 618 — #28
27 Treatment of Carbon
Monoxide Poisoning
with Yoked Prism Lenses
James F. Georgis
CONTENTS
27.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 619

27.2 A Review of Visual System Functioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 620
27.3 Visual Midline Shift Syndrome (VMSS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 622
27.3.1 Symptoms of VMSS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 622
27.3.2 Clinical Testing for VMSS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624
27.4 Treatment of VMSS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625
27.5 Rehabilitative and Supplementary Considerations . . . . . . . . . . . . . . . . . . . . . . . . 626
27.6 Case Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 634
27.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 640
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 640
27.1 INTRODUCTION
Carbon monoxide (CO) poisoning has many clinical manifestations. Malfunctions of
the cardiovascular and nervous systems caused by CO are numerous and varied. This
chapter discusses damage to brain structures that control spatial awareness, balance,
and movement. Treatment of these neurological symptoms requires a multidiscip-
linary approach. The multidisciplinary team often includes the neurologist, primary
care physician, toxicologist, neuropsychologist, psychiatrist, physiatrist, optometrist,
and neuro-otologist. A variety of allied health care professionals are also necessary in
the rehabilitation of the CO-poisoned patient. These include occupational therapists,
vocational rehabilitation specialists, cognitive specialists, and physical therapists.
Seventy percent of all the sensory nerves in the body come from the eyes or
the visual system. This discussion focuses on the role of the visual system in the
rehabilitation of the CO-injured patient. It has been estimated that 30 areas of the
brain are intricately involved in the processing of visual information and integrating
it with other sensory modalities to bring about appropriate actions. The visual process
will be reviewed and specific areas of the brain will be discussed that affect balance
and movement. According to the World Health Organization1 virtually all the cortical
619
8417: “8417_c027” — 2007/9/11 — 12:14 — page 619 — #1

structures and many of the subcortical structures of the brain can be affected by CO
poisoning. It identified common cognitive and motor deficits following CO poisoning.
These included deficits in higher cortical functioning sensory-motor deficits and visual
problems. Visual midline shift syndrome (VMSS) is a result of neurological injury
to subcortical structures. There is strong neuron cross-talk between subcortical and
cortical structures. Yoked prism therapy is the treatment for VMSS. The yoked prism
modulates the neural signals from the subcortical system to improve movement,
balance, and the perception of space.
27.2 A REVIEW OF VISUAL SYSTEM FUNCTIONING

The visual system involves an extreme degree of neural activity. CO poisoning can
destroy this activity in a short period of time. The brain uses 20% of the oxygen we
take up, an extraordinary amount considering that the brain accounts for only 2% of
body mass. A continuous supply of oxygen is essential for brain function. Loss of
oxygen delivery for a period as brief as 10 min can result in neural death.2
Visual information is contained in the light reflected from an object we view. In
order to understand what we see, we need sensory receptors that respond to reflected
light. This light passes through the crystalline lens of the eye, where upon the image
is inverted and focused on the back surface of the eye, the retina. The retina is
composed of millions of photoreceptors. The photoreceptors change the light stimulus
to internal neural signals within the many layers of the retina. These neural signals
are processed and form a bundle, the optic nerve. The optic nerve transmits this
visual information to the central nervous system. The optic nerve divides once it
gets into the brain. Ninety percent of the fibers go to the lateral geniculate nuclei
(LGN) of the thalamus and project directly to the cortex; the remaining 10% of
the fibers go to subcortical structures like the superior colliculus of the midbrain.
However, the fact that these other receiving nuclei are innervated by only 10% of the
fibers does not mean these pathways are unimportant. The human optic nerve is so
large that 10% of the optic nerve constitutes more fibers than are found in the entire
auditory pathway. The superior colliculus plays a big role in visual attention.2 The final
projection to the visual cortex is through the geniculo-cortical pathway terminating
in the primary visual area of the occipital lobe. The retino-geniculo-cortical pathway
contains two different retinal ganglion cells, consisting of the smaller parvocellular
cells (P cells) and larger magnocellular cells (M cells). The P and M cells continue their
separate paths to the occipital cortex where much of the sensory visual information
is processed. The output from the occipital lobe is contained in two major nerve fiber
bundles. One bundle contains P cells, which leaves the occipital cortex, which goes
inferiorly or ventrally into the inferior temporal cortex. Another nerve bundle which
contains M cells leaves the occipital cortex and goes superiorly or a more dorsal path to
the posterior parietal cortex. The ventral or occipito-temporal pathway is specialized
for object recognition, for determining what it is we are looking at. The dorsal or
occipito-parietal pathway is specialized for spatial perception, for determining where
an object is. There are two main questions that must be answered in visual perception.
We must understand “what” we are seeing and “where” we are seeing. The response
8417: “8417_c027” — 2007/9/11 — 12:14 — page 620 — #2

Treatment of Carbon Monoxide Poisoning with Yoked Prism Lenses 621
of neurons in the temporal lobe is activated by fibers from the central vision areas of
the retina or fovea. The response of neurons in the parietal lobe is activated by fibers
from peripheral vision areas of the retina. The fovea has many more retinal cells and
gives us greater visual acuity. We usually look directly at things we wish to identify,
thereby taking advantage of the greater visual acuity of foveal vision.
The dorsal–parietal and ventral–temporal pathways are not isolated from one
another but communicate extensively. We know that the retino-geniculate cortical
tract contains 90% of the fibers in the optic tract. Approximately 10% of the remaining
nerve fibers terminate in the superior colliculus in the midbrain. This is a subcortical
visual pathway that is referred to as the retino-collicular pathway. The superior col-
liculus is an important neural stimulation for spatial awareness and movement. Gerald
Schneider in 1969 at the Massachusetts Institute of Technology found evidence of
the importance of the colliculus in studies of hamsters.3 Those animals with cortical
lesions could not identify visual subjects. This involved damage to the occipito-
temporal pathway, which was affecting the “what” pathway. Hamsters with collicular
lesions exhibited reduced ability to orient toward the stimulus and their motor system
was impaired. They could not move and acted like they were blind. This superior
collicular injury affected the subcortical feedback to the occipito-parietal system and
was affecting the “where” pathway.
As I mentioned earlier, there is strong cross-talk between subcortical and cor-
tical structures. The superior and inferior colliculi receive input from the visual and
auditory pathways and use it to develop a representation of where objects are in
space and to generate eye movements to attend to these objects. The neurons in
the retino-collicular pathway are almost fully myelinated by 3 months after birth.4
The superior colliculus has a virtually normal adult pattern of neuronal lamination
even before birth, in preparation for receiving retinal axons. The retino-collicular
pathway is almost fully operational at birth. This pathway allows the infant to have
visually directed actions in the first few months of life. This pathway is sometimes
referred to as the “ambient system.” Ambient means surrounding on all sides. This
retino-collicular system gets its feed from the retinal cells in the peripheral retina.
Posner states that this pathway is called ambient because the visuomotor behaviors
it produces are directed to the global properties, as converse to their fine details.4
The newborn infant looks at the outline of mother’s face rather than the finite details
of her face like her nose and mouth. This allows the infant to receive visual stim-
ulation from its peripheral visual field and to turn its head or eyes to the object of
interest.
The M cell pathway is sometimes referred to as the ambient system. The P cell
pathway is concerned with fine details and the M cell pathway is concerned with
global orientation. The neural development of the retino-geniculo cortical pathway,
particularly of the P cell subsystem is slower in development. This system is often
referred to as the “focal system.” The fovea in the retina is slow to mature and the
optic nerve fibers innervating the dorsal lateral geniculate body (dLGN) are not fully
myelinated. The dLGN is not fully formed and the visual cortex is not completely
developed. These visual systems do not reach full operation until about 2 years of age.4
Thus, the infant is unable to attend to fine details early in life and the retino-collicular
pathway is the dominant visual pathway until the child is about 2 years old.
8417: “8417_c027” — 2007/9/11 — 12:14 — page 621 — #3

Massive innervations remain between cortical and midbrain structures. These two
structures are constantly communicating in order to plan coordinated actions. Most of
our experiences in the real world involve information from many different modalities.
We plan our actions by receiving information from the visual, vestibular, auditory,
and somatosensory modalities. In order for the visual process to function effectively,
it must have normal ocular structure and the more than 30 brain regions that process
the visual information must be fully operational. Any neurological injury such as
traumatic brain injury, stroke, or CO poisoning would adversely affect the entire
visual system. The visual system would then be unable to integrate visual information
with other modalities like hearing, tactual information, and motor movement.
27.3 VISUAL MIDLINE SHIFT SYNDROME (VMSS)

27.3.1 SYMPTOMS OF VMSS
This syndrome is caused by neurological trauma or CO poisoning to the retino-
collicular pathway in the midbrain. This ambient system gets its feed from the
peripheral retinal neurons. This system organizes spatial information from the visual,
kinesthetic, proprioceptive, and vestibular systems. When there is a mismatch of
sensory information coming to the midbrain from the proprioceptive, vestibular, and
tactile systems, it affects a person’s perception of space. When this mismatch occurs,
the ambient visual system steps in and makes adjustments to alter space perception so
that the body can maintain its balance. The reason that this midbrain is so important is
that it gathers information from our sensorimotor system to organize balance, posture,
and movement. Essentially, the midbrain is responsible for keeping our posture and
letting us know our internal concepts of space. This midbrain functions as a platform
by which higher perceptual processing in the parietal lobe begins to organize. We
know there is direct communication between this spatial collicular subsystem and the
parietal lobe spatial system. Once the midbrain spatial system communicates with
the parietal lobe spatial system, it in turn organizes sensory information going to the
temporal lobe. In other words, the ambient system in the midbrain organizes spatial
information so the focal system can function efficiently. The “where” visual system
must be intact before the “what” visual system can function. We must know where
we are before we know what we are looking at.
In the case of CO poisoning, stroke or traumatic brain injury (diffuse axonal
injuries), the sensorimotor information coming from the right side of the body may
be different from that coming from the left side of the body. This is mismatched
information and the injured patient may feel like the midline of his body is off to one
side. This person is off balance and may place more weight on one side of his body to
maintain balance. The ambient spatial visual system can change its understanding of
space and equalizes the sensorimotor information coming in from the muscles, joints,
and inner ear. Thus, the ambient system recreates space so the patient’s balance is
restored.
The symptoms of CO poisoning are varied and numerous. Penney in 2000 presen-
ted a study of CO-symptoms in the book Carbon Monoxide Toxicity (pp. 408–413).5
In it, data were obtained by e-mail over a period from late 1997 through early 1999.
8417: “8417_c027” — 2007/9/11 — 12:14 — page 622 — #4

Sixty-six correspondents indicated they had sustained chronic CO exposure, that is,
CO exposure lasting 24 h or more. The symptoms are seen in Table 27.1. The total
number of symptoms is 93. Of those symptoms, 31 are common to VMSS. The bolded
symptoms in Table 27.1 are directly connected to VMSS.
TABLE 27.1
Symptoms Noted During Exposure to Carbon Monoxide
Symptoms Symptoms Symptoms
Agitation Flu-like symptom Personality change
Anxiety Flushed Pressure in head
Apathy Forgetful Shortness of breath
Appetite loss Gastrointestinal problems Seasickness/Motion sickness
Ataxia Hair loss Seizure
Attention, loss Hallucinations Shoulder pain
Back pains Handwriting problems Sick feeling
Balance problems Headache Sinusitis
Body ache Hearing problems Skin, cherry red
Bronchitis Hypertension Skin, dryness
Chest tightness/pain Hypoglycemia Sleep problems
Choking Ill, violently Sleepiness
Chronic fatigue In fog Smile, convulsive
Concentration problems Incontinence Speaking problems
Confusion Insomnia Spelling problems
Constipation Iron level low Suicidal
Coolness Irritability Sweats
Coordination Problems Learning problems Syncope, partial/complete
Cough, spells Lethargy Tachycardia
Cramps Libido loss Throat, burning sore
Depression Lightheadedness Tingling legs/arms
Diaphragm pain Lips red Tingling lips
Diarrhea Liver pain Tinnitus
Disorientation Memory loss Tiredness
Dizziness Mood changes Tongue, thickened
Drop things Moodiness Tremor
Dysarthria Muscle ache/pain Twitching fingers
Ear problems Nausea Vertigo
Emotional problems Neck pain Vision problems
Energy level Nerve deafness Vomiting
Extremities cold Numbness Walk, inability to
Eye pain/ache Palpitations Weakness
Fatigue Panic attack Weight loss
Fibromyalgia Paralysis Word-finding problems
Parathesias
Source: From Study A, Penney, D.G., Carbon Monoxide Toxicity, D.G. Penney, ed., CRC Press,
2000.
8417: “8417_c027” — 2007/9/11 — 12:14 — page 623 — #5

TABLE 27.2
Post-traumatic Midline Shift Syndrome
1. Balance problems
2. Bumps into doors, walls, and furniture
3. Walks into other family members while walking with them
4. Trips on stairs and must hold on to the handrail
5. Busy patterns in the carpet, tile, or wallpaper are uncomfortable to look at
6. Very nervous when around crowds because of excessive crowd movement and noise
7. Trouble finding objects on shelves of supermarket or department stores; gets lost in the stores
8. Misplaces objects like keys, glasses, wallet that are located right in front of them
9. Drops things and tends to knock things over when using either hand. Hits their teeth when drinking
out of a glass. Hits lip with eating utensils
10. Handwriting has become illegible
11. Must look at their hands to type on the computer keyboard
12. Trouble driving. Must use excessive concentration in order to stay within the lines of the road
13. Must turn their entire body to see cars coming from their peripheral field. Peripheral moving cars
startle them
14. Trouble judging speed and position of other cars. Trouble changing lanes
15. Must take two or three tries to park the car within the parking space. Hits curbs when making a
turn
16. More trouble driving at night. Bright car headlights confuse them. Tendency to get lost while
driving and often misses turns
17. Photophobia (light sensitive) to outdoor and fluorescent lighting. Bright lights stimulate headaches.
Banks of fluorescent lights in large stores and supermarkets are very bothersome
18. Neck, shoulder, back, hip, and foot pain. Chiropractic, massage therapy, physical therapy only
relieves symptoms for one day or less
19. Loses place while reading. Words overlap and move on the page. Must reread a lot and often uses
finger or a straight edge to follow the line of print
20. Peripheral hallucinations like shadows out to the side of vision and when they look to the side
nothing is there
21. Ringing in ears, ears feel blocked, and trouble hearing. Background noise like the sound of the
television makes it difficult to talk on the telephone
22. Not able to listen to radio or hold a conversation while driving. Trouble with multitasks of any kind
23. Dizziness when they stand up quickly or sit up from a lying down position
24. Trouble sleeping
25. Memory problems
This author has diagnosed and treated approximately 1500 patients with this
VMSS. Table 27.2 lists the most common symptoms of VMSS.
27.3.2 CLINICAL TESTING FOR VMSS

The patient is tested in the sitting and standing position. He is instructed to position
his head and eyes in a straight-ahead position. A pen is held vertically 18 in. in front
of him in front of his right shoulder. The pen is moved in front of his facial plane in
8417: “8417_c027” — 2007/9/11 — 12:14 — page 624 — #6

a parallel pattern from the right shoulder toward the left. The patient is asked to state
when the pen appears to be directly in front of his nose as he follows the pen with only
eye movements. The test is then repeated by moving the pen from the left shoulder
toward the right. The examiner should be seated or standing at random off to the left or
right side at approximately a 30◦ angle to the patient in order to lesson the influence of
the subject’s response relative to the examiner’s position. Motor evaluation of VMSS
is very important in assessing the patient’s posture and movement. The patient is
instructed to walk down a normal hallway and ask which foot he is placing more
pressure on the floor. While the patient is walking, note if his shoulders are level or
if one shoulder is tilted downward.
The next motor evaluation should be directed to the position of the patient’s feet.
Visual midline shift will usually show patients walking on the temporal or outside
edges of one or both feet. Direction of walking is the next motor evaluation. The
patient is instructed to walk down the hallway without trying to stay in the center of
the hallway. Use the following command, “Do not try to stay in the middle of the
hallway. Just float down the hallway, be aware of the walls and tell me if you are
going to the right or left side of the hallway.” VMSS patients have an unequal gait
and place more pressure on one foot. Often these patients have a moderate to severe
shoulder tilt. The foot posture is almost always abnormal and the direction of walk is
rarely straight.
27.4 TREATMENT OF VMSS

Patients with symptoms of VMSS often report them to eye care professionals, that
is, optometrists and ophthalmologists. Eye care professionals often tell them their
problems are not in their eyes and that their eyes appear to be healthy. Patients with
stroke, traumatic brain injury and CO poisoning often experience anxiety. In many
instances these patients are referred to psychologists or psychiatrists in an attempt
to treat their anxiety. The referral for mental treatment is sometimes based on a
diagnosis of visual hysteria. The psychiatrist or psychologist may not recognize that
many of these patients are suffering from syndromes affecting the visual process in
the brain.
In the early acute medical setting of CO injury the eye care is typically ophthal-
mological. In the patients who have survived brain injury, the emphasis is on the
immediate effects of the trauma with an evaluation of the patient’s ocular health.
Neuro-ophthalmologists address the neurological integrity of the visual system and
will order tests to localize injury. Dr. Gianutsos, a research professor of Psychiatry
at New York University Medical Center, states that in the United States ophthalmo-
logists are rarely trained, experienced, or interested in visual system rehabilitation or
function.6 She indicates that an increasing number of optometrists have experience
and interest in rehabilitation which represents a natural extension of the emphasis
within optometry on visual function and training.7
It is this author’s experience that optometrists are the logical eye profession-
als to treat VMSS. Optometrists with an interest in brain injury rehabilitation often
use standard vision therapy techniques to treat brain-injured patients. While these
8417: “8417_c027” — 2007/9/11 — 12:14 — page 625 — #7

techniques are often helpful, the treatment is mostly concerned with eye movement
and eye focusing procedures. VMSS is a brain problem that relates to body movement
and body positioning. Visual therapy techniques do not address the spatial and balance
symptoms of the syndrome.
The treatment of VMSS with yoked prism was developed by William Pad-
ula in 1988. Dr. Padula explains the syndrome and treatment very well in his
many publications.8−11 He discovered the syndrome by studying the research of
Trevarthen,12 who studied primates at Oxford University, England. He developed
a concept that the ambient visual system organizes space and sends information to
cells in the higher brain. Dr. Padula found the use of yoked prisms could rebalance
the ambient visual process. The prism changes the organization of space. The yoked
prism is a wedge of glass or plastic that shifts the image of an object. The patient with
VMSS has a distortion of space and the yoked prism will rebalance the ambient visual
process. The prism shifts the midline in the direction where it belongs. The prism
will change the patient’s posture and weight bearing to equalize his gait. The prism
redirects the concept of midline to enable the injured patient to have better posture
and movement.
The prism is introduced to the patient in small increments and the midline test is
readministered. The prism is added in the opposite direction to the midline shift until
the prism neutralizes the shift. Motor evaluation with the yoked prism in place is a
very important step in refining the power of the yoked prism. When the correct prism
power is used, the patients’ gait will be equal, the foot and shoulder posture will be
normal and the walk will be straight.
Treatment with yoked prisms should be undertaken as soon as possible after the
neurological injury. The other disciplines that are necessary to rehabilitate the injured
patients rely on the patient’s midline treatment. For instance, cognitive therapy is
not as effective if the patients reading ability is compromised by the VMSS. Neuro-
psychological treatment is not effective if the patients’ visual, auditory, and spatial
abilities are not functioning properly. Physiatry pain management is not as effective
in treating joint and muscle pain when the posture is not stabilized by the yoked
prism. Neuro-otological treatment is also hampered by VMSS when there is a mis-
match between the visual and auditory systems. Psychiatry treatment is also affected
when the patient is hampered by spatial and peripheral hallucination symptoms. Voca-
tion rehabilitation specialists are more successful in their recommendations when the
injured patient has good reading and movement skills.
27.5 REHABILITATIVE AND SUPPLEMENTARY

CONSIDERATIONS
A variety of allied health care professionals are necessary in the rehabilitation of the
CO-poisoned patient. Other disciplines find that yoked prism treatment is extremely
beneficial to the patient with midbrain injury. They usually recommend to the patient
that they see a neuro-optometrist who is qualified in the diagnosis of CO poisoning
and treatment with yoked prism lenses. They generally prefer that the patient be
treated with the lenses before thy start their rehabilitative specialty.
8417: “8417_c027” — 2007/9/11 — 12:14 — page 626 — #8

In the next few pages I have included comment from other treating specialists.
These include James M. Gracey, rehabilitative consultant and therapist; Timothy
O. Hall, physiatrist; Susan Gawey-Apgar, speech and language pathologist; Steven
Stockdale, clinical psychologist; Pat McKenna, occupational therapist; and Dennis
A. Helffenstein, a neuropsychologist.
James M. Gracey: A large percentage of my active caseload of more than 250 patients
involves brain injury from trauma or CO exposure. We have been providing rehab-
ilitation and life care planning services for 40 years (see another chapter in this
book). One of the most common and pervasive symptoms, which we observe,
involves visual dysfunction. Table 27.3 shows the problems patients regularly
report. See Chapter 26.
The above problems tend to persist if left untreated, often leading to a further
reduction of home, work, and community functioning. Such patients are referred
to a qualified neuro-optometrist for evaluation and treatment. The combination of
special glasses (yoked prism lenses) and monitoring often results in resolution (or
near resolution) of the problems, thus freeing the patient to focus on the return to
work or school. Such treatment is vital in assisting patients to reengage in their
activities of daily living and work.
James M. Gracey, EdD, CRC, CLCP

Rehabilitation Consultant and Therapist
Certified Life Care Planner
1660 S. Albion Street #1010
Denver, Colorado 80223
Timothy O. Hall: While combining the fields of neurology, orthopedics, and occu-
pational medicine, physiatry provides a unique perspective in medicine. This is of
great importance in the management of brain injury. The multidisciplinary approach
necessary to care for these patients is often managed by physical medicine and
rehabilitation. By incorporating various disciplines, a full range of symptoms can
be addressed and treated.
TABLE 27.3
Visual System Dysfunction Related Problems
• Dexterity, poor handwriting

• Balance, dizziness, tripping
• Reading, poor reading comprehension, moving lines/letters on page
• Photophobia to all kinds of lighting
• Tracking, blurriness, depth perception
• Driving/veering to the left or right/night driving
• Eye discomfort, fatigue, related headaches
• Nervousness in crowds
• Losing/misplacing common items
8417: “8417_c027” — 2007/9/11 — 12:14 — page 627 — #9

In the clinical management of head injury patients over the past 17 years, I have
learned the critical role played by appropriate treatment of midline shift syndrome and
other visual sequelae of brain injury. Many patients present with a complex constella-
tion of symptoms that can be overwhelming in the context of diagnoses and treatment.
Often these symptoms can be better understood and more appropriately treated by
understanding the visual consequences of brain injury. This includes a connection
between postural distortions and chronic myofacial pain, as well as musculoskeletal
headache. It is critical for successful outcomes of these complex presentations to man-
age the visual situation in order to increase the likelihood of success in other areas.
Usually one of my first interventions in patient evaluation involves visual testing.
Treatment of visual distortions is an integral part of the team put together to manage
closed head injury.
Although controversial, it has been my experience that the management of midline
shift and other visual disturbances in this population is critical for success. In my
practice, treatment failures that come my way, often years after injury, are most likely
to improve by not only instituting a multidisciplinary approach to their treatment but
by including management of undiagnosed visual dysfunction.
Timothy O. Hall, MD
559 E. Pikes Peak Avenue, Suite 100
Colorado Springs, Colorado 80903
Susan Gawey-Apgar: Over the past 15 years my practice has consisted of mild trau-
matic brain injured (MTBI) and CO-poisoned patients who have been referred by a
variety of sources. It has been my experience with this population that patients with
MTBI and CO trauma suffer from a variety of physical, cognitive, and emotional
difficulties that affect their day-to-day activities.
Typically, when a patient is referred to my clinic for treatment, a one page doc-
ument is generated identifying them and providing minimal medical information. It
is my job within the first hour with a patient to ask a variety of questions that will
help with the initial phase of treatment. The initial phase of treatment is essential
for MTBI and CO patients because it is the most critical to the patients overall care.
By the time the patient is referred for cognitive rehabilitation, they feel hopeless and
depressed because their previous care has only focused on the physical aspects of their
injury.
The first phase of treatment (there are typically three phases to a patients’program)
consists of analyzing the symptom checklist (Table 27.4). It is of interest that the
checklist for cognitive symptomatology is closely related to the symptoms of post-
trauma midline shift patients.
Post-trauma midline shift syndrome patients and cognitive disorder patients, have
difficulties processing visual data, have poor reading scanning skills, and show a slow
speed of processing, handwriting problems, attentional difficulties, and multitasking
problems. The patient, therefore, benefits from seeing a neuro-optometrist during
phase one of outpatient rehabilitation. Having both specialties working together max-
imizes the effects of cognitive intervention and support the patient with state of the
art medical care.
8417: “8417_c027” — 2007/9/11 — 12:14 — page 628 — #10

TABLE 27.4
Symptom Checklist
Cognitive problems
• Attention and concentration problems
• Short-term memory loss
• Trouble remembering old things
• Word finding
• Understanding what is said
• Understanding what is read
• Making decisions or solving problems
• Slower speed of thinking
• Getting lost or disoriented
• Trouble juggling several things at once
• Disorganized or confused thinking
• Stuttering or slurring
Physical symptoms
• Dizziness
• Coordination of hands, feet, or legs
• Ringing in the ears
• Fatigue
• Jaw pain
Emotional symptoms
• Feelings of sadness or depression
• Crying spells
• Suicidal feelings
• Increased or decreased emotion
• Low motivation
• Change in sex drive
• Irritable and easily frustrated
• Feelings of anxiety or fear
Post traumatic stress syndrome (PTSD)
• Recurrent and intrusive thoughts about the accident
• Nightmares
• Flashbacks
• Anxiety or panic while driving
• Hypervigilance
• Fear
• Does this affect social activities or relationships
CO-poisoned patients like those with MTBI, clearly demonstrate the same profile
on both the cognitive symptom checklist and the VMSS check list (see Table 27.2).
Key to rehabilitation is approaching the patient holistically in order to promote the
maximum rate of medical recovery.
8417: “8417_c027” — 2007/9/11 — 12:14 — page 629 — #11

Susan Gawey-Apgar, MS, CCC-SLP

Speech and Language Pathologist
Neuro-Cognitive Therapist
208 North Corona Street
Steven Stockdale: Electroencephalogram (EEG) neurofeedback has been used in

the treatment of traumatic brain injury, utilizing operant conditioning to normal-
ize brain wave patterns. Many of the symptoms of MTBI, which include physical
symptoms, cognitive symptoms, and psychological-emotional symptoms, correlate
with the symptoms of post-trauma midline shift syndrome. Physical symptoms of
MTBI include sensitivity to noise, tinnitus, dizziness, fatigue, blurred vision, and
headaches. This symptom group overlaps with symptoms from post-trauma midline
shift syndrome.
Many patients that come in for evaluation of MTBI have significant visual prob-
lems that were not there previous to their brain injury. In those cases, they are
referred for evaluation for post-trauma midline shift syndrome. This syndrome can
be corrected by yoked prism lenses.
Although pre- and postsymptom changes in vision have been noted through neur-
ofeedback alone, correcting the midline shift problems greatly enhances the outcome
of the neurofeedback treatment.
Steven Stockdale, PhD

Licensed Clinical Psychologist
EEG Neurofeedback Specialist
2132 North Nevada Avenue
Pat McKenna: When evaluating a person as an occupational therapist (OT), we need

to look at the entire person with regard to what may be affecting them and their ability
to perform tasks. Thus, we need to look at the physical, cognitive, sensory perceptual,
emotional, and psychological factors that might contribute to the person’s difficulties.
When a person has sustained a traumatic injury which may affect the head or neck, this
can have caused damage to not only the joints and muscles, but, also to the sensory
pathways that can include, not only those that contribute to the sensation of pain,
but also those of the sensory perceptual system. This is a less well-known system,
but it is at the core of our abilities to function. This includes the proprioceptive and
kinesthetic systems as well as the vestibular, touch, olfactory, auditory, and visual
systems.
It is these systems which work together or individually to provide us with inform-
ation regarding what is going on around us and happening to us that we need to know,
so we can respond appropriately and effectively. It is our safety net. It is what allows
us to move through the world without falling down, to reach, to get, or replace objects
that we need and even feeding ourselves. It is the system that lets us know we have
touched something hot and that we need to move our hand away fast. It lets us know,
as we walk down stairs, where to place our feet, or walk in the mountains without
falling over on all the rocks and uneven ground. It allows us to learn to drive down
8417: “8417_c027” — 2007/9/11 — 12:14 — page 630 — #12

a highway and stay in the correct lane and to be able to stay out of the way of traffic
or other things that are coming at us. Or, if playing ball, to reach appropriately in
order to catch a ball. It is our silent sentry and helper that allows us to do virtually
everything we do.
This is not to downplay the importance of the joints and muscles and the nerves
that cause them to move. These are things we are all more familiar with and which
are visually very obvious if they are injured. A broken leg would not hold you up. A
torn muscle would not let the body parts move correctly, if at all. These, we have all
seen and understand.
The sensory perceptual system, however, is essential for all of these things to
work. If anyone has gone to an amusement park fun house, it is these senses that
they are “playing with,” You look at a wavy mirror and it distorts your perception of
yourself and your surroundings. You walk on the moving floor that goes up and down
and very soon, you want “out of there,” sometimes becoming nauseous, but typically
were tense and nervous because you could not trust your senses and were afraid
you’d fall.
The ability to be able to automatically trust these senses has been built from
babyhood. At first the baby is just waving its hands around. Then, as this continues,
pathways develop that direct the muscles to respond appropriately to the need, the
feel, the desire of a certain response to occur. The baby can now consistently get food
from the plate to the mouth. They start to try to walk and keep falling. But soon, they
are stable and off they go running. It is this pattern of development that has gone on
in everything we do. To go up and down stairs without watching where our feet go;
to take a hike in the mountains and not pay attention to the minor unevenness of the
surfaces we are walking over.
When a person has a traumatic whiplash or brain injury, these pathways can be
damaged. Our visual system is one of our major sensory systems that these pathways
have learned and depend upon, and it is one vulnerable to injury when the head is
struck or shaken or suffers a lack of oxygen. Now, all of a sudden the information
that the body has trusted to help to determine “where it is in space,” where and how
it needs to move to keep it safe and to perform a task safely, cannot be trusted. The
person tries to walk down a flight of stairs, but finds herself losing her balance and
falling. Where she automatically thought the foot should go, was not correct. They
reach for a glass and knocked it over. They find themselves getting a traffic ticket
for driving with their tires to the left of the centerline. They find themselves getting
a headache or becoming nauseated when they try to read a newspaper—and even
worse, when trying to look at the computer monitor.
Often these problems have been subtle enough that the person is not fully aware
of what is wrong. They often just feel something is weird, and they’re more tired and
have more headaches. They think they must just be getting awkward. Very often they
do not even mention such things to their doctors. They are embarrassed and they do
not realize it could have had anything to do with their head injury.
The role of the OT is to look for any problems that might be present that are
affecting function. Since the OT is typically having the person do these more func-
tional tasks in a more real life type of setting, they have the opportunity to observe
these things happening and to make an association to their head injury. Once these
8417: “8417_c027” — 2007/9/11 — 12:14 — page 631 — #13

types of problems have been observed and a “red flag” is raised, the OT can do some
simple, more specific testing—basic optometric tests to be more clear that, “Yes,
indeed, this person needs to be seen by an optometrist knowledgeable about the types
of problems that can occur following head injury.” Then, on the basis of the results
of the optometrist’s evaluation and the effects of the treatment, and if needed in some
cases, can assist the optometrist with visual training exercises. It is usually a good
collaboration, as both health care professionals are strongly focused on functional
outcomes.
From the perspective of an OT, this is the type of evaluation that needs to be
done very early after an injury. Often, headaches and neck pain are present from the
accident. If the person is tense from his inability to trust that he would not fall, and
so forth, any therapy to try to heal or reduce muscle tension in the neck will not be
successful. And, with constant pain, headache, nausea and a feeling of being unsafe
moving about … depression is hard to avoid.
Unfortunately, these are not problems that are typically looked for early on. If
not addressed early, the person will not heal well and the injuries become more
entrenched and harder to deal with effectively as time passes. Since the visual system
is at the heart of so much of what we depend upon to make our way through life,
when it is not functioning appropriately, it affects the person’s ability to do many of
his basic activities of daily living and work. Driving, shopping, taking the laundry
downstairs, and so forth, all become tasks with which the person now feels tense and
uneasy … sometimes fearful. Thus, they begin to avoid doing them. Since much of a
person’s world of work also depends upon one’s vision, this becomes difficult, if not
impossible for the person to do accurately and productively. As an OT, these things
are issues that we need to help the person deal with. Thus, recognizing the potential
problem and getting them to an optometrist who can appropriately diagnose and treat
it is essential, if we are to help that person return successfully to their previous life,
or, at the very least, to their highest level of function that is possible.
One of the treatments that can often be seen to provide dramatic results is use of
yoked prism lenses. When patients are properly diagnosed and appropriately fitted
with these lenses, we typically see a dramatic response. Patients comment to us that
they “couldn’t believe the difference in how they felt virtually the minute they put
them on.” They could actually “see” the steps and feel like, “Wow! I can relax now
… I don’t feel like I might fall all the time.” And they are not knocking things over
when they reach for them. They’re finding that they can read longer. Granted, not
all things are immediately perfect, but the change is so substantial that now, as a
therapist, I can begin to work better with them on trying to remediate the problems
that are interfering with their ability to do the things they used to do without so much
frustration.
Pat McKenna, Occupational Therapist, Registered
Starting Point
8745 West Fourteenth Avenue, Suite 112
Lakewood, Colorado 80215
Dennis A. Helffenstein: As a neuropsychologist who specializes in evaluating and
treating individuals with acquired brain injuries, I am familiar with the wide-range
8417: “8417_c027” — 2007/9/11 — 12:14 — page 632 — #14

of vision problems they can and often do experience (see another chapter in this
book). Indeed, a number of the tests, which I routinely administer as part of the
test battery, identify and quantify those deficits. In many cases, the visual defi-
cits lead to a significant functional impairment in a wide variety of day-to-day,
leisure, academic, and work-related activities. However, as a rehabilitation neuro-
psychologist, I felt I had little to offer these patients. I would on some regular basis
refer them for vision therapy, but even when treatment was successful, many of
these people were left with profound visual deficits, which continued to limit their
functioning. See Chapter 23.
In the mid-1990s, I received neuro-optometric reports from James Georgis, OD.
It was then that I was introduced to the term, “yoked prisms.” My patients had been
referred to Dr. Georgis for evaluation and treatment at or around the same time that
I was evaluating them. Without exception, each of the patients who Dr. Georgis
treated with yoked prisms returned to me reporting tremendous improvement in their
visual function. More importantly, they were experiencing improved functioning in
their driving, reading, work and leisure activities, and so many other aspects of their
day-to-day lives.
Frequently, I have had the opportunity to retest individuals after they began wear-
ing yoked prisms. On retesting, I routinely administer a number of the vision tests
with their yoked prisms on and then with the yoked prisms off. I was surprised
to find that in almost all cases the deficit identified by the test with their prisms
off is not apparent in test performance when the patient is wearing prism glasses.
The most notable improvements were seen on tests of visual scanning and visual
inattention.
I would also note that in addition to improved visual function, my patients often
report a reduction in the frequency and severity of headaches, improved balance and
gait, and a reduction in cervical and spinal pain (usually because their chiropractic
adjustments are now “holding” better) once they began wearing yoked prisms. In my
experience, it is extremely rare to find a treatment modality that has such an immediate
positive impact on so many aspects of a patient’s functioning.
I now routinely screen for post-traumatic vision syndrome and also administer the
Padula visual midline screening test as part of my test battery. When I identify a visual
midline shift or the patient is reporting visual symptoms suggestive of a hemispatial
inattention or visual midline shift, I consider the use of yoked prisms as a treatment
possibility.
I would like to extend to Dr. Georgis my thanks for his compassionate and
successful treatment of so many of my patients.
Dennis A. Helffenstein, PhD, CRC.

Licensed Clinical Psychologist
Colorado License #1484
Certified Rehabilitation Counselor #00001338
Diplomate, American Board of Vocational Experts #64073
Certified, Health Service Provider in Psychology National Register #41454
3545 American Drive
8417: “8417_c027” — 2007/9/11 — 12:14 — page 633 — #15

27.6 CASE STUDY

Ms. SL was a 50-year-old woman referred for VMSS evaluation by her Colorado-
based neuropsychologist. She reported that she experienced a chronic (long-term,
lower-level) exposure to CO between August 1988 and early November 2000. The
exposure was caused by defects in the construction of her furnace ventilation system.
The vent for her furnace was too small. She also noted that her house was “extremely
tight,” that is, it took 8 h for the air to completely change in her home. This was
a problem because the furnace was a “natural draft” furnace, taking oxygen from
inside the room as opposed to air from an outdoor air source. This tended to create a
negative pressure inside the home, which resulted in downdrafting. That is, the home
was taking air in through the furnace vent. This caused exhaust from the furnace to
migrate into the living area. This also resulted in incomplete combustion of the fuel
gas, which produced CO as well as soot (elemental carbon). Ms. SL noted that from
the time she moved into the home in August 1988 she had a problem with black soot
build-up around the home. For example, she noted that she had a white carpet in
the home which she could “never keep clean.” Negative pressure was present in the
home, particularly when there were fans running inside the home (e.g., bathroom fan,
range hood exhaust fan, or when the clothes dryer was operating). It was noted that
in the winter, downdrafting was more pronounced. There was also a concern that her
water heater may have been vented improperly and may have also been downdrafting.
A reconstruction of how the exposure occurred was preformed by a local expert
who specializes in CO-induced injuries. It was found that the furnace was producing
high levels of CO in the range of 1200–1300 ppm, especially during start-up. A
significant amount of corrosion was noted on the furnace suggesting the presence
of CO. Ms. SL’s exposure occurred on a daily basis during the winter months when
the furnace was operating. She experienced a variety of symptoms consistent with
CO poisoning. These symptoms gradually became worse overtime. She also noted
that during the past summer that she had spent out of her home, she did notice an
improvement in her symptoms. Once the problem was identified, she had both her
furnace and water heater replaced. This occurred in early November of 2000. Ms.
SL noted that her symptoms did begin to abate to some degree when the exposure
ceased.
Ms. SL experienced multiple symptoms during the exposure consistent with
CO poisoning. She reported the following symptoms to her neuropsychologist in
December 2000. These involved multiple “flu-like” symptoms, including nausea,
headache, fatigue/lethargy, and diarrhea. Other physical symptoms included chest
pain/tightness, cough, dizziness, lightheadedness, a metallic taste in her mouth, occa-
sional episodes of vertigo, sleepiness, intermittent tinnitus, restless sleep, and multiple
chemical sensitivity. Motorically, she noted motor weakness, motor incoordination,
muscle spasms and tremors, tingling and numbness in her hands, phonophobia (noise
sensitivity), and balance problems. Visually, she noted difficulties with blurry vis-
ion, photophobia (light sensitivity), double vision, and fluctuation in her visual
acuity, perceiving flashing lights or black dots, and difficulties with depth percep-
tion. Cognitively, she noticed difficulties with attention and concentration, short-term
memory, mental confusion, slowed speed of mental processing, multitasking, and
8417: “8417_c027” — 2007/9/11 — 12:14 — page 634 — #16

more frequent blatant errors in her work. From an emotional and psychological stand-
point, she noted depression, anxiety, and irritability. Ms. SL noted that when she would
spend an extended period of time in the home (e.g., over a weekend without leaving),
her symptoms would gradually increase. She noted that she would frequently “call in
sick” on Mondays.
Educational History: Ms. SL graduated from high school, college, and graduate
school with an approximated grade average of 3.75. She was currently taking another
course in a related field. She noted the new course work was quite difficult for her
and she had to work much harder to obtain good grades. She was currently working
as a college professor at a community college. She served as department chair for
approximately 5 years. She was carrying a full teaching course load, was writing
extensively, and developing extensive instructional materials. When she moved into
her home in 1998 she began to notice a “gradual change for the worse.” She began to
realize, however, that she was working longer and harder hours in order to accomplish
the same amount of work. By 1996 she had to give up some of her outside consulting
contracts, as she was unable to manage both the consulting and teaching. By 1998, she
resigned her position as department chair. Her student ratings (i.e., the ratings that the
students give her upon completion of a course) had deteriorated notably. She noted
that it was in approximately 1996 that she began to note a significant deterioration of
her physical abilities. Whenever she would leave her home for a significant period of
time she would begin to feel better physically. She felt that she was using virtually
100% of her available energy for her work and thus had no social or recreational life.
Neuropsychological evaluation stated Ms. SL had developed a variety of vis-
ion problems. She noted problems with double vision, particularly when she was
fatigued. Her eyes were slow to focus, suggesting problems with accommodation.
She had problems with depth perception such as judging distance. She noted that
she tended to bump into things more frequently and veered off center when walking.
She reported ongoing problems with photophobia (light sensitivity) for both sunlight
and fluorescent light. She was aware of ongoing eye fatigue. Occasionally she would
perceive motion in her peripheral visual field when there was nothing there. She com-
plained of visual distortion in the left lower quadrant of her visual field. She noted
that what she saw was rotated and appeared to pulsate.
Neurological evaluation was conducted in July 1999. The neurologist identified
the following symptoms which included: paresthesias, deficits in short-term memory,
diminished mental sharpness, hand tremors, visual distortions, problems with distract-
ibility, irritability, sleep disturbance, bladder incontinence, fatigue, balance problems,
motor weakness in her arms and hands, decreased fine motor dexterity, blurred vis-
ion, headaches, and problems with verbal fluency. A Magnetic Resonance Imaging
(MRI) and EEG were performed in September 1999. Both scans were determined to
be within normal limits.
Ms. SL underwent a comprehensive evaluation program at a nationally known
medical clinic in March 2000. The fact of her CO poisoning had not yet been
discovered when she was evaluated at the medical clinic. The physician and psy-
chologists at the clinic identified a wide variety of symptoms consistent with CO
poisoning as they conducted their individual evaluations. None of the clinic evaluators
8417: “8417_c027” — 2007/9/11 — 12:14 — page 635 — #17

suspected CO poisoning and they were at a loss in explaining her numerous and wide
variety of symptoms. On the basis of the fact that they could not identify a condition
to account for her symptoms, the ultimate conclusion was that the majority of her
problems related to emotional and psychological issues.
In December 2000, a neuropsychological evaluation in Colorado correctly dia-
gnosed her with chronic CO poisoning. The neuropsychologist recommended that a
cognitive therapist work with Ms. SL using a restorative and compensatory strategy.
He felt it would be helpful for the therapist to work with Ms. SL in developing
strategies to compensate for her residual cognitive deficits. The neuropsychologist
felt Ms. SL was experiencing emotional and psychological distress associated with
the exposure. He felt she could benefit from individual psychotherapy for her feel-
ings of depression, anxiety, and irritability. The neuropsychologist also felt Ms. SL
would benefit from the use of an antidepressant medication in the select serotonin
re-uptake inhibitor class (e.g., Celexa). He stated that the medications in this class
were extremely helpful for individuals who have sustained organic injuries. The med-
ications were recommended to help stabilize her moods, maximize her energy level,
and frustration tolerance. The neuropsychological evaluation found vision problems
consistent with VMSS owing to CO exposure. The neuropsychologist recommended
further evaluation and treatment by a neuro-optometrist.
In the course of the clinical interview with the neuropsychologist, Ms. SL reported
ongoing problems with phonophobia (noise sensitivity). He recommended the use of
noise attenuation earplugs. The neuropsychologist also recommended she consult
with her primary care physician to assist her sleep patterns. He mentioned several
options in this area like homeopathic sleep aids, antidepressant medications such as
Trazadone or Ambien to help regulate her sleep. The neuropsychologist discussed Ms.
SL’s long duration of exposure to CO. He suggested she consult with her physician
regarding the beneficial effects of hyperbaric oxygen (HBO) treatment or normobaric
oxygen therapy. The final recommendation of the neuropsychologist was to reduce
Ms. SL’s workload in order to attend to the needed therapies.
Neuropsychological re-evaluation (1 year) was conducted in December 2001. She
continued to experience ongoing problems with dizziness. Her dizziness and ongo-
ing visual disturbance continued to negatively affect her balance. She felt that her
fine motor speed and dexterity remained less efficient and she did not note any func-
tional improvements in these areas. She reported that her tactile sensitivity bilaterally
was somewhat better. She noted ongoing problems with noise sensitivity and was
unchanged from the time of the last neuropsychological examination. She continued
to have difficulties filtering out background noises, suggesting ongoing problems with
her auditory gating mechanism. She had seen some improvement in her stamina, but
she still was bothered by fatigue.
At the time of the original testing, Ms. SL reported multiple vision problems con-
sistent with VMSS. At the 1 year re-evaluation she reported no improvement in these
problems from the time of the original testing. She was evaluated by several ophthal-
mologists and one neuro-ophthalmologist at a large Midwestern university teaching
eye center. Neither the general ophthalomogist nor the neuro-ophthalmologist found
any problem nor could they offer any advice or recommendations. The neuro-
ophthalmologist stated, “I can discover no organic explanation for the patient’s
8417: “8417_c027” — 2007/9/11 — 12:14 — page 636 — #18

symptoms. I doubt that they are due to CO exposure.” As part of her visit to Colorado
for her 1 year neuropsychological evaluation she was scheduled and referred by the
neuropsychologist to undergo a thorough evaluation with a local neuro-optometrist,
Dr. Georgis.
With regards to her cognitive functioning, Ms. SL reported to Dr. Georgis some
improvement in her concentration abilities, but noted these remained highly incon-
sistent. She lost and misplaced things more frequently and the new learning process
was more difficult for her. She noted that she avoided reading whenever possible
because it was difficult for her.
At the time of neuropsychological retesting, Ms. SL noted ongoing problems
with spatial disorientation and carried a compass in her car to help with spatial ori-
entation. She acknowledged more feelings of anxiety and frustration regarding her
ongoing physical and cognitive defects, and the limitations that those imposed. She
was worried about her job status and future vocational potential. She still had some
difficulty sleeping. She felt the Celexa had helped her get to sleep; however, she
still had difficulty staying asleep. The Celexa had resolved her mood swings and her
frustration, and stress tolerance had improved notably. Since the time of the original
neuropsychological evaluation, Ms. SL had undergone six HBO treatments. She felt
the treatment helped her think more clearly and the tactile sensitivity in her hands had
improved. She felt her moods were more settled and she was less labile. Ms. SL had
met her treatment goals and had been discharged by her psychotherapist.
Ms. SL had reduced her teaching load since her prior neuropsychological eval-
uation. She noted she had to use 100% of her available energy to function in her
reduced academic and teaching load.
As part of the 1-year ongoing assessment of Ms. SL, the neuropsychologist gave a
questionnaire to Ms. SL’s mother and two sisters. They reported Ms. SL had difficulty
in balance, veering off center, judging distances, motor incoordination, penmanship
less legible since the exposure, dropping the chalk while writing on the chalkboard,
grabbing objects in her path while walking and shutting her eyes owing to dizziness.
They noted Ms. SL complains of double vision and she hugged the curb while driving.
They noted she squinted outside owing to photophobia and she had very low energy.
After 1 year between neuropsychological tests, Ms. SL’s performance remained
essentially the same on 24 of the neuropsychological test measures. She made notable
gains on 17 of the neuropsychological test measures when compared to her prior
testing according to the neuropsychologist. However, to some degree, some of the
gains could be attributed to practice effect. When this variable was taken into account,
Ms. SL’s current testing suggested that she was experiencing ongoing inefficiencies
and inconsistencies in her sustained attention and concentration abilities.
Dr. Georgis first saw Ms. SL in December 2001. She reported about her neuro-
ophthalmology examination at the midwest university eye clinic. She said the neuro-
ophthalmologist examined her for 5 min and diagnosed ovarian cancer. She noticed
balance problems beginning about 1992. She was running into the walls in her hallway
and would hit her drawers with her hip. Ms. SL had trouble with depth perception
and she was unable to hit or catch a ball. She noticed she would stumble while
walking and these symptoms increased slowly from 1988. Ms. SL was not comfortable
around crowds of people. She would stand off to the side when attending a family
8417: “8417_c027” — 2007/9/11 — 12:14 — page 637 — #19

gathering. The crowd noise was especially bothersome. She complained of having to
use excessive concentration to drive a car. Her brother and sisters reported she tended
to hug the yellow line. Her siblings could not carry on a conversation with her because
she needed her entire concentration to drive a car. Ms. SL also reported four near car
accidents because she was unable to recognize cars moving in her peripheral visual
field. She also noticed seeing flashes and shadows in her peripheral field and when
she looked to the side nothing was really there. Ms. SL noticed a smoky area in her
visual field and the area would pulsate with her heartbeat. She complained of missing
turns while driving and she misplaced her keys and they were visually in plain sight.
Ms. SL reported tinnitus, noise sensitivity and that her ears felt blocked. She felt like
she had trouble hearing and she had to turn the radio and TV up loud. Some vertigo
was noted while in the prone position and she became very dizzy when she moved
quickly. She also noted some carsickness when riding as a passenger in a moving car.
Ms. SL reported her typing speed and accuracy had deteriorated and she tended to
look at her hands while typing. Other hand–eye coordination problems were apparent
such as dropping things and knocking over objects while cooking. She complained of
severe photophobia especially while diving at night. Fluorescent lights bothered her
at work and she tended to keep her home dark. The computer screen glare was also
bothersome. Ms. SL noted reading was very difficult since her exposure. She used to
enjoy reading, but reading was simply a struggle. She had to reread and tended to use
her finger to follow the line of print. Ms. SL reported hip pain and her right heel had
been very painful.
A visual spatial evaluation found a moderate visual midline shift towards Ms.
SL’s right side. Some body sway was noticed during her motor evaluation. Her gait
was uneven with more pressure placed on her right foot. She walked on the temporal
edge of both feet and walked toward her left side. Ms. SL was put on midline with
a moderate degree of yoked prism. The prism equalized her gait, improved her foot
posture, straightened her walk and eliminated her body sway. The prism improved Ms.
SL’s ability to judge speed, movement, and position of cars moving within six lanes
of moving traffic. Reading evaluation found the prism eliminated the word movement
and word overlap. It also improved her ability to follow the line of print without using
her finger. Photophobia evaluation found a glare control tint that reduced the glare
from the computer screen. Yoked prisms were prescribed for distance and near use
with a photochromic bifocal. A near glare control tint was prescribed for near and
computer work in the classroom.
A phone consultation was completed in March 2002, 2 12 months after receiv-
ing the yoked prism. Ms. SL stated, “Wow and thank you! Dr. Georgis has helped
me the most, in the shortest period of time, than any other doctor.” She said she
was able to walk 2 miles without her foot hurting. She no longer hit walls, dresser
drawers, and did not stumble at all while walking. Ms. SL reported being more com-
fortable around crowds and her peripheral hallucinations had completely resolved.
Ms. SL noted her driving had improved significantly. She was able to stay within
the lines of the road without difficulty and she was able to see other cars within her
peripheral field without difficulty. She was able to talk to her sister while driving
the car and had not had any near accidents since receiving her prism. She repor-
ted her vestibular symptoms had improved a lot. Her noise sensitivity, ear blocking,
8417: “8417_c027” — 2007/9/11 — 12:14 — page 638 — #20

lightheaded feeling in the prone position, dizziness, and motion sickness had com-
pletely resolved. Ms. SL noted her typing speed had increased and she no longer
needed to look at her hands while typing. Her ability to grade papers on the com-
puter was comfortable for up to 6 h and the glare from the computer screen was
nonexistent. She reported her hand–eye coordination had improved and she did not
spill as much while cooking. Ms. SL noted a big improvement in her photophobia
especially while driving at night. She reported her reading had become pleasurable
again. She no longer had to use her finger while reading. However, her reading
volume had continued to be much less since her exposure. Ms. SL noticed her
hip pain had completely resolved and she periodically had some discomfort in her
right heel.
Ms. SL was seen in Dr. Georgis’ office in June 2002, 6 months after receiving
the yoked prism. She was wearing her yoked prism all of her waking hours and
reported, “Fantastic Improvement” from her December 2001 evaluation. However,
she had noticed some pain in her right foot in the last month. Her balance was good
and she reported no trouble with crowd noise. She noted her tinnitus, dizziness, and
lightheaded feeling in the prone position had completely resolved. Her hand–eye
coordination was no longer a problem and the photophobia was totally controlled by
the glare control prism. She noticed some problem with peripheral vision in the last
month. Ms. SL said her reading was better but she was only reading when it was
necessary.
Another visual spatial evaluation found Ms. SL’s visual midline shift had com-
pletely resolved. Her gait was even and her walk was straight. Her balance was very
good and there was no evidence of body sway. Her foot posture was normal. The
yoked prism treatment was discontinued and Ms. SL was instructed to return if any
of the prior symptoms reoccurred.
About 8 months later in February 2003, Ms. SL contacted Dr. Georgis by tele-
phone. She complained of falling three times when walking on uneven surfaces.
She noticed crowd movement was bothersome and her peripheral hallucinations had
returned. She noticed some difficulty judging position and speed of other cars while
driving. She also reported some spatial confusion while driving and was missing turns.
Ms. SL lost her glare control lenses and her photophobia had returned. She noticed
her hand–eye coordination had regressed somewhat. She reported some dizziness,
ear blocking, and trouble hearing around background noise. She was instructed to see
Dr. Georgis as soon as possible and to see a neuro-otologist in Colorado.
Ms. SL was seen by Dr. Georgis 5 months later in July 2003. She complained of
the same symptoms that were reported in the February 2003 telephone consultation.
She was unable to afford the consultation with the Colorado based neuro-otologist.
Visual spatial evaluation found a moderate visual midline shift towards Ms. SL’s
right side. The midline shift was a 20% increase over her February 2001 evaluation.
The midline shift was corrected with yoked prism and Ms. SL was instructed to be
re-evaluated in 6 months.
Ms. SL was last seen in January 2004. She was wearing her yoked prism all
of her waking hours. She reported good balance, good hand–eye coordination and
no dizziness. She noted her photophobia was totally controlled by the glare control
prism. Visual spatial evaluation found a 70% improvement in her visual midline shift.
8417: “8417_c027” — 2007/9/11 — 12:14 — page 639 — #21

Her balance with the decreased prism was very good. Her foot posture was normal
and her walk was straight. Dr. Georgis reduced her yoked prism by 70%.
Ms. SL was seen in January 2004 by her Colorado-based neuropsychologist for
a re-evaluation. She was currently using an antidepressant medication and it was
working well to control her depression. She planned to wean herself off the antide-
pressant in the near future. Ms. SL continued to experience a variety of cognitive
problems, which negatively impacted her day-to-day and work functioning. She had
implemented multiple strategies to compensate for her residual deficits. She had
been unable to identify a cognitive therapist in her area to provide direct cognitive
rehabilitation. Ms. SL reported she continued to work a full load at the community
college. The work required 100% of her available energy. Ms. SL was instruc-
ted to return to Dr. Georgis care on an as needed basis and she was last seen in
November 2004.
27.7 CONCLUSION
CO poisoning is a common health problem worldwide. Neurological damage to brain
structures is a common occurrence. Approximately 30% of the neurological symptoms
of CO poisoning are directly related to spatial awareness, balance, and movement.
VMSS is a result of CO-induced injury to subcortical structures in the midbrain.
These structures affect the injured person’s spatial awareness, balance, and movement.
Yoked prism lenses modulate these subcortical structures through the highly sensory
visual systems. The yoked prism treatment is effective immediately. The balance,
movement, and spatial symptoms are eliminated. The prescription or power of the
yoked prism lenses can generally be reduced gradually, dependent on the adaptability
of the patient’s brain and the severity of the exposure.
References
1. World Health Organization. Carbon Monoxide Environmental Health Criteria 13.
Published under the joint sponsorship of the United Nations Environment Program
and WHO, Geneva, 1979.
2. Gazzaniga, M.S., Ivry, R.B., Mangun, G.R. Cognitive Neuroscience: The Biology of
The Mind. WW Norton and Company, NY. 2nd ed., 2002.
3. Schneider, G.E. Two visual systems. Science, 1969, 163, 895–902.
4. Posner, M.I., Raichle, M.E. Images of the Mind. Sci. Am. Library, NY, 1994.
D.G. Penney, ed., CRC Press, 2000, pp. 393–418.
6. Gianutsos, R. Visual rehabilitation following acquired brain injury. In: Functional
Visual Behavior: A Therapist’s Guide to Evaluation and Treatment Options, Michele
Gentile, Editor, The American Occupational Therapy Association, Inc., Bethesda,
MD, Chapt. 8, pg. 267, 1997.
7. Cohen, A.H., Rein, L.D. The effect of head trauma on the visual system: The doctor
of optometry as a member of the rehabilitation team. J. Am. Optom. Assoc., 1992, 63,
530–536.
8417: “8417_c027” — 2007/9/11 — 12:14 — page 640 — #22

8. Padula, W.V., Shapiro, J.B., Jasin, P. Head injury causing post trauma vision
syndrome. N. Engl. J. Optometry, Dec/Winter, 1998, 16–21.
9. Padula, W.V., Argyris, S. Post-trauma vision syndrome and visual midline shift
syndrome. J. Neurol.Rehabil. , 6, 1996, 165–171.
10. Padula, W.V. Neuro-Optometric Rehabilitation, Publ. by Optometric Extension Pro-
gram Foundation, Inc., Santa Ana, CA, 1st ed., 1998. Chapt. 14, p. 194, Visual
midline shift syndrome.
11. Connor, M., Padula, W. Visual rehabilitation of the neurologically involved person.
In Functional Visual Behavior: A Therapist’s Guide to Evaluation and Treatment
Options, Michele Gentile, Editor, The American Occupational Therapy Association,
Inc., Bethesda, MD, Chapt. 9, pg. 295, 1997.
12. Trevarthen, C.B., Sperry, R.W. Perceptual unity of the ambient visual field in human
commissurotomy patients. Brain, 1973, 96: 547–570.
8417: “8417_c027” — 2007/9/11 — 12:14 — page 641 — #23

8417: “8417_c027” — 2007/9/11 — 12:14 — page 642 — #24
28 Firefighters and Carbon
Monoxide
Kevin J. Reilly, Jr., Frank Ricci, and David Cone
CONTENTS
28.1 Firefighter Fatalities Resulting From Suppression . . . . . . . . . . . . . . . . . . . . . . . . 643

28.2 Postfire and Overhaul . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644
28.3 Workplace/Industry Carbon Monoxide Standards . . . . . . . . . . . . . . . . . . . . . . . . 645
28.4 Investigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 646
28.5 Emergency Smoke Escape Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650
28.6 Emergency Department Screening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650
28.7 Lethal Gas Combinations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651
28.8 Prevention and Awareness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 652
28.9 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 653
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 653
Carbon monoxide (CO) is produced from the incomplete combustion of carbon-based

fuels. It is found at every structure fire and is present most often when an appliance
that uses combustion malfunctions. Firefighting as an occupation carries an extremely
high risk of exposure to CO, owing to fire suppression, postfire overhaul, and response
to routine CO alarm calls. Few other occupations involve the potential of such frequent
exposure to CO. The fire department often provides the first emergency medical care
to victims who have been exposed to CO, and is the first link in the emergency medical
service (EMS) system. This chapter explores how CO impacts this profession and the
civilians who are protected by it. We will bring to light some of the latest technology
available for home and professional use, and will explore the gap between the medical
community and the fire service. The first author of this chapter has written a recent
review of this topic.1
28.1 FIREFIGHTER FATALITIES RESULTING FROM

SUPPRESSION
Smoke can have over 2000 toxic chemicals in it, yet after a review of 105 autopsies,
the results usually indicate CO, hydrogen chloride, acrolein, soot, and hydrogen
cyanide (HCN) as the leading contributing causes of death.2 In the 1990s, 63% of all
nonheart attack deaths inside structure fires were the result of smoke inhalation.3 CO
643
8417: “8417_c028” — 2007/9/11 — 12:13 — page 643 — #1

is considered one of the greatest threats in the fire service and is referred to as the
“silent killer” in smoke.4 Even many professionals are surprised to learn that burns
and crush injuries are overshadowed in number by smoke inhalation. In a National
Fire Protection Association (NFPA) report that examined fire fatalities from 1978 to
1999, smoke inhalation was the leading cause of death for nonheart attacks each year.
Physicians should be aware that a firefighter who has CO poisoning will likely have
other toxic gases in his blood stream. Emergency department (ED) physicians should
test for as many products of combustion as technology allows in smoke inhalation
victims.
Heart attacks account for over 44% of firefighter line-of-duty deaths. There are
many contributing factors to these heart attacks, such as a pre-existing heart condition;
however, the cumulative effect of chronic exposure to CO can act as a catalyst in
initiating a heart attack. Oxygen is essential for the aerobic metabolism of the heart,
and when oxygen transport is impeded by CO in the blood, the heart has to work
harder to keep up, increasing the potential for a heart attack. Additional factors for
firefighters are exhaustion from physically demanding work, and the added heat stress
from their protective clothing. These deaths often occur at the fire station or when the
firefighter returns home. While many firefighter deaths are due to the inherent nature
of the profession, many can be avoided. Education about the hazards of CO and other
products of combustion is key to helping reduce injuries and fatalities. In 2005 alone,
smoke inhalation and respiratory distress accounted for 3390 injuries in the US fire
service.5
Physicians do not routinely measure carboxyhemoglobin (COHb) in heart attack
victims. In the case of firefighters, it is important that blood is analyzed for CO because
of the toxic hazards associated with smoke exposure. When the blood sample of a
heart attack victim is drawn post-CPR, the COHb value will be lower because the
amount of 100% oxygen used on the patient during cardiopulmonary resuscitation
(CPR). This should be considered when calculating the true peak level of COHb at
the time the patient left the environment containing elevated CO concentration.
28.2 POSTFIRE AND OVERHAUL

Overhaul is the process of opening up walls and ceilings of a structure to check for and
extinguish hidden fire, in order to prevent rekindles. This stage of fire operations, when
the building and its contents are still smoldering, often produces high levels of CO.2
As with any hazardous environment, in an area in which higher then allowable CO
levels are suspected, it should be standard practice during overhaul to have a multigas
meter in place to monitor the atmosphere. Being a by-product of combustion, CO is
seldom found by itself, and multigas meters can be used to gauge the presence of
other gases. For example, when a firefighter experiences watery eyes at a structure
fire, acrolein is the likely cause. If the environment is ventilated, then metered, and
found to be clear of CO, the firefighters should not experience watery eyes when they
remove their SCBA (self-contained breathing apparatus) since the acrolein has been
ventilated with the CO.
Some fire department gas sensors only read CO. Firefighters must understand that
if CO is found to be present, it is a likely indication of the presence of other toxic
8417: “8417_c028” — 2007/9/11 — 12:13 — page 644 — #2

Firefighters and Carbon Monoxide 645
gases. Firefighters should wear a personal gas monitor (PGM), and continue to use
breathing apparatus until CO readings in the area fall to within acceptable levels.
Firefighters who smoke, fall into their own high-risk category. Pack-a-day
smokers may have up to 5% COHb. These firefighters are more at risk for at least
two reasons: (1) they are at a higher baseline COHb%, and (2) they are more likely
to be asymptomatic from a CO exposure.
28.3 WORKPLACE/INDUSTRY CARBON MONOXIDE

STANDARDS
The standards for CO concentration are expressed in parts per million (ppm). There
is a direct relationship between ppm and percent (1% = 10,000 ppm). See Chapter 35.
Although the following standards are designed for industry, they play an important
role in risk assessment for fire departments when CO has been stored or transported
and is involved in a motor vehicle or industrial accident. The UN# for CO is 1016. It
is usually transported as a cryogenic liquid.
The current Occupational Safety and Health Administration (OSHA) permissible
exposure limit (PEL) for CO is 50 ppm as an 8-h time-weighted average (TWA)
[29 CFR 1910.1000∗ ]. The National Institute for Occupational Safety and Health
(NIOSH) recommended exposure limit (REL) for CO is 35 ppm as an 8-h TWA
and a ceiling limit (CL) of 200 ppm. The NIOSH recommended immediately dan-
gerous to life and health concentration (IDLH) for CO is 1200 ppm. The IDLH
is the concentration that results in death or irreversible health effects, or prevents
escape from the contaminated environment within 30 min. The American Confer-
ence of Governmental and Industrial Hygienists (ACGIH) has adopted a threshold
limit value (TLV) for CO of 25 ppm, as an 8-h TWA. For more information, see
http://www.cdc.gov/Niosh/carbon2.html.
Underwriters Laboratory standards for residential CO alarms are given in
Table 28.1. CO alarms had to have a reset button. If the alarm sounded and the
inhabitants of the building showed no sign of CO poisoning (nausea, vomiting, gen-
eral body weakness, etc.), they were instructed to press the reset. If the alarm went off
again, they were instructed to vacate the building and call an authority. The selectivity
test criteria remained at previous levels.
After October 1, 1998, UL 2034 listed CO alarms must measure and alarm when
CO is:
• 30 ppm for 30 days

• 70 ppm for no more than 240 min before alarming (may alarm as early as
60 min)
• 150 ppm for no more than 50 min before alarming (may alarm as early as
10 min)
∗ CFR 1910.1000 provides measurement standards for air contaminates and definitions found on an MSDS
(Material Safety Data Sheet). The standards are law and fall under the jurisdiction of the US Department
of Labor.
8417: “8417_c028” — 2007/9/11 — 12:13 — page 645 — #3

TABLE 28.1
UL 2034. Carbon Monoxide Alarm Activation Standards 1992 and 1995.
April 30, 1992, UL 2034 listed CO alarms had to measure and alarm when CO is
• 15 ppm for 8 h before alarming, or
• 100 ppm for no more than 90 min before alarming, or
• 400 ppm for no more than 15 min before alarming
After October 1995, UL 2034 listed CO alarms had to measure and alarm when CO is
• 15 ppm for no less than 30 days, or
• 400 ppm for no more than 15 min before alarming
• 400 ppm for no more than 15 min before alarming (may alarm as early
as 4 min) and have a manual reset that will re-energize the alarm signal
within 6 min if the CO concentration remains at 70 ppm or greater.
Another significant change to the October 1, 1998 CO alarm listing was the addition to
the instructions, stating that individuals with medical problems may consider using
warning devices which provide audible and visual signals for CO concentrations
under 30 ppm. UL 2034 information obtained from www.bacharach-training.com.
CO-IDLH and PELs are based on the responses of healthy adults at resting vent-
ilation. Elderly persons and infants are more susceptible to CO exposure. Specialized
CO detectors are available and should be used in conjunction with these groups.
28.4 INVESTIGATION
Structure fires are not the only situations where firefighters encounter CO, since the
cause and presence of CO at a fire is known. Investigating situations where CO
alarms are activated may be more hazardous, since locating the source of the CO can
be difficult and time consuming (see Table 28.2). Interviewing building occupants
about possible flu-like symptoms, asking what appliances were in use, and asking
about any unusual odors may be helpful. CO is odorless, colorless, and tasteless, yet
in the realm of CO alarm investigation, CO is not the only fire gas that is produced,
and other gases can produce odors. Furnace malfunctions often produce aldehyde
odors that can be compared to rotting fruit. An occupant who experiences watery
eyes and notices a foul odor should exit the home and notify the fire department.
Though the smell in the area and watery eyes may not be directly related to CO, they
may be the result of lacrimators (an uncontrollable response to an irritant that causes
release of tears) and may suggest that CO is also present and needs to be further
investigated.
8417: “8417_c028” — 2007/9/11 — 12:13 — page 646 — #4

TABLE 28.2
Common Investigative Tactics Associated with Finding the Source of a Carbon
Monoxide Problem
• Turn on all fuel-burning appliances
• Close all windows and doors to seal home as tightly as possible
• Set thermostat to activate heating unit
• Run hot water until the water heater turns on
• Start the gas clothes dryer (this device not only creates CO, it also vents air outside the home,
which may create negative pressure inside and induce reverse stacking)
• Turn on house fans. This includes the kitchen and bathroom fan, as well as the electric clothes
dryer and attic fan, if present. These may also contribute to negative pressure and reverse
stacking
TABLE 28.3
Common Sources for the Generation of Carbon
Monoxide in a Residence
• Faulty furnace
• Faulty water heater
• Gas stove
• Gas dryer
• Fireplace
• Attached garage
• Driveway close to an open window or air conditioning unit
• Heavy smoker (approximately 400 ppm in exhaled breath)
Locating the CO problem can be more difficult when a temporary source of CO is

the cause. This might involve a car or lawn mower running near an open window or a
heating ventilation and air conditioning (HVAC) intake unit (see Table 28.3). Water
heaters or furnaces can be tricky too. It may be useful to turn up the thermostat to
activate the furnace, or run hot water to activate the water heater. Meters are then used
to check that the draft and flue pipes of these appliances are working properly. During
any CO investigation, it is important to interview the occupants about what they were
doing prior to the CO alarm activation. CO concentrations outside of buildings in
urban areas might raise CO levels owing to heavy automotive traffic.
Portable detection devices used by fire departments and utility companies have
their limitations. Many meters may take as long as 5–7 min to achieve an accurate
reading. It is important to be familiar with the operating instructions. Proper calib-
ration is critical. The functions and limitations of gas meters vary between different
manufacturers. A little known limitation can be within the sensors themselves. All CO
sensors begin to degrade beyond the calibration range, typically after only 2 years.
As with any tool, it is important to read the manufacturer’s directions and know the
8417: “8417_c028” — 2007/9/11 — 12:13 — page 647 — #5

limitations of the instrument’s use. All meters must be zeroed (cleared) away from
the monitoring site (i.e., before entering the area to be sampled) in order to establish
a baseline.
Technology provides us with additional devices to aid in a CO investigation. One
such device is the data logger. Data loggers are effective when a CO investigation
appears inconclusive. They can be left at the scene of a CO alarm to determine the
CO level and time of that level. This aids in zeroing in on the source of the alarm
activation. Data loggers today are compact and can store time stamped readings for
prolonged periods of time. When retrieved, data loggers are uploaded to computer
software that can determine the time of day an exposure occurs. For example, it may
be used to explain why warming your car up in the garage in the morning will set the
CO alarm off hours later, owing to the time it took the CO to migrate to the household
CO detector.
Traditionally, determining the level of CO exposure in people was limited to
invasive blood testing in a hospital. In the field, first responders relied on symptom
guesswork using complaints such as headache, fatigue/weakness, dizziness, and flu-
like symptoms to estimate CO exposure. Today, breath analyzers can be used in the
field to estimate CO levels in the blood. CO breath analyzers measure the amount
of CO in exhaled breath in ppm, suggesting a level of CO toxicity. Breath analyz-
ers for CO have been available for many years, and were used earlier primarily for
smoking intervention. Recently, EMS units have adopted CO breath analyzers in the
field for their portability and ease of use. Many physicians continue to draw arter-
ial blood samples to determine COHb levels, when venous blood or breath samples
will do. Firefighters associate all invasive methods with pain and react with avoidance.
Venous blood CO testing is just as effective as arterial for CO measurement. Fire-
fighters would be more willing to be tested if venous blood instead of arterial blood
is drawn.
CO has a slightly lower molecular weight than air. The molecular weight of CO
is approximately 28, while air has an average molecular weight of approximately 29.
Reference materials may simplify this by assigning air an arbitrary number of one.
Any substance with a value of less than one will float, and anything that has a value
greater than one will sink. On this basis, CO has a value of 0.97 and this indicates
that CO is slightly lighter than air. However, once CO is mixed with air it cannot
spontaneously separate from it. The key physical characteristic here is that hot air
rises and cold air falls. Accordingly, CO in air should be metered at three different
levels. Unlike smoke detectors which are placed on ceilings, CO detectors should be
placed on walls at head or chest level. CO may be found in closed spaces such as in
closets and attics, necessitating removal by mechanical ventilation.
Socioeconomic factors in urban areas play a part in the rise of CO emergencies.
As heating costs go up, occupants have been found to attempt to heat their homes
by running the gas stove, using kerosene heaters (illegal in most states), or bringing
an outdoor grill inside. Besides the hazards of fire, accumulation of dangerous levels
of CO is likely in all of these scenarios. In addition, absentee landlords may allow
appliances to fall into disrepair with little regard for tenant safety. Illegally constructed
apartments in basements can sometimes cause furnaces to be obstructed, preventing
proper airflow (see Figure 28.1).
8417: “8417_c028” — 2007/9/11 — 12:13 — page 648 — #6

CITY OF NEW HAVEN, CONNECTICUT

DEPARTMENT OF FIRE SERVICE
CARBON MONOXIDE INCIDENT CHECKLIST
Alarm Number: _______________Date: _______________ Dispatch Time: _____________

Company: ____________________ Officer: _________________________________________
* DID ANY PERSON OCCUPYING THE OCCUPANNCY DISPLAY AND SIGNS AND OR SYMPTOMS ASSOCIATED
WITH CARBON MONOXIDE POISONING SUCH AS:
YES NO
Headache .
Nausea .
Dizziness .
Difficulty Breathing . .
Altered Mental Status .
* DID ANY PERSON DISPLAYING SIGNS AND OR SYMPTOMS FEEL BETTER WHEN HE EXITED THE
OCCUPANCY?
YES NO
* DID ANY PERSON EXHIBIT THESE SIGNS AND OR SYMPTOMS BEFORE EXPOSURE TO THIS ENVORONMENT?
YES NO .
* ARE THERE ANY OCCUPANTS PRESENT THAT CAN BE CONSIDERED AS “AT RISK” PERSONS, SUCH AS:
YES NO
Individuals who are pregnant? .
Children? .
Persons with heart of respiratory disease? .
Does anyone smoke in the occupancy? .
Is there any fuel burning appliance present? If yes, List each
_________________________________________________________________________________________________________ .
Was a fireplace and furnace in use at the same time? .

Is there any fireplaces, unvented heaters or coal/wood burning stoves in use at the time, if yes please explain.
______________________________________________________________________________________
Was an oven being used? .

Was an auto running in an attached garage? .
* THE FOLLOWING QUESTIONS REFER TO THE ALARM ACTIVATION:
What time did the CO Detector first sound? __________ am / pm

Is the detectors’ sensor discolored? YES NO
What was the highest digital alarm reading displayed? ______________________
Was the occupancy ventilated after the alarm activated YES NO

Were all appliances shut off by the occupant or NHFD? YES NO
Was a gas grill type unit used near the building? YES NO
* LIST ALL READINGS DETECTED BY THE NHFD

Initial entry: __________ Following ventilation: __________
Ventilation ceased: __________ Followup reading: __________
st nd
Basement: __________ 1 Floor: __________ 2 Floor: __________
rd th th
3 Floor: __________ 4 Floor: __________ 5 Floor: __________
MEASUREMENTS OBTAINED PRIOR TO DEPARTING SCENE:
__________________________
FIGURE 28.1 Sample form used by fire departments during carbon monoxide-related
investigations.
Fire departments must not overlook basements and attics during their investiga-
tions, and have a fiduciary responsibility to summon the appropriate building officials
when improper or illegal conditions are found. Some individuals bring themselves
to the ED with CO poisoning, having bypassed the fire/EMS system. In these cases,
the ED must ensure that this individual is not going home to a toxic environment.
8417: “8417_c028” — 2007/9/11 — 12:13 — page 649 — #7

In urban areas, proper reporting may save the lives of others who may be housed in
the same building. All ED’s should have a mechanism to ensure that the appropriate
agencies are notified.
28.5 EMERGENCY SMOKE ESCAPE DEVICES

There are no current testable North American standards in place for the use of emer-
gency smoke escape devices commonly known as “smoke hoods”. However, the
American National Standards Institute (ANSI) has a written standard ANSI-110 that
is not yet in practice. It is important to know that the proposed standard is very similar
to the existing European EN403 standard that has been in place since 1993. Although
it is a European standard, the EN403 gas challenges are virtually the same as the pro-
posed ANSI-110 standard. Until there is a recognized standard in the United States,
the EN403 standard should be used as a reference when considering the purchase of
a smoke hood, since some devices are of questionable value and may not offer the
protection that they advertise.
Several recognized International Standards specify the minimum requirements
for single use emergency devices intended for use by persons escaping from fires.
These include the European EN403: 2004—Respiratory protection devices for
self-rescue/Filtering devices with hood for escape from fire. ANSI/ISEA 110:2003—
the American National Standard for Air-Purifying Respiratory Protective Smoke
Escape Devices, and the Australian/New Zealand Standard AS/NZS 1716: 2003—
Respiratory Protective Devices.
These standards, whether government mandated or voluntary, are vital to the
safety of the end user and the integrity of the smoke hood industry. Standards rep-
resent a documented agreement, established by a consensus of subject matter experts
and approved by a recognized body that provides rules, guidelines, or character-
istics to ensure that materials, products, processes, and services are fit for their
purpose. Adherence to these standards ensures governance by establishing processes
and practices that promote and ensure integrity, compliance, and accountability. These
performance standards serve as the benchmark for all new technologies and meth-
ods, defining the expected level of performance for all emergency smoke escape
devices.
Tables 28.4A and 28.4B summarize the Test Challenge and Breakthrough
concentrations for determining smoke hood product performance. The breakthrough
time shall not be less than 15 min when tested against the agents referenced in the
tables.
28.6 EMERGENCY DEPARTMENT SCREENING

Screening at the emergency department is essential in reflecting the true statistics
associated with CO poisoning. It is common that an exposure goes undetected or
misdiagnosed. As mentioned earlier in this chapter, there are alternatives for accurate
blood screening, and venous blood sampling is just as effective for CO measurement
as are ABG (arterial blood gas) measurements. Using a CO breath analyzer is an
8417: “8417_c028” — 2007/9/11 — 12:13 — page 650 — #8

TABLE 28.4A
Summary of “Test Challenge” of Smoke Hoods
Test Challenge/Dose Concentrations (ppm)
Gas EN 403:2004 ANSI 110:2003 (1)* AS/NZS 1716:2003

Carbon monoxide 2500 minimum 3000 minimum 2500
Hydrogen cyanide (2)* 400 400 400
Hydrogen chloride 1000 1000 1000
Acrolein 100 100 100
Notes: (1)* = Also includes requirements for cyclohexane and sulfur dioxide; (2)* =
International standards include cyanogen gas in the performance criteria for HCN.
TABLE 28.4B
Summary of “Breakthrough Concentrations” When Testing
Smoke Hoods
Breakthrough (ppm)
Gas EN 403:2004 ANSI 110:2003 AS/NZS 1716:2003

Carbon monoxide 200 (1)∗ 200 250
Hydrogen cyanide (2)∗ 10 10 10
Hydrogen chloride 5 5 5
Acrolein 0.5 0.5 0.5
Notes: (1)∗ = Time weighted average in any single 5-min period; (2)∗ = Total
concentration of HCN and C2 N2 shall not exceed 10 ppm at breakthrough.
excellent method for accurate, noninvasive COHb screening. A blood test can of
course still be utilized to confirm a breath sample and screen for other toxic gases
associated with smoke inhalation. One study found that 23.6% of patients presenting
at a hospital with flu-like symptoms actually suffered from lower-level CO poisoning
(see Reference 6, p. 178).
Consequently, ED personnel, when interviewing patients with flu-like symptoms
or other indicators of CO poisoning, should inquire about the condition of other
household members as well, that is, take proper situational histories. Too often other
household members whose symptoms are not as severe go unnoticed. This could lead
to health problems as a result of low level chronic CO exposure.
28.7 LETHAL GAS COMBINATIONS

Firefighters should be aware of the potential dangers associated with combinations
of toxic gases. A CO reading of 200 ppm, although dangerous, would not by itself be
8417: “8417_c028” — 2007/9/11 — 12:13 — page 651 — #9

considered immediately life threatening, but when combined with minimal amounts
of HCN gas, could form a toxic environment that would be immediately hazardous.
This may be found in postfire overhaul operations, after the main body of fire has
been extinguished.
It is well established that cyanide, primarily in the form of HCN, is an important
and common component of fire smoke. While not as ubiquitous as CO, which can
be found in every fire, cyanide “is to be expected in modern fires,”7 owing to the
widespread use of synthetic materials that generate cyanide upon combustion.
While animal studies have demonstrated that cyanide produced in fire can quickly
be fatal, its true pathophysiologic role in humans in actual fires has been more difficult
to study and quantify. Blood cyanide levels may be low in hot flash fires in which case
heat and oxygen depletion are likely the primary causes of death, but may be high
when the victim is away from the site of fire origin and is not subjected to intense
heat and low oxygen levels.7 The evaluation of a number of fire fatality forensic
databases revealed median cyanide levels ranging from 0.53 mg/L (database of fire
deaths in Glasgow), to 1 mg/L (Dupont Plaza Hotel fire, Puerto Rico, 1986), to 3.14
mg/L (Paris database).7 In most of the databases examined, there was no correlation
between cyanide and COHb levels, though other studies have found correlations,
including an Australian study that found a mean COHb level of 40% and a mean
cyanide level of 1.65 mg/L among 178 fire victims. The correlation coefficient was
0.34 (p < .001).8
From a physiologic perspective, it may be tempting to postulate that the com-
bination of CO and cyanide exposure is more toxic than the simple sum of the two
effects, but the data do not support this conclusion. It was found that subjecting rats
simultaneously to the CO and HCN gas concentrations that individually produced
incapacitation at 5 and 35 min (5706 ppm and 1902 ppm for CO, respectively, and
184 ppm and 64 ppm for HCN, respectively), resulted in incapacitation in only 2.6 and
11.1 min.9 It has been suggested that cyanide, by depressing respirations, may reduce
or prevent the uptake of CO. This hypothesis is supported by observational data from
fire databases, wherein certain fire victims had COHb levels well below lethal, yet
very high blood cyanide levels.7 For example, data from a polyurethane mattress fire
that killed 35 prison inmates in Argentina in 1990 showed COHb levels between 4%
and 18%, but cyanide HCN levels ranging between 2.0 and 7.2 mg/L,10 well above
levels felt to be lethal in humans. For obvious reasons, controlled experimental human
data are lacking.
28.8 PREVENTION AND AWARENESS

Chronically CO-exposed individuals can become asymptomatic over time. An
example of this condition was found in a firefighter who smoked cigars. When
a breath sample was taken from him it was discovered that he had a reading
of 72 ppm (approximately 11.5% COHb), but he had no discernable symptoms.
When this was brought to his attention he went for a medical check up and it was
determined that he was in the early stage of heart disease. As a result he stopped
smoking.
8417: “8417_c028” — 2007/9/11 — 12:13 — page 652 — #10

It is important for firefighters to realize that firefighting as a profession has the

highest rate of heart attack death of any occupation in the US. Regular medical
examinations should be a top priority in the fire service. Being repeatedly exposed
to hazardous environments increases the risk of numerous health problems. There
is a need for a link between the medical community and the fire service to better
understand the effects of CO. More studies need to be conducted in tandem with the
fire service considering the frequency of exposure in this occupation.
28.9 CONCLUSIONS
There are many variables to consider when it comes to understanding the effects of
CO exposure. Acute CO exposure may not even register in the hemoglobin, which is
why it is important for firefighters to monitor the environment in which they work.
• Time is a critical factor to consider when conducting a CO investigation.

• Gas sensors can have varied reaction times, some as long as 7 min or more
to indicate a true ppm reading.
• Air movement can create false positive readings in areas where too many
personnel use a CO monitor. Use by only a limited number of personnel
when investigating a CO alarm is important.
CO is an elusive hazard that often goes undetected, both environmentally and

physiologically. The development of affordable technology such as personal CO
monitors and CO data-logging equipment has enabled more success in CO investiga-
tions. CO, also known as “The Great Imitator,” has many different faces. Symptoms
emulating a flu-like condition, headache, or fatigue can make diagnosis of exposure
tricky to recognize. However, with innovative technology such as a breath analyzer,
there is now a noninvasive means to accurately determine CO levels in the blood. If
there is COHb in the initial screening, a second test is necessary, since the COHb
saturation can be ascending or descending.
Given the vast number of variables associated with CO poisoning, it is important
to recognize that true statistics of this common occurrence are attainable. Members
of the medical community have a valuable resource in members in the fire service,
whose exposure to CO is commonplace. Working together through further research
and studies can produce positive results, which will benefit patients and save lives.
References
1. Reilly, K. Chronic CO poisoning in firefighters. Fire Engineering, 159, June, 2006;
110–114.
2. Jarbo, T. An examination of toxicity hazards associated with the burning of materials
commonly encountered by firefighters, February, 1995. Report to the National Fire
Academy Executive Fire Service Officer Program.
3. Fahy, R.F. NFPA. U.S. fire service fatalities in structure fires, 1977–2000, July, 2002.
National Fire Protection Association, www.nfpa.org.
8417: “8417_c028” — 2007/9/11 — 12:13 — page 653 — #11

4. Alcora, R. Smoke inhalation and acute cyanide poisoning. JEMS, Supplement,

Summer, 2004.
5. NFPA. National fire protection association research and reports on fire ground injuries,
1981 to 2005.
6. Montagna, F.C. Responding to Routine Emergencies, Fire Engineering, Penn Well
Publishing Co., 1999, Tulsa, OK.
7. Alarie, Y. Toxicity of fire smoke. Crit. Rev. Toxicol., 2002; 32: 259–289.
8. Yeoh, M.J., Braitberg, G. Carbon monoxide and cyanide poisoning in fire related
deaths in Victoria, Australia. J. Toxicol. Clin. Toxicol., 2004; 42: 855–863.
9. Chaturvedi, A.K., Sanders, D.C., Endecott, B.R., Ritter, R.M. Exposures to carbon
monoxide, hydrogen cyanide and their mixtures: interrelationship between gas expos-
ure concentration, time to incapacitation, carboxyhemoglobin and blood cyanide in
rats. J. Appl. Toxicol., 1995; 15: 357–363.
10. Ferrari, L.A., Arado, M.G., Giannuzzi, L., Mastrantonio, G., Guatelli, M.A. Hydro-
gen cyanide and carbon monoxide in blood of convicted dead in a polyurethane
combustion: a proposition for the data analysis. Forensic. Sci. Int., 2001; 121:
140–143.
8417: “8417_c028” — 2007/9/11 — 12:13 — page 654 — #12

29 The Purpose and the
Process of Litigation in a
Carbon Monoxide
Poisoning Case
Stephen P. Willison
CONTENTS
29.1 The Purpose—Why Litigate? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656

29.1.1 Why Litigation Works and How It Gets Abused . . . . . . . . . . . . . . . . . . 656
29.1.2 Keeping Focused on The Goal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 657
29.1.2.1 “Compensation” Does Not Mean “Profit” . . . . . . . . . . . . . . 657
29.1.2.2 The Secondary Goals Should be Just That—Secondary 657
29.1.3 The End Game and the Big Picture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658
29.2 The Process—What to Expect and When to Expect it . . . . . . . . . . . . . . . . . . . . 659
29.2.1 Gathering Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
29.2.1.1 Filing Suit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 660
29.2.1.2 The Discovery Process. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 661
29.2.1.3 Trial—Risks and Rewards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 662
29.3 They Call it the Burden of Proof for a Reason . . . . . . . . . . . . . . . . . . . . . . . . . . . . 662
29.3.1 The Standard of Proof . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 663
29.3.2 Two Essential Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 663
29.3.2.1 Liability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 663
29.3.2.2 Damages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664
29.4 The Unique Problems of Carbon Monoxide Poisoning . . . . . . . . . . . . . . . . . . . 666
29.5 Your Life on Trial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 667
29.6 An Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 668
29.6.1 Getting Help Managing the File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 668
29.6.2 Trusting Everyone to Do the Job Assigned . . . . . . . . . . . . . . . . . . . . . . . . 668
29.6.2.1 The Attorney . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669
29.6.2.2 Fact Witnesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669
29.6.2.3 Expert Witnesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669
29.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 670
655
8417: “8417_c029” — 2007/9/11 — 12:13 — page 655 — #1

If you are reading this chapter because you, or someone you care about, has recently
suffered the effects of carbon monoxide (CO) poisoning, you are probably caught in a
whirlwind of frustration, confusion, and anger. This chapter was invited by the editor
to help you calmly and rationally sort out the pros and cons of pursuing litigation,
and if you do pursue it, allow you to make some sense of what to expect. However,
because every case is fact specific, this chapter should never be used to substitute for
the advice of competent legal counsel. Instead, it will lay out for you the purpose and
process of litigation so you can better understand the risks.
29.1 THE PURPOSE—WHY LITIGATE?

Anyone considering litigation needs to honestly ask themselves, “why.” Your motives
now will greatly influence whether you are satisfied with your decision at the end of
the road. If you think it is easy, or you think it will solve your problems, let me save
you some reading—a CO poisoning case may be the hardest thing you ever do, and
you will still have all the same injuries when it is over. On the other hand, it can be a
worthwhile experience if you can keep the right perspective.
29.1.1 WHY LITIGATION WORKS AND HOW IT GETS ABUSED

In many ways, litigation is both the blessing and the curse of our society. It is the
vehicle by which we hold other individuals and companies accountable for their
actions. In fact, it is the only way that anyone other than the government can force
another to change their ways or take responsibility for their misdeeds. Litigation is
why untested products are typically not rushed to market. Litigation is why a man-
ufacturer may choose more expensive, reliable materials over cheap and dangerous
alternatives. Litigation is a significant part of why we can walk into stores, theatres and
other public places with the confident assumption that under normal circumstances
we will not get injured.
At the same time, litigation has been significantly abused. Litigation is why the
next miracle drug is currently not available. Litigation is why automobile insurance
can sometimes cost more than the car. And litigation is why your ladder is covered
with warning labels you probably have never read.
The point is that litigation does expose problems in design, manufacturing, or
training. More importantly, fear of litigation is what motivates a person or company
to fix those problems rather than let them continue. However, because litigation
frequently results in the payment of money, it is particularly susceptible to abuse.
False allegations, exaggerated claims, and an entire cottage industry of people trained
in how to help others cheat have created a counter movement of extraordinary efforts
to prove every lawsuit illegitimate, every plaintiff a gold-digger, and every expert
witness a liar. At the end of the day, we can only hope that truth prevails and the
cheater (plaintiff or defendant) is stopped.
8417: “8417_c029” — 2007/9/11 — 12:13 — page 656 — #2

The Purpose and the Process of Litigation in a Carbon Monoxide Poisoning Case 657
29.1.2 KEEPING FOCUSED ON THE GOAL

Assuming a legitimate cause of action where a plaintiff is actually hurt by someone’s
actual misdeed, the primary goal of litigation is to restore to the plaintiff what someone
else wrongfully took away. Sometimes litigation has additional benefits such as raising
public awareness, getting a defendant to change its ways, or even produce some
degree of punishment. Those secondary goals are very appealing, but in reality they
are usually outside the realm of what you can make happen through the process.
29.1.2.1 “Compensation” Does Not Mean “Profit”

So if the whole point is, “to restore to the plaintiff what someone else wrongfully took
away” let’s start there. In legal circles, this is referred to as compensatory damages.
In a perfect world, the person who did something wrong simply replaces what they
took. In a situation where John Smith destroys your $100 bill, this calculation is easy.
John must give you a replacement $100 bill. Calculating these damages become much
more difficult, however, when we talk about the loss of someone’s life or abilities.
No defendant can bring someone back from the dead. And while a defendant may be
able to pay for certain types of treatment, he cannot erase the fact that you needed
treatment in the first place. In those circumstances where the thing taken cannot
be replaced, money is the substitute. Because money is the substitute, everything
becomes complicated.
When I was in law school I was shocked the first time I heard my professor say
that a plaintiff is “a loser trying to get back to zero,” but it is a true statement. If you
are a plaintiff, it is because you have already lost something. The best case scenario is
where you are compensated in a way that gets you back to zero. Because you probably
valued what you lost more than money, however, you will probably feel you have
been left short. So please understand “compensation” does not mean “profit.” In fact,
it probably does not even mean “restoration.” It is usually more along the lines of
“something is better than nothing.”
So why bother? Well, something really is better than nothing. If CO poisoning
caused one of your loved ones to die, or left you injured, life is going to be hard
enough. We tell our clients to think of it as a sugar coating on a bitter pill. If you have
to live life with the kind of injury that causes you to get lost in your own grocery
store, then you might as well be wearing a comfortable pair of shoes!1
29.1.2.2 The Secondary Goals Should be Just That—Secondary

As we have discussed above, litigation can create change. It can call public attention
to a problem. It can keep others from ending up like you. But those motivations, while
very noble, can be risky. Defendants only change if they want to. Public attention
only gets raised if the media is not obsessed with something different that week. And
1 Yes, a sense of humor makes even this seem easier.
8417: “8417_c029” — 2007/9/11 — 12:13 — page 657 — #3

you probably would not save someone else unless one of those first two mechanisms
work. So if you set out on the 2-year path of litigation, and run up all those expenses,
strictly for the purpose of saving the world, those things you cannot control will drive
you crazy.
So what about raw meat? Is not there a way to punish the bad guy for what he did?
Well in some states there is something called punitive damages. The goal of punitive
damages is to punish very bad behavior. It has nothing to do with what the plaintiff
lost. In fact, punitive damages can be awarded even if the actual loss to the plaintiff is
minimal. Because these type of damages seek to punish a specific defendant, they are
based on the net worth of that defendant. For example, a $1000 punitive award against
a 19-year-old college student might be a crushing blow, but it would mean nothing
to a Fortune 500 corporation. That is why we occasionally hear of jury verdicts in
millions, or even billions of dollars. It is not because one life is worth more than
another, but because a jury needed to get the attention of a very wealthy defendant.
Although punitive damages make a great lead story in the news, it is punitive
damages which are most frequently overturned on appeal or reduced by the trial
judge. We can probably agree it may not be fair for someone who had only a minor
loss, or participated in their own demise, to suddenly become a multimillionaire. At
the same time, in a world where corporations frequently have billions of dollars of
earnings per quarter, is it hard to imagine them being motivated by anything less.
As a result, punitive damages are hard to prove and often restricted by various tort
reform legislation. Some states have completely banned punitive damages. Others
have limited how much can actually be received by the plaintiff. As a result, you
should talk to your lawyer about them, but it is unwise to consider punitive damages
at all in the process of determining whether or not to pursue litigation.
29.1.3 THE END GAME AND THE BIG PICTURE

How you end up feeling about the entire litigation process is greatly dependent on
what you considered when determining whether or not to file a suit. If you seek
anything other than fair compensation, you will be disappointed. If you are motivated
by greed, that will become clear and juries will likely become so offended they may
refuse to compensate you at all for your loss. If you are motivated by revenge, you
will never recover from the fact that the defendant went back to life as normal and
you did not. If you seek anything beyond what is fair, you will find yourself more
stressed, more worried, and more tempted to do and say things that you otherwise
never considered. The only thing worse than being injured by someone else’s misdeed
is voluntarily destroying your integrity to pursue some misguided goal.
Knowing your motivation will also help you make decisions at the end of the
case. With the advice of your lawyer, you will have to decide whether to try the case
or settle, appeal or stop, hire another expert or go with what you have. A clear view
of why you started will help you decide when to stop.
The noble pursuit of secondary goals can cloud your vision when it comes to mak-
ing that decision. More than one plaintiff has shouted, “Never! I will never let them
off the hook until they change their ways or are out of business.” But remember, you
cannot force other people to change. In most cases, the defendant denies responsibility
8417: “8417_c029” — 2007/9/11 — 12:13 — page 658 — #4

even after signing a check. Occasionally you are left in a position where they go free
entirely even when they acknowledge some fault. Then you have a bitter pill that is
just more bitter.
We can all hope to accomplish noble goals out of tragedy. We should all work
toward that always. But you can advocate for change without litigation. In fact, you
might be taken more seriously if you do not sue. On the other hand, you might be
able to use the money you got from the defendant to do it. So separate everything
but financial realities from your decision to pursue the litigation process. Get clear
in your head that you either will or would not make a difference regardless of the
litigation.
Litigation can give you money, but often not satisfaction. How you live and what
you do can give you satisfaction, but often not money. Your decision should be based
on the likelihood of success, potential damages, and the ability to collect what you
win—and nothing else. So if you are still interested, let’s discuss those details.
29.2 THE PROCESS—WHAT TO EXPECT AND WHEN

TO EXPECT IT
Every lawyer has his or her own strategy. What happens at each step of the way is a
matter of individual strategy. Because my opponents will likely also read this book,
we will not discuss my particular strategy here. It is also not the intent of this chapter
to summarize 3 years of law school. But you should understand the basic components
of the process.
29.2.1 GATHERING INFORMATION

In the law, nothing exists without evidence. So after you come to grips with motiva-
tion, step one has got to be figuring out what you can prove. The rules of proof and
evidence can be different from jurisdiction to jurisdiction (see other Willison chapter),
so now is when you need the help of the lawyer who will have to prove it on the basis
of the facts of your case.
Death speaks for itself, but if you are claiming an injury, you have to decide if
you can prove it. Can it be measured and viewed or do we have to take your word for
it? If we have to take your word for it, are there things in your past that might make
taking your word difficult? Are there witnesses who can testify about how you were
before and after? Are there tests that can prove what you are talking about?
Then you have to prove how the CO got to you in the first place. If it was a bad
furnace, do you still have it? If you do not, is there any evidence of the problem still
around, like a repair man, or a photo, or a fire department report? If you have that
evidence, you better also find information about the people who authored it. Do you
have enough information to find them 3 years from now if they move or change jobs?
In fact, being able to locate witnesses is critically important. Someone other than
you needs to record information about what they observed, with as much details as
possible, as soon as possible. This can be in the form of a handwritten statement by the
witness or other means of documenting that information. In addition, it is important
8417: “8417_c029” — 2007/9/11 — 12:13 — page 659 — #5

you have the type of information you will need to find them later if they move. Never
assume you will always be in touch with any person and someone will “never forget”
the details of an event.
Eventually, you will need some way to connect all of the information together to
form the elements of your burden of proof. In many cases, this is done by the testimony
of expert witnesses. Because of factors too numerous to list in this chapter, your
attorney should be involved in that process. For the purpose of gathering important
foundational information, expert witnesses should be brought into the loop as soon
as it is practical.
You and your legal counsel will also need to determine whether you have a right
to sue. Statutes of limitations are laws which limit the amount of time in which you
can file suit. Sometimes this time frame begins to run when you discover the problem.
Other times and places it begins to run as soon as the incident occurs. For example,
depending on your jurisdiction and the theory of liability, the statute of limitations
may begin to run on the date you discovered that CO was entering your home, or
it may have begun to run on the date that the appliance was improperly installed.
In some circumstances, the statute of limitations may have expired before you ever
found out there was a problem. In addition to statutes, the time frame for suing can be
limited by contract or other means. If you do not have a right to sue the guilty party,
you should not waste time, money, and energy on gathering information for the rest
of the case.
Finally, you have to determine whether your “guilty party” is collectable. If your
CO poisoning was caused by a homeless person, for example, winning a one-million
dollar judgment would only be worth the paper it is written on. Even small businesses
may not be collectable unless they are adequately insured. While it may be unsettlingly
pragmatic, it does not make sense to spend $50,000 in litigation expenses if you can
only collect $20,000 from the defendant. It may feel unjust, but it is a necessary thing
to consider before proceeding.
29.2.1.1 Filing Suit

Once you have come to the conclusion that you wish to pursue litigation, and have
gathered enough information to be confident your attorney can prove your case in
court, you must formally begin the lawsuit. This is done by filing a “complaint.” A
complaint is typically a listing of factual allegations which, if proven true, would give
you the right to recover from the named defendant.
In your complaint, your attorney may list several defendants if the facts support
it. In addition, you may have several different theories of liability against a particular
defendant. In fact, most jurisdictions allow you to plead2 alternative (even inconsist-
ent) theories. For example, you have the right to allege in Count I of your complaint
that the defendant acted negligently and then in Count II of the complaint allege that
2 “Plead” or “Plea” are terms which refer to any allegation contained in a pleading. A “pleading” is typically
defined as a complaint or an answer to a complaint.
8417: “8417_c029” — 2007/9/11 — 12:13 — page 660 — #6

he acted intentionally. At trial, only one of them can be proven true, but by pleading
them both you preserve your right to argue either one as further facts unfold.
Once the defendant receives the complaint, they have an obligation to answer
each allegation. By admitting, denying, or saying they do not know either way, you
will immediately be able to see which issues will be the focus of your litigation. In
addition, the defendant has a right to file “affirmative defenses.” These are assertions
of certain legal theories which would get the defendant off the hook even if all the
allegations in your complaint are true. For example, the defendant may allege that
there was some other person who was responsible for the activities you allege in
your complaint. The defendant may have immunity because of some sort of statute
or contractual arrangement. They may also use affirmative defenses to point out
deficiencies in your complaint.
It is the purpose of the complaint, answer, and affirmative defenses to make sure
that all parties involved in the litigation understand all of the reasons why the plaintiff
believes the defendant is guilty and why the defendant believes that he is innocent.
Assuming there is some disagreement after this phase, your case then moves into
“discovery.”
29.2.1.2 The Discovery Process

During the “discovery” phase, both sides get to find out the details of the other’s case.
At this point, each side has the right to demand documents and information from the
other. Each can interview witnesses under oath.3 A defendant can request releases for
your medical information. You can request the design details of the product from the
defendant corporation. In most jurisdictions, the purpose of discovery is to make sure
that there are absolutely no surprises at trial. Both parties should fully understand the
strongest portions of their opponent’s case before making a decision to actually put
evidence in front of the jury.
The discovery phase is by far the longest phase of the case. Parties may fight about
whether they have a legal obligation to provide certain information. There may be
delays in witness schedules, or the availability of attorneys. While the court always
sets a deadline for discovery to be completed, it is not uncommon for one, or both
parties to request an extension to find out more details about a specific issue.
This discovery phase is also when several motions occur.4 This is how parties
ask the judge to sort out which evidence will be allowed to be presented at trial and
whether any portions of the complaint, or affirmative defenses, should be dismissed
by the court.
At some point before trial, the defendant will ask to have your case dismissed. This
is done almost as a matter of routine. You should not take the defendant’s attempts to
have your case dismissed as a sign of the strength or weakness of your case. In light
of the risk of trial, you really cannot blame a defendant for trying. Although it should
3 Statements taken under oath during a lawsuit are called “depositions.”

4 A “motion” is any request to the court (usually by a party) to take some action, order someone else to act
or clarify how the case will proceed.
8417: “8417_c029” — 2007/9/11 — 12:13 — page 661 — #7

always be taken seriously, most judges prefer a case to be tried by a jury than thrown
out on a technicality. Nonetheless, it is a judge’s job to not waste the court’s time if
the plaintiff does not have the evidence required to meet their burden of proof.
29.2.1.3 Trial—Risks and Rewards

At the end of this entire process is your trial. Either party may request to have the
case tried by a jury. If both agree, it can be tried by a judge. Any trial is a risky
proposition. While you may have all the evidence in the world, it is solely up to a
handful of randomly chosen strangers to decide your fate. While you may have spent
years wrestling with issues, they will only have a few days. While most jurors take
their jobs very seriously, some are only interested in whatever path will get them back
to their normal lives the quickest. In a situation such as a CO poisoning injury case,
the quickest route home is to award zero.
At the same time, juries are equally unpredictable for defendants. What may be
perfectly reasonable in industry standards, may outrage the average citizen. Evidence
that the defense team considers to be completely unbelievable may be accepted as
absolute truth by the jury. And of course, the biggest thing working against defendants
in many CO poisoning cases is the fact that every one of the jurors can imagine
themselves in the same situation. Everyone has a furnace which could eventually go
bad. Everyone has a car which may not be properly manufactured. Everyone is the
customer of some utility company.
The jury system is notoriously imperfect. At the same time, it is by far the best
system on earth. It is precisely this unpredictability that causes even the most stead-
fast party to at least consider the possibility of settlement. Every party, plaintiff,
or defendant must calculate the likelihood of the worst and the best case scenario.
Research on previous verdicts in that jurisdiction can begin to narrow what is also
the most likely scenario. All the variables such as the performance of witnesses, the
availability of evidence, the randomness of jurors, and even the health and mood of
the judge, attorneys, parties, witnesses, and jury can influence the outcome of a trial.
Accordingly, most experienced trial attorneys will tell you that if they tried the same
case ten times in the same county they would likely end up with at least two very
different results.
Many courts will require parties to participate in alternative dispute resolution
prior to trial. This can be a form of mediation, arbitration, or even having the case
evaluated by other attorneys from the jurisdiction. It is the goal of every court to get
parties to negotiate an agreement rather than have a jury try the case. Nonetheless, it is
every party’s right to choose to not negotiate. As a result, your entire case preparation
should be focused on trying the case if you have to, but settling if you can.
29.3 THEY CALL IT THE BURDEN OF PROOF FOR A

REASON
Anyone who has ever watched a TV show about lawyers has heard the phrase “burden
of proof.” What this means is that the plaintiff has the obligation to present admissible
8417: “8417_c029” — 2007/9/11 — 12:13 — page 662 — #8

evidence sufficient to prove that the defendant is responsible. The plaintiff not only
has to prove that they were injured, but what caused it, and who is responsible
for that cause. The defendant never has to prove its innocence. In the circum-
stances of a CO poisoning case, this means the plaintiff has to prove all the details
of how they came in contact with a substance that is completely invisible. If the
defendant can convince the jury to not believe just one part of that story, they
go free.
29.3.1 THE STANDARD OF PROOF

The standard of proof refers to the degree of certainty to which the jury must be
convinced before it can say the plaintiff has proved his/her case. In a civil case, such
as a lawsuit over a CO poisoning injury, the standard is a “preponderance” of the
evidence. That simply means that the plaintiff must prove that it is more likely than
not that the proposition is correct. This is remarkably easier to prove than the criminal
standard which is “beyond a reasonable doubt.” In a civil case, a jury can find in favor
of the plaintiff even when they have reasonable doubt as long as it is the “most likely”
cause. On the other hand, if the jury concludes that there are two equally plausible
theories which may have caused the plaintiff’s injuries, the burden of proof has not
been met and the plaintiff will lose.
29.3.2 TWO ESSENTIAL ELEMENTS

In a nutshell, the elements of the burden of proof are liability and damages. Put in
its most basic terms, liability is proving, with evidence, the answers to who, what,
where, when, why, and how. If all that is proven, the plaintiff also has to show provable
damages.
29.3.2.1 Liability
The first element of any litigation is to prove liability. Basically, this means you have
to prove who is legally responsible for your injury. To do that you must first know how
you were injured. It is not enough to know that you have been exposed to CO. You
must be able to prove where that exposure occurred, when that exposure occurred,
and why you were exposed to CO.
It is not enough to just know what caused your exposure, but we must also
determine why that condition occurred. If the CO poisoning is a result of a cracked
heat exchanger in the furnace in your home, it might be the result of a flaw in the
manufacturing. However, it might be the result of a mistake made by a repairman. At
the same time, it might be the fault of the homeowner because a repairman was never
called to do the maintenance necessary on that machine.
Once the where, when, and why questions are answered, you can then determine
who is responsible for that exposure. In a negligence case, this often requires proof of
whether someone has breached a duty they had to someone else. We all have a duty
to act reasonably whenever we do something. This is the standard of a “reasonably
prudent person.” If you drive a car, you must do it reasonably. If you do not, you will be
8417: “8417_c029” — 2007/9/11 — 12:13 — page 663 — #9

held responsible for any unreasonable acts you commit with that car. This is a common
duty we all have. At the same time there are other duties that are imposed on people.
Some duties are imposed upon us by law. For example, laws in your state probably
require a contractor to install a new water heater according to the manufacturer’s
instructions. If the contractor chooses not to, he will likely be held responsible for
anything resulting from his deviation from the instructions. Sometimes we volunteer
for undertaking a duty by contract. For example, if a mechanic agrees to fix a problem
in exchange for money, he has a duty to fix the problem. If he fails to fix that problem,
or makes it worse, he will be responsible for that failure.
Ultimately, liability is a combination of a number of factors which creates respons-
ibility on behalf of one person to another. So you may know that you were injured
by exposure to CO in the month of February in your home coming from a cracked
heat exchanger on your furnace. You may know who manufactured that furnace, but
if that furnace had a 10-year warranty, the manufacturer will not be liable for a heat
exchanger which cracked in the 23rd year. At the same time, a service technician who
failed to notice the crack in the heat exchanger when he performed a safety inspection
might be liable for your exposure to CO even though he did not cause the crack to
occur.5
29.3.2.2 Damages
In order to have a right to bring a lawsuit, you must have suffered a loss. For the most
part, the American civil legal system is set up to compensate for actual harm after it
has occurred rather than prevention of what might happen. After you have answered
the liability question, you still have to prove the “so what” and the “how much.”
In a case of a CO poisoning lawsuit, it is not enough to prove that someone’s
misconduct caused CO to leak into an area where you were at a particular time, you
must be able to prove that it had some effect on you. If the problem was noticed
immediately and you left the area before you suffered any symptoms, you likely have
no basis to bring a lawsuit.
29.3.2.2.1 Proving the injury

Proving that a death occurred as a result of CO poisoning is rather simple, if the
appropriate tests were conducted to determine the cause of death. There are certain
things that happen in the human body when it is exposed to lethal concentrations of
CO which can be readily identified. Because the deceased person stopped breathing,
we can also typically find very high levels of CO in the blood. In a case where a person
does not die, but is injured, proof of CO poisoning becomes much more difficult.
There is an overwhelming lack of understanding in the medical community about
what CO does to human beings, other than cause death. This is a problem from a
proof standpoint, because proving you have symptoms and proving you have been
exposed to CO does not necessarily prove that one caused the other. If you suspect you
5 Of course, in that situation, he would only be responsible for the exposure to carbon monoxide after his
inspection, not the carbon monoxide you were exposed to before he arrived.
8417: “8417_c029” — 2007/9/11 — 12:13 — page 664 — #10

may have been injured by CO poisoning, you need to find someone in the medical
field who understands CO. Let me be clear, this does not mean you should try to
find a doctor to tell you, you have been injured by CO. There are doctors in this
world who, for the right fee, will tell you anything you want to hear. Those doctors
should be avoided at all costs. What you need to find is someone who understands
CO who can truthfully and credibly tell you whether or not your symptoms have been
caused by that exposure. You may go for years frustrating various uninformed medical
professionals who can never determine why you are suffering from headaches and
short-term memory loss before you find one that can scientifically explain to you, and
eventually to a jury, the reason for your symptoms. But someone other than you will
have to provide that connection.
As with any toxin, the damage caused by CO is a function of time and dose. How
long you are exposed to how much CO are important facts in determining whether
your symptoms were caused by the problem. Even if we all agree on those numbers,
there is disagreement in the medical-toxicology field about what CO poisoning can
do. More importantly, what harm CO exposure actually causes can vary widely from
person to person. Unfortunately, however, we do not often agree on either time or
dose. By its very nature, the way CO interacts with the body causes it to disappear
rather quickly. Having a blood test, even the next day, is often useless to determine
the level of exposure.
Because the common effects of CO poisoning are similar to many other medical
conditions, the medical testimony might not go beyond the statements that your
symptoms could have been caused by CO. If that is the case, it may be enough.
In many jurisdictions, the jury is allowed to consider the totality of the evidence to
conclude whether “more likely than not” you have suffered an injury as a result of
your exposure. While there may be other plausible medical explanations for your
symptoms, combining the testimony that it could have been caused by CO with the
fact that you did not suffer those symptoms prior to exposure and the fact that these
symptoms were noticed shortly after the exposure, may be sufficient to meet the
burden of proof on injury. Again, this determination will be very specific to your case
and dependent on the rules of evidence in your jurisdiction.
29.3.2.2.2 Economic damages

Economic damages are those where we can more easily ascertain an amount of money
to replace the loss. For example, if you miss work for six months, it is rather easy to
calculate the amount of your lost wages. The reimbursement for the amount of money
you spent on medical treatments is easy to calculate. It is somewhat more difficult
in a situation where you can never work again and we must calculate the amount of
what would have been your future earnings.
In determining future economic damages experts need to calculate what would
have been your career path. The likelihood of promotions, raises, and job changes
must be considered together with the likelihood of lay-offs, demotions, and even
other future injuries. Experts who are trained to utilize statistics and economic data
can render reasonable opinions about what would have been. Of course, you can also
see how defendants can find experts to point out that while every 20-year old intends
to someday take over the company, only one of them actually will. More importantly,
8417: “8417_c029” — 2007/9/11 — 12:13 — page 665 — #11

while we certainly have the data which allows us to calculate life and work expectancy
we all know that some people live well beyond expectations and others tragically die
too young.
29.3.2.2.3 Noneconomic damages

The second type of compensatory damage is noneconomic. These damages are very
real and important, but always subject to enormous debate. We can all agree in
principal that injured people are entitled to recover for pain and suffering, but what is
the value of a headache? If we can determine the value of a headache, how does that
compare to the value of a migraine? How much should we compensate someone who
can no longer engage in their favorite hobby? What is the value of one more day, or
even one more hour, with a loved one? Is the life of a 5-year old worth more or less
than the life of a 15-year old?
In my experience, almost everyone who suffers one of these losses will tell you
that there is no amount of money that can compensate them for their loss. This is the
problem. While we cannot invent a number big enough to be adequate, the jury at trial
will have to determine a number that is appropriate. It is important to remember that
these damages are only supposed to compensate you for your loss. These damages are
not for the purpose of punishing the defendant. Accordingly, these numbers should
be the same whether it was a little old lady who harmed you or the largest corporation
in the United States.
Be prepared. Because this element of your loss is so uniquely personal, the defend-
ant has the right to probe into some very private details to determine if they are valid.
If you seek recovery for the loss of the companionship of your spouse, the defend-
ant has the right to inquire about the quality of your marriage. If you claim you are
depressed, the defendant will look deeply into your past to see if you were depressed
before the exposure. If you seek damages for no longer being able to play golf, the
defendant even has a right to speak to your golfing buddies to see if you actually
enjoyed it or whether you were planning to give up the game.
In the real world, however, your injury might make a bad marriage worse, people
struggling with depression can go over the edge, and giving up golf should be on
your terms, not someone else’s. In most jurisdictions, making a previous injury or
condition worse is compensable. In fact, many times if the jury cannot sort out the
difference before and after they will be allowed to award the entire value of the
condition to you. As a result, you need to remember that it may be uncomfortable,
but it is important if that is the type of loss you have suffered.
29.4 THE UNIQUE PROBLEMS OF CARBON

MONOXIDE POISONING
CO is odorless, tasteless, and colorless. It does not cause any of your senses to activate
until you are already experiencing the symptoms of damage from exposure. In cases
of acute exposure where someone has a dramatic reaction to a large dose of CO, like
death or unconsciousness, it can be easy to pinpoint the time and place of exposure,
but as you may have already read in other chapters of this book, CO poisoning can
8417: “8417_c029” — 2007/9/11 — 12:13 — page 666 — #12

cause significant damage in much lower doses over long periods of time. Because
nearly every home, car, place of work, and school has something capable of producing
CO, long-term exposures can be very difficult to pin down to a particular place or
time. That job is made even more difficult by the fact that any one of us could have
multiple exposures to CO in various areas. Every time there is a new when or where
it could have a dramatic impact on whether you will ever be able to prove liability
for your exposure.
The same characteristics which make CO difficult to catch inside the human body,
make it even more difficult to catch in a room. Even in situations where lethal levels of
CO exist in a room, the focus is, and should be, on the saving of the lives of individuals
inside. That means first responders are trained to open doors and windows, shut off
the fuel supply, and get affected people outside. Even rescue personnel are trained
to evacuate the building when CO concentration reaches certain levels. As a result,
we may be able to document that the CO concentration exceeded the emergency
evacuation level, but we may never know whether it was double, triple, or a hundred
times that level. This is a problem attorneys must face in CO litigation. Until CO
detectors actually become accurate recorders of information, this will continue to be
a problem.
29.5 YOUR LIFE ON TRIAL

Whether the exposure to CO resulted in injury or death, any litigation will be about
the impact on that life. As a result, the details of that life will be put on trial. The
symptoms known to be caused by CO exposure are, unfortunately, also symptoms
known to be caused by a number of other conditions. If you are alleging that you
suffer from headaches or memory loss, the defendant has the right to examine your
entire medical history to determine if there are other circumstances in your past
which may have caused those symptoms. If you claim that since your exposure
you have been having difficulty with math, the defendant has the right to take the
deposition of your junior high math teacher. If you are claiming that the CO has
caused you to lose control of your moods, you should plan on your ex-husband
being called to testify about every time you were ever moody before the exposure.
As difficult as all this may sound, those are reasonable tactics of legitimate defense
attorneys.
As we all know, however, some lawyers are neither legitimate nor reasonable.
Some lawyers will do anything to win a case regardless of whether it is true, ethical,
or fair. I have personally seen lawyers take one story from a Christmas party and turn
it into allegations of alcoholism. I have seen pediatric records of colds and flu used to
explain away the fact that a 50-year-old man has been coughing for 6 months. I have
seen grieving parents accused of neglect. I have seen a loss of memory painted as a
convenient lie.
For every person who has ever received more than they deserved in litigation,
there are likely far more who have walked away from a legitimate claim because they
choose to no longer put up with the trouble. When deciding whether or not to proceed
with your lawsuit you need to consider this fact. However, I have also found that
8417: “8417_c029” — 2007/9/11 — 12:13 — page 667 — #13

being prepared for this ahead of time and knowing it will eventually come, helps my
clients weather that storm.
You need to have the courage to weather that storm. If you try to cover up any
unfortunate fact, it will make everything worse. When there is a lot of money at risk,
the defendants will eventually find all the information. You do not have to volunteer
information, but if asked you cannot appear to be hiding anything. If you hide a doctor,
a school, a military record, or any other fact, it will instantly look more important than
what it is. It is always easier to explain away a bad fact than to undo the impression
you are being dishonest. It is better for you to perform your best on every test and
physical exam than to give the impression that you are someone who is exaggerating
or malingering. In fact, if you come across as open and honest, any dirty tricks of a
bad defense lawyer may work to your favor rather than his.
29.6 AN OVERVIEW
29.6.1 GETTING HELP MANAGING THE FILE
One of the most common long-terms effects of exposure to CO is short-term memory
problems. People with significant injuries often suffer from depression and CO pois-
oning can also cause emotional instability. These characteristics make the litigation
process particularly difficult. You and your health care professionals need to do an
accurate assessment of whether or not the exposed person is competent to carry on
the tasks of daily living. For the litigation process especially, it is recommended that
an unexposed person who is intimately familiar with the victim or victims, be called
upon to work as your assistant in the litigation process.
Because the litigation is about your life, it involves a seemingly never-ending
process of gathering documents, information, names, phone numbers, and records.
While you need this evidence to prove you have suffered an injury, the fact that
you have suffered this type of injury will make gathering this stuff seem impossible.
When you do a good job, the defendant will claim you are not injured. When you
do a bad job, the defendant will claim you are trying to cover something up. All of
that might be avoided by the appointment of someone to act on your behalf to gather
information.
This person must be in a position such that you do not mind sharing every intimate
detail of your life. At the same time, you need to discuss with your attorney how best to
handle that relationship so that you can preserve the privilege of your attorney/client
communications. If used correctly, this assistant will dramatically reduce your stress.
29.6.2 TRUSTING EVERYONE TO DO THE JOB ASSIGNED

Selecting your attorney is probably as important to the litigation process as choosing
a physician is to having an accurate assessment of your health. Attorneys who under-
stand CO poisoning are even more rare than health care professionals who understand
it. Your town’s typical personal injury attorney does not have the experience neces-
sary to handle the complexities of the case they must prove and the defense attorneys
they will likely face. As a result, you should find whatever attorney you trust most
8417: “8417_c029” — 2007/9/11 — 12:13 — page 668 — #14

and ask them to help you find the attorney who will be your lead counsel in the litig-
ation. Once you find an attorney you can trust who is capable of handling this case,
you need to trust that attorney with the entire process which includes your decision of
whether or not to even pursue litigation. Let’s discuss some key roles in CO poisoning
litigation.
29.6.2.1 The Attorney

Although attorneys may be everyone’s favorite punch line of a joke, good attorneys
make all the difference in a close trial. At the end of the day it is facts which win
a case. Good attorneys cannot invent facts, but bad attorneys can fail to put them
in front of the jury. Attorneys cannot change the testimony of witnesses, but good
attorneys can know how to handle it, or even exclude it under certain circumstances.
These types of trial decisions are enormously complex, based on a number of
factors, and to some degree are a matter of personal style and experience. Accordingly,
an attorney is kind of like the quarterback of a football team. They influence how
things are presented to the jury and what particular parts of the case are emphasized
or de-emphasized. You always have the right to ask your attorney to explain why
decisions are being made. There needs to be open communication and a relationship
of trust between a client and an attorney.
29.6.2.2 Fact Witnesses

Most people who testify at a trial are there for the sole purpose of presenting facts.
Aunt Millie will never be allowed to testify whether she thinks ABC Company ought
to pay you for your injury. However, Aunt Millie may be able to testify about the
differences she saw in you before or after the event. She could also testify about
what she saw, heard, or read. The more fact witnesses are able to stick to just facts
and personal observations, the less susceptible they are to being “beaten up” by the
opposing attorney. It is the job of the attorney, not the witness, to argue which facts
should be given the most emphasis and how they should be used in the litigation
process. When fact witnesses try to do this they quickly loses credibility.
29.6.2.3 Expert Witnesses

Unlike Aunt Millie, expert witnesses are allowed to give opinions at trial. These
opinions must be based on factual evidence in the trial combined with their knowledge,
experience and training. For example, an X-ray showing an arm bone with a line
through it is a fact. A physician’s testimony that the line means that the arm is broken
in that spot is an expert opinion. There are a number of rules affecting the scope and
manner in which expert witnesses can testify (see the other Willison chapter). In a CO
poisoning case, having expert witnesses explain complicated and technical matters
is very helpful to a jury and ultimately critical to the litigation. Most people do not
intuitively understand how toxins affect the brain. Experts are brought in to make
things like this clear for a jury. If you try to do that for yourself, you will play into
the defense strategy of looking like you are manufacturing a case for yourself.
8417: “8417_c029” — 2007/9/11 — 12:13 — page 669 — #15

29.7 CONCLUSION
The litigation process, and all that surrounds it, can be overwhelming to any indi-
vidual, especially those struggling with the effects of being exposed to CO. One needs
to consider and make many difficult decisions before beginning the journey involved
in CO poisoning litigation. To begin with, your considerations should include the
determination of goals and the finding of the right attorney to assist you in meeting
those goals. As the process continues, one needs to gather the information of who,
what, when, where, how, and why the exposure occurred and the person responsible
for bringing those factors together. If sufficient evidence exists, the lawsuit can begin.
Many people want to know how long it will take. That is impossible to answer.
A case could settle at any point. If you go to trial, it depends on how busy the court
is and how long discovery takes. In most places, you should plan on about 2 years
from the time you file suit until you are actually in front of a jury. To the extremes,
it will likely not be less than 1 year, nor more than 3 years. But that is just the trial.
After trial, either party may appeal the decision. That can take another year or two.
Then they might be able to appeal again to the Supreme Court of that jurisdiction.
In the worst case scenario one of the Courts of Appeals might order the parties to go
back and try the case all over again. As you can see, a negotiated settlement is almost
always better than having your day in court. But you will never negotiate a decent
settlement unless you are ready to take your day in court.
We talked a lot about motivations at the beginning of this chapter. Let me end
by bringing this full circle. Once you decide to litigate, the best way to survive the
process is to forget about it. Move on with your life. By the time the complaint is
filed your attorney should have most of the information he or she needs. You will be
called upon from time to time to provide information, but your main roles will be to
testify at a deposition and show up at trial. Over a 2-year period, you may have 30
days where you are personally involved. If you wake up every one of the other 700
days obsessed about the lawsuit, you will drive yourself (and everyone around you)
crazy. The hardest part is to gather the information and make an informed decision
about whether to sue. After you have crossed that bridge, let the process unfold while
you live life. Remember, how you live and what you do brings satisfaction. Your
litigation cannot serve that purpose.
8417: “8417_c029” — 2007/9/11 — 12:13 — page 670 — #16

30 Offering Expert Opinions
in a Carbon Monoxide
Case
Stephen P. Willison
CONTENTS
30.1 A Brief Introduction to the Rules of Engagement . . . . . . . . . . . . . . . . . . . . . . . . . 672

30.2 Roles—Who’s Who and What Do they Do. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 673
30.2.1 Fact Witnesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 673
30.2.2 The Judge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 673
30.2.3 The Jury . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 674
30.3 Your Opinion—the Heart and Soul of Being an Expert Witness . . . . . . . . . 674
30.3.1 Are You Really an Expert? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 675
30.3.2 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 676
30.3.3 Basing Your Opinions on Reality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 676
30.3.4 Scientific Basis of Your Opinion—the Method Prevents the
Madness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 677
30.3.5 Putting It Together—Testifying About Your Opinions . . . . . . . . . . . 679
30.4 The Battle of Experts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 680
30.5 Thinking Long Term . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 681
30.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 682
Some trial attorneys tell jurors that a trial is like assembling pieces of a puzzle. If that
analogy is true, then a good expert witness should be like the picture on the puzzle’s
box. No one likes hearing about the “battle of experts.” After all, the law should be
based on facts not opinions, right? But in reality, the typical juror often does not have
the background or training necessary to understand the importance and relevance of
many of the facts presented to them in a highly technical case. This is especially true
in carbon monoxide (CO) litigation. Terms like “carboxyhemoglobin”, “523 ppm”
and “lateralized cerebral dysfunction” can be critical pieces of the puzzle, but are
useless to a jury unless they are educated about their meaning and place in the case.
So, like the picture on a puzzle box, the expert witness takes those complicated pieces
and helps the jury understand how they fit in and where they go, so the image of what
happened begins to make sense.
671
8417: “8417_c030” — 2007/9/11 — 12:13 — page 671 — #1

This chapter is not intended to be a substitute for law school or an outline of

how to conduct a trial. It also is not intended to be specific to any field. Instead, it
is intended to be a generalized guide for those asked to testify as an expert witness
to understand their role in the process. It should encourage a better dialog between
expert witnesses and trial counsel, not substitute for that dialog. There are many, many
important issues in this area which cannot be covered without specific knowledge of
your case or field.
30.1 A BRIEF INTRODUCTION TO THE RULES OF

ENGAGEMENT
Every court in the United States makes decisions about the admissibility of evidence
on the basis of its Rules of Evidence. All federal courts share one set of rules called
the Federal Rules of Evidence. However, each state has its own set of rules. While
many states model their rules of evidence after the Federal Rules, almost all of them
have some variations. As a result, there are 51 different sets of Rules of Evidence in
this country. In addition to Rules of Evidence, every jurisdiction has its own set of
Court Rules. While the Rules of Evidence determine whether a particular piece of
information is admissible, Court Rules determine whether or not admissible evidence
is actually admitted.
Obviously, we cannot discuss, or even summarize, the thousands of variations.
Because of its broad influence in almost every jurisdiction, an understanding of the
Federal Rules of Evidence1 is a decent foundation for testimony in just about every
court. Understanding the small variances in local rules, however, are critical to your
specific case. For example, the Federal Rules of Evidence allow an expert witness’s
opinion to be based on facts or data “perceived by or made known to the expert.”2 The
Rules of Evidence in the State of Michigan, however, state that the facts or data upon
which an expert bases his opinion “shall be in evidence.”3 So what is the difference?
Everything.
Imagine that a firefighter makes a statement to a local television station at the scene
of an incident that they recorded carbon monoxide levels in excess of 1000 ppm.
The next day, that firefighter retires and disappears. In addition, in his excitement
over retirement, he failed to record that data into his notes. His statement to the
media that the CO concentration was more than 1000 ppm is hearsay.4 Hearsay is
not admissible.5 In a Federal Court (even a Federal Court in Michigan), an expert
witness could still testify about opinions based upon that hearsay because it was a fact
“made known to the expert.” In a State Court in Michigan, an expert witness could
not testify to any opinion based on that hearsay because it would be on the basis of a
fact not in evidence.
1 Sometimes abbreviated as F.R.E.

2 F.R.E. 703.
3 M.R.E. 703.
4 F.R.E. 801.
5 F.R.E. 802.
8417: “8417_c030” — 2007/9/11 — 12:13 — page 672 — #2

Offering Expert Opinions in a Carbon Monoxide Case 673
This is why expert witnesses should not try to be lawyers. Each and every time
you are asked to testify in a case, you should spend time with the attorney finding out
about the rules of evidence, deadlines, disclosure rules, and everything else that may
influence the admissibility of your opinions. In addition, it is critically important to
understand what rulings have occurred in the particular case which may affect your
testimony. A good working relationship between trial counsel and expert witnesses is
a key factor to success. Just because you have testified as an expert a hundred times
over does not mean you should make any assumptions about changes in the rules or
differences between jurisdictions.
30.2 ROLES—WHO’S WHO AND WHAT DO THEY DO

It is your job to help the jury make sense of the various facts that have been presented
to them throughout the trial. However, you will not be the only person involved in
the case. In all likelihood, you will not be the only “picture” presented to the jury as
to what this “puzzle” should look like. Understanding how you fit into the litigation
structure is important to your success.
30.2.1 FACT WITNESSES

Obviously, there would not even be a litigation unless there was some substantial
disagreement between the plaintiff and defendant. In a CO case, the plaintiff is likely
trying to prove that the defendant is responsible for his or her injury or death. The
defendant either reasonably or unreasonably denies responsibility for that injury.
Plaintiffs, defendants, and all non expert witnesses are substantially limited as to
the types of information they can offer in their testimony. As a group, these people
are referred to as fact witnesses. Fact witnesses provide the pieces of the puzzle for
which they have specific knowledge. Examples of fact testimony would include the
following:
• The alarm went off at 2:03 p.m.

• The CO meter read 364 ppm.
• The CO meter was last calibrated on January 15.
• The plaintiff was in the building when we arrived.
With few exceptions, fact testimony is limited to things that were seen, heard, or exper-
ienced by the witness himself. It can be verifiable, like the machine was calibrated
on January 15. It can also be completely subjective like, “I had a headache.”
30.2.2 THE JUDGE

In a civil trial with a jury, the role of the judge is simply to determine what evidence
will and will not be presented to the jury. It is the judge’s job to know, interpret and
enforce the Rules of Evidence we discussed earlier. While judges do not decide the
facts of the case, they can greatly influence the outcome of the case by controlling
what information the jury is allowed to consider.
8417: “8417_c030” — 2007/9/11 — 12:13 — page 673 — #3

Judges are not perfect, but their position does deserve respect. Making objections,
responding to opposing objections, and debating rules should be left to the lawyers
and the judge only. Juries almost always like and trust the judge. If you are disrespect-
ful to the judge in front of them in any way, they will likely start to mistrust you.
Regardless of whether you agree, do whatever the judge tells you. Let the lawyers
fight about it later.
30.2.3 THE JURY

The jury ultimately decides what happened and who is responsible. In doing so, they
are often given a questionnaire to fill out. Each jurisdiction has its own rules for how
they can answer these questionnaires. However, most jurisdictions do not require a
unanimous jury in a civil case.
The members of the jury, more than anyone else, are the people who need to under-
stand and believe your testimony. Although the substance of your opinions should
never change, your method of delivery should be sensitive to your “audience”—the
members of the jury. There is a temptation in a courtroom to want to appear smart.
Impressing the lawyers, the judge, or other experts who may be in the courtroom
might be gratifying, but your testimony will be useless if it is not understood by the
jury. Even worse, your testimony will be harmful if your efforts to be “smart” are
perceived by the jury as “smug.” It is far better for everyone in the room to think
you are an idiot and the jury think you are the best witness at the trial, than for the
opposite to be true.
Typically, lawyers will have some basic information about jurors before you take
the stand. Information like levels of education or professions might be useful, but also
might be dangerous. Just because someone is educated does not mean they are smart.
Just because someone dropped out of school and became a farmer does not mean they
are not reading Plato and Socrates in the evening. Broader stereotypes are even worse.
Information like age, gender, race, or perceived economic class is useless. You are not
trying to convince a class of people, you are trying to convince a particular person.
Body language feedback such as smiles, nods, even rolling eyes and scratching heads
is generally a better indicator than education levels and economic status.
30.3 YOUR OPINION—THE HEART AND SOUL OF

BEING AN EXPERT WITNESS
There are two components to expert testimony. Like two wings on a bird, if either
one is missing, your testimony cannot fly. First, you must be right. Second, you must
have the ability to explain your opinion. Fortunately, these two things feed off each
other. When you are right, it is easy to explain how you reached your conclusion.
Moreover, when your opinion is formed with an understanding that you will have to
explain it later, you have a higher likelihood of coming to the correct conclusion.
Federal Rule of Evidence 702 controls whether expert testimony should be
admitted. It states that:
If scientific, technical or other specialized knowledge will assist the trier of fact to
understand the evidence or to determine a fact in issue, a witness qualified as an expert
8417: “8417_c030” — 2007/9/11 — 12:13 — page 674 — #4

by knowledge, skill, experience, training, or education may testify thereto in the form
of an opinion or otherwise, if:
1. The testimony is based upon sufficient facts or data;
2. The testimony is the product of reliable principles and methods; and,
3. The witness has applied the principles and methods reliably to the facts of
the case.
30.3.1 ARE YOU REALLY AN EXPERT?

In a CO case, it is pretty easy to conclude that there are a number of things necessary
to “assist the trier of fact to understand the evidence or to determine a fact in issue.”
Given this fact, the first question that must be answered by the court is whether a
particular witness, like you, is qualified to testify as an “expert” witness. As the
rule states, an expert can be anyone who would assist the trier of fact by way of
their knowledge, skill, experience, training, or education. You do not have to have a
degree or license to be an “expert” witness. As an example, the US Supreme Court
recognized that a perfume tester maybe an expert in identifying odors on the basis of
years of experience and testing.6 Of course, you can also be an expert witness because
you invented the product, have a doctorate degree or are a world renowned leader
in your field. The point is, the judge and the jury has to have some understanding of
why they should listen to you. We do not want technical cases influenced by “man on
the street” opinions.
As a result, you need to be prepared to present good reasons why you should
be heard. You need to keep a list of any activities you have accomplished, studied,
written, or taught that are relevant to the area in which you intend to testify as an
expert. Every time you do something new, you should add it to your curriculum
vitae immediately. To gain credibility, you should also stay current in your field. Just
because you were the king of your field in 1972 does not mean that you are an expert
in the twenty-first century.
In addition, many jurisdictions require expert witnesses to produce a log of cases
in which they have previously testified as an expert. While this is specific to each
jurisdiction, the Federal Rules require the production of a list of any expert testimony
in the past 4 years and any publications in the past 10 years.7 It is always permissible
for an opposing attorney to make sure the jury is exposed to facts which may sway your
testimony. For example, if you always testify that every product is safe, the jury should
know. If you never treat patients, but only testify as an expert witness, the jury should
know. These facts do not necessarily make your opinions invalid, but they do get
factored into how much credibility your opinions will be given.
You also have to be careful with what you call experience. Letting your ego
run wild or making mountains out of molehills will destroy your credibility. Many
physicians have tried to claim to be an expert in CO only to later reveal that the
6 Kumho Tire Company v Carmichael 119 S. Ct 1167, 1176 (1999).

7 F.R.C.V.P. 26(a)(2)(B).
8417: “8417_c030” — 2007/9/11 — 12:13 — page 675 — #5

entire extent of their training and experience on CO involved reading one half of a
chapter in one textbook, in one semester of medical school 20 years ago. Certainly,
all physicians have training in how the body reacts to various conditions. Certainly,
physicians should have a greater understanding than jurors on how even the rarest
condition will affect a person. When they go the extra step of claiming to be an expert
in carbon monoxide, however, they loose credibility quickly when the background
training and experience is missing.
Good attorneys do their research. I have seen experts have their careers destroyed
because they bragged about degrees they did not earn and positions they did not hold.
If you have published 100 papers on the topic about which you are going to testify, you
should be prepared to be grilled with questions if even one of those papers is contrary
to your current testimony. In fact, some defense attorneys and plaintiff attorneys
keep databases on experts and share deposition transcripts. Never exaggerate your
background because someone eventually will find it.
30.3.2 SCOPE
You should understand the limits of your role in litigation. In all likelihood your job
is not to provide the entire picture of the puzzle, but rather one portion of it. An
expert’s testimony may or may not include a discussion of how his area of expertise
plays out in the person’s life. Always remember—the further you get away from
the solid footing of what you know, the more likely a skilled opposing counsel will
make you look so silly the jury will discount even the core of your opinion.
Sometimes you may be asked to give the ultimate opinion in a case. You may be
the witness called to testify whether a product is safe, or whether a person is injured.
When you serve in this role, your opinion should be based upon all the information you
can get your hands on. Conversely, you may be asked to testify about a very specific
point. For example, you may be asked to simply explain to the jury how CO can cause
short-term memory loss. Perhaps other witnesses will be asked to testify whether it,
in fact, happened. Your role might simply be to answer the hypothetical question.
In this scenario, you may be able to testify based solely upon your background and
experience without reviewing any portion of the case. Understanding your attorney’s
expectation, and communicating your preferences to that attorney, is the best way
to understand how to prepare for your expert testimony, defend your opinions, and
satisfy your client.
30.3.3 BASING YOUR OPINIONS ON REALITY

The Rules of Evidence require your testimony to have some bearing on the facts of the
case. This means it not only has to be relevant8 but it also has to be based on sufficient
facts or data. Most judges in most jurisdictions will not allow an expert witness to
8 “Relevant evidence” means evidence having any tendency to make the existence of any fact that is of
consequence to the determination of the action more probable or less probable than it would be without
the evidence. F.R.E. 401.
8417: “8417_c030” — 2007/9/11 — 12:13 — page 676 — #6

testify about a medical diagnosis of a person they met 15 min before trial in the
lobby of the courtroom. Your opinions must be based upon information gathered by
first hand observation, facts in evidence, hypothetical scenarios, or other information
which experts in your field reasonably rely upon.9
The most common attack on an expert’s opinion will come in the nature of the
facts upon which it is based. For example, if an expert opinion is based upon the
fact that a person was living in a home with a faulty furnace for two years, it would
be critical to know whether that person actually spent time in the home or had been
stationed overseas in the military for that entire period. Gathering information for the
basis of your opinions is a critical part of the process. It can also be difficult. Not
only is the nature of evidence in a CO case easily destroyed,10 there are financial
considerations as to how much the client is willing to pay for the time it takes you to
research these facts.
Simply accepting someone else’s statement of the facts, although tempting, is
a very thin basis for an expert opinion. Sometimes that is necessary for economic
reasons. At the very least, you should review the documents or interview the wit-
nesses who supplied the facts upon which your opinions are based. In addition, once
you have these facts, you can more specifically inquire about areas that may dis-
count those facts, such as previous treatments, other conditions, and so forth. When
you are asked to render an opinion with limited background information, you should
incorporate the scope of that information into your opinion. That way, if your opin-
ion changes one way or another after the addition of facts, you have not lost any
credibility.
30.3.4 SCIENTIFIC BASIS OF YOUR OPINION—THE METHOD

PREVENTS THE MADNESS
If you have ever worked as an expert, or even talked to anyone who has, you would no
doubt have heard people talk about something called Daubert.11 Daubert is actually
the plaintiff’s name in a Supreme Court case that set forth new standards for expert
testimony in 1995. The history and implications of the Daubert decision and the cases
which followed it are much too lengthy to be discussed in this chapter. However, a
brief understanding of its history and its application today is important to any witness
who may be an expert in a toxic exposure case.
Prior to Daubert, the standard for expert testimony was that scientific evidence
must be “generally accepted” in its field before it could be presented to a jury. Of
course, this prevented the law from ever utilizing the latest technology or cutting
edge research. As a result, the Supreme Court changed the standard from general
acceptance to “scientifically valid.” That ruling has now been incorporated into the
9 Werth v Makita Electric Works Ltd. 950 F2nd 643, 648 (1991 CA 10th ), interpreting F.R.E. 703.
10 See “The Purpose and Process of Litigation in the Carbon Monoxide Case” in this book.
11 Dauber v Dow Pharmaceutical 509 U.S. 579 (1993).
8417: “8417_c030” — 2007/9/11 — 12:13 — page 677 — #7

Rules of Evidence which requires that expert testimony must be “the product of
reliable principles and methods.”12
In addition to broadening those standards, however, the court also instituted what
is now referred to as the “gatekeeper” function of the trial court. The Supreme Court
decided in Daubert, that the trial courts have a duty to analyze the proposed expert
testimony before it is presented to a jury to determine whether it is the product of
reliable principles and methods. Daubert specifically identified four types of data
which may be looked at to determine this validity, but it is not an exhaustive list.
Those factors are the following:
1. Whether it is a result of a reliable and repeatable methodology

2. Whether that methodology was subject to peer review or publication
3. The known or potential error rates and the standards of controlling the
technique’s operation
4. Whether the methodology is generally accepted within the scientific
community
Eventually, this “gatekeeper function” was broadened to apply to even nonscientific

expert testimony.13 Other cases have specified that this gatekeeper function can apply
not only to a particular methodology used by an expert, but also to an entire field of
study such as astrology.14
The federal courts have been very consistent and clear about the trial court’s duty
to act as a “gatekeeper.” Unfortunately, the standards by which the gate should be
kept have been very unclear. If a judge wants to keep evidence out, or allow evidence
to come in, the decision surrounding the Daubert case seems to give him cover either
way. The case law has been very clear that this should only be an analysis of the
methodology, but sometimes it appears to hinge on whether the judge is personally
convinced by the expert’s theories.
There is no clear way to predict how a court will decide a Daubert challenge
and, as a result, you should be prepared to explain the methodology by which you
reached any opinion you hold in a case, and explain the research which proves that
method is valid. Put another way, you need to be scientific about your work. If one
were to analyze the historical basis upon which experts have been excluded from
testimony for failing to meet the standard of “reliable principles and methods,” we
would likely find that typically it is the result of the expert’s failure to explain the
validation of the principles or methods utilized during the Daubert hearing rather
than the actual failure of the principles and methods themselves.
It very well may be, that after 20 years in your chosen field you can recognize
certain situations very quickly, just by looking at them. If you ever testify that the
basis of your opinion was “I could just tell by looking at it,” you will likely not
12 F.R.E. 702(2).
13 Kumho Tire Company v Carmichael 119 SCT 1167 (1999).
14 Id. at 1175.
8417: “8417_c030” — 2007/9/11 — 12:13 — page 678 — #8

be allowed to testify as an expert witness. Instead, you need to explain those things
you immediately recognized, the scientific research which supports the fact that those
factors you observed are important, and then explain how putting those two things
together lead to a conclusion, which other experts could repeat. There may be a
disagreement as to whether or not a particular fact exists, but you should be able
to support your conclusion that the existence of the fact has a scientific basis for
being important. If you utilize a particular protocol, you should not only specify that
protocol in your testimony, but identify each step of the process you used in reaching
your conclusions. This is the best way to convince a judge you will not be wasting
the jury’s time. Ironically, it is also the best way to convince a jury your conclusions
are correct.
30.3.5 PUTTING IT TOGETHER—TESTIFYING ABOUT YOUR

OPINIONS
The strategies for testifying in a deposition or in front of a jury can be different
depending on how your attorney wants to handle the discovery process. However,
your goals should be the following:
1. State your opinion so plainly and clearly that even people who have never
heard of your field can understand your conclusion.
2. Explain the reasons you arrived at your opinion in a manner that walks the
listener through the process and in a way that any other conclusion would
seem unreasonable.
3. State factors which you considered and ruled out only in the context of
how they too support your opinion. If you spend too much time focusing
on factors which were not present, the listener can become confused as to
its relevancy or importance.15
The manner in which you testify is important. You do not want to come across arrogant.
You do not want the jury to feel stupid. You also do not want to come across like
some used car salesman desperate to have them buy what you are selling.
Entire books have been written on what style and technique to use in a courtroom.
Basically, they can all be summarized as this: use good manners. The entire courtroom
setting revolves around the question/answer format. Your answers have to be respons-
ive to the question asked. If you wander off into areas you want to discuss without
being asked, the jury may wonder why you did not answer the question asked, and
the attorney may be frustrated that you are not presenting information in the order he
intended.
15 For example, if you ruled out MS as a cause of the plaintiff’s symptoms, you should not describe MS
in detail without tying it back to the facts of this case. So your testimony should be, “MS can cause some
of these symptoms, but it almost always presents with symptom A, which she does not have. Another
symptom of MS is condition B, and the plaintiff also does not have that. The patient also has symptoms
C, D and E and those are never associated with MS.”
8417: “8417_c030” — 2007/9/11 — 12:13 — page 679 — #9

Some experts address every answer to the jury box. That can appear awkward
because you are not having a conversation with the attorney asking the questions.
At the same time, if the question is prefaced by “explain to us . . . ” you absolutely
should take the opportunity to speak directly to the jury.
You should work with your attorney at length before trial to make sure you both
understand what important points will be covered and the order in which you will
cover them. The best expert testimony will appear as an interesting conversation
between lawyer and witness in which both provide information and explain difficult
concepts.
30.4 THE BATTLE OF EXPERTS

In most toxic exposure cases, you will also encounter an opposing expert in your
field. Typically, your opponent will take one of two tactics: you are wrong or your
opinions are incomplete.
“Wrong” can take many different forms. They may allege you have misapplied
scientific principles. They may claim you did not consider certain important facts.
They may claim you have been fed wrong information. They may even claim you are
being dishonest.
The second strategy is to claim that you are improperly reaching a conclusion
without sufficient information. This is what I refer to as the agnostic opinion. This
testimony boils down to, “I don’t know and neither do you.” Because defendants do
not have the burden of proof, this technique is almost exclusively used by defense
experts. This strategy plays upon the jury’s desire to be sure. Experts employing this
strategy will list off a litany of things that we are not aware of such as wind direction
or other toxins the victim may have been exposed to, or the effects of that car accident
the victim was in 20 years ago, and so forth. Believe it or not, there are people in
this world who make significant incomes traveling around the country saying “I don’t
know.”
Accordingly, part of your role in the case may be to discredit the testimony of
this opposing expert based on the facts of the case and the science of your profession.
Ultimately, however, poking holes in an opposing expert’s testimony/opinion, is the
job of the attorney. So, more specifically, your role may involve educating your
attorney as to what the weakest places are in your opponent’s opinion.
Your credibility is enhanced when you are cool under fire. Of course, this is
easier to do when you are confident in your testimony. If opposing counsel makes
outlandish comments, twists your testimony, or tries other silly tricks, solid science
and civil respect will do far more to expose his weakness than any emotional reaction
or wisecrack ever could. Skilled attorneys can sense when you might lose your cool.
If they see the opportunity, they will be happy to make you unravel on the stand.
Juries will forgive the bad behavior of attorneys much quicker than experts, because
attorneys are not supposed to be neutral.
At the same time, you should not be afraid to agree with opposing counsel when
it is appropriate. Juries assume there is some legitimacy to both sides of the story.
If you refuse to acknowledge any bad fact or concede any point, they may conclude
8417: “8417_c030” — 2007/9/11 — 12:13 — page 680 — #10

the rest of your statements are embellished as well. Quite frankly, your testimony
is highlighted when you can answer, “Yes,” “Yes,” “Yes,” “Yes,” “Absolutely not!”
This technique also puts opposing counsel in the uncomfortable position of choosing
to ask you “Why not?” (who knows what you will say) or moving on (he is afraid to
let you talk—you must be right).
30.5 THINKING LONG TERM

As mentioned above, your testimony is never about one case. If you intend to hold
yourself out as an “expert” to be relied upon and trusted by juries, you must give
“expert” advice. Telling clients what they want to hear will bring far more trouble
than even the worst truth. Stretching the bounds of science to help this one case may
haunt you the rest of your career.
At the end of the day, expert witnesses live and die on their reputation and cred-
ibility. Both should be protected at all costs. This is important because when your
opponent cannot tear apart your opinion, and cannot tear apart what your opinion is
based upon, their only choice is to try to tear apart you. These attacks tend to fall
into certain categories. Unfortunately, these attacks work because sometimes they are
true.
The most common personal attack when you testify is financial. This is particularly
effective in blue collar areas or in communities with depressed economies. The easiest
area of attack is your hourly rate. If a juror has a job which pays $15.00 per hour, he
puts that money in his pocket. Accordingly, there can be an assumption that if you
charge $150 per hour, you put ten times as much money in your pocket. Charging
$500 per hour or more, therefore, might appear to be a way to make you look like
you were “bought.”
Assuming you are charging a fair rate for your services, you should not be embar-
rassed about the number. Embarrassment only adds credibility to the attack. To avoid
that feeling, you should be prepared to explain your hourly rate. Perhaps you should
explain that the fees are paid to a business, not you, and you need to pay other employ-
ees and rent and various expenses out of those fees. The ultimate response is that your
client agreed to your rate before you had any information about the case. Clearly, the
rate had nothing to do with the opinion.
If it continues, the best counterattack to assaults on your fees, however, may be
in the form of a simple economics lesson. Perhaps testifying as an expert witness
is equal, or even less profitable, than the fees you can charge by performing other
professional services. Perhaps the market has proven that your rates are reasonable
because even at your hourly rate, you are offered more work than you could handle.
On the other hand, perhaps you charge a higher hourly rate for expert services to
intentionally make sure it does not end up becoming the only thing you do.
Although it is certainly reasonable for an attorney or a jury to inquire about your
level of compensation for the services provided, everyone, from every economic class,
understands the awkwardness of being forced to talk about your personal finances
in public. If it goes too far or becomes too personal, it is possible to garner some
sympathy from the jury with an appropriate retort. For example, you might find an
8417: “8417_c030” — 2007/9/11 — 12:13 — page 681 — #11

opportunity to say, “I think we have established that I have been successful in my

field. Perhaps we could discuss the reasons for that.” Any counterattack, however,
must be both fair and respectful.
Another common attack is to try to twist your relationship with the client or the
attorney. Again, if this is the 100th time you have testified in a case involving that
attorney or client, there should be a legitimate reason for it. Be prepared to explain
that reason. Do not be embarrassed. Perhaps the client keeps hiring you because you
are the best in your field. Perhaps the client hires you because they trust they will get
an honest assessment of the facts from you instead of what they want to hear. Perhaps
you frequently work together because you are geographically close. The fact that you
have worked with an attorney several times might make you more comfortable to
be frank about your opinions and disagree on the assessment of some cases. Even if
you never disagree with that attorney or client, it may simply be because the attorney
is sophisticated enough to understand when it is necessary to hire an expert in your
field.
Personal attacks only work if they are true. If your opinions are honest and your
fees are reasonable, you should have no problem explaining them in court. If you do
have difficulty coming up with a reasonable explanation, perhaps you need to rethink
how you operate. You must always make sure your decisions are based solely upon
the facts of the case and sound science. Once it is proved that you lack integrity, your
usefulness as an expert is over.
30.6 CONCLUSION
Testifying as an expert witness can be an exciting and fulfilling experience. For
many experts, it is an opportunity to be part of making a difference in how things
are made and used. As with anything professional, however, it is important to keep
your perspective. While you may develop opinions on what the outcome of the case
should be, your role is to provide information and education to a jury. Avoiding the
temptation to change from information source to “advocate” is the key to success and
longevity as an expert witness.
8417: “8417_c030” — 2007/9/11 — 12:13 — page 682 — #12

31 Injury Caused by Carbon
Monoxide Poisoning:
Defining Monetary
Damages
Steve Collard
CONTENTS
31.1 What is Health? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684

31.2 Measuring Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684
31.3 Quantifying Less-Than-Perfect Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 687
31.3.1 Years Lived with Disability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 687
31.3.2 Healthy Life Expectancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 687
31.3.3 Life-years Lost to Injury. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 689
31.4 From the Population to a Person . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 690
31.5 Disability and Major Activity: Working . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 692
31.6 Earning Capacity has Two Parts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 693
31.7 Measuring One Person’s Loss of Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 694
31.8 The Standard Vocational Interview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 695
31.9 The Construct of Work. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 696
31.10 General Educational Development (GED) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 697
31.11 Specific Vocational Preparation (SVP). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 697
31.12 Diminishment of Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 699
31.13 Monetary Damages: The Five Steps in the Analysis of Loss of Earning
Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 699
31.14 Case Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 703
31.14.1 Mr. Jones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 703
31.14.2 Mrs. Betty Jones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 714
31.14.3 John Jones. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 717
31.14.4 Cathy Jones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 721
31.15 Dedication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723
683
8417: “8417_c031” — 2007/9/11 — 12:13 — page 683 — #1

31.1 WHAT IS HEALTH?

The well-being of an individual is a function of longevity and morbidity. Morbidity is
defined as those nonlethal aspects of daily function and pain and suffering that affect
people’s lives.1
A measurement device that uses only mortality rates understates the public health
importance of conditions that result in proportionately more morbidity and disability.
Mortality rates are useful in identifying conditions with the worst health outcomes,
but conditions with low mortality rates may have a high burden of disease because
of morbidity.
Overall health of an individual is a function of self-perceived health, the ability to
function, and existential/experiential symptoms and capabilities (see Figure 31.1).2
Awheelchair bound individual who competes in wheelchair basketball games or mara-
thons would probably have a greater self-perceived health status than a wheelchair
bound individual whose level of overall functioning is lower (see Table 31.1).
Standardizing health measurement tools allows investigators insight in defining
the effects of impairment causing disability on the individual in his or her environment.
31.2 MEASURING HEALTH

In monitoring and determining the health of the population in the United States, in
the early 1900s the leading causes of death were from infectious diseases. Today,
the leading cause of death of persons age 1–44 are from injuries as a result of motor
vehicle accidents.3
The overall health of a population was once measured as a function of the infant
mortality rate. In 1993, Dr. Alan Lopez of the World Health Organization (WHO)
prepared estimates of child-death by cause, which were consistent with the death totals
provided by demographers at the World Bank. Concurrently, a World Bank effort was
preparing consistent mathematical estimates for adult mortality by cause. Ensuring
these consistencies was a major advance, and is a precondition for systematic attempts
to measure disease burden.1
A goal of the WHO has been the gathering of timely and reliable health inform-
ation, because the decisions related to spending billions of dollars annually require
valid health statistics. From the community levels of health services to the national and
international levels, information is gathered to determine the effectiveness of health
strategies. These data help in monitoring the global epidemics, the potential pandemic
H5N1 bird flu, and provide a continuous assessment of public health approaches to
disease and injury prevention and control.
The first nationwide surveillance system to measure the burden of any disease
was implemented in 1950, when malaria surveillance was developed. The second
and third national surveillance systems were implemented in 1955, due to shortages
of the vaccine for polio and in 1957, influenza.
Making a surveillance system that identifies and responds to health problems
has led to the beginning of more scientific study of the morbidity and mortality from
dozens of conditions. It is a challenge, which, given widespread use will lead to better
decisions on spending health dollars that ensures a more equitable future health for
everyone.
8417: “8417_c031” — 2007/9/11 — 12:13 — page 684 — #2

Injury Caused by Carbon Monoxide Poisoning: Defining Monetary Damages 685
Overall health
Existential/
Self-perceived Abilityto experiential
(self-rated) function symptoms and
health capabilities
Pain
Physical function
Energy/vitality
Mobility
Daily activities Emotional
Social/role Happiness/
function mood
With family Depression/

and friends anxiety
Self-image
Major life role
Recreation Sensory
Cognitive Vision
function
Hearing
Memory
Problem
solving
FIGURE 31.1 Concept of overall health.2
In 1988, by the World Bank, the Global Burden study began with Phase 1, called
“Health Sector Priorities Review.” This was begun as an attempt to measure the
significance to public health of individual diseases and what was known about the cost
and effectiveness of relevant interventions for their control. The goals of this research
were to, (1) include nonfatal diseases and injuries in the analysis of international health
policy; (2) to decouple epidemiological assessment from advocacy so that estimates
of the mortality or disability from a condition are developed as objectively as possible;
and (3) to quantify the burden of disease in a manner that allows for cost-effective
8417: “8417_c031” — 2007/9/11 — 12:13 — page 685 — #3

Mr. Smith's life expectancy would have been to age 77.7 had he
not died due to carbon monoxide poisoning at age 50.
Mr. Smith was 50 years old at the time of his death.
Assuming survival, Mr. Smith would have continued working, benefitting society through
his taxes and expenditures, and he would have continued providing services to his family
27.7 years of life lost and friends and chores and tasks for his household. The monetary losses to his estate
exclude his personal consumption, however.
0 10 20 30 40 50 60 70 80 90
FIGURE 31.2 Life expectancy and years of life lost.
analysis.1 Dr. Christopher Murray of Harvard University introduced the term DALY
(disability-adjusted life years) as a common measure of effectiveness for the review.1
The DALY is comprised of two components: The number of years of life lost
(YLL) and years lived with disability (YLD). Years of life lost is the difference in
years of the age at death owing to a specific cause from the cohort’s statistical life
expectancy. A man who dies prematurely at age 50 would lose 27.7 years of expected
life, because his life expectancy at age 50 is 77.7 years.1 (see Figure 31.2)
Years lost to disability calculations take into account the severity of disability
owing to a given disease as well as the average length of time the disability persists.
The level of disability owing to a given condition is revealed as a disability weight.
Disability weights range from 0.0 at perfect health to 1.0 for death.1 Murray and Lopez
systematically reviewed published and unpublished data to estimate the incidence,
prevalence, and duration of 483 disabling sequelae of 107 diseases and injuries.
Internal consistencies were ensured using a software program that identified consistent
parameters.
The severity of disability in the YLD calculations was measured in a deliber-
ate manner. First, panels of health professionals and disease experts in more than
100 diseases or injuries, were drawn from the WHO, the International Agency for
Research in Cancer, the World Bank, the United States Centers for Disease Control
and Prevention, and from numerous academic institutions.
The professionals provided estimates (on the basis of published and unpublished
studies) of the duration of the disease and of incidence, remission, case-fatality,
prevalence, and the death rates. An average handicap was eventually derived for
22 indicator conditions. The different dimensions of a given nonfatal health out-
come can be described as occurring under one of the indicator conditions, like
8417: “8417_c031” — 2007/9/11 — 12:13 — page 686 — #4

physical manifestations (deafness or below-the-knee amputation, for example),

neuropsychiatric manifestations (psychosis, depression), manifestations related to
social or group interaction consequences (vitiligo), or pain manifestations (severe
migraine, angina, or sore throat).1
31.3 QUANTIFYING LESS-THAN-PERFECT HEALTH

31.3.1 YEARS LIVED WITH DISABILITY
The US Department of Health and Human Services (DHHS) publication # April 7,
1995 is titled “Years of Healthy Life.” In this publication, a definition of health and
well-being is as follows:
Symptoms of health and well-being usually involve the assessment of physical and
psychological sensations, such as pain and feelings of anxiety, which are not directly
observable. Physical functioning may be measured in terms of being confined to bed,
couch, or chair due to health reasons, or in terms of health-related limitations in mobility.
Social functioning may be measured in terms of an individual’s limitation in performing
one’s usual social role, whether it is work, housework, or school. Health perceptions are
assessed in terms of subjective evaluations of health and satisfaction with health. Social
opportunity includes resilience and coping and can be measured in terms of social impact
due to health. When symptoms and subjective complaints; mental, physical and social
functioning; general health perceptions; and social opportunity are combined to describe
health, the resulting multidimensional concept is generally referred to as Health-Related
Quality of Life (HRQL).3
When the US DHHS implemented its Healthy People 2000, National Health
Promotion and Disease Prevention Objectives for the Nation, one goal was increasing
the span of healthy life. In 1987, the average life expectancy for the average 51 year
old male was 25.1 years. In 2000, the number of years of life expectancy for same
increased to 27.1 years, and most recent data published in 2006, based on 2003 data
shows the life expectancy increase to 28.5 years for same.4
The DHHS’s Healthy People, 2010 program has two goals: (1) to increase the
quality as well as the years of healthy life (YHL) of the US population; and, (2) to
eliminate health disparities in the US population.
31.3.2 HEALTHY LIFE EXPECTANCY

A table titled “Difference Between Healthy Life and Full Function Healthy Life at
Single Years of Age by Race and Sex, United States, 1996” quantifies in terms of years
the effect of morbidity on life expectancy.5 Researchers at the National Center for
Health Statistics defined 30 health states on a scale of 1.0 (perfect health) to 0.10 (poor
health). The 30 states consist of six stages of health combined with a self-perceived
health scale of five categories. The five levels of self-perceived health are “excellent,”
“very good,” “good,” “fair,” and “poor.” According to the values, a year of perfect
health implies one full year of healthy life. Other health states result in less than one
full year of healthy life.
8417: “8417_c031” — 2007/9/11 — 12:13 — page 687 — #5

The six stages of health are as follows:
1. Not limited • Not limited (includes unknowns) regardless of

age; this category includes unknown role regard-
less of a person’s age
2. Limited in other activities • Limited in other activities regardless of age, or
• Limitation in activity and 65–69 years of age but
able to perform activities of daily living (ADLs)
and able to perform instrumental activities of
daily living (IADLs)
3. Limited in major activity • 64 years of age and younger—limited in amount
or kind of major activity
• 65 years and older—major activity is considered
to be ADL and IADL activities; therefore people
in this age group cannot fall in this category
4. Unable to perform major • 64 years of age and younger—Unable to perform
activity major activity
• 65 years and older—major activity is considered
to be ADL and IADL activities; therefore people
in this age group cannot fall in this category
5. IADL • 0–17 years of age—not applicable. Proxy
respondents were used to obtain this informa-
tion about children; unable to perform their major
activity is the most severe functional limitation to
which they can be assigned
• 18–64 years of age—unable to perform routine
needs without the help of other persons and
unable to perform or limited in major activity
• 65 years of age and older—unable to perform
routine needs without the help of other persons.
6. ADL • 0–4 years of age—not applicable. Proxy respond-
ents were used to obtain this information about
children; unable to perform their major activity
is the most severe functional limitation to which
they can be assigned
• 5–64 years of age—unable to perform personal
care needs without the help of other persons and
unable to perform or limited in major activity
• 65 years of age and older—unable to perform
personal care needs without the help of other
persons
8417: “8417_c031” — 2007/9/11 — 12:13 — page 688 — #6

Using the National Health Interview Survey (NHIS), researchers categorize age
groups, numbers of individuals within the groups, and the overall score for each
health state. The resulting scores, by age groups, can then be applied to a life table
onto the hypothetical cohorts in the life tables. A healthy life expectancy is less, when
measured in terms of years, than a life expectancy, because of the mixture of perfect
health and less-than-perfect health states found in the data.
31.3.3 LIFE-YEARS LOST TO INJURY

The effect of preventive measures on increasing the overall health of our society
is well known, but what is less well known is that the same type of measurement
leading to preventing diseases is being applied by researchers in analyzing motor
vehicle crashes to prevent injuries.
Air bags and bicycle helmets are two examples of such research, the results of
which mathematically can be confirmed as beneficial to society, and that citizens are
probably going to remain healthier overall for using them than for not. The money
being spent to prevent saves money over the long-term and increases postaccident
quality of life outcomes.
When a motor vehicle crash is reported, lots of information is gathered. Even
before the law was passed requiring air bags, researchers with the National Highway
Traffic Safety Administration knew that the devices would reduce loss of life and
reduce the impact of head trauma (the combination of an air bag in addition to a lap
and shoulder belt reduces the risk of serious head injury by 81%, compared with 60%
reduction for belts alone).6
The authors of a study for the National Highway Traffic and Safety Administration
titled “The Economic Impact of Motor Vehicle Crashes 2000” analyzed data from the
National Automotive Sampling System Crashworthiness Data System. The economic
cost of crashes includes costs for medical aspects, emergency services, vocational
rehabilitation, market productivity, household productivity, insurance administra-
tion, workplace costs, legal costs, travel delay, property damage, and psychosocial
impacts.6
A functional capacity index (FCI) was developed to standardize the loss of ability
to perform different life functions, over the long-term, after sustaining injury in a
motor vehicle crash. The FCI scale ranges from zero (being no loss of function) to
one (full loss of function). The FCI measures across major activity (work, school,
or homemaker), instrumental activities of daily living (IADLs) (getting around the
home and community), and activities of daily living (ADLs) (feeding and self-care).
Clinical validation studies confirm the FCI does measure functional outcome.
To measure the long-term cost to society, the FCI for a given injury is multiplied
times the person’s life expectancy, which results in a measure called life-years lost
to injury (LLI). The LLI values for all kinds of injuries can be aggregated over a
statistical population to measure the effect different types of injury have on society
as a whole.
For example, Tom, a 40 year old male, sustains back injury and cannot bend
or lift as a result. The FCI for this injury is 0.49. This means that compared to a
normal, healthy person’s level of function, Tom’s is reduced by 49%. At age 40,
8417: “8417_c031” — 2007/9/11 — 12:13 — page 689 — #7

his life expectancy is 37 years. His resulting LLI is 18 years. Of his potential time
living with unlimited functioning, half of his life functioning now is as a result of
injury.
31.4 FROM THE POPULATION TO A PERSON

Typically, the medical outcome of a patient is expressed in a percentage as an impair-
ment rating. A zero impairment rating implies no limits, and as the impairment
percentage rating goes up, the implication is that the degree of the severity of the
condition is increased, and assumed difficulties in functioning are increased.
Medical doctors routinely provide impairment ratings, but there is a transparent
stretch of logic, usually, with connecting a quantity or diminished quantity of ability
to function from an impairment rating. Occasionally, a functional capacity evaluation
is done, which reveals the effect of the impairment on the person’s ability to perform
a given function, and for the amount of time or tolerance for doing that particular
task.
In 1980, the International classification of impairments, diseases and handicaps
(ICIDH) was developed as a part of a global health measurement system. With this,
the positives and negatives of different medical treatments for similar sequelae could
be more readily analyzed to help improve health care, in consideration to both overall
costs and patient outcomes. After 9 years of international revision efforts coordinated
by the WHO, the World Health Assembly on May 22, 2001, approved the Inter-
national Classification of Functioning, Disability and Health,and its abbreviation
of “ICF.”
The ICIDH uses a linear progression to describe the framework of a nonfatal
health outcome from disease (injury) to pathology to manifestation to impairment
to disability to handicap. However, the new ICF is structured around the following
broad components:
1. Body functions and structure

2. Activities (related to tasks and actions by an individual) and participation
(involvement in a life situation)
3. Additional information on severity and environmental factors
Functioning and disability are viewed as a complex interaction between the health
condition of the individual and the contextual factors of the environment as well as
personal factors. The picture produced by this combination of factors and dimensions
is of “the person in his or her world.” The classification treats these dimensions as
interactive and dynamic rather than linear or static. It allows for an assessment of the
degree of disability, although it is not a measurement instrument. It is applicable to
all people, whatever their health condition. The language of the ICF is neutral as to
etiology, placing the emphasis on function rather than condition or disease.
In Appendix 2 of the ICF Checklist for the examiner, there is a guide to help
when interviewing the respondent to probe about problems in functioning and life
8417: “8417_c031” — 2007/9/11 — 12:13 — page 690 — #8

activities, in terms of the distinction between capacity and performance. This guide
is a good summary of the goal of a standard vocational interview:
The first probe tries to get the respondent to focus on his or her capacity to do a task or
action, and in particular to focus on limitations in capacity that are inherent or intrinsic
features of the person themselves. These limitations should be direct manifestations of
the respondent’s health state, without the assistance. By assistance we mean the help
of another person, or assistance provided by an adapted or specially designed tool or
vehicle, or any form of environmental modification to a room, home, workplace and
so on. The level of capacity should be judged relative to that normally expected of the
person, or the person’s capacity before they acquired their health condition.
The second probe focuses on the respondent’s actual performance of a task or action in
the person’s actual situation or surroundings, and elicits information about the effects
of environmental barriers or facilitators. It is important to emphasize that you are only
interested in the extent of difficulty the respondent has in doing things, assuming that
they want to do them. Not doing something is irrelevant if the person chooses not
to do it.7
The WHO defines disability as any restriction or lack (resulting from an impair-
ment) of ability to perform an activity in the manner or within the range considered
normal for a human being. The Americans with Disability Act adds the condition that
the restriction or lack substantially limits the amount or kind of functioning.
The NHIS defines Impairments as, Chronic or permanent defects, usually static
in nature, that result from disease, injury or congenital malformation. Impairments
represent decrease or loss of ability to perform various functions, particularly those
of the musculoskeletal system and sense organs.
Activity limitations are defined in terms of a person’s ability to perform a major
activity. Major activities are defined differently for different age groups as follows:
1. For children under 5 years old, the major activity is ordinary play.
2. For persons age 5–17, the major activity is attending school.
3. For persons 18–69, the major activity is working or keeping house.
4. For persons 70 and older, the major activity is self-care, without needing
assistance in performance of ADLs or IADLs.
ADLs (self-care) usually include bathing, dressing, toileting, bed or chair trans-
fer, feeding self, getting around the home, and continence. IADLs usually include
performing customary household chores and tasks, handling money, getting around
the community, shopping, using the telephone, and preparing meals.
The NHIS defines Chronic Health Condition as a “condition that a respondent
describes as having persisted for three or more months, or one that is on the NHIS
list of conditions always classified as chronic, no matter how long the person has had
the condition.” An “activity limitation” is defined as “being limited in an activity that
a person would otherwise be expected to perform.” Activity restrictions are defined
in terms of “bed-days,” “school-loss days,” “work-loss days,” and “cut-down days.”
8417: “8417_c031” — 2007/9/11 — 12:13 — page 691 — #9

31.5 DISABILITY AND MAJOR ACTIVITY: WORKING

The Department of Commerce, Bureau of the Census report called Labor Force Status
and Other Characteristics of Persons With a Work Disability, Current Population
Reports, Series P-23, Number 160, defines disability with a similar definition to the
NHIS definition. Persons identified with a disability in the Current Population Survey
(CPS) are defined as those having a health problem or disability which prevents them
from working or which limits the kind or amount of work they can perform.
Interestingly, the health state #3 from the NHIS
(3) Limited in major activity 64 years of age and younger—limited in

amount or kind of major activity.
is very similar to the CPS definition of work disability.

The National Center for Health Statistics defined 30 health states. The derived
health state value for an individual who reports excellent health but is limited in
work is 0.81, with a health state value of 1.00 being equal to excellent health and no
limits. The derived health state value for an individual who reports very good health
but is limited in work is 0.74. The derived health state value for an individual who
reports good health but is limited in work is 0.67. The derived health state value
for an individual who reports fair health but is limited in work is 0.48. The derived
health state value for an individual who reports poor health but is limited in work is
0.34. Since it is these health state values that are applied to the life expectancy data in
computing healthy life expectancy, wouldn’t it be similarly appropriate to consider
applying health state values to work life expectancy?
The CPS is the primary source of information on labor force characteristics of
the population of the United States. The sample is scientifically selected to represent
the civilian noninstitutional population.8 The annual demographic survey, or March
Supplement, provides the primary source of detailed information on the income and
work experience of US workers by disability status, age, sex, and level of educational
attainment.
Two facts derived from the CPS data are (1) Persons meeting the definition of
work disability on average earn less than nondisabled counterparts, regardless of age,
education or gender; and (2) Persons meeting the definition of work-disability on aver-
age report lower participation and employment rates than nondisabled counterparts,
regardless of age, education, or gender.8
In the CPS framework for work-disability, it is acknowledged that some respond-
ents will report impairments, functional limitations, or disabilities in life activities
other than work. This jibes with health state #2 from the NHIS
(2) Limited in other activities Limited in other activities regardless of age, or

Limitation in activity and 65–69 years of age but
able to perform ADLs and able to perform IADLs
8417: “8417_c031” — 2007/9/11 — 12:13 — page 692 — #10

The derived health state values are greater in this category across the board when
compared to those in the #3 health state.
TABLE 31.1
Perceived Health Status
Activity Limitation Excellent Very Good Good Fair Poor
Not limited 1.00 0.92 0.84 0.63 0.47
Limited-other 0.87 0.79 0.72 0.52 0.38
Limited-major 0.81 0.74 0.67 0.48 0.34
Unable-major 0.68 0.62 0.55 0.38 0.25
Limited in IADL 0.57 0.51 0.45 0.29 0.17
Limited in ADL 0.47 0.41 0.36 0.21 0.10
Less-Than-Perfect-Health status can be measured as a burden of a condition on the

individual level and when applicable as a function of reduced labor force participation
and employment rates. Both measurements result in a loss of time, either in years lost
to disability (YLDs) or reduced worklife expectancy.
Disability weights and healthy life expectancy data provide empirical evidence
supporting the observed evidence of the CPS. Work-disabled status is a health state
that is more of a burden on the individual than perfect health and more likely to result
in a shortened worklife. Compelling evidence supports using a diminished worklife
probability in calculations determining loss of future earning capacity.
31.6 EARNING CAPACITY HAS TWO PARTS

Earning capacity represents an individual’s ability or power to earn money. It’s that
one figure that best represents the person’s lifetime ability. It may or may not be
synonymous with actual earnings. In most instances, a mature worker experiences
actual earnings that are congruent with earning capacity. A younger individual or
somebody like a lawyer just passing the bar exam rarely has earning capacity defined
by actual earnings. In such circumstance, a proxy, or substitute, representing the
individual’s earning capacity can be used.
Earning capacity is more commonly reduced, rather than destroyed, as a function
of occupational disability. The occupationally disabled person’s age, education, pre-
vious work history, skill level, general learning ability, and severity of impairment
combine to produce either a destruction or reduction of earning capacity.
On the back of the Kool-Aid packet, the instructions state you’ll need a two-quart
capacity pitcher. Earning capacity, conceptually, is similar to the Kool-Aid pitcher-
picture, in that the amount of Kool-Aid in the pitcher doesn’t define the capacity of
the pitcher, unless it is filled to the top.9 A mature worker like a 57-year-old legal
assistant with 17 years in the same firm would probably have actual earnings that
were representative of capacity.
8417: “8417_c031” — 2007/9/11 — 12:13 — page 693 — #11

When occupational disability puts a crack in the Kool-Aid pitcher, the loss of
capacity has two parts. The earning capacity on an annual basis is only one factor.
The other issue that must be considered when defining loss of lifetime capacity to work
and to earn money is called “worklife expectancy.” Worklife expectancy addresses
the issues of being alive in the future, being a labor force participant in the future and
being employed in the future.
The government has annually collected data on the employment potential of dis-
abled versus nondisabled persons, all other factors held constant, since 1981 through
the CPS. The name of the study is The Labor Force Status and Other Characteristics
of Individuals by Age, Education, Sex, and Disability status.
The CPS provides the pulse-beat of who’s working in America by demographic
characteristic. Employment rates are routinely stated on radio/TV news and in print
media. In addition, studies and data which have corroborated these data are found in
the Decennial Census and in the American Community Surveys. The various surveys
early on had slightly different definitions of disability and sought different answers,
but a recent trend has standardized the definitions. Quantifying less than perfect health
with work disability status is an evolving construct and measurements should always
be improved accordingly.
The CPS time series data reveal that, just like nondisabled males with less than
a high school diploma have a lower employment rate than nondisabled males with a
high school diploma, who have lower employment rates than those college educated,
the disabled, across the board, are less-frequent participants in the labor force and
are less likely to find employment than the nondisabled counterpart, and have lower
employment rates.
While econometrical models which project future work life expectancy, like
the Increment-Decrement Model, have an element of statistical validity, two-
state (employed and unemployed) and three state (employed, unemployed and
inactive) models fail to capture the diminishment of functioning over time as
experienced by disabled workers. As the time-series CPS data show, those per-
sons with disability departing the labor force at a greater rate than nondisabled
counterparts, the increment–decrement models would reveal the same future work
life for a disabled person if he or she were actually working at the time of
measurement.
31.7 MEASURING ONE PERSON’S LOSS OF CAPACITY

As Dr. Murray writes in the Global Burden of Disease, “. . . we believe it is preferable
to make an informed estimate of disability due to a particular condition than to have
no estimate at all. The absence of an estimate fosters the tacit assumption that there
is no problem. For example, it may well be that the continued neglect of primary
and secondary prevention and rehabilitation of disability is related to the lack of data
on its magnitude, especially when compared to the information available about life
lost due to premature mortality. Ex Cathedra statements without supporting empirical
evidence do not contribute constructively to informed policy debate, and should be
avoided.”1
8417: “8417_c031” — 2007/9/11 — 12:13 — page 694 — #12

31.8 THE STANDARD VOCATIONAL INTERVIEW

Trying to quantify the effects of impairment on one person’s ability to work and
earn money in the future—in other words, trying to measure something that hasn’t
happened—still can be an informed, reliable estimate. Using a standard vocational
interview, a vocational-rehabilitation professional can reveal an individual’s demo-
graphic characteristics in his or her environment, as they existed before becoming
impaired and with impairment.
Basic demographic information of the client, like date of birth, race, gender, level
of educational attainment, work history, skills and abilities are collected in the stand-
ard vocational interview, allowing the analyst to categorize the client. An important
consideration is prior health status, or the existence of preexisting condition(s) which
may or may not have limited the client’s ability to work and earn money.
Investigating the client’s life through four domains, social, occupational, practical,
and emotional-psychological in their preinjury and postinjury lives provides relevant
information to the effect of the given impairments on this individual’s ability to
function, in his or her environment.
The SF-36 Health Survey was constructed for the Medical Outcomes Study to
survey health status, and has been tested and validated extensively. Designed for use
in clinical practice and research, health policy evaluations, and general population
surveys, the SF-36 includes one multi-item scale that assesses eight health concepts:
(1) limitations in physical activities because of health problems; (2) limitations in
social activities because of physical or emotional problems; (3) limitations in usual
role activities because of physical health problems; (4) bodily pain; (5) general mental
health (psychological distress and well-being); (6) limitations in usual role activities
because of emotional problems; (7) vitality (energy and fatigue); and (8) general
health perceptions (see Table 31.2).
The SF-36 is only a guide for the interviewer to delve into the client’s world
through his or her social functioning, emotional/psychological functioning, prac-
tical functioning (household tasks, ADLs and IADLs), and occupational functioning.
Questions are slightly modified in terms of functioning preinjury and postinjury.
TABLE 31.2
An SF-36 Data Set
Scales Mean SD
Limitations in physical activities because of health problems 70.61 27.42
Limitations in usual role activities because of physical health problems 52.97 40.76
Limitations in social activities because of physical or emotional problems 55.78 40.71
Energy/Fatigue/Vitality 52.15 22.39
Emotional Well-being 70.38 21.97
Social Functioning 78.77 25.43
Pain 70.77 25.46
General Health 56.99 21.11
8417: “8417_c031” — 2007/9/11 — 12:13 — page 695 — #13

The first question of the SF-36 asks the same five-part self-perceived health ques-
tion used in the standardized definition of overall health. The interviewer, however,
needs to know the change in perceived health status, and would ask the client this
question first in the context of pre-incident, and then ask about current perception,
with impairments:
1. “In general, you would say that your overall health, prior to being exposed
to CO poisoning, was either “excellent,” “very good,” “good,” “fair,” or
“poor”?
2. “Now, as a result of being exposed to CO poisoning, how would you rate
your overall health, “excellent,” “very good,” “good,” “fair,” or “poor”?
31.9 THE CONSTRUCT OF WORK

Work is organized in a variety of ways. As a result of technological, economic,
and sociological influences, nearly every job in the economy is performed slightly
differently from any other job. Every job is also similar to a number of other
jobs.
In order to look at the millions of jobs in the US economy in an organized way,
the Dictionary of Occupational Titles (DOT) groups jobs into “occupations” based
on their similarities and defines the structure and content of all listed occupations.
Occupational definitions are the result of comprehensive studies of how similar jobs
are performed in establishments across the nation and are composites of data collected
from diverse sources. The term “occupation,” as used in the DOT, refers to this
collective description of a number of individual jobs performed, with minor variations,
in many establishments.10
Worker functions are described as the ways in which a job requires a worker to
function in relationship to data, people, and things, as expressed by mental, inter-
personal, and physical worker actions. In the DOT, every job is assigned the three
worker functions that best characterize the worker’s primary involvement with data,
persons, and things.
Data functions are an arrangement of different kinds of activities involving inform-
ation, knowledge, or concepts. Some are broad in scope and some are narrow in
scope. Components of data are synthesizing, coordinating, analyzing, compiling,
computing, copying, and comparing.
People functions are activities that have little or no hierarchical arrangement bey-
ond the generalization that the least level is taking instructions/helping. Components
of people are mentoring, negotiating, instructing, supervising, diverting, persuading,
speaking-signaling, serving, and taking instructions-helping.
Things functions can be divided into relationships on the basis of the worker’s
involvement with either machines or equipment. Components of things are setting up,
precision working, operating-controlling, driving-operating, manipulating, tending,
feeding-off bearing, and handling.
From the general guide of an interview, though, a more complete picture needs
to be coordinated with the “worker characteristics components” applied to the more
8417: “8417_c031” — 2007/9/11 — 12:13 — page 696 — #14

than 12,000 job titles in the DOT. The Selected Characteristics of Occupations (SCO)
is a companion volume to the US Department of Labor’s DOT. Worker characteristics
include job analysis components which reflect worker attributes needed to contribute
to successful job performance. The seven worker characteristic components are as
follows:
• General Educational Development (GED)

• Specific Vocational Preparation (SVP)
• Strength
• Physical demands
• Environmental conditions
• Temperaments
• Aptitudes
31.10 GENERAL EDUCATIONAL DEVELOPMENT (GED)

This has three divisions, each with six levels:
Reasoning development (applying common-sense understanding at the lowest
level to applying logical or scientific thinking to a wide range of intellectual prob-
lems at the uppermost level), mathematical development, and language development.
The six levels of Mathematical and Language Development are based on school cur-
riculum taught in the United States, and the higher the assigned value is, equates to
a higher level of education.
31.11 SPECIFIC VOCATIONAL PREPARATION (SVP)

This is defined as the amount of time required by the typical worker to learn the
techniques, acquire the information, and develop the facility needed for average
performance in a specific job-worker situation. SVP includes vocational educa-
tion, apprenticeship training, in-plant training, on-the-job training, and essential
experience from other jobs on the career ladder. The 9-level scale of SVP is as
follows:
SVP 1: Short demonstration only

SVP 2: Anything beyond a short demonstration up to and including 1 month
SVP 3: Over 1 month up to and including 3 months
SVP 4: Over 3 months up to and including 6 months
SVP 5: Over 6 months up to and including 1 year
SVP 6: Over 1 year up to and including 2 years
SVP 7: Over 2 years up to and including 4 years
SVP 8: Over 4 years up to and including 10 years
SVP 9: Over 10 years
8417: “8417_c031” — 2007/9/11 — 12:13 — page 697 — #15

The strength needed to perform a job is expressed by one of five terms over the
period of the average work-day needed (constantly, frequently, occasionally, or not
present).
Sedentary Work, for example, requires that a worker be capable of exerting up to
10 pounds of force occasionally (activity exists up to 1/3 of the time) or a negligible
amount of force frequently, to lift, carry, push, pull, or otherwise move objects,
including the human body. Sedentary work involves sitting most of the time, but may
involve walking or standing for brief periods of time. Jobs are sedentary if walking
or standing is required only occasionally and all other sedentary criteria are met.
Light Work, requires that a worker be capable of exerting up to 20 pounds of force
occasionally, or up to 10 pounds of force frequently, (activity exists from 1/3 to 2/3 of
the time) or a negligible amount of force constantly to move objects. Even though the
weight lifted may be only a negligible amount, a job should be rated “light work”:
• When it requires walking or standing to a significant degree

• When it requires sitting most of the time but entails pushing or pulling of
arm or leg controls
• When the job requires working a production rate pace entailing the constant
pushing pulling of materials even though the weight of those materials is
negligible
Medium Work requires that a worker be capable of exerting 20–50 pounds of force
occasionally, or 10–25 pounds of force frequently, or greater than negligible up to 10
pounds of force constantly (activity exists 2/3 or more of the time) to move objects.
Heavy Work requires a frequent lift, carry, push, pull of weight up to 50 pounds
with an occasional lift, carry, push, pull of weight up to 100 pounds.
Very Heavy Work requires exerting with lift, carry, push, or pull of weight in
excess of 100 pounds occasionally, or in excess of 50 pounds frequently, or in excess
of 20 pounds constantly.
Physical demands analysis is a systematic way of describing the physical activities
the job requires, over the period of the average work-day the activity is needed (con-
stantly, frequently, occasionally, or not present). The physical demands in addition to
strength are climbing, stooping, kneeling, crouching, crawling, reaching, handling,
fingering, feeling, talking, hearing, tasting, and smelling, near acuity, far acuity, depth
perception, accommodation, color vision, and field of vision.
Environmental conditions are the surroundings in which a job is performed and
consists of fourteen factors over the period of an average work-day needed (constantly,
frequently, occasionally, or not present). It involves exposure to weather, extreme
cold, extreme heat, wet and humid, noise intensity level (five levels), vibration,
atmospheric conditions, proximity to moving parts, exposure to electrical shock, and
working in high, exposed places.
Temperaments are defined as the “personal traits” or the adaptability require-
ments made on the worker by a specific job situation. Eleven temperaments are
A—working alone or in isolation from others; D—directing, controlling and plan-
ning activities of others; E—expressing personal feelings; I—influencing people
in their opinions, attitudes, and judgments; J—making judgments and decisions;
8417: “8417_c031” — 2007/9/11 — 12:13 — page 698 — #16

P—dealing with people; R—performing repetitive and continuous short-cycle work;

S—performing effectively under stress; T—attaining precise set limits, tolerances,
and standards; U—working under specific instructions; and V—performing a variety
of tasks.
A Standard vocational interview reveals the client in his or her environment.
Utilizing the DOT and the SCO, the analyst is able to quantify with a reasonable
degree of certainty that individual’s capacity to work. Within this framework, the
analyst can ascertain the diminution of one individual’s capacity.
31.12 DIMINISHMENT OF FUNCTION

Being limited in the amount or kind of work can usually be (but not always) observed
within the construct of work delineated in the DOT. Occasionally, though, the standard
vocational interview will reveal a client who has not missed a day of work, but
it is revealed that the other three life domains are substantially impacted. It’s not
reasonable to assume that a person with disability, who is giving all his life’s energy
to occupational functioning, while his life’s energy toward functioning in the other
three domains is diminished, will enjoy the same longevity in the labor force he
potentially once had.
While the client’s self-reported limitations provide a general overview for the
analyst’s consideration, when combined with the medical provider’s, a more specific
overview is derived. In physical injury situations, a medical provider like a physician
with a specialty in rehabilitation medicine will assess and report through a functional
capacity evaluation of the patient’s diminishment of functioning from preinjury levels,
and will often opine on future medical care along with known outcomes from research
which indicate the client will probably suffer exacerbations that will cause need for
future interventions. These exacerbations and subsequent medical treatments will
probably interrupt functioning in the four domains. In brain injury situations, a com-
prehensive neuropsychological evaluation can reveal specific losses which are easily
correlated to worker functions described in the DOT. Frequently, the impact of injury
affects physical, cognitive, and emotional-psychological functioning.
As the individual describes a lower level of overall health due to loss of ability to
function, along with pain, cognitive dysfunction or emotional-psychological distress,
the normal effects of a human’s aging body will come abnormally sooner on a damaged
body which is less prepared for the diminishment.
31.13 MONETARY DAMAGES: THE FIVE STEPS IN THE

ANALYSIS OF LOSS OF EARNING CAPACITY
Just as an attorney uses statutes as a framework for a given client’s circumstance, the
analyst assessing loss of future earning capacity uses a five-step framework. The Five
Steps are
Step I: Preinjury earning capacity

Step II: Preinjury work life expectancy
8417: “8417_c031” — 2007/9/11 — 12:13 — page 699 — #17

Step III: Postinjury earning capacity

Step IV: Postinjury work life expectancy
Step V: Present value calculation
Step I, the preinjury earning capacity, is the “size of the individual’s Kool-aid
pitcher.” It is that dollar amount that best represents that person’s lifetime ability to
earn money. It may or may not be the same as the person’s actual earnings. Typically,
a mature worker in his mid-to-late 40s will be earning at capacity, but rarely will a
younger worker’s actual earnings represent lifetime ability.
A person’s earning capacity is a function of educational attainment, intelligence,
skill, gender, and disability status. When actual earnings do not fairly represent
earning capacity for the person, a substitution, or proxy can be used.
Numerous data sources are available which reveal average annual earnings of
persons by demographic characteristics or by job title. Computer software exists
which can sort job titles from the DOT by skill level and intelligence level. The earning
capacity of a child with average intelligence, for example, could be predicated upon
the average annual earnings of similar persons with average intelligence.
The preinjury worklife expectancy, Step II, is the number of years into the future
we would reasonably expect the person to be alive, participating in the labor force, and
employed. Consideration must be given to the person’s disability status in calculating
preinjury worklife expectancy.
The US Bureau of the Census, under the US Department of Commerce, collects
population-related data. The Bureau of the Census’ CPS is conducted monthly. Since
1981, the March Supplement to the CPS has been collecting disability-related data.
This information can be reviewed at the Department of Commerce’s Internet address
HTTP://STATS.BLS.GOV/CPSAATAB.HTM. March surveys are a combination of
the basic CPS and a Supplement which delves into work status and disability. Surveys
are conducted by interviewers in approximately 57,000 households throughout over
700 sample areas which comprise nearly 2,000 counties, independent cities, and
minor civil divisions. The sample is continually updated to account for new residential
construction.
The Labor Force Status and Other Characteristics of Persons With a Work-
Disability: 1981–1988, Series P-23, No. 160, issued in July of 1989 notes the basic
concept of disability and the relationship of the basic concept to the operational
concept adopted for the March Surveys. A person has a disability if he or she has
a limitation in the ability to perform one or more of the life activities expected of
an individual within a social environment. The primary way for this concept to be
operationalized in the March CPS is to ask whether any household member has a
health problem or disability which prevents them from working or which limits the
kind or amount of work they can do.
Some of the persons who do not have a work disability do have impairments,
functional limitations, or disabilities in life-activities other than work. The term
“impairment” indicates a physiological, anatomical, cognitive loss or abnormal-
ity. The term “functional limitation” indicates a restriction in a physical functional
activity (such as walking, reaching, or hearing), an emotional functional activity
8417: “8417_c031” — 2007/9/11 — 12:13 — page 700 — #18

(maintaining satisfactory personal relationships), or a cognitive functional activity

(solving problems). Persons with a given level of functional limitation may or may
not have a work disability, depending on the individual environment and the reaction
of the individual.
An individual is considered to have a work-related disability if one or more of the
following conditions are met:
1. Identified by the question, “Does anyone in this household have a health

problem or disability which prevents them from working or which limits
the kind or amount of work they can do?”
2. Identified by a question that asks, “Is there anyone in this household who
ever retired or left a job for health reasons?”
3. Did not work in the survey week because of a long-term illness or disability
which prevents performance of any kind of work?”
4. Did not work at all in previous year because ill or disabled.
5. Under 65 years of age and covered by Medicare.
6. Under 65 years of age and a recipient of Supplemental Security Income.”
The data are represented in three categories (1) With no work disability; (2) With
a work disability; and (3) With a severe disability. The categories are broken into age
groups in ten-year increments from age 16 up to age 74 for both males and females,
and by level of educational attainment (in four subgroups: less than a high school
diploma, high school diploma, 13–15 years of education, and 16 or more years of
education).
What the data show are the participation in the labor force rates and employment
rates by age, sex, education, and disability or nondisability status. Employment rates
are calculated relative to those who are considered labor force participants (those
who are either employed or actively seeking employment). Another nationally known
household survey, the Survey of Income and Program Participation, has a compilation
of employment rates by type and degree of disability status, but the population of this
group is not representative of the overall general population. The decennial census and
the American Community Survey additionally show diminished employment rates in
different domains owing to the effect of disability.
As noted above, category (2), those with a work disability, includes all six (1 to 6)
conditions. Inspection of the data shows it is heavily weighted by the low participation
rates for those in the labor force described in conditions “3” through “6” which
comprise the data in category (3), with a severe disability. Analysis of the data in
category (2) that remains after extracting the category (3) data reveals a subset of
persons considered as category (4), with a moderate work disability.
When considering future working status for all persons, the certainty of life should
be reduced by the measurable probability of surviving. The chance a person could die
is calculated using US DHHS, National Center for Health Statistics, Vital Statistics
of the US Life Tables.
A worklife expectancy considers the probability in the future of surviving, along
with the probability of being a labor force participant, and the probability of being
employed. A worklife expectancy for an individual in category (4), with a moderate
8417: “8417_c031” — 2007/9/11 — 12:13 — page 701 — #19

work disability is less than a work life expectancy for a nondisabled cohort, regardless
of that person’s age, sex, or years of education. The effects of work disability are
greatest on those in the population with the least education, workers with less than a
high school diploma.
Two important facts exist for persons who meet the definition of occupational
disability: On average, those persons earn less than nondisabled counterparts, and on
average, those persons have lower participation and employment rates, and therefore
have a reduced worklife expectancy.
Step III, the postinjury earning capacity is typically, but not always, less than
preinjury earning capacity. It could be congruent with preinjury capacity, but will
never exceed it. If an individual with disability benefits with retraining which
increases his postinjury earning capacity, clearly that individual had the capacity
prior to injury to retrain as well. The difference is that the newly retrained person
with disability could have been retrained before, without the burden of disability,
and the analyst would now have to question if the disabled person with retrain-
ing would reach the compensation level of the nondisabled counterpart in the same
occupation.
Step IV, the postinjury worklife expectancy, like Step II, is the number of years
into the future we would reasonably expect the person to be alive, participating in
the labor force and employed. The CPS data would be used in conjunction with
the analyst’s observation of the effect of the impairment on that individual in that
individual’s environment to define the appropriate category of disability, moderately
disabled, average disabled, or severely disabled. In some cases, the analyst might
employ a continuum of employability as an opinion that the given individual in his
or her specific environment might be somewhat better than the average for a given
category.
Step V, the present value calculation, considers the amount of money needed today,
which in a prudent investment will replace the loss over time. Typically, courts of law
require the analyst to reduce the loss to present value. This language, however, refers
to the algebraic calculation of simplifying (reducing) a polynomial equation, and
doesn’t necessarily imply that the present value will be less than the actual summation
of the loss. Rather, the present value of a growing lump sum considers both future
growth in relation to future interest accruing on the components, and the resulting
figure could be equal to, greater, or less than the summed figure. In calculating the
present value of a growing lump sum, the formula

PV = CF × {(1 + GROWTH RATE)/(1 + INTEREST RATE)}n
where CF represents the annual loss, cash flow, and n represents the exponent
of time.
As empirical evidence doesn’t relate the variables, from this formula three assump-
tions can be made as follows: (1) if it is assumed the rate of growth of wages will be
greater than the assumed interest rate, the numerator is greater than the denominator,
and the PV will require more money being invested today; (2) if it is assumed the
denominator is greater than the numerator, the PV will require less money being
8417: “8417_c031” — 2007/9/11 — 12:13 — page 702 — #20

invested today, and (3) if it is assumed that the growth rate equals the interest rate,
one recognizes a number divided by itself is 1, and the present value is equal to the
summation of the loss (also known as a total offset assumption).
31.14 CASE REPORT

The Jones Family: Mr. Jones and his wife Betty, and their two children, John and
Cathy
Dear Mr. Attorney:
You have asked me to perform analyses to measure the monetary losses, including
loss of future earning capacity, for Mr. and Mrs. Jones, and their two minor children’s
loss of adult future earning capacity, if any, as a result of impairments resulting
from chronic carbon monoxide (CO) poisoning in 2005. In conducting the analyses,
I interviewed the family on August 18 and 19, 2006. In addition, I reviewed the
following information forwarded to me by your office:
• Complaint
• Plaintiff’s answers to interrogatories
• Deposition transcripts of Mr. and Mrs. Jones
• Mr. Jones’ income tax returns for 2001–2005
• John’s school records from middle school
• Cathy’s school records from elementary school
• Medical records from rehabilitation hospital
• Medical records from hospital
• Medical records from Dr. Maher
• Medical records from Dr. Paul
• Medical records from Dr. Gelbman
• Neuropsychological evaluation reports from R. Hoffman; PhD
31.14.1 MR. JONES

The interview reveals the following for Mr. Jones:
Mr. Jones’s date of birth is August 18, 1969. He is 37 years old. Mr. Jones was
graduated from high school and attained a master of business administration degree
from the University. Over his worklife, he has functioned as an investment analyst
(DOT # 160.267-026), and had been promoted to partner, and his responsibilities
included supervising 15 analysts and their assistants.
This work is skilled (SVP = 8) in nature and sedentary in terms of physical
demands, and requires a constant physical demand of near acuity (clarity of vision
at 20 inches or less). The GED for reasoning, mathematics, and language are all at
level 5, the second highest level. To perform the job successfully, the worker must
possess above average aptitudes of (intelligence), verbal aptitude (the ability to under-
stand meanings of words and ideas), numerical aptitude (the ability to perform basic
arithmetic operations quickly and accurately), and clerical perception (the ability to
perceive pertinent detail in verbal or tabular material).
8417: “8417_c031” — 2007/9/11 — 12:13 — page 703 — #21

His generalized work activities included the following12 :
• Analyzing data or information—Identifying the underlying principles,

reasons, or facts of information by breaking down information or data
into separate parts.
• Getting information—Observing, receiving, and otherwise obtaining
information from all relevant sources.
• Interacting with computers—Using computers and computer systems
(including hardware and software) to program, write software, set up
functions, enter data, or process information.
• Processing information—Compiling, coding, categorizing, calculating,
tabulating, auditing, or verifying information or data.
• Communicating with supervisors, peers, or subordinates—Providing
information to supervisors, coworkers, and subordinates by telephone, in
written form, e-mail, or in person.
Mr. Jones was a cub scout leader for his son, John’s group. They enjoyed camping
and backpacking, and would frequently take their ski boat on camping trips to state
parks outside of scout functions with Betty and Cathy. Mr. Jones and Betty enjoyed
twice a year-long-weekend trips to Las Vegas, without the children.
Mr. Jones performed all outside yard and car care work, did home repairs, kept
the gutters clean, and maintained a vegetable and herb garden.
In December 2004, the Jones family moved into a new home with natural gas
fueling the forced air furnace and water heater. By spring, 2005, the entire family
felt as if they had prolonged flu. Mr. Jones found it difficult to focus at work with his
constant headaches. He began having difficulty following through on projects, and
became easily frustrated with his coworkers and supervisors. His vision was blurry
and he had difficulty reviewing the reams of paperwork he once easily reviewed and
analyzed.
He fell from the roof after a dizzy spell while cleaning the gutters, and fortunately,
the shrubs below broke his fall enough to prevent serious injury from occurring. The
family physician, Dr. Maher, suspected something besides the flu was affecting the
family at this point, as all had similar complaints of lethargy, fatigue, headaches,
cognitive dysfunction, and emotional outbursts, when none of this existed before
they moved.
After a thorough home inspection, an improperly installed ventilation system
on the furnace was discovered, which was leaking directly beneath Mr. and Mrs.
Jones’ bedroom. Dr. Maher referred Mr. Jones to a toxicologist, Dr. Smith, and a
neurologist, Dr. Paul. The latter found Mr. Jones’ EEG (Electroencephalogram) and
QEEG (Quantitative Electroencephalogram) to be abnormal. Dr. Smith referred the
Jones family to R. Hoffman, Ph.D., for neuropsychological testing.
Brain injury involves a change in the structure, physiology, and chemistry of the
brain which is caused directly or indirectly by physical, chemical or force insult to
brain tissue which results in a change in the way in which the brain acquires and
processes information, and produces command signals.
8417: “8417_c031” — 2007/9/11 — 12:13 — page 704 — #22

The results of testing for Mr. Jones revealed problems with word finding, memory,
sequencing, multitrack thinking (i.e., multitasking) and differentiation abilities. The
report continues with Mr. Jones’s attention and concentration abilities assessed at a
level rated impaired, and memory and executive functioning in the impaired range. He
also has deficits in abstract reasoning and in systematic thinking. Mr. Jones’s scores
in the language domain were average.
Executive function refers to the human brain’s ability to be aware of its
environment, following through with tasks, self-motivation, self-correction and self-
initiation. Multitrack thinking and the speed at which thoughts are processed are
considered executive function. Components of executive functioning include:
• Word-finding (finding the word you want and speaking it in a reasonable

length of time)
• Sorting (the brain’s ability to separate truth from projection and its ability
to retrieve information)
• Discernment (the ability to be tactful or to keep social behavior appropriate)
and details (the ability to notice small differences)
• Sequencing (ability to perform tasks in a logical or productive progression)
• Follow-through (ability to take an event all the way through to completion)
• Multitrack thinking and differentiation (doing more than one thing at
a time, thinking more than one thought at a time, or juggling sev-
eral different thoughts or activities simultaneously, as required in adult
thinking)
Data functions are an arrangement of different kinds of activities involving inform-

ation, knowledge or concepts. In Mr. Jones’s job as a supervisor/investment analyst,
his data functioning involved both coordinating and analyzing prior to injury. In
coordinating, one determines time, place and/or sequence of operations or activities
on the basis of analysis of data; executing determinations or reporting on events. As
the single point of contact for responsibility on a job, Mr. Jones wrote reports, met
with supervisors and customers, and managed a team of coworkers and their assist-
ants. In analyzing, one examines and evaluates data, and frequently presents alternate
actions in relation to the evaluation. Because of his cognitive impairments, Mr. Jones
has difficulty coordinating and analyzing at his prior level.
As per people functioning, Mr. Jones’ level was supervising. Supervising is
determining or interpreting work procedures for a group of workers, assigning spe-
cific duties to them, maintaining harmonious relations among them, and promoting
efficiency. A variety of responsibilities are involved, such as training workers, eval-
uating workers’ performance, assisting workers in solving work problems, initiating
and recommending personnel actions like hiring, firing, promoting, transferring or
disciplining; and maintaining records of performance of these duties.
Mr. Jones had to take advantage of short-term disability insurance. Eventually, he
returned to working but at a much reduced level of functioning, performing at only a
small percentage of his former ability, and at a lower rate of pay. Because of his prior
success with the company (and a benevolent employer), a job was created especially
8417: “8417_c031” — 2007/9/11 — 12:13 — page 705 — #23

for him that entailed only compiling information (data), and helping (people), and
involved limited contact with coworkers he formerly supervised.
Mr. Jones was administered the SF-36. He reports his current overall health as
“fair.” He reports being limited a lot with vigorous activities, limited a little with
moderate activities, and physical functioning causes his bodily pain to increase. He
reports not accomplishing as much as he would have liked to, cut-down days, and
missing numerous days from work when his headaches were severe.
Mr. Jones’s life expectancy at age 37 is 40.0 years. His healthy life expectancy at
age 37 is 35.9 years, but this does not consider the 19.2 years diminution of health he
sustained as a result of CO poisoning. Prior to injury, Mr. Jones says that he enjoyed
excellent overall health, and did not have a condition or impairment that limited the
amount or kind of work he could do (see Figure 31.3).
Mr. Jones’ social functioning is impacted by his chronic fatigue to the extent he
cannot participate in the many activities he once enjoyed. He is trying with all his
energy to function at work so he can keep a roof over the family. He knows, also, that
the frequency and the type of leisure the family once enjoyed will not be as affordable,
and this, too, is adding to his clinical depression.
Emotional and psychological functioning, occupational functioning, practical
functioning, and social functioning are the four life domains. Mr. Jones’ health-
related quality of life is significantly below average, overall, as shown below through
the results on the eight scales of his SF-36 (see Table 31.3).
Although nonpecuniary damages like “pain and suffering,” and “loss of pleasure
of life (hedonic damages)” have not traditionally been allowed in expert testimony,
it seems that as disability becomes more quantifiable, courts will become more apt to
consider that in life-years-lost to disability, time living with disability includes time
not working as having a value.
An assumed worklife to age 67 is 30 years
13.9 years of worklife for the average

disabled cohort
21.53 years of worklife probability for the average

moderately disabled cohort
26.48 years of worklife probability for the average nondisabled cohort
0 5 10 15 20 25 30 35
FIGURE 31.3 Mr. Jones worklife expectancy.
8417: “8417_c031” — 2007/9/11 — 12:13 — page 706 — #24

TABLE 31.3
Mr. Jones’ SF-36 Data Set
Scales Mean SD
Limitations in physical activities because of health 38.99 70.61 27.42
problems
Limitations in usual role activities because of physical 0 52.97 40.76
health problems
Limitations in social activities because of physical or 0 55.78 40.71
emotional problems
Energy/Fatigue/Vitality 10 52.15 22.39
Emotional Well-being 52 70.38 21.97
Social Functioning 25 78.77 25.43
Pain 55 70.77 25.46
General Health 40 56.99 21.11
Two controversial aspects of hedonic damages calculations are the monetary value
of a statistical human life, and the value of the life to the individual. What is the value
of a statistical human life? Is life valued greater when an individual is young and
reckless and has freedom to play and experiment, or when the individual is closer to
the end of life, and each day has unique preciousness? Could an average jury decide
the value of a statistical human’s life, and use the percentage of life-years lost to
disability as the time metric?
Different methods to quantify the value of a statistical life used by experts
in Hedonic damages exist. When the value of a statistical life becomes main-
stream/accepted, risk companies’actuaries will be better able to project funding needs
for statistically projected losses. Until then, we will focus on the pecuniary damages.
The five steps in analyzing Mr. Jones’loss of future earning capacity are as follows:
Step 1, his preinjury earning capacity is reasonably represented by his actual
earnings for 2005, $98,000 (For simplicity’s sake, fringe benefits are not considered
as part of Mr. Jones’ total compensation/earning capacity);
Step 2, Mr. Jones’ preinjury worklife expectancy is like a non-disabled male’s
with a college degree, reduced by the likelihood of living and working from age 37
through age 74, or 26.48 years;
Step 3, his postinjury earning capacity is reasonably represented by his annual
rate of earnings in his new position, $55,000;
Step 4, Mr. Jones’ postinjury worklife expectancy, at best, given the employer
accommodations, is like an average moderately disabled male’s with a college degree,
reduced by the likelihood of living and working from age 37 through age 74, or 21.53
years; and,
Step 5, the present value calculation reveals a lifetime loss of future earning
capacity in a range of $1,102,432–$1,410,772, using a range of an assumed net
discount rate of 1.6% to a total offset assumption.
A second analysis is conducted based on the assumption that Mr. Jones would
lose his job for any reason, and would have to compete for comparable work with
8417: “8417_c031” — 2007/9/11 — 12:13 — page 707 — #25

non-disabled cohorts. Additionally, without the employer’s benevolence, Step 3, his

postinjury earning capacity would be more like that of the average bookkeeper’s, or
$30,000 per year. Clearly, his future employability becomes less likely under the given
circumstance, so Step 4, Mr. Jones’ postinjury worklife expectancy would be more
like an average disabled cohort’s, or 13.9 years. Step 5, the present value calculation
for the second analysis reveals a lifetime loss of future earning capacity in the range
of $1,722,092–$2,177,092, using a range of an assumed net discount rate of 1.6% to
a total offset assumption.
The following spreadsheet shows the underlying data and calculations for Mr.
Jones’ loss of future earning capacity and his household services replacement costs.
The numbers of life survivors in column C are from the most recent data from 2003
released in 2006, and the work probability data, E-H are a 10-year average, 1995–2004
(see Tables 31.4 and 31.5).
Around the household, Mr. Jones is no longer able to fully perform chores and tasks
as he once was. He cannot do the heavier, more physically demanding activities like
yard work, and is advised to stay off the roof because he gets dizzy. Even when doing
lighter demanding activities, he has to take frequent breaks, and he tries conserving
his energy for work.
From the National Human Activity Pattern survey conducted by the Environ-
mental Protection Agency, Economic Demographers from Expectancy Data in Prairie
Village, Kansas compiled numerous tables of categories of household members’
activities, outside the realm of occupational functioning.11
The statistical cohort of Mr. Jones, a married male working full-time with two
dependent children in the household, averages 15.6 h per week of household produc-
tion. Taking Mr. Jones’ self-reported activities from the standard vocational interview
but using the average statistical cohort’s production rate assures the activities were
performed while avoiding over-reporting the number of hours. Although the self report
confirms that activities were done, instead of using as the baseline the self-reported
amount of time spent performing activities, using the average cohort’s derived time
spent performing activities makes a fairer baseline.
Though we are not paid for doing household chores and tasks, there is a value
for doing them. Even without the benefit of having the video of Mr. Jones’ “Truman
Show,” and regardless of the quality of his work, this replacement value would be
at least minimum wage, and could be as much as the opportunity cost of trading
an extra hour of occupational functioning for household services functioning. Mean
wages from Standard Occupational Classification codes as reported in Occupational
Employment Statistics can be used as a proxy to assess an average hourly value of
time spent by activity category.
As Mr. Jones can no longer do half the activities he performed prior to injury, and
the remaining activities take him twice as long to do as they used to, his estimated
residual functioning is 25% of his preinjury level, implying 11.7 h per week need
replacement. Using the national median replacement cost of $11.63 per h, approxim-
ately $136 per week would be needed. When the resulting $7076 per year figure is
calculated over his formerly healthy life expectancy of 35.9 years, the present value
of household services replacement costs are in a range of $192,651–$254,017 (see
Table 31.6).
8417: “8417_c031” — 2007/9/11 — 12:13 — page 708 — #26

TABLE 31.4
Data and Calculations for Mr. Jones’ Loss of Future Earning Capacity and His Household Services Replacement Costs
A B C D E F G H I J K L
Number Mr. Jones’
Mr. Number Mr. Jones’ Probability Probability Probability Probability Probability Probability
of Probabil- Probability
Jones’ of Life Probabil- of working of working of working of Living of Living of Living
Peri- ity of of Living
age Survivors ity of as as Average, as Severely and and and
ods working as and
beginning Life Moderately Disabled Disabled Working, Working, Working,
Non- Working,
at age 37 Disabled Nondis- Moderately Average Severely
Disabled
abled disabled disabled disabled
1 37 96,045 0.9980 0.96 0.857 0.542 0.129 0.958 0.855 0.541 0.129
2 38 95,852 0.9959 0.96 0.857 0.542 0.129 0.956 0.853 0.540 0.128
3 39 95,648 0.9935 0.96 0.857 0.542 0.129 0.954 0.851 0.538 0.128
4 40 95,424 0.9910 0.96 0.857 0.542 0.129 0.951 0.849 0.537 0.128
5 41 95,182 0.9882 0.96 0.857 0.542 0.129 0.949 0.847 0.536 0.127
6 42 94,916 0.9852 0.96 0.857 0.542 0.129 0.946 0.844 0.534 0.127
7 43 94,627 0.9821 0.96 0.857 0.542 0.129 0.943 0.842 0.532 0.127
8 44 94,321 0.9785 0.96 0.857 0.542 0.129 0.939 0.839 0.530 0.126
9 45 93,984 0.9748 0.956 0.85 0.556 0.123 0.932 0.829 0.542 0.120
10 46 93,629 0.9707 0.956 0.85 0.556 0.123 0.928 0.825 0.540 0.119
11 47 93,228 0.9662 0.956 0.85 0.556 0.123 0.924 0.821 0.537 0.119
12 48 92,799 0.9615 0.956 0.85 0.556 0.123 0.919 0.817 0.535 0.118
13 49 92,342 0.9563 0.956 0.85 0.556 0.123 0.914 0.813 0.532 0.118
14 50 91,848 0.9508 0.956 0.85 0.556 0.123 0.909 0.808 0.529 0.117
15 51 91,318 0.9449 0.956 0.85 0.556 0.123 0.903 0.803 0.525 0.116
16 52 90,756 0.9387 0.956 0.85 0.556 0.123 0.897 0.798 0.522 0.115
17 53 90,154 0.9323 0.956 0.85 0.556 0.123 0.891 0.792 0.518 0.115
18 54 89,540 0.9251 0.956 0.85 0.556 0.123 0.884 0.786 0.514 0.114
19 55 88,851 0.9177 0.814 0.594 0.354 0.099 0.747 0.545 0.325 0.091
20 56 88,142 0.9094 0.814 0.594 0.354 0.099 0.740 0.540 0.322 0.090
21 57 87,340 0.9013 0.814 0.594 0.354 0.099 0.734 0.535 0.319 0.089
Injury Caused by Carbon Monoxide Poisoning: Defining Monetary Damages
22 58 86,569 0.8921 0.814 0.594 0.354 0.099 0.726 0.530 0.316 0.088
23 59 85,678 0.8820 0.814 0.594 0.354 0.099 0.718 0.524 0.312 0.087
8417: “8417_c031” — 2007/9/11 — 12:13 — page 709 — #27

(Continued)
709
710
Table 31.4
(Continued)
Number Mr. Probability
Mr. Number Mr. Probability of working Probability Probability Probability Probability
of Jones’ Probability
Jones’ of Life Jones’ of working as Average of working of Living & of Living & of Living
Peri- Probabil- of Living
age Survivors Probabil- as Disabled, as Severely Working, Working, and
ods ity of & working
beginning ity of Moderately Nondis- Disabled Moderately Average Working,
Life Nondis-
at age 37 working Disabled abled disabled disabled Severely
abled
as Non- disabled
Disabled
24 60 84,710 0.8709 0.814 0.594 0.354 0.099 0.709 0.517 0.308 0.086
25 61 83,640 0.8595 0.814 0.594 0.354 0.099 0.700 0.511 0.304 0.085
26 62 82,549 0.8467 0.814 0.594 0.354 0.099 0.689 0.503 0.300 0.084
27 63 81,324 0.8335 0.814 0.594 0.354 0.099 0.678 0.495 0.295 0.083
28 64 80,050 0.8191 0.814 0.594 0.354 0.099 0.667 0.487 0.290 0.081
29 65 78,674 0.8040 0.454 0.231 0.188 0.019 0.365 0.186 0.151 0.015
30 66 77,222 0.7879 0.454 0.231 0.188 0.019 0.358 0.182 0.148 0.015
31 67 75,675 0.7706 0.454 0.231 0.188 0.019 0.350 0.178 0.145 0.015
32 68 74,010 0.7523 0.454 0.231 0.188 0.019 0.342 0.174 0.141 0.014
33 69 72,257 0.7326 0.454 0.231 0.188 0.019 0.333 0.169 0.138 0.014
34 70 70,359 0.7117 0.278 0.174 0.15 0 0.198 0.124 0.107 0.000
35 71 68,359 0.6902 0.278 0.174 0.15 0 0.192 0.120 0.104 0.000
36 72 66,286 0.6667 0.278 0.174 0.15 0 0.185 0.116 0.100 0.000
37 73 64,037 0.6422 0.278 0.174 0.15 0 0.179 0.112 0.096 0.000
38 74 61,682 0.6166 0.278 0.174 0.15 0 0.171 0.107 0.092 0.000
39 75 59,223
40 76 56,590 26.48 21.53 13.90 3.129
8417: “8417_c031” — 2007/9/11 — 12:13 — page 710 — #28

TABLE 31.5
Data and Calculations for Mr. Jones’ Loss of Future Earning Capacity and His Household Services Replacement Costs (cont.)
First Analysis Second Analysis
M N O P Q R S T U
Preinjury Postinjury Difference Assuming Assuming Postinjury Difference Assumed Assumed
EC $98,000 EC $55,000 in Pre–post Growth Rate Discount Rate at EC $30,000 in Pre–post Growth Rate Discount
multiplied multiplied expected at 1.1% 2.7% Multiplied expected at 1.1% Rate at 2.7%
by I by J Earnings by K Earnings
$93,891 $47,040 $46,851 $47,366 $46,121 $16,227 $77,664 $78,518 $76,454

$93,691 $46,940 $46,751 $47,785 $45,306 $16,193 $77,499 $79,213 $75,103
$93,472 $46,830 $46,642 $48,198 $44,496 $16,155 $77,317 $79,897 $73,760
$93,235 $46,712 $46,523 $48,605 $43,691 $16,114 $77,121 $80,571 $72,426
$92,974 $46,581 $46,393 $49,002 $42,890 $16,069 $76,905 $81,229 $71,098
$92,692 $46,439 $46,252 $49,390 $42,094 $16,020 $76,672 $81,873 $69,778
$92,392 $46,289 $46,103 $49,772 $41,304 $15,968 $76,424 $82,506 $68,469
$92,061 $46,124 $45,938 $50,139 $40,515 $15,911 $76,150 $83,115 $67,161
$91,331 $45,574 $45,757 $50,492 $39,727 $16,260 $75,071 $82,839 $65,178
$90,941 $45,379 $45,562 $50,829 $38,941 $16,191 $74,750 $83,391 $63,888
$90,523 $45,170 $45,352 $51,152 $38,158 $16,116 $74,406 $83,921 $62,603
$90,077 $44,948 $45,129 $51,460 $37,378 $16,037 $74,040 $84,426 $61,324
$89,594 $44,707 $44,887 $51,747 $36,599 $15,951 $73,643 $84,898 $60,046
$89,077 $44,449 $44,628 $52,014 $35,821 $15,859 $73,218 $85,336 $58,769
$88,529 $44,176 $44,353 $52,263 $35,046 $15,761 $72,767 $85,744 $57,497
$87,942 $43,883 $44,059 $52,487 $34,271 $15,657 $72,285 $86,112 $56,226
$87,343 $43,584 $43,759 $52,703 $33,507 $15,550 $71,792 $86,466 $54,973
$86,671 $43,249 $43,423 $52,873 $32,732 $15,431 $71,240 $86,745 $53,701
$73,208 $29,982 $43,226 $53,213 $32,076 $9,746 $63,462 $78,124 $47,092
$72,542 $29,709 $42,833 $53,309 $31,289 $9,657 $62,885 $78,265 $45,937
$71,902 $29,447 $42,455 $53,420 $30,530 $9,572 $62,330 $78,428 $44,822
$71,162 $29,144 $42,018 $53,452 $29,745 $9,474 $61,688 $78,475 $43,670

$70,358 $28,815 $41,544 $53,429 $28,951 $9,367 $60,991 $78,441 $42,504
8417: “8417_c031” — 2007/9/11 — 12:13 — page 711 — #29

(Continued)
711
712
Table 31.5
(Continued)
First Analysis Second Analysis
M N O P Q R S T U
Preinjury Postinjury Difference Assuming Assuming Postinjury Difference Assumed Assumed
EC $98,000 EC $65,000 in Prepost Growth Rate Discount Rate at EC $30,000 in Prepost Growth Rate Discount
multiplied multiplied Expected at 1.1% 2.7% Multiplied Expected at 1.1% Rate at 2.7%
by I by J Earnings by K Earnings
$69,469 $28,451 $41,019 $53,335 $28,140 $9,248 $60,221 $78,303 $41,313

$68,563 $28,079 $40,484 $53,218 $27,340 $9,128 $59,435 $78,131 $40,139
$67,545 $27,663 $39,883 $53,005 $26,514 $8,992 $58,553 $77,818 $38,927
$66,487 $27,229 $39,258 $52,748 $25,692 $8,851 $57,636 $77,441 $37,720
$65,345 $26,761 $38,583 $52,412 $24,858 $8,699 $56,645 $76,948 $36,494
$35,773 $10,215 $25,558 $35,100 $16,209 $4,535 $31,238 $42,901 $19,812
$35,056 $10,010 $25,046 $34,775 $15,637 $4,444 $30,612 $42,504 $19,112
$34,285 $9,790 $24,495 $34,384 $15,055 $4,346 $29,939 $42,026 $18,401
$33,472 $9,558 $23,914 $33,938 $14,469 $4,243 $29,229 $41,482 $17,685
$32,594 $9,307 $23,286 $33,411 $13,870 $4,132 $28,462 $40,837 $16,952
$19,391 $6,811 $12,579 $18,247 $7,376 $3,203 $16,188 $23,482 $9,492
$18,803 $6,605 $12,198 $17,889 $7,041 $3,106 $15,697 $23,020 $9,060
$18,165 $6,381 $11,784 $17,472 $6,696 $3,000 $15,164 $22,484 $8,617
$17,497 $6,146 $11,351 $17,014 $6,349 $2,890 $14,607 $21,895 $8,170
$16,799 $5,901 $10,898 $16,516 $6,001 $2,775 $14,024 $21,253 $7,722
$1,410,772 $1,102,432 $2,177,971 $1,722,092
8417: “8417_c031” — 2007/9/11 — 12:13 — page 712 — #30

TABLE 31.6
Household Services Analysis
Household Services Analysis

V W X Y
Number of Periods Annual Assumed Growth Assumed Discount
in Full-functioning Replacement Rate at 1.1% Rate at 2.7%
Life Expectancy Cost
1 $7,076 $7,154 $6,965

2 $7,076 $7,232 $6,857
3 $7,076 $7,312 $6,750
4 $7,076 $7,392 $6,645
5 $7,076 $7,474 $6,541
6 $7,076 $7,556 $6,440
7 $7,076 $7,639 $6,339
8 $7,076 $7,723 $6,240
9 $7,076 $7,808 $6,143
10 $7,076 $7,894 $6,048
11 $7,076 $7,981 $5,953
12 $7,076 $8,068 $5,861
13 $7,076 $8,157 $5,769
14 $7,076 $8,247 $5,679
15 $7,076 $8,338 $5,591
16 $7,076 $8,429 $5,504
17 $7,076 $8,522 $5,418
18 $7,076 $8,616 $5,334
19 $7,076 $8,710 $5,251
20 $7,076 $8,806 $5,169
21 $7,076 $8,903 $5,088
22 $7,076 $9,001 $5,009
23 $7,076 $9,100 $4,931
24 $7,076 $9,200 $4,854
25 $7,076 $9,301 $4,778
26 $7,076 $9,404 $4,704
27 $7,076 $9,507 $4,631
28 $7,076 $9,612 $4,559
29 $7,076 $9,717 $4,488
30 $7,076 $9,824 $4,418
31 $7,076 $9,932 $4,349
32 $7,076 $10,042 $4,281
33 $7,076 $10,152 $4,214
34 $7,076 $10,264 $4,149
35 $7,076 $10,377 $4,084
35.9 $6,368 $9,442 $3,618
$254,017 $192,651
8417: “8417_c031” — 2007/9/11 — 12:13 — page 713 — #31

31.14.2 MRS. BETTY JONES

The interview reveals that Mrs. Jones was born on October 12, 1961. She is a 35-
year-old woman who was graduated from high school, and subsequently attained
an associate of science degree in accounting. As far as her worklife, Mrs. Jones
states she began working at age 15, and occasionally worked two jobs concurrently,
even while attending school full time. She worked part-time as a sales person (sales
person, women’s apparel, DOT # 261.357-066), mostly because she liked wearing
new fashions, and with her employee discount, the job basically supported her clothing
wants. She functioned as a bookkeeper (DOT # 210.382-014) during her 2 years of
college, and then after successfully completing on-the-job training, got a job as a
claim examiner (DOT # 241.267-018) (see Figure 31.4).
The work she’s performed is semiskilled to skilled in nature and sedentary to light
in terms of physical demands. Her career path was voluntarily interrupted around the
time her second child, Cathy, was born, but she had planned to return to work when
Cathy began middle school, which would be calendar school year 2008–2009.
A claim examiner needs above-average levels of general learning ability (intel-
ligence), verbal aptitude, and clerical perception. Physical demands needed on a
frequent basis include reaching, handling, fingering, talking, hearing, and near acuity.
A claim examiner’s occupation-specific tasks include the following:
• Adjust reserves and provide reserve recommendations to ensure reserving

activities consistent with corporate policies.
• Communicate with reinsurance brokers to obtain information necessary
for processing claims.
• Conduct detailed bill reviews to implement sound litigation management
and expense control.
Betty’s life expectancy at age 35 is 46.4 years
Betty’s full functioning life expectancy would be 39.7 years
Betty’s health-adjusted life expectancyis 29.7 years, meaningout

of her 46.4 year life expectancy, 64% is with burden of disability
0 5 10 15 20 25 30 35 40 45 50
FIGURE 31.4 Betty’s Jones life and healthy life.
8417: “8417_c031” — 2007/9/11 — 12:13 — page 714 — #32

• Confer with legal counsel on claims requiring litigation.

• Contact and/or interview claimants, doctors, medical specialists, or
employers to get additional information.
• Enter claim payments, reserves and new claims on the computer system,
inputting concise yet sufficient file documentation.
• Examine claims investigated by insurance adjusters, further investigating
questionable claims to determine whether to authorize payments.
• Investigate, evaluate and settle claims, applying technical knowledge and
human relations skills to effect fair and prompt disposal of cases and to
contribute to a reduced loss ratio.
• Maintain claim files, such as records of settled claims and an inventory of
claims requiring detailed analysis.
• Pay and process claims within designated authority level.
• Prepare reports to be submitted to company’s data processing department.
• Present cases and participate in their discussion at claim committee
meetings.
• Report overpayments, underpayments, and other irregularities.
• Resolve complex, severe exposure claims, using high service oriented file
handling.
• Supervise claims adjusters to ensure that adjusters have followed proper
methods.
• Verify and analyze data used in settling claims to ensure that claims are
valid and that settlements are made according to company practices and
procedures.
Mrs. Jones’ free time as a homemaker/mother was probably busier than it was
during her college-working days, as she wore all the different hats, juggled all the
changing schedules and planned the leisure activities, all things considered, for the
family. Like Mr. Jones, Mrs. Jones and Cathy were den mother and brownie for a
local troop, and they shared lots of time doing many activities.
Mrs. Jones reports she was able to keep an impeccably clean home. She regularly
visited the farmer’s market for fresh vegetables, in addition to the tomatoes and
peppers and herbs she grew at home. She kept roses and changed seasonal flowers in
her flower beds. She was in a ladies bowling league on Tuesday nights, and visited
her mother at least once per week.
Dr. Paul found abnormal findings on Mrs. Jones’ EEG and QEEG. He suspected
Mrs. Jones’ bilateral hand tremors and chronic pain were due to damage in her cere-
bellum, and had her evaluated by a neurologist specializing in motion and movement
disorders. Dr. Hoffman did neuropsychological testing.
The results of testing for Mrs. Jones revealed problems with discernment (the
ability to be tactful or to keep social behavior appropriate), and details (the abil-
ity to notice small differences); and multitrack thinking and differentiation (doing
more than one thing at a time, thinking more than one thought at a time, or juggling
several different thoughts or activities simultaneously, as required in adult think-
ing, i.e., multitasking) abilities. She is very depressed, and he is doing counseling
and cognitive retraining. She states they tried giving her numerous psychotropic
8417: “8417_c031” — 2007/9/11 — 12:13 — page 715 — #33

medications, but she had negative reactions to each one, and they all put her
in a fog.
Mrs. Jones was referred to Dr. Craig, a Physiatrist specializing in Rehabilitation
Medicine. Dr. Craig performed numerous physical capacities tests, and provides a
reliable and valid measurement of Mrs. Jones’ physical functioning. Additionally, Dr.
Craig delineates the future medical care and affiliated costs which Mrs. Jones will
probably need in the future, based on his knowledge of outcomes for persons in the
similar condition.
Consistency is noted when comparing the notes from the standard vocational
interview to Dr. Craig’s functional capacity evaluation. Mrs. Jones has a significant
loss of ability to do things like reaching (extending her hands and arms), handling
(seizing, holding, grasping, turning, or otherwise working with hand or hands. Fingers
are involved only to the extent that they are an extension of the hand, such as to turn
on a switch or shift gears), and fingering (picking, pinching, or otherwise working
primarily with fingers rather than with the whole hand or arm, as with handling) due
to chronic joint pain, tremors and muscle fatigue.
She is unable to lift, carry, or push/pull significant weight without causing her
pain to increase and minor activities exhaust her energy. A Home Health Aide assists
Mrs. Jones with performing her ADLs and IADLs, and performs the food shopping,
food-preparation and cooking. A housekeeper comes in twice per week and main-
tains the laundry and housecleaning. Dr. Craig opines she will need this assistance
for life.
Prior to CO Poisoning, Mrs. Jones’ states her overall health was Excellent. Her
Life expectancy at age 35 is 46.4 years, and her Full-Functioning Life Expectancy at
age 35 is 39.7 years; however, this doesn’t consider the negative health impact CO
poisoning has on her present overall health (see Figure 31.5). At present, Mrs. Jones
rates her overall health as good, but she is limited in terms of performing her ADLs.
The results of her SF-36 are below (see Table 31.7).
Mrs. Jones had the capacity to work and earn money like a Claim Examiner before
her exposure to CO Poisoning, and could have returned to doing it at any time. The
Five Steps analyzing Mrs. Jones’ loss of future earning capacity are as follows:
Step 1, her preinjury earning capacity is reasonably represented by the actual
earnings that accrues to Claim Examiners, or $41,670;
Step 2, Mrs. Jones’ preinjury worklife expectancy is like a nondisabled female’s
with 13–15 years of education, reduced by the likelihood of living and working from
age 35 through age 74, or 24.19 years;
For Step 3, her postinjury earning capacity, and Step 4, her postinjury worklife
expectancy it is assumed Mrs. Jones is 100% occupationally disabled as a result of a
combination of impairments, including exertional and nonexertional limitations; and,
Step 5, the present value calculation reveals a lifetime loss of future earning
capacity in a range of $792,809–$1,008,179, using a range of an assumed net discount
rate of 1.6% to a total offset assumption.
Mr. Cooley’s Life Care Plan delineates the cost for Mrs. Jones’ future medical
care ($8,750 per year) and home care/assistance needs ($41,184 per year). When these
figures are calculated through Mrs. Jones’ Life Expectancy, she will need between
$1,740,402 and $2,168,321, stated in terms of present value.
8417: “8417_c031” — 2007/9/11 — 12:13 — page 716 — #34

When these figures are added to her loss of future earning capacity, her combined
losses are in a range of $2,533,212–$3,176,500, stated in terms of present value (see
Table 31.8).
Betty’s worklife expectancy chart
Betty’s assumed worklife toage 67 is 33 years
A severely disabled cohort of Betty’s would

have a future worklife of 2.63 years.
An average disabled cohort would have a

9.27 year worklife expectancy
An average disabled cohort would have a 9.27

year worklife expectancy
At age 35, Betty’s nondisabled worklife expectancy is 24.19 years
0 5 10 15 20 25 30 35
FIGURE 31.5 Betty Jones’ worklife expectancy chart.
TABLE 31.7
Mrs. Jones’ SF-36 Data Set
Scales Mean SD
Limitations in physical activities because of health 33.3 70.61 27.42
problems
Limitations in usual role activities because of physical 0 52.97 40.76
health problems
Limitations in social activities because of physical or 0 55.78 40.71
emotional problems
Energy/Fatigue/Vitality 10 52.15 22.39
Emotional Well-being 52 70.38 21.97
Social Functioning 25 78.77 25.43
Pain 32.5 70.77 25.46
General Health 40 56.99 21.11
31.14.3 JOHN JONES

John Jones celebrated his 12th birthday on August 21, 2006. Dr. Hoffman’s neuro-
psychological evaluation of John’s abilities reveals that, as a result of CO poisoning,
he has a postinjury level of intelligence that is one standard deviation below his
8417: “8417_c031” — 2007/9/11 — 12:13 — page 717 — #35

718
TABLE 31.8
Loss of Future Earnings Data
Number Mr. Probability Probability
Mr. Number Mr. Probability Probability Probability Probability
of Jones’ Probability of Living of Living
Jones’ of Life Jones’ of Work of work of Living of Living
Peri- Probabil- of work and and
Age Survivors Probabil- Average Severely and and
ods ity of Moderately Working, Working,
Beginning ity of Disabled Disabled Working, Working,
Life Disabled Moderately Average
At age 35 Work as Nondis- Severely
Non Disabled Disabled
abled Disabled
disabled
1 35 98,073 0.9990 0.818 0.653 0.354 0.13 0.817 0.652 0.354 0.130
2 36 97,977 0.9980 0.818 0.653 0.354 0.13 0.816 0.652 0.353 0.130
3 37 97,875 0.9968 0.818 0.653 0.354 0.13 0.815 0.651 0.353 0.130
4 38 97,763 0.9956 0.818 0.653 0.354 0.13 0.814 0.650 0.352 0.129
5 39 97,644 0.9943 0.818 0.653 0.354 0.13 0.813 0.649 0.352 0.129
6 40 97,512 0.9928 0.818 0.653 0.354 0.13 0.812 0.648 0.351 0.129
7 41 97,364 0.9911 0.818 0.653 0.354 0.13 0.811 0.647 0.351 0.129
8 42 97,204 0.9894 0.818 0.653 0.354 0.13 0.809 0.646 0.350 0.129
9 43 97,030 0.9875 0.818 0.653 0.354 0.13 0.808 0.645 0.350 0.128
10 44 96,848 0.9854 0.818 0.653 0.354 0.13 0.806 0.643 0.349 0.128
11 45 96,642 0.9832 0.844 0.676 0.319 0.084 0.830 0.665 0.314 0.083
12 46 96,424 0.9807 0.844 0.676 0.319 0.084 0.828 0.663 0.313 0.082
13 47 96,184 0.9782 0.844 0.676 0.319 0.084 0.826 0.661 0.312 0.082
14 48 95,933 0.9755 0.844 0.676 0.319 0.084 0.823 0.659 0.311 0.082
15 49 95,666 0.9725 0.844 0.676 0.319 0.084 0.821 0.657 0.310 0.082
16 50 95,375 0.9693 0.844 0.676 0.319 0.084 0.818 0.655 0.309 0.081
17 51 95,064 0.9659 0.844 0.676 0.319 0.084 0.815 0.653 0.308 0.081
18 52 94,730 0.9622 0.844 0.676 0.319 0.084 0.812 0.650 0.307 0.081
19 53 94,367 0.9584 0.844 0.676 0.319 0.084 0.809 0.648 0.306 0.081
20 54 93,992 0.9541 0.844 0.676 0.319 0.084 0.805 0.645 0.304 0.080
8417: “8417_c031” — 2007/9/11 — 12:13 — page 718 — #36

21 55 93,569 0.9495 0.669 0.476 0.201 0.055 0.635 0.452 0.191 0.052
22 56 93,124 0.9443 0.669 0.476 0.201 0.055 0.632 0.450 0.190 0.052
23 57 92,615 0.9392 0.669 0.476 0.201 0.055 0.628 0.447 0.189 0.052
24 58 92,108 0.9331 0.669 0.476 0.201 0.055 0.624 0.444 0.188 0.051
25 59 91,508 0.9264 0.669 0.476 0.201 0.055 0.620 0.441 0.186 0.051
26 60 90,858 0.9190 0.669 0.476 0.201 0.055 0.615 0.437 0.185 0.051
27 61 90,133 0.9116 0.669 0.476 0.201 0.055 0.610 0.434 0.183 0.050
28 62 89,398 0.9030 0.669 0.476 0.201 0.055 0.604 0.430 0.182 0.050
29 63 88,562 0.8940 0.669 0.476 0.201 0.055 0.598 0.426 0.180 0.049
30 64 87,674 0.8840 0.669 0.476 0.201 0.055 0.591 0.421 0.178 0.049
31 65 86,692 0.8734 0.275 0.181 0.131 0.005 0.240 0.158 0.114 0.004
32 66 85,658 0.8622 0.275 0.181 0.131 0.005 0.237 0.156 0.113 0.004
33 67 84,553 0.8499 0.275 0.181 0.131 0.005 0.234 0.154 0.111 0.004
34 68 83,351 0.8368 0.275 0.181 0.131 0.005 0.230 0.151 0.110 0.004
35 69 82,062 0.8228 0.275 0.181 0.131 0.005 0.226 0.149 0.108 0.004
36 70 80,693 0.8072 0.145 0.084 0.067 0 0.117 0.068 0.054 0.000
37 71 79,169 0.7910 0.145 0.084 0.067 0 0.115 0.066 0.053 0.000
38 72 77,579 0.7735 0.145 0.084 0.067 0 0.112 0.065 0.052 0.000
39 73 75,860 0.7549 0.145 0.084 0.067 0 0.109 0.063 0.051 0.000
40 74 74,033 0.7349 0.145 0.084 0.067 0 0.107 0.062 0.049 0.000
41 75 72,070
42 76 24.19 18.52 9.27 2.63
43 77
44 78
45 79
46 80
46.4 81.4
8417: “8417_c031” — 2007/9/11 — 12:13 — page 719 — #37

719
preinjury level, as was found on standardized school tests. The Perdue Pegboard test
reveals he has a one standard deviation decrease in his ability to manipulate small
pegs, diminishing his handling and fingering ability.
John’s Life Expectancy is 63.5 years at age 12. His Full-functioning Healthy Life
Expectancy would have been 56.8 years, but for the CO Poisoning. His Disability-
Adjusted Life Expectancy, considering his overall health is very good, but he is
obviously limited in his major activity, work, as an adult, compared to what he
could have done, 74%, is 46.99 years. Thus, 16.51 years are lost to disability (see
Figure 31.6).
Step 1, John’s preinjury adult earning capacity is like that of an average nondis-
abled male with at least 1–3 years of college and at most a college degree, or
approximately $69,937 per year;
Step 2, John’s preinjury worklife expectancy, beginning at age 18 and reduced
through age 74 by the probabilities of Life and Working for nondisabled males with
at least 1–3 years of college and at most a college degree, is 40.83 years; (see
Figure 31.7)
Step 3, John’s postinjury adult earning capacity is at least like that of an average
disabled male with a high school diploma, and at most like that of an average disabled
male with 1–3 years of college, which is approximately $40,771 per year;
Step 4, John’s postinjury worklife expectancy beginning at age 18 and reduced
through age 74 by the probabilities of life and working for disabled males with at
least a high school diploma and at most, disabled males with one to three years of
college, or 32.24 years;
Step 5, John’s loss of future earning capacity is in a range of $1,035,692–
$1,541,334, stated in terms of present value.
From age 12, 26% of John’s life expectancy, or 16.51 years

is with burden of disability
John’s full-functioning, healthy life expectancy is 56.8 years
John’s life expectancy at age 12 is 63.5 years
0 10 20 30 40 50 60 70
FIGURE 31.6 John’s life and health-adjusted life expectancy.
8417: “8417_c031” — 2007/9/11 — 12:13 — page 720 — #38

An average nondisabled cohort of John’s has

a 40.83 year worklife probability
An average moderately disabled cohort of John’s

has a 32.24 year worklife probability
An assumed worklife from age 18 to retirement age 67 is 50
0 10 20 30 40 50 60
FIGURE 31.7 John’s worklife chart.
31.14.4 CATHY JONES

Cathy Jones celebrated her 9th birthday on September 8, 2006. Dr. Hoffman’s neuro-
psychological evaluation of Cathy’s abilities reveals that, as a result of the CO
poisoning, she has not sustained a cognitive decline in functioning. However, the
Perdue Pegboard test reveals she has a one standard deviation decrease in her ability
to manipulate small pegs, diminishing her handling and fingering ability.
Cathy’s Life Expectancy is 71.7 years at age 9. Her Full-functioning Healthy Life
Expectancy would have been 63.5 years, but for the CO poisoning. Her Disability-
Adjusted Life Expectancy, considering her overall health is very good, but she is
obviously limited in her major activity, work, as an adult, compared to what she
could have done, 74%, is 53.06 years. Thus, 18.64 years of Cathy’s life are lost to
disability.(see Figure 31.8)
Step 1, Cathy’s preinjury adult earning capacity is like that of an average nondis-
abled female with at least 1–3 years of college and at most a college degree, or
approximately $46,128 per year;
Step 2, Cathy’s preinjury worklife expectancy, beginning at age 18 and reduced
through age 74 by the probabilities of life and working for nondisabled females with at
least 1–3 years of college and at most a college degree, is 37.37 years;(see Figure 31.9)
Step 3, Cathy’s postinjury adult earning capacity is at least like that of an average
disabled female with 1–3 years of college, and at most like that of an average disabled
female with a collage degree, which is approximately $39,884 per year;
Step 4, Cathy’s postinjury worklife expectancy beginning at age 18 and reduced
through age 74 by the probabilities of life and working for moderately disabled females
with at least a 1–3 years of college and at most a college degree, or 30.47 years;
8417: “8417_c031” — 2007/9/11 — 12:13 — page 721 — #39

18.64 years, or 26 %of cathy’s life is lived with

the burden of disability
Cathy’s full-functioning healthy life expectancy would have been 63.5 years
Cathy’s statistical life expectancy at age 9 is 71.7 years
0 10 20 30 40 50 60 70 80
FIGURE 31.8 Cathy Jones’ life and health-adjusted life expectancies.
Cathy’s statistical worklife probability as

average moderately disabled is 30.47 years
Cathy’s statistical worklife probability as

nondisabled is 37.37 years
The period from Cathy’s 18th year to a retirement age

67 years later is 50 years
0 10 20 30 40 50 60
FIGURE 31.9 Cathy Jones’ worklife chart.
Step 5, Cathy’s loss of future earning capacity is in a range of $341,096–$508,492,

stated in terms of present value.
Social Scientists are reminded of the poem by John Saxe, The Six Blind Men of
Indostan, who each, after inspecting a different body part of an elephant, described
the elephant differently from his own, limited perspective. Disability, as a Gestalt,
has aspects from the health economist, mental and physical health-care providers,
8417: “8417_c031” — 2007/9/11 — 12:13 — page 722 — #40

social workers, and occupational-rehabilitational professionals. By better understand-

ing the “healthy elephant,” social scientists can continue improving quality of care
provided and hopefully ensuring a better quality of life outcome for persons with
disability.
31.15 DEDICATION
This chapter is dedicated to the memory of Don Vogenthaler, Rh.D. A mentor, friend
and colleague for more than a decade, Don’s dedication to assisting survivors of
traumatic brain injury was inspiring and lead me to become actively involved with
facilitating the recovery of survivors in a similar group. Don was a caring advocate for
the many wrongfully-injured persons he touched throughout his career. He is greatly
missed, although his endeavors will continue with at least this analyst, and with the
readers, if they, too are challenged to care.
References
1. Murray, C.J. and Lopez, A.D.,Global Burden of Disease and Injury, Copyright 1996,
World Health Organization, NY and Geneva.
2. Fryback, D.G., “Methodological Issues in Measuring Health Status and Health-related
Quality of Life for Population Measures: A Brief Overview of the “HALY” Family of
Measures”, Appendix C, Summarizing Population Health - Directions for the Devel-
opment and Application of Population Metrics, National Academy Press, Washington
DC, 1998.
3. Pennifer Erickson, Ronald Wilson, and Ildy Shannon. “Years of Healthy Life”,
U.S. Department of Health and Human Services, Public Health Service, Centers
for Disease Control and Prevention, National Center for Health Statistics,
www.health.gov/healthypeople, April, 1995.
4. National Center for Health Statistics, United States Abridged Life Tables, vital stats
1987, 2000 and 2003.
5. Expectancy data, Economic Demographers, Shawnee Mission, Kansas, “Difference
between expected life and full function healthy life at single years of age. . .”
6. The Economic Impact of Motor Vehicle Crashes, 2000, U.S. Department of
TransportationNational Highway Traffic Safety Administration, May, 2002.
7. International Classification of Functioning, World Health Organization, 2001.
8. Department of Commerce, Bureau of the Census report Labor Force Status and Other
Characteristics of Persons With a Work Disability, Current Population Reports, Series
P-23, Number 160.
9. This descriptive analogy quoted with permission from Dr. Donald Vogenthaler, a dear
friend and mentor.
10. United States Department of Labor Office of Administrative Law Judges Law Library,
http://www.oalj.dol.gov/PUBLIC/DOT/REFERENCES/DOTPARTS.HTM
11. Expectancy Data, The Dollar Value of a Day: 1999 Dollar Valuation. Shawnee
Mission, KS, 2001, Table 1, Average Hours of Activities in a Week by Persons
Employed Full-time.
12. Occupation Information Network, O-Net Online, http://online.onetcenter.org/link/
summary/13-2051.00
8417: “8417_c031” — 2007/9/11 — 12:13 — page 723 — #41

8417: “8417_c031” — 2007/9/11 — 12:13 — page 724 — #42
32 My Carbon Monoxide
Poisoning: A Victim’s
Story
Joseph A. Cramer
I’m Joe—an everyday guy. I’m six foot two, 200 pounds give or take, and chunk up a
bit in the winter. A real family guy, a loyal friend, and trusting. I work and play hard.
I had thinning hair by age 30, brown eyes,and calloused hands. When the chips are
down I’m there with a hand to help. After a beer or two I can join the guys in a round
of jokes, but I know when to go home.
I have stories to tell about every age—from running away from home at age 5
and walking around Lake Michigan at age 20. From being smitten with my 9th-grade
teacher and giving up on becoming a priest, to learning what it was like when all hell
breaks loose in my marriage. Working hard to provide the good life. A good husband
and a good father.
Hunting and fishing, especially fishing were mainstays of my leisure time if I
wasn’t occupied fawning over my family, teaching my kids the proper way to squash
ants. I loved racing sports cars, watching “Wild World of Sports” on TV, and watching
my family have fun. Little things were important. Walking the dog every morning,
riding a bicycle, spending time with my kids on Saturday mornings, and surprising my
wife with flowers. I worked hard so I could do all of that. Nothing meant anything
without my family alongside.
That is me, nothing more, nothing less. That all went away for 2 years after carbon
monoxide (CO) poisoning.
I had worked at an auto care facility for almost 2 years. Occasionally, when
the doors were closed for long periods because of cold weather, the air inside grew
stagnant and had an odd smell. On most workdays, by early afternoon I would feel
rather silly and laugh at most anything —funny or not. Other employees were affected
similarly, but not to the same degree. I was a bit of a workaholic and spent long hours
at the facility. On those days, by the time I arrived home however, I’d developed a
splitting headache and usually had a terrible temper, which was not at all my usual
temperament (nor had I ever suffered from headaches to any noticeable degree).
Starting in the fall of 1977 my symptoms became more frequent and more severe.
So much so, that my wife brought it to my attention because I was starting to shout
at her and the kids once in a while for little or no reason. We still didn’t relate it to
work as it got worse and worse. I was working 5.5 or 6 days a week, 12–16 h a day
725
8417: “8417_c032” — 2007/9/11 — 12:13 — page 725 — #1

and I was inhaling fumes all the while. Never having enough time to clear my system
of the poison, it was taking its toll, little by little it had become chronic. As friends
and relatives started coming over less and less because of my appalling behavior, it
never entered my mind that it could possibly be CO poisoning at work. I knew the
building was equipped with an air makeup system for just such possibilities.
My family wanted to know what was happening. It was like I was becoming a
different person right before their eyes. It seemed that no matter how hard I breathed, I
couldn’t get enough oxygen into my lungs. I constantly forgot why I was in a particular
place and what I was supposed to do. My cognitive skills began to diminish to an
almost ridiculous level of understanding. My body was slowly falling asleep, sort of
like when you hit your funny bone on your elbow, and I started to feel numb all the
time—not just at work. I thought that soon it would go away. It didn’t. And I didn’t
know how to stop it.
On the final day, the last thing I can remember is somebody putting a locker where
I didn’t think it belonged. I freaked out and tried to smash it with a hammer. I didn’t
know what was happening to me, but it was a bad thing and I tried to fight it somehow.
Next I started to yell at a customer; I even swore at him and I never did those sorts of
things. I had been working about 18 h, nonstop, that day. My tongue started to swell
and I couldn’t talk right. There was no oxygen, nothing was working, and I had to
find a way to put air in my lungs—I had to keep fighting! I asked a fellow employee
to help me, but I don’t think he understood what I was asking. I swore at him, too,
just before he grabbed me.
Although I have no personal memory of it, it’s been explained to me that at this
point in time I was driven home by a couple of fellow employees. They left me there
with my wife and children. My wife, Sandy, called our doctor and for the next 24 h it
remained a mystery as to what was happening to me and what course to take. Finally
I was asked to speak to the doctor. My tongue was swollen and I spoke gibberish. He
ventured a guess that perhaps I had CO poisoning and had my wife bring me to the
hospital for tests where his hypothesis was confirmed.
My mind was a goo of mishmash and lightning flashes. My head hurt so badly
it felt like it would explode. I pushed on my temples as hard as I could to try and
stop the pain. A lady was talking to me in a loud voice, but I couldn’t understand
what she was saying. I looked at her but it appeared like she was a long way off in a
long dark tunnel. There was a little boy by her feet. I tried to talk to her and tell her
about my head breaking apart. The words didn’t come out of my mouth in the way
I was thinking them. My tongue was still swollen and wouldn’t do what I wanted it
to. Everything slowly went dark and I couldn’t see anymore although my eyes were
open. There was more noise and someone cried.
Now I felt odd, very odd. Nothing hurt. Actually I felt nothing at all. I was floating.
My feet weren’t touching anything. I felt I must be flying in the air, but I couldn’t
see where I was. I didn’t know where I was or even if I was anywhere at all … I
wasn’t real anymore … Nothing was real … It was a dream … . I couldn’t breath
… I thought I must be dead … . Everything was gone… It was like I’d never been
born .… There was nothing!
I am lying down and a woman in a white dress was trying to jab a needle into
my wrist. I explained that it hurt terribly, but she didn’t appear to hear me and kept
8417: “8417_c032” — 2007/9/11 — 12:13 — page 726 — #2

My Carbon Monoxide Poisoning: A Victim’s Story 727
jabbing. My whole body felt like it was made of cement and attempts at movement
were useless. My head hurt even more and I tried to sleep. Noises woke me up
sometimes and I found them extremely annoying. Someone gave me a pill and a
drink and soon my head stopped hurting and I fell asleep again.
There were people around me sometimes, but I had no idea who they were. A
man in a light gray suit woke me up and asked my name. I couldn’t tell him. He said
I was Joe. That name sounded OK to me. He told a lady I could go home and I went
with her. We didn’t talk much, but she asked a lot of questions I couldn’t answer. I
felt myself getting upset. When we got to her house, she gave me a pill and I took
it with a glass of water. After a while my mouth got very dry, but I was not upset. I
didn’t know how to think or speak understandably. It had been explained to my wife,
Sandy that even after 24 h there still remained almost a lethal amount of CO in my
blood, but the symptoms should go away after a few days of rest. (editor’s note: This
often does not occur, as you will see in this instance).
My memory was almost entirely gone except for a few basics. I did start talking
understandably when my tongue began working again, but much of it was nonsense;
speaking about lights and sounds no one else could see or hear, but gradually evolving
into an acceptable type of communication, though it could still certainly not be called
articulate. That took a great deal longer. That’s what I’ve been told, but can only
vaguely remember.
It was a truly frightening time for my family. I had no feelings other than placid
and annoyance. I took trips to various doctors and received several tests including a
CAT scan along with more blood tests. If I remember correctly, perhaps an electrocar-
diogram or something of that nature with electrodes glued all over my head and upper
torso. They didn’t show much, if anything at all that we were told of. There was really
nothing that could be done, they told me. I was given a drug called “Elavil” to keep
me calm. I took as many as four aspirin at a time, as many as five times a day to numb
the headache, but that also made my ears ring. Medical doctors treated me; however
they could and after a few weeks seemed to be both surprised and disappointed that
my memory had not returned. As time went on, it seemed as though some in the
medical field thought I was being stubborn by not getting better—that I was actually
refusing to improve by my own choice.
In the very beginning of this ordeal, some of my behavior was nothing short
of juvenile. I couldn’t pass a refrigerator without putting my hand into the freezer.
It was fascinating to feel the cold on my hand and wonder where it came from.
I also wondered where everything went when I flushed the toilet. Crude perhaps,
but interesting. Sandy kept telling me to stay away from the stove because I’d burn
myself. I watched her light matches and blow them out. Eventually, I lit one myself
and burned my hand to see what it felt like. It hurt like hell!
The following is my wife Sandy’s comments about that period in my life:
“I opened the door and two guys that Joe worked with told me that Joe wasn’t
acting right and they couldn’t control him. They thought they’d better bring him home
till he got better. Then they left.
I looked at Joe and his eyes were blank, as if he didn’t know where he was. I
was scared. I didn’t see the guy that went to work the day before. (He’d worked all
night and most of that day —about 18 h). I held the door open and he came in, his
8417: “8417_c032” — 2007/9/11 — 12:13 — page 727 — #3

eyes began to dart all over the place. He was acting mean. I backed away, he moved
in quick movements, grabbing things as if to keep his balance. Going from room to
room looking quickly and moving to the next. I asked him what was going on. He
looked at me and his mouth opened but no words came out, only sounds I’d never
heard before. I backed away again. I thought, “Who was this?”
Joe kept moving all around; when he saw himself in the mirror he kept sticking
his tongue out and scraping it on his teeth. When our son went up to him and started
to talk Joe yelled odd sounds and moved his legs toward him, making him run away
crying. It was horrible. I didn’t know what to do. Then suddenly Joe became calm and
quiet, acting exhausted. I led him to the bed and he lay down and went to sleep. I was
still scared but felt that when he woke up it would be all right. I didn’t sleep at all that
night. My home didn’t feel safe anymore. Joe had always protected us. Not that night.
The next morning Joe woke up and held his pants like a little kid, looking around.
I took him to the bathroom and he went. He didn’t seem able to talk, made odd sounds,
eyes still looked glazed and distant, darting back and forth, still exhausted. He looked
weird and disoriented. He wasn’t better. I called our family doctor and did my best to
tell him what was going on. He asked to talk to Joe. I gave Joe the phone. He made
noises into the phone and dropped it. I picked it up and the doctor asked me where
Joe had been when this happened. I told him and he said to call the poison control
center and get to the hospital for tests.
I called a babysitter for the kids and when she got there I managed to get Joe in the
car and drove to the hospital. (I cannot remember exactly how I got Joe to do all this).
Blood was drawn for tests and a doctor told me Joe had enough Carbon monoxide in
his body to kill most people. He said that within a day or so he’d be back to being
himself. I’m not sure, but I think they gave Joe a shot. Then he became irritable and
they gave him a pill. He calmed down and I took him home. I kept asking Joe questions
but he didn’t answer. It was like he didn’t understand me, but at least he was calm and
not mean.
The next few weeks it was just, give him pills to keep him calm and try to teach
him to talk. He could eat and went to the bathroom but that was about it. When he
began to speak in a couple days it was about lights at first and I couldn’t see them.
Then we could talk but it was like being with a little kid. He didn’t know anything.
He was mean sometimes, sometimes not. He certainly didn’t get better in a few days.
When we went back to my family doctor he told me to start teaching Joe about his
life and then he would get better. It took 2 years.”
For lists of changes and symptoms that occurred after the CO poisoning, see
Appendix 1.
The first weeks were filled with one new learning experience after another, and
not all of them nice. My son, Stephen was playing with a toy matchbox car and I
stepped on it. It hurt. I screamed at Stephen, grabbed the car and smashed it with a
hammer until it was flat, Stephen crying all the time. It happened so quickly, it took
Sandy by surprise and she couldn’t stop me. The damage was already done. Another
lesson—I was a very different and had to be watched more closely. It was frightening
for my entire family. What would happen next? She took me back to the doctor.
My general practitioner suggested that I walk and ride a bicycle to build up my
lung capacity and keep pumping oxygen into my system. It would also keep me
8417: “8417_c032” — 2007/9/11 — 12:13 — page 728 — #4

occupied and I’d have less chance of becoming upset. Walking was not difficult, but
my kids went with me so I wouldn’t get lost. I had no memory of my neighborhood
and if the house was not within sight I was apt to go in any direction.
Riding a bicycle proved to be a much greater challenge. Although I had been an
avid cyclist all of my life, my sense of balance was gone and had to be developed
because whenever I got on the bike it would fall over before I’d get it started. It
looked rather ridiculous—my being a full-grown adult and all, falling over on a bike.
When I finally accomplished the balance to an acceptable level, there arose another
small detail we hadn’t considered. I still had no sense of direction or location and
got lost the first trip. After that I rode with a note pinned to my pocket reading, “If
found please contact …” with a phone number to reach my wife, Sandy and a small
explanation of my predicament.
For a reason beyond my understanding, if I was gone for more than a few hours I
could often be found somewhere close to the house I was born in—most often sitting
on the curb in front of it. How I managed to find my way there was a mystery.
Because I was not responding to medical treatment or perhaps the lack of any
medical treatment available, I was sent to a psychiatrist. He asked me all kinds of
confusing questions I didn’t understand and couldn’t answer. He asked me how I felt
a lot of times. I had no idea. The meetings went like that. I really don’t remember
what we talked about. I just know it annoyed me (I can distinctly remember swearing
at him inside my head, perhaps even out loud) and after several visits I was told that
perhaps this refusing to remember was due to a difficulty I had with my parents, more
specifically my mother. Now, that was just plain crazy. Even I knew it wasn’t that. I
never went back to him.
It was as though there was no possible way that CO poisoning could cause what I
was going through. (Editor’s note—This is such a common response of physicians) I
was continually examined and treated, but with disbelief about my situation. Appar-
ently, there were no effective tests or procedures available. Sandy explained later
that she felt as if everyone was simply guessing. When she spoke with our G.P., he
pretty much confirmed her feeling. I wasn’t going to change. Even Sandy was getting
depressed. Where do we go from here?
By the following fall, when the snow came, I had been told what it was but the
sight and feel of it was an unforgettable experience, almost as if I walked outside
into a different world—all cold and clean and fluffy. I played in it like a little child.
Stephen and Brenda and I got along well playing in the snow like that. They taught
me how to slide and throw snowballs and build a snowman. We needed that type
of activity together badly. It was like being on a huge, cold cloud. I didn’t like the
shoveling so much, however, the snow did teach me that recreation was cathartic, so
we played whenever we could.
The specter of my unpredictable reactions always loomed around the next corner,
however. During one snowstorm my neighbor suggested I use his snow blower to
clear my driveway and sidewalk. Fine, it sounded good to me. He got it started and
off I went. I pushed it into a snow pile too fast and stopped the engine. I pulled and
prodded and could not get it going. With the wind blowing snow down my neck
and my ears feeling all prickly with frost bite I wound up on my neighbors front
porch accusing him of tricking me with a broken snow blower. He wouldn’t come
8417: “8417_c032” — 2007/9/11 — 12:13 — page 729 — #5

out the door, but talked me through getting the thing started again. I finished but I
was also half-frozen and irritated. When I went into the house, my son Stephen, was
all excited and wanted to go out and play in the snow. I scolded him and told him to
quit bothering me. Devastated, he began to cry and ran to his bedroom. I chased him
down the hallway yelling all the way until he slammed the door. I’d done it again. I
saw what I had done and went down into the basement to be alone and make sure I
stayed away from everybody for a while.
When the next summer came, Sandy and I spent a couple of weekends away
without the kids; they stayed at Grandma’s house. The first was spent canoeing with
close friends on a river in Northern Michigan called the Pine. We figure’d I wouldn’t
get lost because the river only goes one way. It went rather well actually except for a
couple things. I was teased a lot by everyone there, because I stepped into the river at
a stop to cool-off and darn near drowned in 8 ft of water. It was muddy and I couldn’t
see how deep it was. “Still water runs deep,” they said. OK, lesson learned. At the
end of the trip, Sandy and I arrived first and stood on the bridge waiting for the other
canoes. The water under the bridge was quite still. When the canoes came around the
bend with our friends, I dove off the bridge to swim out to them. BANG! I hit my
head on the bottom. The water was only a couple feet deep. I’m lucky I didn’t break
my neck, although I did crack a couple vertebrae and bit the end of my tongue off.
Another lesson learned: “Not all still water runs deep.”
Later in the summer, another friend had a birthday celebration at his cottage, also
in Northern Michigan and invited us to join his weekend party. We had a great time.
No tense moments that I recall. I woke up Sunday morning and got into the car to go
to church about 5 miles away. I left and didn’t come back for over 4 h. I got lost and
never did make it to church. I finally got back by stopping at a grocery store to ask
for help and as luck would have it, the owner knew where the party was and led me
back. Now that was good for a few laughs and thankfully I didn’t panic.
The following 2 years seemed to be an effort in futility; with many useless attempts
to find a medical solution. Outside of our own learning activities, nothing seemed
to help. After over a year, I went to a holistically inclined doctor, an osteopath
recommended by an acquaintance. He took my case under his wing because he found
it interesting. I don’t even think he charged me. After he examined and tested me, his
opinion was that my brain was or had been swollen just like my tongue had been. It
was his diagnosis that the effects were similar to those following a stroke because of
the lack of oxygen. Also, because of the length of time since the accident, he didn’t
believe there was any drug or therapy that could help other than to keep me calm
like Elavil was doing. Only time and the body’s natural ability to heal would make
a difference. He suggested I follow a specific diet free of impurities and continue to
exercise as much as possible. He recommended that I hone my cognitive skills by
working out word and math puzzles so my brain could create new pathways. Other
than that, I’d pretty much have to learn to live with what I had.
I began the puzzles with a passion. In the beginning it was hilarious. I couldn’t
solve a thing and had to ask someone for every answer. It was the strangest thing.
Here we knew I had a vocabulary somewhere in my head because I could talk. I
had perfected swearing, almost to a fine art. But ask me a question about a word or
number problem and I drew a blank. My thinking process was all screwed up.
8417: “8417_c032” — 2007/9/11 — 12:13 — page 730 — #6

This doctor also believed I should find a job working with people for a while
in order to work on my social skills along with my regular work of detailing cars.
A close friend, Virgil helped get me a job selling sports cars at the same place he
worked and he also briefed me on many things such as driving a car and reading and
understanding maps so I didn’t get lost. It took no time at all to learn the driving part.
I had a real knack for that. Then I could drive a car back and forth to work instead
of my bicycle. My boss, Dave was a real comedian and introduced me to being the
brunt of jokes, constantly. He said he’d never had a retarded salesman before. The
laughing was good, Dave helped me learn to laugh. Some other people had treated
me as though I was retarded and avoided me. Dave said it out loud and treated me
like a friend.
I had a strong attraction to sports cars (I had driven and raced them in the past),
so working in car sales was a real plus. I was earning money and learning how to
interact with other people when my mind shut off. Eventually, however, I left the car
lot to continue my own business full time. Dave had allowed me to clean cars in his
service department at night, but that was not enough. I couldn’t get over the fear of
anxiety attacks in front of people. When I described the anxiety problem to one of my
doctors, he suggested I should carry a paper bag in my pocket and when the attack
came on to put the bag over my mouth and keep breathing my own air until it was
over. I think I tried it three or four times. It lessened the reaction, but I still believed I
was going to die. After a while I forgot about the bag. In looking back, I realize that
leaving the car lot, very much against my wife’s better judgement, was not such a
good idea because for a while the anxiety attacks and depression increased until I had
other employees around me again. The socializing had been way more important than
I’d accepted. It kept my mind working where as physical labor, although therapeutic
in its own way, came almost automatically, instinctively.
While working one night, I didn’t take any Elavil, got upset and threw a buffer
through a 6 in. thick plasterboard wall from my shop into the next. The owner of the
other shop came over carrying the buffer and with a smile on his face said, “Didn’t
take your med’s did ya?” His attitude helped me through the moment. Reminded me
of Dave.
Meanwhile my wife was left with the responsibility of rebuilding me as a person.
Daily, she showed me pictures and told me stories of the past. In the beginning many
of these sessions were not very nice at all because of my volatile temperament. I’d
become angry and aggressive when told of certain things or places, but there was
no warning ahead of time, it wasn’t always the same things and, of course, I had
no understanding or explanation for it. When visiting my mother’s house on some
occasions I’d get irritable and refuse to walk into the front yard (my father had died
in my arms in the front yard, but I had no memory of it then, just an uneasy feeling)
and still other times I strolled across the yard without hesitation. Slowly I was able
to develop an alter memory based on what she told me along with a rather flimsy
grasp of feelings I gathered from watching the reaction of others to situations. I was
an excellent mimic.
Sandy and my kids, Brenda and Stephen all helped me with reading, so I eventually
gained knowledge that way. It wasn’t the reading so much as the understanding of
what I had read. Stephen was only four and five during that time, but he knew or
8417: “8417_c032” — 2007/9/11 — 12:13 — page 731 — #7

sensed there was something wrong with his dad, so he sat on my lap and recited
words from stories he had memorized and turned pages for me. Kids are amazing
sometimes.
Those people that I came into contact with either believed I was ill and tried to
understand or didn’t believe it, and thought I should have gotten over it within a few
days. The signs of my illness were not physical, and were therefore, perhaps not
acceptable. So many times I’d try to hide my problem by being a jerk. Sandy was led
to believe and later explained to me that the overall general consensus, even in the
medical community (even though I feel they did what they could) was that the effects,
whatever they may be, should be over with in a matter of days or weeks. But still the
problems continued. Worst of it all, was the swearing and my being an obnoxious
jerk. I knew I was not reacting very well to a lot of situations, but didn’t know how
to stop it.
Occasionally, I’d panic when it seemed my mind wasn’t working at all. An
example of it was one day I was sitting behind my desk in my office, which is
12 ft. long and 8 ft. wide. I needed to go out the door for some reason (it was on the
other side of my desk). I couldn’t see anything else in the room, only my desk and
the door. What to do? I climbed over the desk instead of simply walking around it. I
had to get out. Nothing else mattered. If it happened when I was around other people
I would try to find a place to be alone until it passed. A bathroom maybe or just going
outside. I had to escape, otherwise at times, I’d get so upset I’d throw up. That would
mess me up for a few days, every time. I hated that.
There was this constant impulse that I had to fight off anything that was unknown
to me. I had little ability to defend myself with words, but I realized early on that
I had a strong physical ability. Subduing that impulse to be physical was not my
easiest task.
Although I was an excellent student and learned quickly and could function in
many capacities, there seemed to be a problem in that I could not differentiate between
my own thoughts and what I had heard from others. As an example; my brother told
me the story of how he had caught a huge fish and the fight it put up, how long it
was and how much it weighed. A short time later, I described the identical story to
a mutual friend, only now it was my fish. I believed it in my heart and could not
remember the previous conversation with my brother.
I continued to develop my extreme habit of swearing and cursing along with
yelling for little or no reason. I especially did not cope well when a conversation
would speed up and I wasn’t allowed enough time to organize my thoughts and
match them to what I had learned. There were those times when I’d get upset and
throw things, too. Thank God, it was never at a person. I never physically hurt anyone.
All of these things were a complete turn around from my natural personality. I fought
with severe bouts of depression, wishing I were dead, and also almost crippling events
of anxiety—highlighted by the feeling of being suffocated, without enough oxygen
no matter how hard I breathed.
I worked entire days alone, locked up in my shop, keeping myself away from
other people including my family. I knew I was “different,” and was afraid I would
hurt someone or their feelings, or that my mind would grow blank, or any of a dozen
other fears that would crop up at any time. I did not know how to react to many
8417: “8417_c032” — 2007/9/11 — 12:13 — page 732 — #8

common situations, including the death of my brother-in-law. I was completely lost

in feelings I couldn’t name or describe. Whatever they were they made me want to
disappear from the situation and never come back. More Elavil and aspirin did not
help. I decided not to go to a psychiatrist again because I didn’t feel he believed me
the first time. On a friends suggestion, I finally went to a priest and speaking with
him did help. At the very least it got me through the crisis at the time. Being in or
around a church had a calming affect on me. At the funeral I saw those around me,
his wife and children, siblings and parents, suffering loss. I wasn’t aware of how to
make it better, but I felt the need to do something. I was completely at a loss of what
to do, so I said almost nothing. I felt I was responsible for their anguish, and I had
caused so many other problems because I wasn’t normal.
I have to say that redeveloping a relationship with my wife in some ways was
a real treat. Laying on a blanket in the back yard after dark and the kids were in
bed, looking at the stars and listening to Sandy’s stories of back when I was a good
guy both as a father and a husband was wonderful. I had always been there for my
family—teaching our children the important things of life and helping with their skills
like making things, fishing, child’s games and, the all-important art of squishing ants
with your thumb. I would take my family everywhere with me, even to work. I loved
being with my family. She told me all of these things and more (it helped me like
who I had been, but I also grew to dislike, very much, who I had become).
I had difficulty with mathematics and completely lost my ability to understand
music or play a guitar, both of which I had done quite well before the accident. The
order of musical notes simply did not make sense to me. My daughter, Brenda helped
me tremendously with the math part—sitting for long stretches on the living room
couch correcting me and making sure I grasped a calculation and then moving on.
She was 12 at the time. I honestly don’t know how it would have gone without the
help of my family. Not nearly as well though, I think.
The accident apparently came about because of the air make up unit at the car
care facility not functioning properly, if at all. Most of the difficulties I experienced
in the beginning such as equilibrium, body numbness, eye focus, speaking, and dex-
terity gradually diminished over a period of 6 months to where they were almost
unnoticeable. Life as a family member and business owner was difficult but manage-
able because of the intense efforts of my family and friends who taught me who I
was.
The feelings and understanding that went with the knowledge were nowhere near
where they should have been, though. It was merely survival, not a walk in the park.
I fell deeper into depression over my continuous fits of anger and inability to do
anything right. Then towards the middle of the 2nd year, I experienced a sort of
awakening. All of the learning efforts sort of began to come together. It started with
my venturing out of the shop in search of customers. We were about to go broke
because there wasn’t much business just walking through the door. It was the middle
of January and we had no money for bills, or food, or gas for the car. It was a dark
time. I walked to work one morning in a blowing snowstorm feeling very depressed.
Sandy and I had talked the night before about maybe losing our home and everything
else we had. She wondered where we would go. I promised her I’d find a way to work
it out. Instead of going into the shop I went to the bus stop and took a bus to a car lot
8417: “8417_c032” — 2007/9/11 — 12:13 — page 733 — #9

about 8 miles away. I asked the manager if he had any cars to detail. He said no, so
I walked to the next car lot and asked the same question. No again. I walked 6 miles
through the snowstorm from car lot to car lot until around dinnertime a manager said
I must be crazy to be walking around like that. I said I probably was crazy, but I
could clean cars like nobody’s business. He gave me a car and I cleaned it. When he
saw how well I did, he sent me a bunch more. I didn’t sleep for 72 h, but I cleaned
enough cars to get us out of trouble. When I saw Sandy’s face as I gave her the money
it changed me. I’d finally done something good and I wanted to see that face again.
Something clicked!
The next day I hired my sister-in-law, Karen to help me with the business and
watch over me. She understood my problems and also needed a job. It worked. If
I lost focus or became depressed, she’d find ways to snap me out of it. A joke or
word of advice, a cup of her lousy coffee, or a call to Sandy—whatever it took. Even
without the long-term memory, I began to excel as a person and a businessman. Life
took a new turn. I learned to ask a question if I didn’t understand something. Bit by
bit I gained confidence in myself. My disability became less and less obvious. I began
to experience happiness, a real emotion. It beat the heck out of depression, too. Life
was better at work and at home, especially for my family.
A problem arose when workmen’s compensation refused to pay for the doctor bills
and tests any longer. The effects of CO poisoning were not supposed to last that long;
therefore, I was technically better and required no further treatment. Nothing had
worked so far anyway. I hired an attorney, an excellent attorney, and he jumped into
the case with both feet. When we had the hearing, he told me to be quiet and let him
talk. Although I wasn’t understanding what was happening, I sat quietly. Afterwards
I fired him.
When I first went to him, I explained that all I wanted was for the company to pay
my doctor and hospital bills. I felt I had made that clear. When we went to the hearing
he told the attorneys and the judge we were demanding hundreds of thousands of
dollars for this and that and suffering and stuff like that. Sandy and I had already
discussed that issue and she was certain we didn’t believe in that sort of thing. We
just wanted the bills paid. After firing my attorney, I called the company’s attorney
and they paid the bills that same day. We chose not to pursue it any further. I signed
a waiver and it was done. We were now on our own. Agree or not agree, it’s our way.
The memory problems lasted over 2 years, until midspring of the second year.
I was driving home from work in the late afternoon and got stuck in a traffic jam
caused by an accident. The weather was pleasant so the windows were down and a
breeze flowed into the car and brought with it the smell of lilacs from a grove located
alongside the road. My eyesight became blurred as if I was looking into a TV screen
when the station goes off the air and I saw, as if it was actually happening, my old
house when I was 11 years old, filled with lilacs that I had brought from a field. After
my head cleared, I continued home and told Sandy about it. She didn’t know what it
meant, so I called my mother and she explained that when I was 11, my uncle, her
brother had died and she went to church to pray for him. When she got home I had
filled the entire house with lilacs because I didn’t want her to cry.
A few weeks later, I was in a store shopping for onions and the smell triggered
another memory of picking onions with my brother and father when I was about
8417: “8417_c032” — 2007/9/11 — 12:13 — page 734 — #10

five or so. The memories were so extreme that I had to sit right down on the floor of
the grocery store—the clerk thought I was dying or something. It became manageable
in a short time and I went home, leaving the clerk to wonder what happened to the
crazy guy. My mind was no longer paralyzed.
Over the next few months it happened often and with more frequency. Eventually
my memory became closer to normal, and in some ways greatly enhanced. That
was over 26 years ago. During that ordeal there is no doubt in my mind that all of
the puzzles and riddles I attempted to solve created unique pathways in my brain
function patterns causing my perspectives to be altered in a sometimes indelicate
fashion. Occasionally, when a problem is brought up during normal conversation, I
have to refrain from giving my input even when asked. My perspectives are no longer
necessarily what people want to hear. I don’t judge them; they are simply there and
I can’t control them. I don’t often share those extreme thoughts except with those I
am closest to. Those perspectives refine, in my mind, the issues I hear into a specific
black and white area with no gray in between.
At other times, in my mind, a sentence I speak or hear becomes broken up and
moved around and forms an entirely different meaning. I am surprised I’m telling you
about them now, but I feel it’s important to say. There are always consequences we
cannot change. Those perspectives still cause me concern. I doubt they would have
been there normally. I’m still learning to keep my mouth shut and deal with them,
one at a time. As it is, I’ve used that perspective to my advantage for 26 years by
conducting an annual treasure hunt made up of riddles to celebrate the return of my
memory and help create awareness of the danger of CO and the reality of amnesia.
It’s been quite successful with over 40,000 participants at times. I do what I can with
what I’ve got and where I am.
There are still gaps. I still cannot play a guitar or understand much about music
(that may be my gift to the world). Yet, even now from time to time, I will reach a point
in a conversation and draw a complete blank. It still comes on without provocation
or warning, but happens less every year and doesn’t last long, usually. I’ve learned
to deal with it and I don’t usually panic over it. The swearing and temper took a long
time to go away much to the dismay of my family and friends and me when I’d realize
what I was doing. They had become a habit I guess. I’m a pretty calm person now, but
believe me, I had to do a lot of apologizing for a few years. And the anxiety attacks
kept up for about 4 years. They’ve been gone ever since.
Because I have seen the consequences of behavioral problems with swearing and
temper, for the last 18 years I’ve attempted to spend time whenever I can with people
that are alone and without friends because of those same problems. I visit, play cards,
and listen to their swearing and temper tantrums, leave and come back another day.
Sometimes I call and simply listen to them talk of their problems. I offer odd jobs
when I can to some that are able, but out of work. We talk then. I visit as many as I
can on Thanksgiving morning, and bring them some sort of little gifts, usually some
stupid toy for laughs, a few donuts, and a bottle of wine or juice and we play cards
and share a few stories. My hope is it makes a difference in their lives and gives them
a sense of worth. I’ve been there and understand that sometimes that’s as good as it
gets. But I can remember the loneliness that invaded my very soul when I’d drive
others away with my words or actions. It was the people that stayed that gave me life.
8417: “8417_c032” — 2007/9/11 — 12:13 — page 735 — #11

My life has been altered in many ways because of the CO poisoning, but except
for the 2 years of horror to me and my family while my memory was gone, is as close
to normal as one might expect. Is it a burden any longer? No, I really don’t think
so. We laugh, we cry, we have problems and dilemmas, we find solutions and move
on. The most incredible outcome of all of this is I now have emotions. For better or
worse, they are my greatest gift. There are still occasions when something will trigger
a memory from long ago and that in turn will trigger an emotion. Happy, sad, angry,
melancholy, what have you. It’s a treasure I did not have for a long, long time.
My memory has become much keener than before and I seem to remember my
past in much finer detail than most others. I am most grateful for that. I do think
rather slowly for whatever reason. Some common ordinary situations I respond to
out of habit, but if it’s not something I deal with almost daily I can take a considerable
amount of time to form an opinion. It would be interesting to find out what’s going
on in there where the thinking is done. And, I’m not at all good at swearing any more
(except of course … well, we’ll leave it at that). I do have a decent sense of humor,
too. That’s a good thing. Just in case I screw up.
That’s what happened when I didn’t die from carbon monoxide poisoning.
APPENDIX 1. CHANGES/SYMPTOMS AFTER THE CO

POISONING, JOSEPH A. CRAMER
Before I was brought home
Silly, I’d laugh at almost anything
Headache
Argumentative/confrontational
Uncontrollable, unpredictable aggressively violent behavior
Speech difficulties, talking incoherently
Eyes glazed
Heavy breathing
Immediately upon being brought home

Angry, mean, out of control
Screamed and hollered at everyone
Repeatedly looked in the mirror at my tongue, uttering something about it
being thick
Talking was difficult and unintelligible
Spoke about lights and sparks
Couldn’t feel anything, numb
Uncooperative, didn’t appear to understand anything
Calmed down, looked exhausted,
Went to the hospital after 24 h
8417: “8417_c032” — 2007/9/11 — 12:13 — page 736 — #12

Unable to answer questions, became frustrated when asked

(Given a pill to calm me down)
No memory, did not respond to familiar surroundings or people
Inability to explain or describe anything
Arterial blood gases (ABG) indicated CO poisoning at an extreme level
Sedated and sent home with my wife
At home after the hospital

Behaved erratically, sometimes almost comatose, sometimes overactive
Wandered aimlessly, enthralled with normal things, appeared to be learning
Depressed and anxious, erratic behavior continued
Elavil began to take effect and behavior became calm
Little sense of balance, wobbled or fell when I closed my eyes
After 3 days I could talk more understandably, childish sentences
After about 2 weeks at home

Began to learn about his surroundings
Knew enough to ask questions
Smiled, but for no reason
Walked in a straight line instead of wandering
Got lost if not in sight of home
After about 4 weeks at home

Knew how to read and to make sense of what I read
Rode a bicycle without falling over
Learning became easier
No long-term memory
During the time of my memory loss

Angered easily
Swore constantly
Depressed
Anxiety
I knew how to work and play, but didn’t know why until it was explained
Indecisive
Cognitive skills absent
Math and music skill deficits
8417: “8417_c032” — 2007/9/11 — 12:13 — page 737 — #13

Personality remained opposite (i.e., different)

I felt as though I was being controlled by another person but I couldn’t see him
How I have changed since before the CO poisoning

Cannot understand music or play a guitar (I could play quite well before)
Can now remember minute details from the past, which I couldn’t before
Extremely sensitive to odors and lack of fresh air
Don’t deal well with others in control
Have an overdeveloped sense of responsibility toward others, that is, extremely
empathetic
Slower thinker—I have to consider all perspectives
Difficulty with crowds and close places
Editor’s note: This is Joe’s story. Be aware that everyone does not respond in the same
way to CO poisoning. Some people are more sensitive to CO and some more tolerant.
For some the damage may be more serious cognitive-memory deficits, while for others
it may be greater psychiatric or sensory-motor deficits, and yet for others it may be
constant miserable physical symptoms and gross neurologic effects. Most CO victims
sustain none, or at the most, very small decrements in long-term memory, unlike Joe.
Also for most CO victims, damage in the cognitive-memory area is irreversible. It is
clear from his story, that Joe was luckier than most CO poisoning victim’s in terms
of eventual recovery of most of his functionality. It is this very strong recovery by
Joe that has allowed him to present his story for you to hear.
8417: “8417_c032” — 2007/9/11 — 12:13 — page 738 — #14

33 Noninvasive
Measurement of Blood
Carboxyhemoglobin
with Pulse CO-Oximetry
Neil B. Hampson
CONTENTS
33.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 739

33.2 New Pulse CO-Oximetry Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 740
33.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 743
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 743
33.1 INTRODUCTION
Carbon monoxide (CO) poisoning causes symptoms that range from headache,
nausea, vomiting, and dizziness to loss of consciousness, pulmonary and cardiac
failure, and even death. Since the milder symptoms of CO poisoning are nonspecific,
patients may be misdiagnosed with conditions such as viral illness, food poisoning, or
motion sickness, depending upon the circumstances of the exposure. Diagnosis of CO
poisoning requires both clinical awareness and biological confirmation of exposure.
Among the many mechanisms of toxicity of CO,1−5 its effect on hemoglobin has
been known for over a century. When inhaled, CO binds to hemoglobin in red blood
cells transiting the pulmonary capillaries, forming carboxyhemoglobin (COHb).
Because CO binds to hemoglobin in sites normally used to transport oxygen, the
result is a decrease in the oxygen content of arterial blood and an associated reduction
in peripheral oxygen delivery.
Since CO binds to hemoglobin much more avidly than oxygen, COHb remains
in the circulation for hours and is a biomarker that can be measured to document
recent exposure to CO. Until recently, determination of an individual’s COHb level
required drawing a blood sample and measuring it in a laboratory with a benchtop CO-
oximeter or estimating it by measuring exhaled CO.6 Laboratory CO-oximeters use
multiple wavelengths to spectrophotometrically distinguish and quantify the various
hemoglobin species present (oxy-, deoxy-, carboxy- and methemoglobin).
739
8417: “8417_c033” — 2007/9/11 — 12:13 — page 739 — #1

33.2 NEW PULSE CO-OXIMETRY EQUIPMENT

Pulse oximetry is a technique widely utilized for the immediate evaluation of a
patient’s oxygenation status. The technology provides an instantaneous, noninvasive,
in vivo estimate of arterial hemoglobin saturation with oxygen. Conventional pulse
oximeters transmit two wavelengths of light through tissue (typically 660 and 940
nm), measuring changes in absorbance at each wavelength over time and calculating
the functional saturation of hemoglobin with oxygen (SpO2 ).
When arterial hemoglobin is partially saturated with CO, pulse oximetry measure-
ments have been shown to overrepresent true arterial fractional hemoglobin saturation
with oxygen in both animals7 and humans.8 COHb and O2 Hb have similar absorp-
tion characteristics at 660 nm, but not at 940 nm (Figure 33.1). As a result, pulse
oximeters measure COHb similarly to O2 Hb. Although some have suggested that
the two species are measured identically, this is not true and the difference becomes
apparent at high COHb levels.8
A new pulse CO-oximeter, developed by Masimo Corporation9 (see Addendum),
utilizes eight wavelengths of light and is able for the first time to provide a noninvasive
measurement of COHb (“SpCO”) in seconds, in addition to conventional oximeter
variables SpO2 and pulse rate. The accuracy of the device has been demonstrated by
the manufacturers up to 40% SpCO, with a range of ± 3% around the measurement
(Figure 33.2).9
Because the instrument is relatively new, only a few independent studies of it
are available. In a series of 31 patients reporting to a pulmonary function laboratory
for testing, arterial blood analysis by laboratory CO-oximetry confirmed the stated
10
1 Methemoglobin
Extinction coefficient
Oxyhemoglobin
Reduced
.1 hemoglobin
Carboxyhemoglobin
.01
600 640 680 720 760 800 840 880 920 960 1000
Wavelength (nm)
FIGURE 33.1 Absorption spectra of four hemoglobin species.
8417: “8417_c033” — 2007/9/11 — 12:13 — page 740 — #2

Noninvasive Measurement of Blood Carboxyhemoglobin with Pulse CO-Oximetry 741
Test results for

Masimo Rad-57 versus blood sample
40.0
35.0
Masimo Rad-57 SpCO (%)

30.0
25.0
20.0
15.0
10.0
5.0
0.0
0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0
Reference HbCO from blood sample (%)
FIGURE 33.2 Pulse CO-oximetry SpCO measurements versus simultaneous laboratory CO-
oximetry COHb levels in normal volunteers (From Masimo Corporation website. Rad-57 Pulse
CO-oximeter. Available at: http://www.masimo.com/rad-57/index.htm. Accessed: September
26, 2005.)
accuracy of the device.10 However, the range for COHb in that population was
only 0.8–9.3%. In a clinical laboratory study, ten volunteers breathed 500 ppm CO
until COHb was raised to 15%.11 SpCO correlated with CO-oximeter COHb with a
precision of 2.19%.
The pulse CO-oximeter has also been used to measure baseline COHb levels
in patients presenting to another pulmonary function laboratory for assessment of
pulmonary diffusing capacity (DLCO).12 The SpCO value obtained was then util-
ized to “correct” the measured DLCO when severity of impairment was graded by
the interpreting physician. In another study, the pulse CO-oximeter was utilized in
an ambulatory research setting to measure the blood COHb levels of smokers and
nonsmokers exposed to second-hand cigarette smoke.13 In a case report, the device
was used to continuously monitor COHb during treatment of a victim of CO poisoning
resulting from smoke inhalation.14 At an actual initial COHb of 35%, the SpCO was
39%. After 150 min of normobaric 100% oxygen, the COHb was 5% with an SpCO of
6%. This same group has since reported on using the device to screen patients present-
ing to their emergency department (ED).15 Over 1700 patients had SpCO measured
at triage. Not surprisingly, they found that self-reported smokers exhibited higher
SpCO readings than nonsmokers (5.3% ± 3.8% versus 2.9% ± 2.7; p < .00001). More
importantly, they identified three cases of unsuspected CO poisoning that were con-
firmed through laboratory analysis. In all clinical studies to date, the device has been
found to be convenient and easy to use.
It is felt that the 40,000 cases of CO poisoning diagnosed each year in US EDs
underestimate the actual incidence and that many more cases are either not seen in
an ED or are not diagnosed when seen.16 Because clinicians have traditionally only
8417: “8417_c033” — 2007/9/11 — 12:13 — page 741 — #3

ordered blood measurement of COHb when the condition was suspected, it is likely
that there has been a tendency to measure COHb in the more symptomatic patient or
in those whose exposure history was known. EDs, emergency medical support (EMS)
providers and other first-responders will begin using the new pulse CO-oximeter soon.
Since EMS providers and paramedics commonly use a pulse-oximeter to measure
SpO2 at the scene, one can predict that many instances of elevated SpCO will be
discovered among patients without a classic history or recognized exposure to CO.
A suggested triage and management plan for patients with elevated SpCO levels has
recently been published to address this issue (Figure 33.3).17
Furthermore, many hospitals have not had the ability to measure COHb until
now. One recent study of a four-state region found that less than one-half of the acute
care hospitals had laboratory CO-oximetry available.18 This is due to the expense of
the instrument, as suggested by the fact that hospitals without CO-oximeters tend to
be located in smaller communities. Since pulse CO-oximeters are significantly less
expensive, their availability will undoubtedly contribute to increased diagnosis of CO
poisoning. Even though most hospitals without CO-oximetry report that they currently
send blood samples to other laboratories for COHb measurement, the attendant delay
appears to affect timeliness of diagnosis and management. In the same study, over
90% of CO-poisoned patients referred to a regional hyperbaric oxygen treatment
facility came from hospitals able to measure COHb. Since hyperbaric treatment is
Measure SpCO
0–3% >3%
Loss of consciousness or
No further medical evaluation
neurological impairment
of SpCO needed
or SpCO > 25%?
Yes No
Transport on 100% oxygen

for ED evaluation.
SpCO > 12% SpCO < 12%
Consider transport to hospital
with hyperbaric chamber
Transport on 100% oxygen Symptoms of

for ED evaluation CO exposure? *
Yes No
No further medical evaluation

Transport on 100% oxygen of SpCO needed.
for ED evaluation Determine source of CO
if nonsmoker
FIGURE 33.3 Algorithm for individuals possibly exposed to carbon monoxide, on the basis
of pulse CO-oximetry SpCO measurement (From: Hampson, N.B., Weaver, L.K. Noninvasive
CO measurement by first). Responders: A suggested management algorithm. J. Emerg. Med.
Serv. 2006, 24(suppl.): 10–12.
* Common symptoms of CO exposure include nausea, vomiting, headache, dizziness,
weakness, and loss of consciousness.
8417: “8417_c033” — 2007/9/11 — 12:13 — page 742 — #4

Noninvasive Measurement of Blood Carboxyhemoglobin with Pulse CO-Oximetry 743
felt to be more effective when administered early,19 rapid identification of poisoned

individuals is of great importance.
33.3 CONCLUSION
The new pulse CO-oximeter represents a major advance in field and ED screening
of individuals for CO exposure and poisoning. Because many of these will initially
be discovered to have an elevated SpCO level by first-responders, it is important that
triage and management protocols be available. As use of the device increases in all
venues, the number of individuals diagnosed with CO poisoning each year is likely
to increase dramatically.
References
1. Piantadosi CA. Carbon monoxide intoxication. In: Vincent JL, ed. Update in Intens-
ive Care and Emergency Medicine. New York, NY: Springer-Verlag NY Inc; 1990;
460–471.
2. Zhang J, Piantadosi CA. Mitochondrial oxidative stress after carbon monoxide
hypoxia in the rat brain. J. Clin. Invest. 1991; 90: 1193–1199.
3. Thom SR. Carbon monoxide-mediated brain lipid peroxidation in the rat. J. Appl.
Physiol. 1990; 68: 997–1003.
4. Thom SR. Leukocytes in carbon monoxide-mediated brain oxidative injury. Toxicol.
Appl. Pharmacol. 1993; 123: 234–247.
5. Thom SR, Bhopale VM, Fisher D, Zhang J, Gimotty P. Delayed neuropathology after
carbon monoxide poisoning is immune-mediated. Proc. Natl. Acad. Sci. USA. 2004;
101: 13660–13665.
6. Cunnington AJ, Hormbrey P. Breath analysis to detect recent exposure to carbon
monoxide. Postgrad. Med. J. 2002; 78: 233–237.
7. Barker SJ, Tremper KK. The effect of carbon monoxide inhalation on pulse oximetry
and transcutaneous PO2 . Anesthesiology 1987; 66: 677–679.
8. Hampson NB. Pulse oximetry in severe carbon monoxide poisoning. Chest 1998;
114: 1036–1041.
9. Masimo Corporation website. Rad-57 Pulse CO-oximeter. Available at:
http://www.masimo.com/rad-57/index.htm. Accessed September 26, 2005.
10. Mottram CD, Hanson LJ, Scanlon PD. Comparison of the Masimo Rad57 pulse oxi-
meter with SpCO technology against laboratory CO-oximeter using arterial Blood.
Resp. Care 2005; 50: 1471.
11. Barker SJ, Curry J, Morgan S. Measurement of COHb and MetHb by pulse oximetry:
A human volunteer study. Anesthesiology 2006; in press.
12. Mahoney AM, Stimpson CL, Scott KL, Hampson NB. Noninvasive measurement
of carboxyhemoglobin levels for adjustment of diffusion capacity measured during
pulmonary function testing. Am. J. Respir. Crit. Care Med. 2006; 3: A720.
13. Hampson NB, Ecker ED, Scott KL. Use of a noninvasive pulse CO-oximeter to
measure blood carboxyhemoglobin levels in bingo players. Resp. Care 2006; 51:
758–760.
14. Plante T, Harris D, Monti J, Tubbs R, Jay GD. Carbon monoxide poisoning detected
and monitored continuously and noninvasively: A case report. Resp. Care 2005; 50:
1480.
8417: “8417_c033” — 2007/9/11 — 12:13 — page 743 — #5

15. Chee KJ, Suner S, Partridge RA, Sucov A, Jay GD. Noninvasive carboxyhemoglobin
monitoring: Screening emergency department patients for carbon monoxide exposure.
Acad. Emerg. Med. 2006; 13 (Suppl. 1): 179.
16. Hampson NB. Emergency department visits for carbon monoxide poisoning.
J. Emerg. Med. 1998; 16: 695–698.
17. Hampson NB, Weaver LK. Noninvasive CO measurement by first responders: A
suggested management algorithm. J. Emerg. Med. Serv. 2006; 24 (Suppl.): 10–12.
18. Hampson NB, Scott KL, Zmaeff JL. Carboxyhemoglobin measurement by hospitals:
Implications for the diagnosis of carbon monoxide poisoning. J. Emerg. Med. 2006;
31: 13–16.
19. Hampson NB, Mathieu D, Piantadosi CA, Thom SR, Weaver L. Carbon monox-
ide poisoning: Interpretation of randomized clinical trials and unresolved treatment
issues. Undersea Hyperb. Med. 2001; 28: 157–164.
8417: “8417_c033” — 2007/9/11 — 12:13 — page 744 — #6

34 Chronic Carbon
Monoxide Exposure:
How Much Do We
Know About it?—an
Update
Alastair W.M. Hay
CONTENTS
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 750
Editor’s Note: This chapter was invited as an update on the study of chronic carbon
monoxide (CO) poisoning published by the CO Support group in the late 1990s (Hay
et al., 2000).1
Exposure to CO and the health consequences for those exposed acutely and
severely are well documented.2−5 However, the effects of chronic exposure to CO
gas are far less well known and less frequently the subject of research. This short
treatment reviews new developments on the subject and thoughts about our study a
decade out.
It was nearly 70 years ago that the first report on the effects of repeated exposure
to sublethal concentrations of CO was reported.6 In his study of 97 individuals, Beck
recorded the fact that his subjects had been exposed to CO over periods ranging
from several months up to 18 years. Seven principal symptoms were reported by the
subjects he examined, all over 40 years of age. These included headaches, vertigo,
nervousness, palpitations, and neuro-muscular pain (see Reference 7). It is not unusual
to see these same symptoms in individuals who have been acutely poisoned by CO.
Regrettably when poisoning with CO is discussed, it is invariably assumed that the
exposure has only been acute.
Perceptions about exposures to CO are changing, but slowly. In the past the view
prevalent amongst clinicians was that unless exposure to CO rendered an individual
unconscious, the consequences for that person’s health were minimal. All that might
be required was fresh air, or if poisoning was severe a whiff of pure oxygen. But the
745
8417: “8417_c034” — 2007/9/11 — 12:13 — page 745 — #1

evidence indicates that exposure to concentrations of CO, insufficient to render a per-

son unconscious, will damage the brain and result in neurological sequelae.8,9 Chronic
exposure to CO has now also been reported to result in damage to the brain,10,11 some
of which appears after a delay. In his fine review of the literature, Weaver states that
“lower-level CO exposures can cause headache, malaise, and fatigue and can result
in cognitive difficulties and personality changes.”12
If acute CO exposure will cause neurological sequelae in individuals, then the
consequences of repeated exposure to CO may be of even more consequence. In the
vast majority of published studies, carboxyhemoglobin (COHb) saturations are not
known. For individuals chronically exposed to CO, there is even less likelihood of
some measurement of COHb being made. Their absence has frequently frustrated
victims who have attempted to discover the cause of their continuing ill health.
More problematic still, for those chronically exposed to CO, is the lack of aware-
ness in the medical profession that such exposures can have serious consequences for
health.
To address some of the gaps in what was known about the effects of chronic CO
exposure, we conducted a questionnaire study on 77 individuals, who were known
to have been exposed to CO, over periods ranging from several days to more than 20
years.1,13 The results were reported separately for two groups who were chronically
exposed, a cohort of 65 people who never suffered loss of consciousness (LOC), and
a group of 12 people who did. For those who had LOC there was a period leading
up to the episode in which they collapsed when they had been exposed to lower, and
invariably, increasing concentrations of CO.
Sixty years after Beck6 reported the consequences of chronic exposure to CO, the
situation has not changed much in the United Kingdom. Although there is increasing
recognition of the risks from CO with some 30 deaths occurring annually in the United
Kingdom, there is less concern about the effects of chronic exposure. The results of
our survey indicated that there was a continuing and unrecognized problem associated
with chronic exposure to CO. Most physicians did not recognize the symptoms of
CO poisoning, and in consequence, did not diagnose it. Many individuals suffered
for years as a result of their exposure to the gas, and our survey indicated that many
people continued to suffer symptoms years after the exposure had stopped. Many
respondents to our questionnaire indicated that they had experienced a wide range of
symptoms for up to 2 years after exposure ended.1,13
One weakness of our study was in relation to the evidence about the duration
of symptoms during exposure to CO. Most respondents complained of a range of
symptoms, and some stated that they had had these symptoms for many years. Without
any objective evidence of actual exposure to CO, it was difficult to be certain that
individuals who complained that they had experienced these symptoms for, say, 20
years prior to CO exposure being discovered, were, in fact, suffering the effects solely
of CO exposure. But this is true for many individuals who are exposed to CO. Given
that the symptoms of exposure are nonspecific and similar to those of an infection,
although long-lived, clinicians invariably look for a viral or bacterial cause.
Unique to our survey, we were able to ascertain that the majority of respondents
had been exposed to CO because the problem had been identified (i.e., discovered) by
gas engineers/technicians inspecting gas fires (i.e., heaters), central heating boilers,
8417: “8417_c034” — 2007/9/11 — 12:13 — page 746 — #2

Chronic Carbon Monoxide Exposure 747
and the flues and chimneys which vent the exhaust gases into the atmosphere. Physi-
cians were the least likely group to identify the problem. In the majority of the cases
reported in our study,1,13 it was a gas engineer who condemned a fireplace or a flue
as dangerous, and as the source of CO exposure. We thus had confirmation for the
vast majority of our respondents, that they were indeed exposed to CO. However,
we were unable to say how much of the gas they inhaled, or what their likely COHb
saturations might have been. For many, COHb concentrations were likely to have
been over 20%, given the reports of headaches, however, we could not be certain of
the range, nor their duration. For the subjects who suffered LOC as a result of their
exposure, the evidence suggests that their COHb concentrations would have been
40% or greater.2,14
In his report on the investigation of clinical symptoms in 97 people repeatedly
subjected to sublethal concentrations of CO over periods ranging from several months
up to 18 years, Beck6 noted the symptoms cited above. In addition, his subjects
also complained of nervous and mental symptoms, including feelings of depression,
restlessness, anxiety, and fears. Also reported were experiences of introspection,
emotional upheaval, mental retardation with memory defects, and confusion. For a
further discussion of these symptoms, see Penney, 2000.7
Feelings of weakness and an inability to walk properly were mentioned by some.
Drowsiness and insomnia were recorded frequently, and approximately one-third
of patients complained of paresthesia (“bugs walking on the skin”), chiefly in the
extremity. In the 97 individuals concerned, 7 complained of speech defects, but the
exact nature of the defects is not specified. Disturbances of the vasomotor system were
also reported, and these resulted in morbid flushing, local sweating, cold extremities
(thermoregulatory dysfunction); and a purplish congestion of the hands and feet. The
principal neuromuscular complaint was a pain which was either felt as a dull pain,
or as acutely spasmodic in nature. A dull aching pain often occurred in the back,
shoulders, epigastrium, lower abdomen, and chest. Of 97 patients, 10 complained of
dysuria. Beck would have no difficulty recognizing the spectrum of symptoms in the
cohort of our study.1,13 Complaints by the respondents to our questionnaire, indicate
a spectrum of symptoms all too similar to those reported by Beck’s own patients. It
was clear to us that the problem had not disappeared.
Myers et al.10 document seven cases where the evidence indicates chronic expos-
ure to CO occurred. For a further discussion of these symptoms, see Penney, 2000.7
Symptoms similar to those reported by Beck6 are described as well as the results of
neuropsychological testing of all the subjects. These tests demonstrate damage to
the brain sufficient to cause serious distress to the subjects. Their assessment led the
authors to suggest that chronic CO exposure should be suspected if a virus-like ill-
ness persists with a significant neuropsychiatric overlay and a history of some form
of “gas” exposure. Sick cohabitants, visitors, or pets (and pets dead) are also signi-
ficant in the view of the authors.10 The authors note that the CO Neuropsychometric
Screening Battery (CONSB) which was developed for assessing the effects of acute
exposure, is of little value for the assessment of those with chronic poisoning. A far
better approach is a detailed neuropsychological evaluation that includes visual, spa-
tial, motor, intellectual, and perceptual testing. They note that with hyperbaric oxygen
treatment (HBOT) functional ability is improved, but by a mechanism which is not
8417: “8417_c034” — 2007/9/11 — 12:13 — page 747 — #3

understood. Psychiatric difficulties also tend to persist in chronically exposed patients,

but are treatable with psychotherapy and medications for anxiety and depression.10
Similar cases are reported in two other publications,11,15 one of which reports
the outcomes of magnetic resonance imaging in the assessment of brain damage in
four patients chronically-exposed to CO.15 The author reports that magnetic res-
onance imaging (MRI) scans are often within normal limits despite clinical and
neuropsychological evidence of intellectual impairment.
Another report16 documents the case of a crane driver at a smelter, who developed
permanent symptoms after 20 years of exposure to CO. Also reported are the effects of
exposure from a faulty oilfired central heating system, and the long lasting symptoms
in four members of one family. The authors also note that symptoms of chronic CO
poisoning which include headache, dizziness, and tiredness, usually disappear after
some weeks or months in people who have been exposed, but that in some individuals
they become permanent.
Infants are also at risk from chronic CO exposure. The case of a three-and-a-half
year old girl who was exposed repeatedly to CO has been described.17 The child was
admitted to a hospital when a month old, discharged after she improved, readmitted at
2 months when her mother noted she had breathing difficulties which were so severe
that the father had to administer mouth-to-mouth breaths. After treatment, the child
was again discharged with the mother receiving advice on treatment for apnea. At 3
months of age she was readmitted again with respiratory distress. Areview of her blood
results revealed slightly elevated COHb, prompting an investigation of the home
environment where it was found that the mother had been using a kerosene heater in
the house. On questioning it was apparent that other members of the household and
visitors too had experienced headaches which resolved when they left the house. Two
interesting biochemical findings were revealed when the child was hospitalized: serum
potassium and lactic acid levels were elevated, both of which resolved spontaneously.
The failure to recognize CO as the cause of symptoms in the cohort which we
studied,1,13 meant that few of our subjects received treatment that was of much benefit.
Only one-third of the patients who experienced LOC as a result of their CO exposure
received HBOT in our cohort. Two of the subjects who were never unconscious, but
were chronically exposed to CO, also received HBOT over a period of time after their
exposure to CO was recognized.
Medical treatment of the symptoms in patients chronically exposed to CO and
those who were unconscious, following what would appear to have been a prolonged
exposure period prior to their collapse, was far from satisfactory. Our interviewees
complained of continuing pain. There is an urgent need to identify treatment proto-
cols, in addition to the psychological, for subjects suffering from the effects of CO
poisoning. As well as the need for more effective treatment, there is a more pressing
requirement which is the need to help health care professionals recognize patients
who are exposed to CO. It is important to explain to the medical community that
although many individuals who are acutely exposed to CO may recover relatively
quickly from their ordeal, there are some who will experience continuing illhealth
for a long period of time. For those who are chronically exposed, recovery will be
slow and prolonged. Crucially, there is now evidence that HBOT may improve the
functionality of those chronically exposed to CO.10
8417: “8417_c034” — 2007/9/11 — 12:13 — page 748 — #4

Our data indicate that many of those chronically exposed to CO were likely
to have had COHb saturations in excess of 20%. Some individuals are likely to
be at considerable risk if they are repeatedly exposed to CO at those and even to
lesser COHb levels. Those who are physiologically stressed either by exercise or
through medical conditions are likely to be susceptible to lower levels of CO. In
individuals where there is any impairment of oxygen transport to the brain, there is
likely to be increased susceptibility to increased CO. This group includes those with
atherosclerotic lesions, noninsulin diabetes, and cerebral microvascular pathology.18
Women experiencing complications in pregnancy such as pre-eclampsia where there
is cerebral vasoconstriction that fails to respond to the normal stimuli that cause
vasodilation, could be at risk. And the elderly where there may already be a degree of
vascular compromise (as in Alzheimer’s) could be more sensitive to elevated CO. The
effect of CO exposure in individuals also taking various medications (particularly of
a psychoactive nature) or who are abusing drugs, is unknown. The effects in some
of these individuals, however could be anticipated because of CO’s effects on the
brain.18
Of no lesser importance is the risk to the developing fetus. With evidence sug-
gesting that exposures to CO of 150–200 ppm, leading to COHb saturations of
15–25%, cause birth weight reductions, delayed behavioral development, and disrup-
ted cognitive function in laboratory animal species19 the risk for the human fetus is
all too clear. Evidence from studies on smoking implicate CO at lower concentrations
as one agent responsible for lower birth weight babies, however cigarette smoke con-
tains numerous other agents which may also contribute to the effect and which make
the role of CO more difficult to disentangle in these circumstances.18 The fascinating
studies of low birth weight in over 100,000 live births in Los Angeles correlated with
ambient CO concentrations at or below the US Environmental Protection Agency
(US EPA) threshold limit standard of 9 ppm, suggest that the developing fetus may
be exquisitely sensitive to chronic CO exposure.20
The manner in which the source of CO was identified in our cohort bears fur-
ther examination. Apart from the small role which the medical profession played
in the diagnosis of the problem, it was evident that regular servicing of appliances
will not always guarantee that individuals escape poisoning by CO. For over 40%
of the respondents in both our chronically exposed group, and in the group of LOC
subjects, regular servicing of an appliance did not identify that there was a prob-
lem with it. The problem occurred after the equipment was serviced. How long
after the servicing the problem appeared was not known. It was also apparent from
the returns to the questionnaire, that many appliances were not serviced routinely.
The best advice that can be given to individuals to prevent exposure to CO, is to
ensure that all fuel-using appliances are serviced regularly, and that they ensure
that the flues or chimneys which vent exhaust gases to the atmosphere, are cleaned
regularly—at least once a year. Use of CO detectors and alarms provides a third
tier of protection. Ultimately, however, equipment should be designed so that it will
cease operation when there is any risk of significant concentrations of CO build-
ing up in the living space. It is not sufficient to have equipment turn off simply if
there is a failure of adequate oxygen to ensure complete combustion. Devices that
monitor excessive CO production could be wired to shut-down heating appliances.
8417: “8417_c034” — 2007/9/11 — 12:13 — page 749 — #5

For more on equipment self-monitoring of CO, see the chapter by Drs. Galatsis and
Wlodarski (Chapt. 10). Adoption of this kind of technology would prevent excessive
production of CO in heating devices and thus reduce the dangers of CO exposure
to users.
It is only through continuing discussion with victims of CO poisoning, the medical
community, and the fuel industry which provides the much needed heating for our
homes, that the problem of exposure to CO whether acute, or chronic, will be solved.
Much remains to be done to understand the effect of CO on the brain and other organs,
what it may do in vulnerable individuals and what treatments might ameliorate its
effects.
References
1. Hay, A.W.M., Jaffer, S., Davis, D. Chronic carbon monoxide exposure: The CO
Support study. In Carbon Monoxide Toxicity, D.G. Penney, ed., CRC Press, NY,
2000, Chapt. 19, pp. 419–437.
2. World Health Organisation. Carbon Monoxide. Environmental Health Criteria,
No. 13. Geneva, 1979.
3. Lowe-Ponsford, F.L., Henry, J.A. Clinical aspects of carbon monoxide poisoning.
Adverse drug reaction. Acute Poisoning Review, Oxford University Press, Oxford,
U.K. 1989, 8, 217–240.
4. Smith, J.S., Brandon, S. Morbidity from acute carbon monoxide poisoning at three
year follow up. Brit. Med. J., 1, 318, 1973.
5. Garland, H., Pierce, J. Neurological complications of carbon monoxide poisoning.
Quart. J. Med., New series 36, 144, 445, 1967.
6. Beck, H.G. Carbon monoxide asphyxiation: a neglected clinical problem. JAMA, 107,
1025, 1936.
D.G. Penney, ed., CRC Press, NY, 2000, Chapt. 18, pp. 393–418.
8. Chambers, C., Hopkins, R.O. Weaver, L.K. Cognitive and affective outcomes com-
pared dichotomously in patients with acute carbon monoxide poisoning. Undersea
Hyperbaric Med., 48, 2006.
9. Hopkins, R.O. Neurocognitive and affective sequelae of carbon monoxide poisoning
In Carbon Monoxide Poisoning, D.G. Penney, ed., CRC-St. Francis Press, 2007,
Chapt. 22.
10. Myers, R.A., Defazio, A., Kelly, M.P. Chronic carbon monoxide exposure: a clinical
syndrome detected by neuropsychological tests. J. Clin. Psych., 54, 555, 1998.
11. Knobeloch, L., Jackson, R. Recognition of chronic carbon monoxide poisoning.
Wisconsin Med. J., 98,26, 1999.
12. Weaver, L.K. Environmental emergencies. Carbon monoxide poisoning. Crit. Care
Clin., 15, 297–320, 1999.
13. Hay, A.W.M., Jaffer, S., Davis, D. Carbon Monoxide Support. Effects of chronic
exposure to CO:Aresearch study conducted by CO Support. Technical Paper. October,
1997. 47 pp, appendices.
14. Winter, E.M., Miller, J.N. Carbon Monoxide Poisoning. JAMA, 236, 1502, 1976.
15. Prockop, L.D., Carbon monoxide brain toxicity: clinical, magnetic resonance ima-
ging, magnetic resonance spectroscopy, and neuropsychological effects in 9 people.
J. Neuroimag. 15, 144, 2005.
8417: “8417_c034” — 2007/9/11 — 12:13 — page 750 — #6

16. Tvedt, B., Kjuus, H., Kronisk CO-Forgiftning. Eruken av. Generator gass under
den annen verdenskrig og nyere forskning. [Chronic CO poisoning. Use of gen-
erator gas during the Second World War and recent research]. Tidsskrift for Den
NorskeLageforening, 117, 2454, 1997.
17. Foster, M., Goodwin, S.R., Williams, C., Loefflerj. Recurrent acute life-threatening
events and lactic acidosis caused by chronic carbon monoxide poisoning in an infant.
Pediatrics, 104, 34, 1999.
18. Raub, J.A., Benignus, V. Carbon monoxide and the nervous system. Neurosci.
Biobehav. Rev. 26, 925, 2002.
humans. In Carbon Monoxide, D.G. Penney, ed., CRC Press, NY, 1996, Chapt. 6,
pp. 109–144.
20. Ritz, B., Yu, F. The effect of ambient carbon monoxide on low birth weight among
children born in Southern California between 1989 and 1993. Environ. Health
Perspectives, 107, 17–25, 1999.
8417: “8417_c034” — 2007/9/11 — 12:13 — page 751 — #7

8417: “8417_c034” — 2007/9/11 — 12:13 — page 752 — #8
35 Essential Reference
Tables, Graphs, and
Other Data
David G. Penney
CONTENTS
35.1 Physical Characteristics of Carbon Monoxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . 753

35.2 History of Carbon Monoxide Poisoning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 754
35.3 Carbon Monoxide Concentration Inhaled and Blood CO Saturation . . . . 755
35.4 Carbon Monoxide Wash-Out Rate, Peter Tikuisis, Ph.D. . . . . . . . . . . . . . . . . . 759
35.5 Scientific Methodology in Clinical Toxicology . . . . . . . . . . . . . . . . . . . . . . . . . . . 763
35.6 Hyperbaric Oxygen Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 764
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 764
35.1 PHYSICAL CHARACTERISTICS OF CARBON

MONOXIDE
TABLE 35.1
Physical Characteristics of Carbon Monoxide
Molecular weight 28.01
Critical point −145◦ C at 43.5 atm
Melting point −207◦ C
Boiling point −192◦ C
Specific gravity relative to air 0.968
Density
At 0◦ C, 760 mm Hg 1.25 g/L
3.54 mL per 100 mL
At 20◦ C 2.32 mL per 100 mL
At 25◦ C, 760 mm Hg 1.25 g/L
At 37◦ C 2.14 mL per 100 mL
Explosive limits in air 12.5–74.2% (by volume)
Solubilitya
At 0◦ C 3.54 mL/100 mL water
At 25◦ C 2.14 mL/100 mL water
Conversion factors
At 0◦ C, 760 mm Hg 1 mg/m3 = 0.800 ppm
1 ppm = 1.250 mg/m3
At 25◦ C, 760 mm Hg 1 mg/m3 = 0.874 ppm
1 ppm = 1.145 mg/m3
a Volume of CO indicated is at 0◦ C, 760 mm Hg.

Source: Adapted from Criteria for a Recommended Standard … Occupational
Exposure to Carbon Monoxide, 1972. U.S. Department of Health, Education and
Welfare; Maynard and Waller, In Air Pollution and Health, Academic Press 1999;
www.coheadquarters.com/CO1.htm.
753
8417: “8417_c035” — 2007/9/11 — 12:13 — page 753 — #1

35.2 HISTORY OF CARBON MONOXIDE POISONING
TABLE 35.2
Benchmarks in the History of Carbon Monoxide Poisoning (Jain, 1990)4
Aristotle, third century BC “Coal fumes lead to heavy head and death”
Cicero (106–43 BC) Rome Coal fumes were used for suicide and execution
Paracelsus (1493–1541) Wrote the first treatise on diseases of miners
Van Helmont (1577–1644) Experimented on himself with “woodgas” from a pot of
charcoal and nearly died
Rammazzini (1633–1714) Wrote De Morbis Arteficum (Diseases of miners). Pointed
out the danger of gases from burning coal
Clayton (1688) Distilled coal gas from coal
Priestly (1772) Described a combustible gas that burned with a blue flame
(carbon monoxide)
Harmant (1775), France First clinical description of coal gas poisoning
Murdock (1792), England (1794), Prussia Proposed the use of coal gas for illumination
First regulations for protection against coal gas poisoning
Cruickshank (1800) Showed that carbon monoxide is an oxide which can be
converted to CO2 by exploding it with oxygen
LeBlanc (1842) Identified CO as the toxic substance in coal gas
Chenot (1854) First explanation of the mode of action of CO
Claude Bernard (1857), France Showed that CO produces hypoxia by reversible
combination with hemoglobin
Hoppe (1857), Germany Showed that CO changes the color of blood to bright red
Linus and Limousin (1868) First to try oxygen therapy for CO poisoning
Haldane (1895) Showed that rats survived CO poisoning when placed in
oxygen at 2 atm pressure
Saint-Martin and Nicloux (1898) First demonstration of endogenous CO
Mosso (1901) Suggested the use of hyperbaric oxygen for CO poisoning
Warburg (1926) CO shown to depress respiratory chain enzymes
End and Long (1942) Treated CO poisoning in experimental animals using HBO
Migeote (1949) Detection of CO in the atmosphere
Smith and Sharp (1960) First clinical use of HBO in CO poisoning
1980s/1990s Introduction of CO detector/alarms in living/work spaces,
catalytic converters required on motor vehicles
8417: “8417_c035” — 2007/9/11 — 12:13 — page 754 — #2

Essential Reference Tables, Graphs, and Other Data 755
35.3 CARBON MONOXIDE CONCENTRATION

INHALED AND BLOOD CO SATURATION
TABLE 35.3
Carboxyhemoglobin Equilibrium at a Barometric
Pressure of 1 atm.
CO Inhaled (ppm) COHb Saturation (%)
1 0.49
3 0.81
5 1.14
7 1.46
9 1.78
10 1.94
30 5.03
50 7.92
70 10.65
90 13.22
100 14.45
300
500 45.40
700 53.77
900 59.91
1,000 62.41
3,000 83.26
5,000 89.23
7,000 92.06
9,000 93.72
10,000∗ 94.31
30,000 98.03
50,000 98.81
70,000 99.15
90,000 99.33
100,000 99.40
300,000 99.80
500,000 99.88
700,000 99.91
900,000 99.93
* 10,000 ppm = 1%.
8417: “8417_c035” — 2007/9/11 — 12:13 — page 755 — #3

Hb b
CO COH
6.5% 49.4%
30 ∝ 5 t=∝ Hb
% t= .8% CO
5,00 ppm
2.0 t = ∝ 44
3,0 0 ppm
Blood COHb (%)
00 pm
∝6
m
Hb
t= .4% CO
pp
p
t = ∝ 39
10,000
pm
00
0
p
0
50 m
.4% CO
Hb
2,
20 1, 0 pp ppm t = ∝ 32
m
, 20 ,000 0 pp ppm
1 1 80 600 ppm
500 ppm t = ∝ 22.6% COHb
400
10 m
300 pp
at t = 14.0%COHb
200 pp
m Equilibrium value
100 ppm CO in inspired air
1 2 3 4 5
Exposure time (h)
FIGURE 35.1 Uptake of carbon monoxide by humans under resting conditions. COHb sat-
uration after “infinite” exposure time (steady state conditions) is shown on each line. (Redrawn
plot of Forbes, W.H., Sargent, F., Roughton, F.J.W., Am. J. Physiol, 143, 594–608, 1945.)
Hb
CO COHb t=∝
Hb 9.4%
CO 56.
5% t = ∝4 32.8%
% =∝ COHb
2.0 t 4.8%
∝6 t = ∝4
t=
C OHb
9.4%
t = ∝3
0.50%
0.30%
%
1.0%
%
20
% t=∝
15
12
10
0.
0.
0.
OHb
0.
C 22.6%
2.8%
8% t = ∝3
0
0.
6% t=∝
0.0 % t = ∝24.6% COHb
5
0.0 14.0%
0.04%
0.03% COHb
e at t = ∝24.6%
Equilibrium valu
0.02%
ired air
0.01% CO in insp
Minutes of exposure according to scales given below
200 250 REST

200 PULSE 70
200
L.T. ACT.
150 PULSE 80
100 LT WORK
PULSE 110
20 30 40 50 60 70 80 90
10
min
HD WORK
PULSE 135
FIGURE 35.2 Uptake of carbon monoxide by humans at various levels of activity-rates of

ventilation (From Forbes, W.H., Sargent, F., Roughton, F.J.W., Am. J. Physiol, 143, 594–608,
1945.)
8417: “8417_c035” — 2007/9/11 — 12:13 — page 756 — #4

20
18
16
Hb
14 t = ∝14.0% CO
100
% CO Hb (increase)
12
10
8 t = ∝7.6% CO Hb
50
t = ∝6.1% CO Hb
6 40
t = ∝4.7% CO Hb
30
4 t = ∝3.9% CO Hb
25
t = ∝3.2% CO Hb
20 t = ∝2.4% CO Hb
2 15 t = ∝1.6% CO Hb
10
Pulse 70 0
6 l/min 0 1 2 3 4 5 6 7 8 9 10 11 1213 14 15 16 17 181920 21 2223 24 Rest
Pulse 80 Light
10 l/min 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 activity
Pulse 110
20 l/min 0 1 2 3 4 5 6 7 8 9 Walking
Pulse 135 Hard
30 l/min 1 2 3 4 5 6 7 work
Time (h)
FIGURE 35.3 Uptake of carbon monoxide at air concentrations up to 100 ppm and various
levels of activity—from Maynard, R.L., Waller, R., In Air Pollution and Health, Academic
Press 1999 (Redrawn from data of Forbes, Sargent, and Roughton, 1945)
8417: “8417_c035” — 2007/9/11 — 12:13 — page 757 — #5

1000
800
600
500
400
Se
de
300
nt
ar
y
CO concentration in air (ppm)
200 LW
H
W
100
80
60
50
40
30
20
10
10 20 30 40 60 80 100 200 300 400 500
Duration of exposure (min)
FIGURE 35.4 Length of time to achieve 5% carboxyhemoglobin (COHb) by humans

engaged in light and heavy work (From Criteria for a Recommended Standard … Occupational
Exposure to Carbon Monoxide, 1972. U.S. Department of Health, Education and Welfare.)
“Light work” is VA = 18 L/min; DL = 40 ml/min/mm Hg. “Heavy work” is VA = 30 L/min;
DL = 60 ml/min/mm Hg.
8417: “8417_c035” — 2007/9/11 — 12:13 — page 758 — #6

35.4 CARBON MONOXIDE WASH-OUT RATE, PETER

TIKUISIS, PH.D.
Assumptions:
• Predictions based on CFK model
• M = 218 (Haldane ratio)
• Vco = 0.007 mL STPD·min−1
• Wt = 70 kg, blood volume = 5500 mL, total hemoglobin concentration =
0.154 g·mL−1 blood
• DLCO = 20, 25, 30 mL STPD·min−1 ·Torr−1 for Rest, 2 × Rest, and 3 ×
Rest, respectively
• VA = 6, 12, 18 mL STPD·min−1 for Rest, 2 × Rest, and 3 × Rest,
respectively
• COHb start = 25%
TABLE 35.4
Depletion Factor; Hyperbaric Condition = 1580 mmHg (2 × Normobaric)
O2 level → 21% Normobaric 100% Normobaric 100% Hyperbaric
Activity → Rest 2 × Rest 3 × Rest Rest 2 × Rest Rest
Time (min) Depletion Factor (%)
0 0 0 0 0 0 0
10 2 3 5 12 18 22
20 4 7 9 23 33 39
30 6 10 14 33 45 53
40 8 13 18 41 55 63
50 10 16 21 49 63 71
60 11 19 25 55 70 78
90 17 27 35 70 83 89
120 21 34 44 80 91 95
150 26 41 52 86 95 98
180 30 47 58 91 97 99
210 34 52 64 94 98 99
240 38 57 69 96 99 100
270 42 61 73 97 99 100
300 45 65 76 98 100 100
330 48 68 80 99 100 100
360 51 71 82 99 100 100
390 54 74 85 99 100 100
420 57 77 87 99 100 100
450 59 79 88 100 100 100
480 62 81 90 100 100 100
510 64 83 91 100 100 100
540 66 85 92 100 100 100
570 68 86 93 100 100 100
600 70 87 94 100 100 100
8417: “8417_c035” — 2007/9/11 — 12:13 — page 759 — #7

TABLE 35.5
Remaining Factor (= 100–Depletion Factor); Hyperbaric Condition =
1580 mmHg (2 × Normobaric)
Time (min) Remaining Factor (%)

0 100 100 100 100 100 100
10 98 97 95 88 82 78
20 96 93 91 77 67 61
30 94 90 86 67 55 47
40 92 87 82 59 45 37
50 90 84 79 51 37 29
60 89 81 75 45 30 22
90 83 73 65 30 17 11
120 79 66 56 20 9 5
150 74 59 48 14 5 2
180 70 53 42 9 3 1
210 66 48 36 6 2 1
240 62 43 31 4 1 0
270 58 39 27 3 1 0
300 55 35 24 2 0 0
330 52 32 20 1 0 0
360 49 29 18 1 0 0
390 46 26 15 1 0 0
420 43 23 13 1 0 0
450 41 21 12 0 0 0
480 38 19 10 0 0 0
510 36 17 9 0 0 0
540 34 15 8 0 0 0
570 32 14 7 0 0 0
600 30 13 6 0 0 0
8417: “8417_c035” — 2007/9/11 — 12:13 — page 760 — #8

TABLE 35.6
Multiplication Factor (= 100/Remaining Factor)—to be Applied Against a
Measured Value at Some Time After Start of Washout to Determine Initial
Value; Hyperbaric Condition = 1580 mmHg (2 × Normobaric)
Time (min) Remaining Factor (%)

0 1.00 1.00 1.00 1.00 1.00 1.00
10 1.02 1.03 1.05 1.14 1.22 1.28
20 1.04 1.08 1.10 1.30 1.49 1.64
30 1.06 1.11 1.16 1.49 1.82 2.13
40 1.09 1.15 1.22 1.69 2.22 2.70
50 1.11 1.19 1.27 1.96 2.70 3.45
60 1.12 1.23 1.33 2.22 3.33 4.55
90 1.20 1.37 1.54 3.33 5.88 9.09
120 1.27 1.52 1.79 5.00 11.11 20.00
150 1.35 1.69 2.08 7.14 20.00 50.00
180 1.43 1.89 2.38 11.11 33.33
210 1.52 2.08 2.78 16.67 50.00
240 1.61 2.33 3.23 25.00
270 1.72 2.56 3.70 33.33
300 1.82 2.86 4.17 50.00
330 1.92 3.13 5.00
360 2.04 3.45 5.56
390 2.17 3.85 6.67
420 2.33 4.35 7.69
450 2.44 4.76 8.33
480 2.63 5.26 10.00
510 2.78 5.88 11.11
540 2.94 6.67 12.50
570 3.13 7.14 14.29
600 3.33 7.69 16.67
Narrative: The factors listed in the above table will yield the initial %COHb when multiplied
by the %COHb measured at a specific time after the end of an exposure to CO. For example,
if the measured COHb = 15% at 3 h after an exposure to CO has ended, then the ini-
tial value = 1.43 × 15% = 21.5% for a resting individual breathing air. Note that these val-
ues are approximate and are out-of-range or suspect if the resultant initial COHb exceeds 50%.
Further Note: The last caveat arises because errors in measurement at low values of COHb can cause
wide variations the initial values (e.g., if measured COHb = 5 ± 2% at 2 h on 100% O2 at rest, then the
initial value can range from 5 × 3 = 15% to 5 × 7 = 35%. The effect worsens as the multiplication factor
increases. Note: For additional information see Tikuisis, P. Modelling the uptake and elimination of carbon
monoxide. In: Carbon Monoxide. Penney DG (ed), CRC Press, Boca Raton, FL, 1996, pp 45–67:
8417: “8417_c035” — 2007/9/11 — 12:13 — page 761 — #9

Ambient
PCO CO body stores
Exogenous
Alveolar Endogenous
PCO
COHB 8 mLCO COMB 1.5 mL Metabolism

0.2%/h
Production 0.3 mL/h CO–X <0.5mL CO→CO2
Hgb Catab
0.1 mL/h
Other
Intravascular Extravascular
FIGURE 35.5 Carbon monoxide body stores—(From James F. Coburn, Ann. NY Acad. Sci
174, 11–22, 1970.)
% Hb O2
100
90
80
70
75 50 25 10 0 % COHb
60
50
40
30
20
10
10 20 30 40 50 60 70 80 90
P O2 mm Hg.
FIGURE 35.6 Shift of the oxyhemoglobin dissociation curve in the presence of COHb—
(From Rogers, M.C., Helfaer, M.A. Handbook of Pediatric Intensive Care., 1999.
8417: “8417_c035” — 2007/9/11 — 12:13 — page 762 — #10

35.5 SCIENTIFIC METHODOLOGY IN CLINICAL

TOXICOLOGY
Criteria for the Occupational and/or Environmental-Relatedness of Disease:
“Sir Bradford-Hill’s criteria for causation were developed for use in the field
of occupational medicine, but have been widely applied in other fields.9,10,11 These
criteria serve as a general guide, and are not meant to be an inflexible list. Not all
criteria must be fulfilled to establish scientific causation. The key criteria required
to establish causation are commonly, (1) temporal relationship, (2) specificity, (3)
biological plausibility, and (4) coherence.
The US Supreme Court in its Daubert decision12 addresses this issue and con-
cludes that the expert has to provide scientific opinion on the basis of methods and
techniques generally used, and relying on peer reviewed publications for testabil-
ity of his opinion that causation is more likely than not owing to the toxicant in
question (>50% probability). The expert does not necessarily need to rely on epi-
demiology or scientific certainty, or provide scientific proof, but rather to follow the
guidelines applied by the US Supreme Court Daubert decision. The expert must know
the accepted methods to establish causation as described in the scientific and medical
literature and established originally by Sir Bradford-Hill.” (from: OEM Meducation
Fact Sheet 3)
TABLE 35.7
Scientific Methodology in Clinical Toxicology
1. Gather the facts—history, interview, and so forth
2. Do the facts fit the literature?—consistency with past knowledge
3. Exclude improbable/unlikely causes—use differential diagnosis
4. Temporal relationship—EFFECT follows CAUSE, not the other way around
5. Is the effect occurring in only one individual, or has it occurred to several people simultaneously?
immediate, nonselective, latent period
6. Consistency and unbiasedness of findings—self-report and other-report
7. Dose-reponses—specific relationship of toxin strength to response of biological system
8. Strength of association—frequency that factor is found in “disease”; correlation coefficient
9. Specificity—only one factor is isolated and stimulates the condition
10. Coherence—do the facts fit the picture?
11. Biological plausibility
Source: Adapted from Sir-Bradford-Hill, Proc. Royal Soc. Med., 9, 295–300, 1966; Bradford-Hill, A.,
Proc. Royal Soc. Med., 58, 295, 1966; Rom, W.N., In Textbook of Environmental and Occupational
Medicine, 1992.
8417: “8417_c035” — 2007/9/11 — 12:13 — page 763 — #11

35.6 HYPERBARIC OXYGEN THERAPY
TABLE 35.8
Disorders Approved for Hyperbaric Oxy-
gen Therapy
• Decompression illness
• Air embolism
• Clostridial myonecrosis
• Osteomyelitis
• Acute traumatic ischemias (compartment syndrome)
• Skin grafts and flaps (compromised)
• Radiation tissue damage
• Smoke inhalation
• CARBON MONOXIDE POISONING
• Cyanide poisoning
• Thermal burns
• Anemia caused by excessive blood loss
Source: From Rogers, M.C., Helfaer, M.A., Handbook of

Pediatric Intensive Care. Lippincott Williams & Wilkins,
Philadelphia, PA, 1999.
References
1. Criteria for a Recommended Standard … Occupational Exposure to Carbon Monox-
ide, 1972. U.S. Department of Health, Education and Welfare.
2. Maynard, R.L., Waller, R. Carbon monoxide. In Air Pollution and Health, Academic
Press, NY, Chapt. 33, 1999.
3. Available at: www.coheadquarters.com/CO1.htm.
4. Jain, K.K. Carbon Monoxide Poisoning. Warren H. Green, Inc., St. Louis, MO, p.
377, 1990.
5. Forbes, W.H., Sargent, F., Roughton, F.J.W. The rate of carbon monoxide uptake by
normal men. Am. J. Physiol., 143, 594–608, 1945.
6. Tikuisis, P. Modelling the uptake and elimination of carbon monoxide. In: carbon
monoxide. D.G. Penney, Ed., CRC Press, Boca Raton, FL, 1996, pgs. 45–67.
7. Coburn, J.F. The carbon monoxide body stores. Ann. NY Acad. Sci., 174, 11–22,
1970.
8. Rogers, M.C., Helfaer, M.A. Handbook of Pediatric Intensive Care. Lippincott
Williams & Wilkins, Philadelphia, PA, 1999, p. 156.
9. Bradford-Hill, A. The environment and disease: association or causation? Presidents
Address. Proc. Royal Soc. Med., 9, 295–300, 1965.
10. Bradford-Hill, A. Criteria for causation in occupational and/or environmental-related
disease. Proc. Royal Soc. Med., 58, 295, 1966.
11. Rom, W.N. Causation. In Textbook of Environmental and Occupational Medicine,
2nd ed., 1992, Little, Brown, Boston, MA.
12. Daubert versus Merrill Dow Pharmaceuticals—509 U.S. 579, 113 Supreme Court
2786, 1993.
8417: “8417_c035” — 2007/9/11 — 12:13 — page 764 — #12

Index
Note: Page numbers in italics refer to figures and tables
A occupational exposure standards, for CO,

Absorption, 141, 204 291, 645
Academic testing, 523 American Gas Association (AGA), 138
Activities of Daily Living (ADL), 419, 485, American Industrial Hygiene Association
504, 593, 607, 632, 689, 691 (AIHA), 292
Activity limitation, 691 American Insurance Association, 218
Activity restrictions, 691 American Lung Association, 252
Acute CO poisoning, 346, 348, 350, 351, 355, American National Standards Institute (ANSI),
360, 391, 458–459, 460, 462, 463, 464, 44, 90, 93, 102, 103, 118,
482, 538, 653, 746; see also Poisoning 138, 148
HBO therapy for, 382, 484–485 ANSI-110, 650
basic science overview, 395–398 ANSI 119.2, 60, 66
effects, 394–395 ANSI Z21.1, 54, 102–103, 112, 115,
human evidence, 398 120, 127
Affective disorders, in CO-poisoned patients ANSI Z21.47, 148
depression and anxiety, 487–488 standards, for warning labels
Kluver–Bucy syndrome, 488 ANSI 535.2, 207
obsessive-compulsive disorder, 488 ANSI A13.1, 208
Airbag deployment, 30 workplace area warnings and labeling,
Air-free CO measurement, 52–53, 103, 211–213
104–106 American Society for Testing and Materials
Air movement, 201, 653 (ASTM)
Air Pollutants Exposure Model (APEX), 14 workplace area warnings and labeling,
Air quality control region (AQCR), 7, 9 211–213
Ala Moana Shopping Center American Society of Heating, Refrigeration,
motor vehicle exhaust in, 22 and Air Conditioning Engineers
Alarm, CO, 120, 215, 217, 251, 271–274, 280, (ASHRAE), 48, 54
315, 749 Amyl nitrite, 357
domestic CO alarms, 253–263 Anatomical imaging findings
UL standards, 646 CT, 323, 352, 358, 359, 384, 449, 458–459,
effectiveness, in marine environment, 472, 473, 485, 486, 571
183–184 MRI, 323, 356, 358, 359, 384, 449, 450,
onboard, failure of, 190 452, 459, 461, 462, 463, 472, 473,
Allopurinol, 351–352 485, 486, 497, 506, 570, 571, 610,
Ambient visual system, 621, 622, 626 635, 748
American Boat and Yacht Council (ABYC), Anthony Farr and Stacy Beckett Boating
158, 187 Safety Act (2004), 182
American College of Emergency Anxiety, in CO-poisoned patients, 487–488,
Medicine, 398 516, 625, 637, 731
American College of Emergency Apoptosis, 347–348
Physicians, 380 Aripiprazole, 361
American Community Survey, 694, 701 As-measured CO measurement, 103–106
American Conference of Governmental and Assembly Bill (AB) 2222, 182
Industrial Hygienists (ACGIH), 32 Audit trail, 297
765
8417: “8417_c036” — 2007/9/11 — 18:55 — page 765 — #1

766 Index
Australian Medical Association (AMA), C

265, 266
Automobile Cabin cruisers, 170–171, 177–178
air quality monitor (AQM), 254, California Air Resources Board (CARB),
263–265, 264 181, 292
commuter exposure, 12–13, 26 Canadian Gas Association (CGA), 44
Auxiliary power generators, 62–64 Canadian Standards Association
(CSA), 138
Carbogen, 343, 356
Carbon monoxide (CO), 6, 130, 144, 251, 313,
437; see also individual entries
B accumulation, within an enclosure, 87–90
Back drafting, 64–65 rate of dissipation, 89–90
in fireplace flue, 66 concentration inhaled
of flue gases, 83 and blood co-saturation, 755–758
Barbecue Industry Association, 18 as drowning risk factor, 165–166
Battle of experts, 680–681 in human body, 334, 664, 667
Benzodiazepine, 361 poisoned patients, case study for, 606–611
Boating Accident Report Database ADL, 419, 485, 504, 593, 607, 632,
(BARD), 160 689, 691
Boat-related CO poisoning, 164–165, LCP issues
177–180; see also Nonfatal care providers/residential care, 610
boat-related CO poisoning complications, 611
medical characteristics for, 164–166 diagnostic studies, 610
Brain-derived neuropeptides, 355 laboratory, 610
Brain injury, 514, 603, 704; see also Mild medical care, 609
traumatic brain injury; Traumatic brain medication, 610
injury in children recreation, 610
mechanism, 478 therapeutic modalities, 609
structural brain injury, evidence of, 571 transportation, 610
Brain SPECT imaging, 450–451 outcome, 608–609
after CO poisoning, 457 TSA, 607–608
chronic lower-level CO exposure, Carbon Monoxide Headquarters (COHQ),
findings of, 464 285–286
delayed CO-induced encephalopathy, goals, 285–286
findings of, 464–471 Carbon Monoxide Neuropsychological
neuroimaging modalities, 458 Screening Battery (CONSB), 357, 359,
anatomical imaging findings, 458–459 403, 503, 504, 747
functional imaging findings, 460–463 Carboxyhemoglobin (COHb), 159, 164–166,
Breath analyzer, 648, 650–651, 648, 653 273, 289, 317, 321, 343, 480, 484, 644,
British Gas, 281–282 746, 747, 748, 749
Bromocriptine, 361 in blood, noninvasive measurement of
BS EN 50291, 252 with pulse co-oximetry, 739
Building Officials and Code levels, 290
Administrators International, Inc. half-life, 294
(BOCA), 139 measurement limitations, in marine
Building Performance Institute (BPI), environment, 164
118, 120 Caspase-1, 348
Buoyancy, of CO, 200–201 Catecholamine, 347, 353
Burden of proof, 662–666 Central furnaces, 135–136, 140, 148
damages, 664 Cerebrolysin, 355
economic damages, 665–666 Certified Life Care Planners (CLCP), 606
on injury, 664–665 Charcoal
noneconomic damages, 666 briquets, CO poisoning from, 18
liability, 663–664 grills, 18, 19
standard of proof, 663 Chemical properties
Butane, 50, 141, 142, 143, 202 of CO, 67–68
boiling point, 141–142 of LP-gas, 52
8417: “8417_c036” — 2007/9/11 — 18:55 — page 766 — #2

Index 767
Children during pregnancy, 360

brain injury in neuroimaging studies, 358–360
mechanism, 478 historical perspective, 342–344
structural brain injury, evidence of, 571 management of sequelae, 361–362
CO toxicity in, see Toxicity mechanisms, of CO toxicity, 344–348
treatment, for CO poisoning, 360–361 neuroprotective treatments
Chimney allopurinol, 351–352
dynamics, 82–83 brain-derived neuropeptides, 355
Chronic CO exposure, 571, 573, 577, HBO therapy, 348–351
585–586, see also Exposure hypothermia, 355, 425
update on study, 745–750 insulin, 352–354
Chronic CO poisoning, 320, 443–446, 464, NAC, 351–352
472, 497, 551; see also Poisoning NBO, 321–322, 348, 349, 350, 356,
case study 379–380, 396, 397, 400, 401, 406
educational history, 540–541 NMDA receptor antagonists, 347,
exposure information, 540 354, 376
initial testing, 541–542 visual system functioning, 620–622
neuropsychological re-evaluation, VMSS, 620
542–545 case study, 634–640
patient demographics, 540 clinical testing, 624–625
persisting physical symptoms, 541 symptoms, 622–624
characteristics, 443 treatment, 625–626
cognitive sequelae, 498–502 with yoked prism, 626
demographics, 552–553 rehabilitative and supplementary
discovery of problem, 445–446 considerations, 626–633
emotional and affective sequelae, 498–502 Clothes dryers, 65–66, 135
Helffenstein study CO-caused morbidity, 241–243
battery, of neuropsychological tests, 508 recent national studies of nonfatal CO
MMPI-2, 531–536 poisoning, 242–243
neuropsychological testing outcome, 516 CO-caused mortality, 236–240
norms, 509 studies, 237–240
persisting symptoms, reported, 511–516 Cognitive issues, of CO poisoning,
reported symptoms, during chronic CO 593–595, 595
exposure, 509–511 case study, 606–611
sample and exposure data, 506 executive dysfunction, 594
vocational outcomes, 536–538 sensory-motor dysfunction, 594
incidence, 497–498 short-term memory dysfunction, 594
misdiagnosing, 446 vision and information processing
physical sequelae, 498–502 dysfunction, 595
self-report questionnaires, 558–564 Cognitive sequelae, CO poisoning, 478
source, 551–552 DNS, 481–482
symptomatic evaluation data, 553–558 functional imaging, 486
utilizing neuropsychological assessment, HBO effect on, 484–485
502–506 lower level, 483–484
Chronic Health Condition, 691 markers, severity and outcome, 484
City Technology, 253 neuroimaging findings, comparison,
CO sensor, 261–262, 262 485–486
Clean Air Act (CAA), 7, 8, 9, 14, 30, 181, PNS, 481–482
189, 246 Cognitive therapy, 626
Clinical effectiveness, of HBO therapy, CO Hot Pot™ , 110–111, 112–115, 125
378–382 Combustion, 68, 131–132
negative trials, 378–379 CO as product of, 199–200
positive trials, 379–382 initiate and perpetuate basic requirements
Clinical treatment, for CO poisoning, 341, 619 to, 68–69
approach to patient with CO poisoning, principles, 143–144
355–361 complete combustion, products of,
children, 360–361 143–144
8417: “8417_c036” — 2007/9/11 — 18:55 — page 767 — #3

768 Index
Combustion (Continued) Hyperbaric Oxygen” (SR Thom),

incomplete combustion, products of, 144 402–404
Community exposure Department of Commerce, Bureau of the
standards and guidelines, for CO, 291–292, Census, 692, 700
293 Department of Environment, 277, 281
Commuter exposures, 25–31 Department of Trade and Industry, 281–282
airbag deployment, 30 Depression, in CO-poisoned patients, 8, 319,
defective exhaust systems, 26–29 355, 392, 409, 487–488, 515
drive-up facilities, 29–30 Descriptive characteristics, of marine CO
motor vehicle emission control program, poisoning, 161–163
30–31 Detectors, CO, 251, 305, 749
parking garages, 29 as preventive medicine
service stations, 29 Chicago experience, 307–308
Compensatory damages, 657 clinical implication, 308
Complete combustion, products of, 143–144 current and future directions, 309–310
Computed tomography (CT), 323, 352, 358, effectiveness, 308–309
359, 384, 449, 458–459, 472, 473, 485, initial standard setting, 306–307
486, 571 technology, 306
Consumer products, 204–205 effectiveness, in marine environment,
CO-related accidents investigation 183–184
case study, 151–154 Differential diagnosis, 296
root cause, 148–151 Diffuse axonal injury (DAI), 572
CO level estimation, 151 Diffusion capacity for CO (DLCO), 741
determination, 151 Digit Symbol Incidental Memory Test, 502
follow-up inspections, 149–151 Dilution, 201–202
initial discovery, 149 Dipyridamole, 343
CORGI, 282 Direct vent system, 137
Coronary artery disease, 8 Disability, 691, 700
COSTAR® , P-1, 210 estimating, 694
Council of American Building Officials and functioning, 690
(CABO), 139 and working, 692–693
Council on Environmental Quality (CEQ), 11, Disability-adjusted life years (DALY), 686
12, 13 Dissipation, 89–90, 204
CSA 6-19-01, 252 Dizocilpine, see MK-801
CSA Star, 138 Domestic water heater, 65
Current Population Survey (CPS), 692, 693, “Do no harm”, 214
694, 700 Draft diverter, 132–133
Custom and practice, of CO safety safety problems, 133
information, 224, 226–227, 228 Drive-up facilities, 29–30
Cytochrome c, 343, 344, 345, 359, 478 Drowsiness, 747
D E
Dantrolene, 361 Earning capacity, 605, 693–694
Data functions, 696, 705 loss, analysis of, 699–703
Data loggers, 14, 15, 648 loss of future, 703–723
“Death Zone”, 171 Echocardiogram (ECHO), 345
Decennial Census, 694, 701 Eco-Balls, 272
Defective exhaust systems, 26, 29 Econometric models, 694
Delayed administration, of HBO therapy, 383 Economic Demographers from Expectancy
Delayed neurological sequelae (DNS), 349, Data, 708
358, 359, 361, 375, 384, 396, 403, 413, Elavil, 727
414, 423, 458, 462, 481–482, 486 Electrocardiogram (EKG), 290, 345, 356, 396
“Delayed Neuropsychologic Sequelae after Electrochemical gas sensors, 253, 261–263
Carbon Monoxide Poisoning: Electroencephalogram (EEG), 349, 350, 358,
Prevention by Treatment With 402, 461, 486, 630, 635, 704
8417: “8417_c036” — 2007/9/11 — 18:55 — page 768 — #4

Index 769
Electronic ignition system, 134 defective exhaust systems, 26–29

Emergency department (ED), 644, 649–650, drive-up facilities, 29–30
651, 741–742 motor vehicle emission control program,
screening, 650 30–31
Emergency medical service, 187, 643, 649, parking garages, 29
742 service stations, 29
Emergency medical support, 742 duration assessment, 292, 294
Emergency medical technicians (EMTs), 295 methylene chloride exposure, 31–32
Emergency smoke escape devices, 650 nonoccupational exposure, 31
Emission occupational exposure, 32
from gas ranges, 99–128 microenvironmental exposures, 17
trends, 8–10 commuter exposures, 25–31
Emotional and behavioral issues, of CO occupational exposures, 20–22
poisoning, 596–598 recreational exposures, 23–25
case study, 606–611 residential exposures, 17–20
Emotional and psychological functioning, 706 shopping center exposures, 22
EN403 standard, 650 NAAQS, guidelines of, 7–8
Engineering control research and occupational exposure standards and
development, 184–187 guidelines, 291
Environmental conditions, 698 OSHA and industry standards for, 54
“Environmental health gap”, 243 types, 440
Ethane, 142, 202
Ethical and philosophical reasons, for safety
information, 214–216, 217, 218
Excess air combustion, 69, 71–72, 103, 104, F
105, 143, 144, 314
Excess fuel combustion, 69, 72 Fact witnesses, 669
Exchange transfusion, 343 opinion, in CO case, 673
Excitatory amino acids (EAA), 347–348, 354, Failure modes, of gas appliances, 146–148
376 False alarms, 307–308
Executive functioning, 538, 572, 585, 705 Fatal exposures, 19–20
miscellaneous tests, 529–530 Faulty heating equipment, 19
Exit velocity, in CO producing equipment, Federal Rules of Evidence (FRE), 672,
202–203 674–675, 702
Expert opinions, in CO case Figaro, 253
battle of experts, 680–681 Firefighters and CO
components, 674 emergency department screening, 650
expert witnesses, 674–680 emergency smoke escape devices, 650
fact witnesses, 673 fatalities from suppression, 643–644
judge, 673–674 investigation, 646–650
jury, 674 lethal gas combinations, 651–652
Expert witnesses, 288, 289, 660, 669, 671, overhaul and postfire, 644–645
672–673, 674, 675–676 prevention and awareness, 652–653
long-term thinking, 681–682 workplace/industry CO standards, 645–646
opinions, on reality, 676–677 Fireplaces, 66
scientific basis, 677–679 FiS, 253
scope, 676 SB series sensor, 255, 256, 257–258
testifying, strategies for, 679–680 Fixed-site monitoring (FSM) stations, 8, 9, 12,
Exposure, 54, 289 13
chronic CO exposure Flame rollout switch, 134
update on study, 745–750 Flue gases, 82
COHb levels, 290 back drafting, 83
community exposure Fluid Attenuation Inversion Recovery
standards and guidelines, for CO, (FLAIR), 459, 571
291–292, 293 Flu symptoms, 319, 446, 498, 505, 634, 646
commuter exposures, 25–31 Focal visual system, 621, 622
airbag deployment, 30 Food and Drug Administration (FDA), 221,
222, 357
8417: “8417_c036” — 2007/9/11 — 18:55 — page 769 — #5

770 Index
Fuel, 68–69 Gracey, J.M., 627

burning devices, 44, 60 “The Great Imitator”, 653
combustion, 44, 72 Gulf of Mexico CO poisoning, 169
fuel gases, 140–143
input adjustments, 140
Fume hoods, 64
Functional capacity evaluation (FCE), H
600–601, 690, 699, 716
Functional capacity index (FCI), 689 Haber’s Law, 292
Functional imaging findings, 460 Hall, T.O., 627–628
fMRI, 461 Halstead–Reitan tests, 522
MRS, 359, 449, 450, 461–462, 472, 482 Health, 684–690
PET, 358, 359, 385, 449, 450, 460–461, less-than-perfect health, quantifying
472, 473, 486, 504 healthy life expectancy, 687–689
SPECT, 323, 345, 358, 359, 385, 449, 450, life-years lost to injury, 689–690
451, 452, 453, 457, 459, 460, 461, years lived with disability, 687
462–463, 464–471, 472, 473, 486, measuring, 684–687
497, 538, 539–540 stages, 688
Functional limitation, 700–701 Health and Safety Executive (HSE)
Functional Magnetic Resonance Imaging department, 282–284
(fMRI), 461 Health Care Financing Administrations
(HCFA), 422
Health-Realted Quality of Life (HRQL), 687
Healthy life expectancy, 687–689
G Healthy People 2010, 243, 687
Heating, ventilation, and air conditioning
Gamma aminobutyric acid (GABA), 353 (HVAC) system, 252, 253, 263, 647
Gas appliance Heating/gas appliance system, 145–146
certification, 138 combustion and ventilation air supply, 146
design, 130–131 fuel supply, 146
failure modes, 146–148 service personnel, 146
heating/gas appliance system, 145–146 venting system, 146
industry, 130 Heating value, of gas, 143
installation requirements, 138 Heavy work, 698
manufacturer’s installation instructions in, Hedonic damage calculations, controversial
139–140 aspects of, 707
safety devices, 134–135 Helffenstein, D.A., 632–633
types, 135–138 Helffenstein study
Gas burners battery, of neuropsychological tests, 508
combustion equations, 50–53 MMPI-2, 531–536
Gas flow, 72–73 neuropsychological testing outcome
Gasoline engine, 53–54, 209 academic testing, 523, 525
Gas refrigerators, 65 executive function test, 529–530
Gas sensors, 653 Halstead–Reitan tests, 522, 523, 524
Gas Research Institute, 102, 108, 111–112 index score, 516–518
Gas Research Institute of Des Plaines, 108, information processing speed, 527–528
122 IQ tests, 518–522
Gawey-Apgar, S., 628–630 language comprehension, 530–531
“Gel cell” detectors, 306 memory testing, 522–523, 524, 525
General educational development (GED), 697 motor skills, 528–529
Generator tailpipe exhaust, 83–85 visual–visual perceptual testing, 525–527
Glasgow Coma Scale (GCS), 164, 166, 384 norms, 509
Glen Canyon National Recreation Area persisting symptoms, reported, 511–516
(GCNRA), 158, 159, 160, 165 cognitive symptoms, 512, 515
Glen Canyon NPS decision logic, for patient physical symptoms, 511, 513
triage, 188 psychological/behavioral symptoms, 513,
Gliosis, 347 515–516
Glutamate, 347, 354, 376 visual symptoms, 512, 513–515
8417: “8417_c036” — 2007/9/11 — 18:55 — page 770 — #6

Index 771
reported symptoms, during CO exposure, Hypoxic hypoxia, 316–317

509–511 Hypoxic ischemia, 344
sample and exposure data
admission criteria, 506–507
CO exposure level, 507
duration and frequency of exposure,
507–508 I
sample demographics, 507
source and location exposure, 507 Immediately Dangerous to Life and Health
time since exposure, 508 (IDLH), 291, 645
vocational outcomes, 536–538 Impairment rating, 690
Helmholtz resonance, 81 Impairments, 691, 700
Heme Incomplete combustion, products of, 144
iron, 458 Increment–Decrement Model, 694
oxygenase, 347 Index score, 508, 516–518
proteins, 345, 376 Induced draft system, 137–138
Hibachis, 19 Industrial products and processes, 205
High limit control, 134 Information website, CO, 274–277
History of CO poisoning, 754 Infra-red (IR) sensors, 253, 256–261
Houseboats, 171–174, 178–179 Injury by CO poisoning
Household cooking gas appliances, 135 case report, 703
Household Cooking Gas Appliances (ANSI diminshment of function, 699
Z21.1-1993), 102–103, 112, 127 disability and working, 692–693
Human exposure earning capacity, 693–694
to CO concentrations, 10–17 general educational development, 697
Hydrocarbon gas, 31, 43, 51, 53, 141, 142, health, 684–690
143, 144; see also Butane; Natural gas; less-than-perfect health, quantifying,
Propane 687–690
Hydrogen cyanide (HCN), 643, 652 measuring, 684–687
Hydrogen peroxide, 343 stages, 688
Hydroxycobalamin, 357 monetary damages
“Hyperbaric or Normobaric Oxygen for Acute earning capacity loss, analysis of,
Carbon Monoxide Poisoning: a 699–703
Randomised Controlled Clinical Trial” from population to person, 690–691
(CD Scheinkestel), 406–412 specific vocational preparation, 697–699
Hyperbaric oxygen (HBO) therapy, 241–242, standard vocational interview, 695–696
295–296, 321–322, 344, 348–351, 375, work construct, 696–697
400, 401, 406, 482, 764 Insomnia, 747
for acute CO poisoning, 391, 393 Inspection and maintenance (I/M) programs, 9
basic science overview, 395–398 Institute for Environment and Health, 497
effects, 394–395 Instrumental activities of daily living (IADLs),
human evidence, 398 689, 691
clinical effectiveness, 378–382 Insulin, 352–354
negative trials, 378–379 International Academy of Life Care Planners
positive trials, 379–382 (IALCP), 605
CO effects, at cellular level, 376–377 International Agency for Research in Cancer,
delayed administration, 383 686
functional ability, 747, 748 International Association of Plumbing and
future directions, 384–385 Mechanical Officials (IAPMO), 139
indications, 382–383 International BoatBuilders Exposition, 187
during pregnancy International Classification of Diseases (ICD),
preposed indications, 359, 360 236
repeated treatment, 384 codes, for CO poisoning, 237
“Hyperbaric Oxygen for Acute Carbon International Classification of Functioning,
Monoxide Poisoning” (LK Weaver), Disability and Health (ICF), 690–691
412–420 International classification of impairments,
Hypothermia, 355, 425 diseases and handicaps (ICIDH), 690
International Code Council, Inc. (ICC), 139
8417: “8417_c036” — 2007/9/11 — 18:55 — page 771 — #7

772 Index
International Conference of Building Officials ethical and philosophical reasons, 214–216,

(ICBO), 139 217, 218
Investigation of CO poisoning, 287 litigation-driven safety information,
case study, 297–299 227–230
duration assessment, 292, 294 need, 216–219
exposure level regulations, 219–224, 225
COHb levels, 290 voluntary standards, 219, 220, 221
community exposure standards and workplace area labeling, 211–213
guidelines, 291–292 The Labor Force Status and Other
occupational exposure standards and Characteristics of Individuals by Age,
guidelines, 291 Education Sex, and Disability status,
factors, 296 694
signs and symptoms, 288–289 Lake Cumberland CO poisoning, 174, 178
treatment, 295 Lake George Inlet CO poisoning, 170
Lake Havasu CO poisoning, 179–180
Lake Minnetonka CO poisoning, 169
Lake of the Ozarks CO poisoning, 174
J Lake Powell CO poisoning, 158, 159, 160,
161, 165–166, 168–169, 171, 173–174,
Job coach, 601 177, 178, 179, 180, 183, 184, 189
Judge Large vehicle propulsion engines, 62
opinion, in CO case, 673–674 Leakage location, into an enclosure, 90
Jury, 662, 663, 665, 666 smoke testing, 91
opinion, in CO case, 674 considerations, 91–92
negative pressure testing, 92
positive pressure testing, 92
typical smoke testing, 93–96
K Lees-Haley Fake Bad Scale (FBS), 546–547
Karg Field Protocol, 118 Less-severe CO poisoning, 483–484, 565
development, 108–121 Less-than-perfect health, 693
oven bake burners, field testing, 115–121 quantifying, 687
range top burners healthy life expectancy, 687–689
field testing, 109–112 life-years lost to injury, 689–690
laboratory testing, 112–115 years lived with disability, 687
for measuring CO emissions, from gas Lethal gas combinations, 651–652
ranges Lidocaine, 343
burning testing, preparation for, 124–125 Life Care Plan (LCP), for CO-poisoned
failed burner, 126 patients, 605–606
oven bake burner testing, 125–126 case study, 609–611
range top burner testing, 125 care providers/residential care, 610
safety, during emission testing, 124 complications, 611
visual inspection and customer education diagnostic studies, 610
oven inspection, 123–124 laboratory, 610
range top inspection, 122–123 medical care, 609
objectives, 102 medication, 610
Ketamine, 354 recreation, 610
Kidde, 253 therapeutic modalities, 609
Kluver–Bucy syndrome, 488 transportation, 610
Kool Aid packet, 693–694 development, 606
Life-years lost to injury (LLI), 689–690
Light work, 698
Lipid peroxidation, 288, 346–347, 351,
L 376, 377
Liquefied petroleum (LP)
Labeling, 206; see also Safety information, butane, 50, 141, 142, 143, 202
for CO propane, 50, 52, 123, 124, 141, 142, 143,
custom and practice, 224, 226–227, 228 199, 202, 540
design, 207–209
8417: “8417_c036” — 2007/9/11 — 18:55 — page 772 — #8

Index 773
Litigation, in CO poisoning identification, 159–161

assistant in process, 668 case identification, 161
attorney, 668–669 at Lake Powell, 160, 161
burden of proof, 662 at nationwide, 160–161
damages, 664–666 medical characteristics, 164–166
liability, 663–664 prevention efforts, 180–188
standard of proof, 663 ski boats, 166–169, 177
expert witness, 669 Masimo Corporation, 740
fact witness, 669 Masimo Rad-57, 446–448
gathering information process, 659 Masimo Rainbow SET™ , 447
discovery phase, 661–662 Masimo SET® , 447, 448
filing suit, 660–661 McKenna, P., 630–632
trial–risks and rewards, 662 Mckenzie, Lord, 283
life on trial, 667–668 Medical characteristics, of marine CO
problems, 666–667 poisoning, 164–166
purpose boat-related CO poisoning, 164–165
goal, 657–658 CO, as drowning risk factor, 165–166
motivation, 658–659 COHb measurement limitations, 164
work and abused, 656 nonfatal boat-related CO poisoning, 166
Litigation-driven safety information, 227–230 Medicare Services Advisory Committee of the
Long-term community adjustment themes Department of Health and Aged
for CO-poisoned patients, 603–605 Care, 399
Lopez, Alan, 684, 686 Medium work, 698
Loss of consciousness (LOC), 3, 166, 359, Memory impairment, 361, 480–481, 585
395, 397, 480, 484, 488, 501, 746, 747 Memory testing, 522–523
Low-pressure boilers, 136–137 Methane, 70, 141, 142, 202, 314
LP-gas Methylene chloride exposure, 31–32
physical and chemical properties, 52 nonoccupational exposure, 31
vapor pressures, 53 occupational exposure, 32
Microchemical (MiCS), 253, 254–155
Micro-Electro-Mechanical Systems, 255
Microenvironmental exposures, 17
M commuter exposures, 25–31
airbag deployment, 30
Magnetic resonance imaging (MRI), 323, 356, defective exhaust systems, 26–29
358, 359, 384, 449, 450, 452, 459, 461, drive-up facilities, 29–30
462, 463, 472, 473, 485, 486, 497, 506, motor vehicle emission control program,
570, 571, 610, 635, 748 30–31
Magnetic resonance spectroscopy (MRS), 359, parking garages, 29
449, 450, 461–462, 472, 482 service stations, 29
Malonylaldehyde, 346 occupational exposures, 20–22
“Managing Carbon Monoxide Poisoning with recreational exposures, 23–25
Hyperbaric Oxygen” (JC Raphael), residential exposures, 17–20
405–406 fatal exposures, 19–20
Manufacturer’s installation instructions, in gas nonfatal exposures, 18–19
appliances, 139–140 shopping center exposures, 22
proper fuel input adjustments, 140 Migration, 200–202
Marine CO detector evaluation, 183–184 Mild poisoning, 289
Marine CO poisoning, 157 Mild traumatic brain injury (MTBI), 628, 629,
airborne CO concentrations, on and near 630; see also Brain injury; Traumatic
boats, 175–180 brain injury
evaluation criteria, 176 Minnesota Multiphasic Personality
investigation methods, 175 Inventory (MMPI-2), 506, 531–536,
Lake Havasu study, 179–180 546, 547
cabin cruisers, 170–171, 177–178 Minnesota Multiphasic Personality
descriptive characteristics, 161–163 Inventory-Adolescent (MMPI-A), 531
future directions, 188–190 Mirage software464
houseboats, 171–174, 178–179
8417: “8417_c036” — 2007/9/11 — 18:55 — page 773 — #9

774 Index
Misconceptions, CO N-acetylcysteine (NAC), 351–352

detection, 313, 314, 315 National air monitoring stations (NAMS), 9
miscellaneous, 322–323 National Ambient Air Quality Standards
physiology, 316–317 (NAAQS), 9, 10, 14, 15, 18, 19, 20,
presence, 313 26, 33
properties, 313, 314 guidelines, 7–8
symptoms, of CO poisoning, 318–320 National Association of State Boating Law
treatment and outcome, 320–322 Administrators (NASBLA), 182
MK-801, 354 National Case Listing, 161
Mobile homes, of CO National Center for Environmental Health,
accidents statistics, 59 237, 238
codes and standards, 59–60 Air Pollution and Respiratory Health
fuel-fired appliances Branch, 237
combustion process, 68–69 National Center for Health Statistics (NCHS),
excess air combustion, 71–72 19, 236, 241, 305, 687, 692
excess fuel combustion, 72 Vital Statistics of the US Life Tables, 701
physical and chemical properties, 67–68 National Center for Injury Prevention and
thermochemical equations, 69–71 Control (NCIPC), 236, 238
Moderate poisoning, 289 National Council of State and Territorial
Monetary damages Epidemiologist (CSTE), 243–246
earning capacity loss, analysis of, 699–703 National Electronic Injury Surveillance
Monox, 253 System (NEISS), 240, 241, 243
Morbidity, 241–242, 471, 478, 488–489, 684 National Fire Protection Association (NFPA),
limitations, 245–246 44, 90, 93, 644
More severe CO poisoning, 565 NFPA 54, 139, 140, 150
Mortality, 236, 246, 377, 684 NFPA 501 c, 59, 60
limitations, 244–245 National Health Interview Survey (NHIS),
national CO mortality, studies of, 237–240 689, 691, 692
Motor homes, of CO, 60 National Highway Traffic Safety
interior devices Administration, 689
clothes dryers, 65–66 National Hospital Ambulatory Medical Care
domestic water heaters, 65 Survey (NHAMCS), 241, 242
fireplaces, 66 National Hospital Discharge
gas refrigerators, 65 Survey 2002, 246
oil-fired space heaters, 65 National Human Activity Pattern Survey
ovens, 64–65 (NHAPS) study, 17, 18, 20, 22, 25, 32,
ranges, 64 33, 708
internal combustion engines, 61 National Institute for Occupational Safety and
auxiliary power generators, 62–64 Health (NIOSH), 3, 20, 54, 160, 179,
large vehicle propulsion engines, 62 218, 220, 291
Motor vehicle ceiling limit (CL), 645
emission control program, 30–31 immediately dangerous to life and health
exhaust, in Ala Moana Shopping Center, 22 concentration (IDLH), 645
exhaust gas, 265–267 recommended exposure limit (REL), 645
Multiple Cause-of-Death Mortality Database, National Marine Manufacturers Association
236 (NMMA), 183, 216
Murray, Christopher, 686, 694 National Oceanographic and Atmospheric
Myelin basic protein (MBP), 346–347, 377, Administration (NOAA), 78
397, 478, 570 National Park Service (NPS), 158, 159, 160,
Myleoperoxidase, 346 178, 187, 188
Myoglobin binding, 345 National surveillance system, for CO,
243–246, 684
case definitions, 243–244, 244
morbidity limitations, 245–246
N mortality limitations, 244–245
National Vital Statistics System (NVSS), 236,
NAAQS Exposure Model (NEM), 14 237, 239, 240, 245, 246
N-acetyl-aspartate (NAA), 453, 461
8417: “8417_c036” — 2007/9/11 — 18:55 — page 774 — #10

Index 775
National Workgroup on Carbon Monoxide N-methyl-d-aspartate (NMDA) receptor

Surveillance, 243, 244 antagonists, 347, 354, 376
Natural gas, 100, 109, 123–124, 130, 140, 142, “Non-Comatose Patients with Acute Carbon
143, 145, 199, 200, 314 Monoxide Poisoning: Hyperbaric or
boiling point, 141 Normobaric Oxygenation?” (JL
NEISS All Injury Program (NEISS-AIP), 241, Ducasse), 401–402
242 Nondispersive infrared reference (NDIR), 9,
Neuroimaging 13, 306
after CO exposure, 449 Nonfatal boat-related CO poisoning, 166; see
location and symmetry of lesions also Boat-related CO poisoning
in brain regions, 450–451 Nonfatal exposures, 18–19
sequential studies, 450 Nonoccupational exposure
modalities, 458 methylene chloride, 31
anatomical imaging findings Nor-binaltrophimine, 361
CT, 458–459 Normobaric oxygen (NBO), 321–322, 348,
MRI, 459 349, 350, 356, 379–380, 396, 397, 400,
functional imaging findings, 460 401, 406
fMRI, 461 NPS Emergency Medical Services (EMS),
MRS, 461–462 158, 160, 643, 648, 742
PET, 460–461 medical management, innovations in,
SPECT findings, in CO poisoning, 187–188
462–463
studies, 358–360
Neuro-otological treatment, 626 O
Neuroprotective treatments, for CO poisoning
allopurinol, 351–352 Obsessive–compulsive disorder, 488
brain-derived neuropeptides, 355 Occupation, 696
HBO therapy, 348–351 Occupational Employment Statistics, 708
hypothermia, 355, 425 Occupational exposure, 32
insulin, 352–354 microenvironmental, 20–22
NAC, 351–352 standards and guidelines, for CO, 291
NBO, 321–322, 348, 349, 350, 356, Occupational functioning, 706
379–380, 396, 397, 400, 401, 406 Occupational Health and Safety
NMDA receptor antagonists, 347, 354, 376 Administration (OSHA), 3, 44, 54,
Neuropsychological re-evaluation 222, 226, 252, 291
circumstances, 542 permissible exposure limit (PEL), 645
results, 543–545 Office of Air Quality Planning and Standards
self-reported symptoms, 542–543 (OAQPS), 9
test scores comparison, 543 Ohio survey, 100–101, 102, 126
Neuropsychological testing, for chronic CO Oil-fired space heaters, 65
poisoning, 502–506 Optical gas sensors, see Infra-red sensors
academic testing, 523, 525 Oven, 64–65
executive function test, 529–530 Oven bake burners, 123
Halstead–Reitan tests, 522, 523, 524 emissions, measurement of, 125–126
index score, 516–518 field testing, 115–121
information processing speed, 527–528 Overhaul and postfire, 644–645
IQ tests, 518–522 Oxidative stress, 347–348, 377, 478
language comprehension, 530–531 Oxygen, 644
memory testing, 522–523, 524, 525 Oxygen depletion sensor (ODS), 44, 45, 215,
motor skills, 528–529 217–218
proper fuel input adjustments, 140 application, to small burners, 48–50
visual–visual perceptual testing, Oxygen therapy, 342–344
525–527
Neuropsychological treatment, 626
Nitric oxide synthase (NOS) inhibitor, P
347, 348
Nitrogen, 143 Paresthesia, 747
Parking garages, 29
8417: “8417_c036” — 2007/9/11 — 18:55 — page 775 — #11

776 Index
Parkinsonism, 361 thermal pressure, 76

Pearson Assessment website, 546, 547 total pressure, 73–75
People functions, 696 ventilator opening, 80
Perdue Pegboard test, 720, 721 wind rose, 78
Peroxynitrite, 346 diagnosis, 437, 444
Perry Lake CO poisoning, 173 chronic CO poisoning, 443–446
Persistent neurocognitive sequelae (PNS), Masimo Rad-57, 446–448
358, 359, 400, 413, 414, 464, 481–482 problems, 438, 439–442, 443
Personal exposure monitors (PEMs), 13 situation history, 440
Personal gas monitor (PGM), 645 emotional and behavioral issues, 596–598
Pew Environmental Health Commission, 243 history, 754
Phenyl-n-tert-butyl-nitrone (PBN), 351 investigation, see Investigation of CO
Physiatry pain management, 626 poisoning
Physical characteristics of CO, 67–68, 753 marine, see Marine CO poisoning
Physical demands, 698, 714 nonfatal boat-related CO poisoning, 166
Physical issues, of CO poisoning, 595–596 sequelae management, 361–362
case study, 606–611 Portable monitors, 11
Pilot flame, 45, 68–69, 134 Positron Emission Tomography (PET), 358,
Poisoning 359, 385, 449, 450, 460–461, 472, 473,
acute, see Acute CO poisoning 486, 504
boat-related CO poisoning, 164–165, Postinjury worklife expectancy, 702
177–180 Post-Traumatic Vision Syndrome
case study, 606–611, 725 (PTVS), 514
chronic, see Chronic CO poisoning Practical functioning, 706
cognitive issues, 593–595, 595 Pregnant patients
case study, 606–611 treatment, for CO poisoning, 360
executive dysfunction, 594 Preinjury earning capacity, 700
sensory-motor dysfunction, 594 Preinjury worklife expectancy, 700
short-term memory dysfunction, 594 Present value calculation, 702
vision and information processing Prevention efforts, for marine CO poisoning,
dysfunction, 595 180–188
cognitive sequelae, 478 boat manufacturing/recall authority USCG
DNS, 481–482 regulations for, 180–181
functional imaging, 486 CO detector/alarms
HBO effect on, 484–485 case-based data, 183
lower level, 483–484 evaluation, 183–184
markers, severity and outcome, 484 engineering control research and
neuroimaging findings, comparison, development, 184–187
485–486 marine engine emissions, EPA regulation
PNS, 481–482 for, 181–182
combustion theory application NPS EMS medical management,
air flow around and through 187–188
enclosures, 77 state legislative action, 182
back drafting of flue gases, 83 Probabilistic NEM for CO (pNEM/CO), 14,
chimney and flue dynamics, 82–83 15, 17
cyclical inflow and outflow of Procaine hydrochloride, 343
enclosure, 81 Propane, 50, 52, 123, 124, 130, 142, 143, 199,
deflector, 80 202, 540
exhaust gas flow, 72–73 boiling point, 141
flow through bent or displaced exhaust Proper fuel input adjustments, 140
pipes, 85–86 Psychiatry treatment, 626
generator tailpipe exhaust, 83–85 Public health surveillance, for CO, 233,
Helmholtz resonance, 81 234–236
living quarters, inflow and outflow, 82 CO-caused morbidity, 241–243
missing exhaust system, 86 CO-caused mortality, 236–240
scoop, 80 components, 235
stagnation pressure, 75–76 national surveillance system, 243–246, 684
8417: “8417_c036” — 2007/9/11 — 18:55 — page 776 — #12

Index 777
Public perceptions, about CO, 325 postacute multidisciplinary treatment

best action, 336–338 planning, 599–600
electric generator use, 327–329 tools, for community reintegration, 600
emitting devices, 339 FCE, 600–601
exhaust gas danger, conditions affecting, supported employment, 601
332–333 work hardening, 601
experienced CO poisoning, 338–339 Rehabilitation counselor, 592–593, 595, 596,
greater hazards, 334–335 597, 598, 599, 600, 601, 602, 603, 607,
propane radiant heater use, 329–330 608–609
property, 338 Rehabilitation evaluation, for CO-poisoned
recreational powerboat safety, 331–332 patients, 592–593
safe use indoors, 333 case study, 606–611
time duration, in body, 338 Residential exposures, 17–20
worst poison, 335–336 fatal exposures, 19–20
Pulse CO-oximetry equipment, 446–448, nonfatal exposures, 18–19
740–743 Room air ventilation, for humans
Pulse oximetry, 740 oxygen uptake estimation, for setting, 45–47
Punitive damages, 658 Room heaters, 136
Pyridoxalated hemoglobin-polyoxyethylene Rules of Evidence, 672–673
conjugate (PHP), 343
S
Q Safe Practices for Boat-Towed Watersports
Quantitative Electroencephalogram (QEEG), Act, 182
349, 704 Safety and monitoring devices, 44–45
Quantitative Magnetic Resonance Imaging Safety information, for CO, 206; see also
(QMRI), 358, 459, 463, 497 Labeling; Warning
Quantum Group, 253, 258, 263 custom and practice, 224, 226–227, 228
biomimetic CO gas sensor, 258–259, ethical and philosophical reasons, 214–216,
260, 261 217, 218
litigation-driven safety information,
227–230
need, 216–219
R regulations, 219–224, 225
voluntary standards, 219, 220, 221
“Randomized Prospective Study Comparing written instructional information on,
the Effect of HBO Versus 12 h NBO in 209–211
Noncomatose CO Poisoned Patients: Saxe, John, 722
Results of the Interim Analysis” (D. Scientific methodology, in clinical toxicology,
Mathieu), 404–405 763
Range top burners Sedentary work, 698
emission, measurement of, 125 Segami Corporation, 464
field testing, 109–112 Selected Characteristics of Occupations
inspection, 122–123 (SCO), 697, 699
laboratory testing, 112–115 Select Serotonin Re-uptake Inhibitor (SSRI),
Reactive oxygen species (ROS), 346, 347, 516, 636
348, 349, 458 Self-contained breathing apparatus (SCBA), 3,
Recreational exposures, 23–25 317, 644
at indoor sporting events, 24–25 Semiconducting metal oxide (SMO) gas
on vehicles, 23 sensors, 253, 254–256
Regulations, for CO safety information, Sensor systems, 263–265
219–224, 225 Sensor technologies, for CO gas detection,
Rehabilitation counseling, for CO-poisoned 253–263, 263
patients, 598–599 electrochemical gas sensors, 253, 261–263
job development and prevocational optical gas sensors, 253, 256–261
planning, 602 SMO gas sensors, 253, 254–256
8417: “8417_c036” — 2007/9/11 — 18:55 — page 777 — #13

778 Index
Sequential neuroimaging studies, 450 Standard of proof, 663

Service stations, 29 Standard vocational interview,695–696, 699
Severe poisoning, 289, 309, 397 State and local air monitoring stations
SF-36 Health Survey, 695–696 (SLAMS), 9
Shasta Lake CO poisoning, 171 State and Territorial Injury Prevention
Shopping center exposures, 22 Directors Association (STIPDA), 244
Signal Extraction Technology (SET)® , 446; Stockdale, S., 630
see also Masimo Rainbow SET™ ; Suicide(s), 471
Masimo car exhaust as means of, 265–267, 392
SET® Supervising process, 705
Simulation of Human Activity and Pollutant Supported employment, 601
Exposure (SHAPE), 14, 15 Surveillance system, to health problems, 684
Single photon emission computed tomography Survey of Income and Program
(SPECT), 323, 345, 358, 359, 385, Participation, 701
449, 450, 451, 452, 453, 486, 497, 538,
539–540
brain imaging, 457, 459, 460, 461, T
462–463, 472, 473
for delayed CO-induced encephalopathy, Teak surfing, 168, 169, 177, 182, 190, 331
464–471 Tedlar™ bags, 12, 23
Single-zone mass balance model, 106–108 Temperaments, 698–699
Ski boats, 166–169, 177 Thermal pressure, 76
Small engine exhaust gases Thermochemical equations, 69–71
CO accumulation from, 53–54 Thermocouple, 44, 45, 48, 49, 134
Smoke alarm, 273, 274 Things functions, 696
“Smoke kills” campaign, 273 Tiarks, Frank, 275
Smoke testing, 91 Time, 653
considerations, 91–92 Total pressure, 73–75
negative pressure testing, 92 Toxic Exposure Surveillance System (TESS),
positive pressure testing, 92 241, 242
typical smoke testing, 93–96 Toxicity
Social functioning, 706 in children
Socioeconomic factors, in urban areas, 648 brain injury
Sodium nitrite, 357 mechanism, 570–571
Sources, of CO structural brain injury, evidence of, 571
common investigative tactics, for case study, 577–586
location, 647 development effects
consumer products, 204–205 age as critical variable, 574–575
generation in residence, 647 developmental neuropsychological
industrial products and processes, 205 approach, 575–576
Southern Building Code Congress pediatric vulnerability, 573–574
International, Inc. (SBCCI), 139 neurobehavioral symptoms, 571–572
Space heating appliances symptoms, 569–570
central furnaces, 135–136 white matter injury, 572
low-pressure boilers, 136–137 mechanisms of, 344–348
room heaters, 136 apoptosis, 347–348
wall furnaces, 136 hypoxic ischemia, 344
water heaters, 137 lipid peroxidation, 346–347
special design appliances, 137–138 myoglobin binding, 345
SpCO, 740, 741 oxidative stress, 347–348
Special design appliances, 137–138 vascular oxidative stress, 346
Specific gravity, of gas, 142–143 Tracer gas model
Specific vocational preparation (SVP), for predicting CO accumulation
697–699 in small structures, 47–48
Stagnation pressure, 75–76 Transferable Skills Analysis (TSA), 607–608
Standard Occupational Classification Traumatic brain injury (TBI), 572, 603, 604,
codes, 708 605; see also Brain injury; Mild
traumatic brain injury
8417: “8417_c036” — 2007/9/11 — 18:55 — page 778 — #14

Index 779
“Trial of Normobaric and Hyperbaric Oxygen safety problems, 133

for Acute Carbon Monoxide failure modes, 147
Intoxication” (JC Raphael), 400–401 Vent safety shutoff system, 134–135
Tsongas protocol, 116, 118, 120 Very heavy work, 698
Visual Midline Shift Syndrome
(VMSS), 620
U case study, 634–640
clinical testing, 624–625
U-50488H, 361–362 symptoms, 622–624
UL 2034, 184, 252, 305–306, 307, 309, treatment, 625–626
645–646 with yoked prism, 626
Underwater and Hyperbaric Medical Society rehabilitative and supplementary
(UHMS), 380, 384, 419 considerations, 626–633
Underwriters Laboratory (UL), 44, 305–306 Visual system
CO warning label, 221 function, 620–622
UL 2034, 184, 252, 305–306, 307, 309, Visual–visual perceptual testing,
645–646 525–527
United States Centers for Disease Control and Vocational outcomes, 536–538
Prevention, 6, 160, 161, 236, 237, 239, Voluntary standards, for CO warnings, 219,
240, 241, 242, 243, 329, 686 220, 221
United States Coast Guard (USCG), 158, 159,
160, 161, 177, 185, 187
boat manufacturing/recall authority, W
regulation for, 180–181
United States Consumer Product Safety Wall furnaces, 136
Commission (CPSC), 19, 44, 100, 108, Warning, 206; see also Safety information,
180, 218, 222, 225, 240, 241, 243, 252, for CO
305, 306 custom and practice, 224, 226–227, 228
CO deaths, estimates of, 145 design, 207–209
United States Department of Energy (DOE) layout, 208
program, 100–101, 108, 110, 122 ethical and philosophical reasons, 214–216,
United States Department of Health, 217, 218
277–280, 399 litigation-driven safety information,
United States Department of Health and 227–230
Human Services (DHHS), US, 422, need, 216–219
687, 701 regulations, 219–224, 225
United States Environmental Protection voluntary standards, 219, 220, 221
Agency (EPA), 3, 6, 7, 8, 9, 10, 14, 15, written instructional information,
33, 54, 154, 181, 189, 218, 239, 292, 209–211
315, 334, 708, 749 workplace area warnings, 211–213
marine engine emissions, regulation for, Wash-out rate, of CO, 759–762
181–182 Water Fuse, 272
United States Public Health Service, 227 Water heater, 137
USCG Office of Boating Safety Recreational chronic CO poisoning due to, 505, 540
Boating Product Assurance Division, domestic water heater, 65
180, 183, 184 Web-based Injury Statistics Query and
Reporting System (WISQARS), 236
Wide-ranging Online Data for
V Epidemiologic Research
(WONDER), 236
Vascular oxidative stress, 346, 458 Wind rose, 78, 79
Ventilation, for humans requirements, in small Work construct, 696–697
structures, 48 Worker characteristics components,
room air ventilation 696–697
oxygen uptake estimation, for setting, Worker functions, 696
45–47 Work hardening, 601
Venting system, 131–133, 146 Working Group on Motor Vehicle Exhaust
draft diverter, 132–133 Suicide, 266
8417: “8417_c036” — 2007/9/11 — 18:55 — page 779 — #15

780 Index
Worklife expectancy, 694 X

Workplace area warnings and labeling
general OSHA warning signage, Xanthine oxidase inhibitor, see Allopurinol
211–212
industry warning criteria, 213 Y
specific OSHA exposure warning signage, Years lived with disability (YLD),
212–213 686, 687
Workplace/industry CO standards, Years of life lost (YLL), 686
645–646 Yoked prism therapy, 620
Work-related disability, 701 for VMSS, 626, 638, 639, 640
World Bank, 684, 685, 686 rehabilitative and supplementary
World Health Assembly, 690 considerations, 626–633
World Health Organization (WHO),
3, 236, 252, 619, 684, 686, Z
690, 691
guidelines, for CO, 8 Z21/83 Committee, 138, 148
Written instructional information, on safety Zero impairment rating, 690
information, 209–211 Ziprasidone, 361
8417: “8417_c036” — 2007/9/11 — 18:55 — page 780 — #16

SY CK-BRAIN
03/27/06
Baseline data versus adult normals I
+4
+3
+2
Left lateral view Posterior view Interactive view-no cerebelhm
-2
-3
-4
-5
COLOR FIGURE 20.1 A SPECT scan of cortical function after carbon monoxide poisoning.
The color scale (left side) displays normal perfusion in gray, subnormal perfusion in green and
blue, and hyperperfusion in red. In other cases hypoperfusion is found in the frontal areas.
Thus, the abnormalities found vary from one patient to another. (Credit to J. Michael Uszler)
8417: “8417_c037” — 2007/9/12 — 14:33 — page 1 — #1

Finding No. 1(cont.):
Toxic injury from carbon monoxide poisoning
"CO Normal
poisoned"
COLOR FIGURE 21.2 Example of diffuse neuronal injury two years after acute carbon
monoxide poisoning.
Mild Medium Severe
COLOR FIGURE 21.3 SPECT scans of three patients with mild, medium and severe
cognitive defects two years after acute carbon monoxide poisoning.
8417: “8417_c037” — 2007/9/12 — 14:33 — page 2 — #2

Adult male Adult female
COLOR FIGURE 21.4 SPECT scans of male and female patients two years following acute
carbon monoxide poisoning with identical carboxyhemoglobin levels (34.5% vs. 34.9%)
LC IG TM SP
RL BAW BRW
COLOR FIGURE 21.5 Superior, transverse views of SPECT perfusion findings in isolated
lentiform nuclei of seven patients, two years following acute carbon monoxide poisoning.
Yellow areas in the color plates are areas of abnormally decreased perfusion.
8417: “8417_c037” — 2007/9/12 — 14:33 — page 3 — #3

ROI label # Elts Volume Maximum Minimum Mean Standard
deviation
Caudate nucleus—Left 975 0.3 % 71.1 % 26.9 % 42.6 % 9.3 %
Caudate nucleus—Right 975 0.3 % 74.1 % 37.9 % 52.9 % 7.7 %
Left lateral view Posterior view Interactive view
COLOR FIGURE 21.6 Six SPECT isolation views of caudate nuclei of a patient two years
following acute carbon monoxide poisoning. All areas other than red in the color plates represent
areas of abnormally decreased perfusion.
8417: “8417_c037” — 2007/9/12 — 14:33 — page 4 — #4

Carbon Monoxide Poisoning - D. Penney (CRC, 2008) BBS

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Carbon Monoxide Poisoning - D. Penney (CRC, 2008) BBS

Uploaded by

Copyright:

Available Formats

CARBON

8417: “8417_c000” — 2007/9/25 — 20:28 — page i — #1

Boca Raton London New York

CRC Press is an imprint of the

8417: “8417_c000” — 2007/9/25 — 20:28 — page iii — #3

8417: “8417_c000” — 2007/9/25 — 20:28 — page v — #5

8417: “8417_c000” — 2007/9/25 — 20:28 — page vii — #7

8417: “8417_c000” — 2007/9/25 — 20:28 — page ix — #9

8417: “8417_c000” — 2007/9/25 — 20:28 — page xi — #11

Rob Aiers Kosmas Galatsis, Ph.D.

Thomas M. Dydek, Ph.D., D.A.B.T., P.E. Alastair W.M. Hay, Ph.D.

Peter G. Flachsbart, Ph.D., A.I.C.P. Dennis A. Helffenstein, Ph.D.

8417: “8417_c000” — 2007/9/25 — 20:28 — page xiii — #13

Gunnar Heuser, M.D., Ph.D. Joshua A. Mott, Ph.D.

Gary Hutter, Ph.D.

Jane Brown McCammon, B.S., M.S. Robert E. Schreter, B.S., P.E.

8417: “8417_c000” — 2007/9/25 — 20:28 — page xiv — #14

Christian Tomaszewski, M.S., M.D., Stephen P. Willison, B.S., J.D.

8417: “8417_c000” — 2007/9/25 — 20:28 — page xv — #15

Table of Contents for Carbon Monoxide Toxicity, 2000 . . . . . . . . . . . . . . . . . . . . . . . . . xxi

Table of Contents for Carbon Monoxide, 1996 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxv

8417: “8417_c000” — 2007/9/25 — 20:28 — page xvii — #17

8417: “8417_c000” — 2007/9/25 — 20:28 — page xviii — #18

8417: “8417_c000” — 2007/9/25 — 20:28 — page xix — #19

8417: “8417_c000” — 2007/9/25 — 20:28 — page xx — #20

8417: “8417_c000” — 2007/9/25 — 20:28 — page xxi — #21

8417: “8417_c000” — 2007/9/25 — 20:28 — page xxii — #22

8417: “8417_c000” — 2007/9/25 — 20:28 — page xxiii — #23

8417: “8417_c000” — 2007/9/25 — 20:28 — page xxv — #25

8417: “8417_c000” — 2007/9/25 — 20:28 — page xxvi — #26

Better Education of Physicians: This is essential if CO-poisoned patients are

8417: “8417_c001” — 2007/9/11 — 12:17 — page 1 — #1

New Approaches to Treating Acute, Severe CO Poisoning: There appears to be

Requiring Proper Warnings on Combustion Equipment: The lack of proper and

8417: “8417_c001” — 2007/9/11 — 12:17 — page 2 — #2

Realization that Brain Damage Resulting from CO Poisoning Is not Dependent

Carbon Monoxide Disaster Management: Release of CO during certain kinds

8417: “8417_c001” — 2007/9/11 — 12:17 — page 3 — #3

foolhardy to run a portable generator inside a garage or house, why is it not

8417: “8417_c001” — 2007/9/11 — 12:17 — page 4 — #4

8417: “8417_c002” — 2007/9/11 — 12:16 — page 5 — #1

8417: “8417_c002” — 2007/9/11 — 12:16 — page 6 — #2

2.2 STANDARDS AND GUIDELINES FOR EXPOSURE

8417: “8417_c002” — 2007/9/11 — 12:16 — page 7 — #3

Maximum Concentrations Averaging Times

9 ppm (10 mg/m3 ) 8h

2.3 TRENDS IN CARBON MONOXIDE EMISSIONS

8417: “8417_c002” — 2007/9/11 — 12:16 — page 8 — #4

8417: “8417_c002” — 2007/9/11 — 12:16 — page 9 — #5

2.4 HUMAN EXPOSURE TO CARBON MONOXIDE

8417: “8417_c002” — 2007/9/11 — 12:16 — page 10 — #6

by dividing the person’s integrated exposure by the period of integration (t1 − t0 ).

8417: “8417_c002” — 2007/9/11 — 12:16 — page 11 — #7

1. The CO concentrations measured by the county’s fixed-site monitors were

8417: “8417_c002” — 2007/9/11 — 12:16 — page 12 — #8

8417: “8417_c002” — 2007/9/11 — 12:16 — page 13 — #9

following model advocated by Fugas40 and Duan.41

8417: “8417_c002” — 2007/9/11 — 12:16 — page 14 — #10

8417: “8417_c002” — 2007/9/11 — 12:16 — page 15 — #11

8417: “8417_c002” — 2007/9/11 — 12:16 — page 16 — #12

2.5 MICROENVIRONMENTAL EXPOSURES

2.5.1 RESIDENTIAL EXPOSURES

8417: “8417_c002” — 2007/9/11 — 12:16 — page 17 — #13

2.5.1.1 Nonfatal Exposures