Professional Documents
Culture Documents
Carbon Monoxide Poisoning - D. Penney (CRC, 2008) BBS
Carbon Monoxide Poisoning - D. Penney (CRC, 2008) BBS
MONOXIDE
POISONING
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
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
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
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
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
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
Chapter 20
Neuropsychological Evaluation of the Carbon Monoxide-Poisoned Patient
Dennis A. Helffenstein
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
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
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
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).
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.
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.
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.
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
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
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.
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
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.
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
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%.
Ci (t )
Concentration
t0 t1
Time
FIGURE 2.1 Exposure of person i to air pollutant concentration (C) as a function of time t.
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
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:
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
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
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
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
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.
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
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.
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
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
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).
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.
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
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.
NA = not available
Carbon Monoxide Poisoning
Exposure to Ambient and Microenvironmental Concentrations 29
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.
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.
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
2.7 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.
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
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.
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.
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.
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.
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.
Environ. Health, 50, 207, 1995.
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.
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.
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
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
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.
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.
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:
The mathematical variables used in the following equations and calculations are
defined as follows with their associated units denoted in [… ..]:
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:
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.
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
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
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
chamber is well mixed and that CO is not absorbed or adsorbed into the tent material
or its contents.
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
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.)
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.
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,
Source: Handbook Propane–Butane Gases, 4th ed., Chilton Company, Los Angeles, CA, PP. 22,23 1962.
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:
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:
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.
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.
3.10 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.
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.
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
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
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.
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
E H
B F G
D
I
J K
C
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
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.
4.4.2.2 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
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.
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.
Fresh air
Leakage air
Dryer exhaust
Exhaust gas
Dryer exhaust back flow
Clothes dryer
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
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.
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.
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.
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
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.
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).
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
ht = hs + hv + f (4.7)
where
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
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.
2
Um
h = ∗ δ ∗ sin φ (4.8)
1096.5
Where
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.
1.00
0.80
0.60
0.40
0.20
0.00
0 10 20 30 40 50 60
Wind speed MPH
Where
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.
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).
A B
C E
A B G
− H
+ −
C + −
+ − −
+
+
F
D
E
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.
1 c2∗ S
fo = ∗ ∗ (4.10)
2 π lc∗ V
Where
Where
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
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.
Wall
Trapped exhaust gas
Direction of flow
of exhaust gas
Motor home
Entrained exhaust
Tailpipe
Exhaust discharge
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.
Total volume of exhaust plus leakage air into the enclosure (CFM).
where
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.
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)
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.
where
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
1200 β 2, n
Co concentration (ppm)
900
600
300
0
0 20 40 60 80 100
tn
Time (min)
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.
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.
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
FIGURE 4.19 Looking aft second leak area, 2 min 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.
FIGURE 4.21 Second leak, totally obscured cabinet, 3 min elapsed time.
DEN*100
12
CO2
10
CO
8
O2
6
4
2
0
0 1 2 3 4 5 6
Equivalence ratio
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.
CONTENTS
99
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
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
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
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.
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
(a) (b)
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.
where
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
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 .
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
5.1.5.1 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.
(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.
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
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.
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
CO as-measured (ppm)
5
Chamber purged
4
0 15 30 45
Time after startup (minutes)
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.
CO as-measured (ppm)
250
40
CO air-free (ppm)
200
30
150 CO as-measured ppm
(scale at rightaxis)
20
100
50 10
0 0
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Time after startup (minutes)
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
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.
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.
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
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.
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.
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
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.
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.
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.
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)
2000
15–30 min (ppm)
1500
R = 0.56
1000
R2 = 0.31
500
0
0 200 300 400 500 600 700 800
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.
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).
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.
period has expired, calibrate the instrument before use. Zero the instrument according
to the manufacturer’s recommendations before testing.
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
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.
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.
CONTENTS
129
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.
• 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.
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
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
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.
• 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.
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.
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.
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.
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.
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.
ESIG N
D
ER D
C
TIFIE ®
FIGURE 6.2 CSA Star.
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
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
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
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
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.
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
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.
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
system. As a minimum, the following external components or factors, along with the
gas-fired appliance make up the system
The following failure modes can result in incomplete combustion causing the
appliance to produce CO:
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.
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.
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
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.
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.
(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
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.
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.
CONTENTS
157
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.
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).
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
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
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
other poisonings (higher in the summer instead of higher in the winter) and reflects
the pattern of boat use.
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
Zippered
canopy
Swim opening
platform
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%.
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.
Surfer
Displacement wave
(b)
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
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%.
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%.
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.
Living
quarters/
boat cabin
Swim
step Extended deck Engine compartment
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”.
Swim platform
Used by permission.
Copyrigh 2007 Boat Ed
www.boat-ed.com
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
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.
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.
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.
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
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
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.
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
[Requirements.]
[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.
Nevada followed in 200439 Oregon in 200540 and Washington state41 and Utah42
in 2006, each state adopting and embellishing the NASBLA model language.
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.
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
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
NO YES
Patient exhaled COHb level > 20%
(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.
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.
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:
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
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.
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
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
http://www.cdc.gov/niosh/surveyreports/pdfs/ectb-171-27a.pdf
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
http://www.cdc.gov/niosh/surveyreports/pdfs/ectb-171-32a.pdf
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
http://www.cdc.gov/niosh/surveyreports/pdfs/ectb-171-36a.pdf
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
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.
CONTENTS
197
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.
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.)
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.
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
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
reduced to 50 ppm. This exponential dilution decay mechanism allows for significant
migration before the gas is rendered harmless (Figure 8.1).
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.
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
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.
Arecent development on some chemical containers for lawn, garden, and pesticide
products is a “fold out” label that provides additional space for written information.
Signal word
Alert symbol
Hazard
identification
Result of
ignoring
warning How to avoid
hazard
Additional
information
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.
This manual also contains a text under the title “Safety Information” that states
Under “Maintenance Safety,” there are additional “Safety Precautions” that state
(It should be noted that the exhaust emissions are met in part by the use of a catalytic
converter system.)
_____________________________________________________________________________________________
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.
• 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.
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
Additional information is provided about how to avoid this hazard, what to look
for in detecting this hazard, and other precautions.
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.
DANGER
ASBESTOS
CANCER AND LUNG DISEASE HAZARD
AUTHORIZED PERSONNEL ONLY
DANGER
CADMIUM
CANCER HAZARD
CAN CAUSE LUNG AND KIDNEY DISEASE
AUTHORIZED PERSONNEL ONLY
RESPIRATORS REQUIRED IN THIS AREA
DANGER
CONTAINS CADMIUM
CANCER HAZARD
AVOID CREATING DUST
CAN CAUSE LUNG AND KIDNEY DISEASE
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
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.
FIGURE 8.8 Warning information from Office of Gas Safety Level 1, Wool House, 369
Royal Parade, Parkville, Victoria 3052.
FIGURE 8.9 Warning from National Marine Manufacturers Association, 200 East Randolph
Drive, Suite 5100 Chicago, IL 60601-6528, Available at: www.nmma.org.
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 :
Carbon Monoxide
Fire and Explosion
Electrocution
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
!WARNING
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).
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.
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
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
“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 8.14 Society of Automotive Engineers (SAE) “Test Procedure for Measuring Car-
bon Monoxide Concentrations in Vehicle Passenger Compartments,” SAE J989, Approved
1968, 1971 Handbook.
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.
FIGURE 8.16 ANSI Z 21.49—1989 “Gas Fired Gravity and Fan Type Vented Wall Furnaces.”
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
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.
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.
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
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
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:
Section (f) contains detailed labeling requirements addressing the use of:
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.
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.
WARNING
CARBON MONOXIDE
FIGURE 8.21 State of Tennessee OSHA F:\TN Department of Labor and Workforce
Development.htm “Special Emphasis Program—Carbon Monoxide, The Hidden Killer.”
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.
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.
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.
FIGURE 8.24 Title 15—Commerce and Trade Chapter 36—Cigarette Labeling and Advert-
ising §1331. Congressional declaration of policy and purpose.
FIGURE 8.25 109th U.S. Congress (2005–2006) S. 2084: Portable Generator Safety Act,
Introduced December 13, 2005.
FIGURE 8.26 Consumer Product Safety Commission, 6 CFR Part 1500. Available at:
http://www.cpsc.gov/businfo/frnotices/fr95/95-40785.html.
CAUTION
CARBON MONOXIDE
MAY BE PRESENT
a Commission judge agreed with the Secretary that both OSHA’s standards and industry
custom and practice plainly distinguish between lifelines and lanyards.
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
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
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.
“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.
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:
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
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….’
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.
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.
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
3. Completeness
4. Placement
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.
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.
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.
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.
CONTENTS
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
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.
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
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.
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
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
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.
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.
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
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.
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.
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
Suicides
12 120
CDR per1,000,000 person-
CO Emissions control timeline
10 100
fleet.
4 100
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.
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
FIGURE 9.3 The nonfire-related, carbon monoxide poisoning pyramid, 2002, U.S.
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).
CONTENTS
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
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.
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
optimized for various toxic gases. However, for aftermarket vehicle CO detectors,
optical sensors from Quantum have dominated.
σ = kcn
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).
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.
(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
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).
(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)
I = Io e(−kc l)
(a)
1014
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)
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).
(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).
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.
Capillary diffusion
barrier
Sensing
electrode
Separator
Counter
electrode
Current
collectors
Electrolyte
Sensor pins
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).
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.
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:
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:
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).
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.
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.
CONTENTS
271
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
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.
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.
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,
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.
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
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
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.
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.
11.7 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
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,
CONTENTS
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
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.
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.
C×t =k
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 TEEL-1 83 15 min
AIHA TEEL-2 83 15 min
AIHA TEEL-3 330 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.
TEEL-2, the maximum concentration in air below which it is believed nearly all individuals
could be exposed without experiencing or developing irreversible or other serious health
effects or symptoms that could impair their abilities to take protective action.
TEEL-3, the maximum concentration in air below which it is believed nearly all individuals
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.
ERPG-2, the maximum concentration in air below which it is believed nearly all individuals
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.
ERPG-3, the maximum concentration in air below which it is believed nearly all individuals
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.
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.
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
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
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.
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.
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.
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.
CONTENTS
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
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
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
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.
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
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.
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.
CONTENTS
313
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).
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.
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.
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.
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.
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.
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.
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.
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.
CONTENTS
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
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
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
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
TABLE 15.3
When Using a Recreational Powerboat, What is Safe (one only)?
Percent of Total
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.
∗ “medical”—all adults, MI and FL, with self-reported medical, biological occupations.
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
TABLE 15.4
What Weather Condition Might Greatly Affect the Levels of Exhaust Fumes
on Board a Boat?
Percent of Total
TABLE 15.5
Which One of the Following May Safely be Used Indoors?
Percent of Total
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
TABLE 15.7
Which Poison Debilitates or Accidentally Kills the Most People Each Year in
the USA?
Percent of Total
risky behaviors, especially, sitting on the transom with the engine running by the
Florida youth.
TABLE 15.8
If Carbon Monoxide Exposure is Suspected, What is the Best Advice?
Percent of Total
Male Female Range Mean Age A B C D E
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
Florida (rows 1–4 public; 5–8 private)
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.
*“medical”—all adults, MI and FL, with self-reported medical, biological occupations.
with boat speed and direction, greatly influence the concentrations of CO found on
boat decks, in cabins and immediately behind the boat.
TABLE 15.9
How Long Does Carbon Monoxide Stay in the Human Body?
Percent of Total
Male Female Range Mean Age A B C D E
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
Florida (rows 1–4 public; 5–8 private)
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.
TABLE 15.10
Which One of the Following is a Property of Carbon Monoxide?
Percent of Total
Male Female Range Mean Age A B C D E
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
Florida (rows 1–4 public; 5–8 private)
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.
* “medical”—all adults, MI and FL, with self-reported medical, biological occupations.
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.
TABLE 15.11
Have YOU Experienced the Toxic Effects of Carbon Monoxide Poisoning?
Percent of Total
A = Yes, once; B = Yes, more than once; C = No; D = No, but I know of someone who has; E = I don’t
know.
* “medical”—all adults, MI and FL, with self-reported medical, biological occupations.
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.
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
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.
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.
CONTENTS
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
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.
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
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
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
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
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
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.”
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.
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.
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.
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
16.5.1 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
TABLE 16.1
Proposed Indications for Hyperbaric Oxygen Therapy in
Pregnancy
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
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.
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.
4. Smith, J.S., and Brandon, S. Morbidity from acute carbon monoxide poisoning at
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.
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.
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.
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.
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.
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.
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.
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.
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
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.
222. Seger, D., and Welch, L. Carbon monoxide controversies: neuropsychologic testing,
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.
230. Smith, J.S., and Brandon, S. Morbidity from acute carbon monoxide poisoning at
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.
CONTENTS
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
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.
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.
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
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.
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.)
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
17.8 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.
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.
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.
59. Smith, J.S., and Brandon, S. Morbidity from acute carbon monoxide poisoning at
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
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.
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.
CONTENTS
391
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
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.
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.
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.
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.
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:
• 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
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
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)
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,
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).
• 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,
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.
• 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
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
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.
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
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
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
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:
• 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
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.
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).”
• 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
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.
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
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.
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.
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.
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
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.
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.
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.
CONTENTS
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
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
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.
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.
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
Delayed presentation COHb may be low
+100% oxygen COHb may be low
• Chronic (>24 h) Immediate presentation COHb high to low
Delayed presentation COHb may be 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.
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
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.
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]
For every single case of chronic carbon monoxide poisoning reported / successfully
diagnosed, there are 10 cases that go unreported/undiscovered/undiagnosed.
Cycle of misdiagnosis
Chronic exposure to CO
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
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.
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.
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
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
7. Penney, D.G. Effects of carbon monoxide exposure on developing animals and
humans. In: Carbon Monoxide, D.G. Penney, Ed., CRC Press, NY, 1996, pp. 109–144.
CONTENTS
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
SY CK-BRAIN 03/27/06
+4
Baseline data versus adult normals I
Right lateral view Anterior view Superior view
+3
+2
-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)
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.
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
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.
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.
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.
33. 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 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.
CONTENTS
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
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.
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).
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.
TABLE 21.1
SPECT Areas of Decreased Perfusion Seen in Studies of
Carbon Monoxide Poisoning
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
Basal ganglia Decreased
2004 Pach23 Parietal cortex Decreased
Frontal lobes Decreased
Basal ganglia 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.
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.
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
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).
"CO Normal
poisoned"
FIGURE 21.2 (See color insert following page 422). Example of diffuse neuronal injury
two years after acute carbon monoxide poisoning.
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.
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.
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.
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.
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.
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.
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.
CONTENTS
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
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
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.
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.
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.
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
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.
22.5 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
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.
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.
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.
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.
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.
CONTENTS
495
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),
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
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).
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.
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.
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
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).
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
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
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.
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.
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.
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.
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.
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,
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.
TABLE 23.6
Persisting Physical Symptoms, Helffenstein Study (N = 21)
Symptom % of Participants Reporting
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
TABLE 23.7
Persisting Visual Symptoms, Helffenstein Study (N = 21)
Symptom % of Participants Reporting
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)
Symptom % of Participants Reporting
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
TABLE 23.9
Persisting Psychological/Behavioral Symptoms, Helffenstein
Study (N = 21)
Symptom % of Participants Reporting
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.
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:
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:
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,
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
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
TABLE 23.14
WAIS-III Subtests: Percentage of
Participants who Scored in the
Below Average to Impaired Range
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.
TABLE 23.15
WAIS-III Summary Scores: Percentage of
Participants who Scored in the Below
Average to Impaired Range
TABLE 23.16
Halstead–Reitan Tests: Percentage of Participants who Scored in each Range
of Performance
Mild/ Mod./
Above BA/ Mild Moderate Moderate Severe
Average Average Bdl. Impairment Impairment Impairment Impairment
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.
TABLE 23.17
Halstead–Reitan Tests:Percentage
of Participants who Scored in the
Below Average to Impaired Range
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
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/
Above BA/ Mild Moderate Moderate Severe
Average Average Bdl. Impairment Impairment Impairment Impairment
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
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.
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
Average to Impaired Range
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.
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
Average Average Bdl. Impairment Impairment Impairment Impairment
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.
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.
TABLE 23.25
Other Motor Functions: Percentage of
Participants who Scored in the Below
Average to Impaired Range
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
TABLE 23.26
Miscellaneous Tests of Executive Function: Percentage of Participants Scoring
in each Range of Performance
Mild/ Moderate/
Above BA/ Mild Moderate Moderate Severe
Average Average Bdl. Impairment Impairment Impairment Impairment
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.
TABLE 23.27
Executive Function Tests: Percentage
of Participants who Scored in the
Below Average to Impaired Range
PASAT 62
Behavioral Dyscontrol Scale 60
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/
Above BA/ Mild Moderate Moderate Severe
Average Average Bdl. Impairment Impairment Impairment Impairment
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.
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
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
TABLE 23.30
MMPI-2 Content Scales Sum-
mary (N = 19)
Scale T-Score Mean T-Score Range
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
TABLE 23.32
MMPI-2 Restructured Clinical
Scales Summary (N = 16)
Scale T-Score Mean T-Score Range
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
FIGURE 23.2 Group MMPI-2 Restructured Clinical (RC) scales profile of 16 patients
chronically exposed to CO-Helffenstein Study
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.
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
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
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.
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
Behavioral Dyscontrol Scale 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
never been diagnosed with any type of learning disability nor had she ever received
special education or remedial services.
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.
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.
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
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
References
1. 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 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.
15. 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.
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-
chological correlates of carbon monoxide exposure: A case report, Environ. Health
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.
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.
CONTENTS
551
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,
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.
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
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
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
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
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).
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.
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
Smokers
Chronic Carbon Monoxide Poisoning: A Case Series
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
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
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
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
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.
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
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.
CONTENTS
569
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.
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
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
25.7.1 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
TABLE 25.1
Summary of Psychometric Test and Clinical Data for Four Case Studies
Cases
(Continued)
Cases
TABLE 25.1
(Continued)
Cases
TABLE 25.1
(Continued)
Cases
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
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.
Environ. Health, 27, 349, 1973.
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.
Neuroimaging, 15, 144, 2005.
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.
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.
Neuroimaging, 6, 161, 1996.
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.
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.
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.
CONTENTS
591
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.
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
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
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)
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.
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.
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
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.
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.
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.
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.
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.
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.
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.
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
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.
26.5.5.6 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.
26.5.5.8 Transportation
The cost of mileage reimbursement for travel to medical and rehabilitative appoint-
ments is included in the LCP.
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.
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.
dementia.
Life Care Plan
Medications∗
Item or Service Duration Frequency Purpose Estimated Cost Supplier/Vendor
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
CONTENTS
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
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.
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.
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.
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.
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.
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.
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.
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.
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
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.
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.
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
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
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.
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
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
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
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,
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.
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.
5. Penney, D.G. Chronic carbon monoxide poisoning. In Carbon Monoxide Toxicity,
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.
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.
CONTENTS
643
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.
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.
∗ 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.
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
• 200 ppm for no more than 35 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
• 100 ppm for no more than 90 min before alarming, or
• 200 ppm for no more than 35 min before alarming, 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.
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)
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).
* 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
_________________________________________________________________________________________________________ .
__________________________
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.
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.
TABLE 28.4A
Summary of “Test Challenge” of Smoke Hoods
Test Challenge/Dose Concentrations (ppm)
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)
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.
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.
28.9 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.
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.
CONTENTS
655
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.
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.
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.
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.
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.
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.”
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.
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.
29.3.2.1 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
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.
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.
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.
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.
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.
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.
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.
29.7 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.
CONTENTS
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
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.
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.”
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.
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.
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.
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.
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.
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).
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:
12 F.R.E. 702(2).
13 Kumho Tire Company v Carmichael 119 SCT 1167 (1999).
14 Id. at 1175.
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.
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.”
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.
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).
30.6 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.
CONTENTS
683
Overall health
Existential/
Self-perceived Abilityto experiential
(self-rated) function symptoms and
health capabilities
Pain
Physical function
Energy/vitality
Mobility
Social/role Happiness/
function mood
Self-image
Major life role
Recreation Sensory
Cognitive Vision
function
Hearing
Memory
Problem
solving
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
Mr. Smith's life expectancy would have been to age 77.7 had he
not died due to carbon monoxide poisoning at age 50.
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
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
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.
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.
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.
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
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.”
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
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.
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
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”?
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:
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”:
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;
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
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
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
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).
• 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
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.
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:
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.
0 5 10 15 20 25 30 35
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
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
Table 31.4
(Continued)
A B C D E F G H I J K L
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
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
Table 31.5
(Continued)
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
TABLE 31.6
Household Services Analysis
0 5 10 15 20 25 30 35 40 45 50
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
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.
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).
0 5 10 15 20 25 30 35
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
TABLE 31.8
Loss of Future Earnings Data
A B C D E F G H I J K L
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
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.
0 10 20 30 40 50 60 70
0 10 20 30 40 50 60
Cathy’s full-functioning healthy life expectancy would have been 63.5 years
0 10 20 30 40 50 60 70 80
0 10 20 30 40 50 60
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
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
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
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
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
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
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.
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
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
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
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.
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.
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.
CONTENTS
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
10
1 Methemoglobin
Extinction coefficient
Oxyhemoglobin
Reduced
.1 hemoglobin
Carboxyhemoglobin
.01
600 640 680 720 760 800 840 880 920 960 1000
Wavelength (nm)
35.0
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
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
Yes No
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.
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.
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.
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
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
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.
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.
7. Penney, D.G. Chronic carbon monoxide poisoning. In Carbon Monoxide Toxicity,
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.
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.
19. Penney, D.G. Effects of carbon monoxide exposure on developing animals and
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.
CONTENTS
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
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
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
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
100 LT WORK
PULSE 110
20 30 40 50 60 70 80 90
10
min
HD WORK
PULSE 135
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)
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)
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
TABLE 35.5
Remaining Factor (= 100–Depletion Factor); Hyperbaric Condition =
1580 mmHg (2 × Normobaric)
O2 level → 21% Normobaric 100% Normobaric 100% Hyperbaric
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)
O2 level → 21% Normobaric 100% Normobaric 100% Hyperbaric
Activity → Rest 2 × Rest 3 × Rest Rest 2 × Rest Rest
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:
Ambient
PCO CO body stores
Exogenous
Alveolar Endogenous
PCO
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.
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.
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
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.
765
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
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
+3
+2
-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)
"CO Normal
poisoned"
COLOR FIGURE 21.2 Example of diffuse neuronal injury two years after acute carbon
monoxide poisoning.
COLOR FIGURE 21.3 SPECT scans of three patients with mild, medium and severe
cognitive defects two years after acute carbon monoxide poisoning.
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.
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.