Professional Documents
Culture Documents
Ebook Modern Epidemiology Fourth Edition Timothy L Lash Online PDF All Chapter
Ebook Modern Epidemiology Fourth Edition Timothy L Lash Online PDF All Chapter
Ebook Modern Epidemiology Fourth Edition Timothy L Lash Online PDF All Chapter
Timothy L. Lash
Visit to download the full and correct content document:
https://ebookmeta.com/product/modern-epidemiology-fourth-edition-timothy-l-lash/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
https://ebookmeta.com/product/modern-epidemiology-4th-edition-
kenneth-rothman-timothy-l-lash/
https://ebookmeta.com/product/applying-quantitative-bias-
analysis-to-epidemiologic-data-2nd-edition-timothy-l-lash/
https://ebookmeta.com/product/ethics-and-epidemiology-2nd-
edition-steven-s-coughlin-tom-l-beauchamp/
https://ebookmeta.com/product/modern-occupational-diseases-
diagnosis-epidemiology-management-and-prevention-ki-moon-bang/
Endurance Sports Medicine A Clinical Guide Timothy L.
Miller
https://ebookmeta.com/product/endurance-sports-medicine-a-
clinical-guide-timothy-l-miller/
https://ebookmeta.com/product/equity-and-trusts-fourth-edition-
edition-bradley-l-jones/
https://ebookmeta.com/product/veterinary-public-health-
epidemiology-veterinary-public-health-epidemiology-zoonosis-one-
health-krishna-gopal-narayan/
https://ebookmeta.com/product/epidemiology-for-dummies-amal-k-
mitra/
https://ebookmeta.com/product/mathematical-models-in-
epidemiology-brauer-fred/
Modern Epidemiology
FOURTH EDITION
Timothy L. Lash
O. Wayne Rollins Distinguished Professor of Epidemiology and Chair
Department of Epidemiology
Rollins School of Public Health, Emory University
Atlanta, Georgia
Tyler J. VanderWeele
John L. Loeb and Frances Lehman Loeb Professor of Epidemiology
Departments of Epidemiology and Biostatistics
Harvard T.H. Chan School of Public Health
Boston, Massachusetts
Sebastien Haneuse
Professor of Biostatistics
Department of Biostatistics
Harvard T. H. Chan School of Public Health
Boston, Massachusetts
Kenneth J. Rothman
Distinguished Fellow, Research Triangle Institute
Research Triangle Park, North Carolina
and
Professor of Epidemiology
Boston University School of Public Health
Boston, Massachusetts
Executive Editor: Sharon Zinner
Development Editor: Sean McGuire
Editorial Coordinator: Cody Adams, Julie Kostelnik
Production Project Manager: Catherine O
Design Coordinator: Steve Druding
Manufacturing Coordinator: Beth Welsh
Prepress Vendor: TNQ Technologies
9 8 7 6 5 4 3 2 1
Printed in Mexico
ISBN-13: 978-1-4511-9328-2
shop.lww.com
Contents
PART I
Foundations
4 Measures of Occurrence
Timothy L. Lash and Kenneth J. Rothman
PART II
Study Design and Interpretation
7 Cohort Studies
Kenneth J. Rothman, Tyler J. VanderWeele, and Timothy
L. Lash
8 Case-Control Studies
Timothy L. Lash and Kenneth J. Rothman
10 Field Methods
Lauren A. Wise and Patricia Hartge
PART III
Data Analysis
22 Time-To-Event Analysis
Sebastien Haneuse
26 Analysis of Interaction
Tyler J. VanderWeele, Timothy L. Lash, and Kenneth J.
Rothman
27 Mediation Analysis
Tyler J. VanderWeele
29 Bias Analysis
Timothy L. Lash
31 Agent-Based Modeling
Brandon D.L. Marshall
PART IV
Special Topics
33 Reproductive Epidemiology
Clarice R. Weinberg, Allen J. Wilcox, and Anne Marie
Jukic
34 Psychiatric Epidemiology
Katherine M. Keyes, Sharon B. Schwartz, and Ezra S.
Susser
35 Clinical Epidemiology
Jan P. Vandenbroucke and Henrik Toft Sørensen
36 Molecular Epidemiology
Claire H. Pernar, Konrad H. Stopsack, and Lorelei Mucci
37 Genetic Epidemiology
John S. Witte and Duncan C. Thomas
39 Social Epidemiology
Jay S. Kaufman
40 Environmental Epidemiology
Irva Hertz-Picciotto and Stephanie M. Engel
41 Occupational Epidemiology
David Richardson
42 Nutritional Epidemiology
Walter C. Willett and Frank B. Hu
43 Pharmacoepidemiology
Sebastian Schneeweiss and Krista F. Huybrechts
Index
Preface and Acknowledgments
Ma hew P. Abdel MD
Professor of Orthopedic Surgery
Mayo Clinic College of Medicine
Consultant
Department of Orthopedic Surgery
Mayo Clinic
Rochester, Minnesota
James W. Buehler, MD
Clinical Professor & Interim Chair
Department of Health Management & Policy
Dornsife School of Public Health
Drexel University
Philadelphia, Pennsylvania
John S. Wi e, PhD
Professor
Department of Epidemiology and Biostatistics
University of California San Francisco
San Francisco, California
PART I
Foundations
CHAPTER 1
The Scope of Epidemiology
Inductivism
Modern science began to emerge around the 16th and 17th centuries,
when the knowledge demands of emerging technologies stimulated
inquiry into the origins of knowledge. An early codification of the
scientific method was Francis Bacon’s Novum Organum, which, in
1620, presented an inductivist view of science. In this philosophy,
scientific reasoning is said to depend on making generalizations, or
inductions, starting from observed pa erns and progressing to
general laws of nature; the observations are said to induce the
formulation of a natural law in the mind of the scientist. Thus, an
inductivist would have said that Jenner’s observation of lack of
smallpox among milkmaids induced in Jenner’s mind the theory that
cowpox (common among milkmaids) conferred immunity to
smallpox.
Inductivist philosophy was systematized in the canons of John
Stuart Mill,37 which evolved into inferential guidelines that are still
in use today, such as those of Bradford Hill.38 These guidelines are
discussed in detail in Chapter 2.
Refutationism
Inductivist philosophy provided numerous important insights.
Nonetheless, in the 18th century, the Sco ish philosopher David
Hume described a disturbing deficiency in inductivism. Hume
showed that an inductive argument carried no logical force in the
deductive sense familiar in mathematics; instead, such an argument
represented an assumption that certain events would in the future
follow the same pa ern as they had in the past. Thus, to argue that if
one threw a ball in the air, it would eventually slow down and
reverse course to fall to earth because that had always been observed
in the past corresponded to an assumption that the pa ern observed
to date (objects thrown upward slow down, reverse course, and
return to earth) would continue into the future. In the same way, to
argue that cowpox caused immunity to smallpox because no one got
smallpox after having cowpox corresponded to an assumption that
the pa ern observed to date (no smallpox after cowpox) would
continue into the future.
Hume pointed out that, even for the most reasonable sounding of
such assumptions, there was no logical necessity behind the
inductive argument. Of central concern to Hume was the issue of
causal inference and failure of induction to provide a foundation for
it39
Frequentism
In light of the limits of investigative methods, our inferences and our
store of accepted facts will inevitably contain errors to some
unknown degree. This conclusion, labeled fallibilism, was reached by
the American pragmatic philosopher Charles Sanders Peirce in the
late 19th century.47 To address these limitations, the geneticist and
statistician R. A. Fisher developed a theory and methodology of
inference based on experiments designed to provide known
frequencies of results under explicit, ideal conditions.48, 49
A pivotal requirement of Fisher’s frequentist methodology was that
the experiment incorporate a device or instrument that generated
treatment assignments (allocations) in a random fashion, with
known frequencies for all possible assignment pa erns. This known
behavior of the assignment instrument (also known as the allocation
mechanism) could then be used to compute the probabilities of
subsequent observations under a specific hypothesis about the effect
of treatment, which formed the basis of Fisher’s statistical tests.
Typically, the hypothesis chosen for the computation—the test
hypothesis—was that the treatment had no effect, which today is
called the causal null hypothesis. Other hypotheses could be used
instead, but experimental design and subsequent computations are
simplest for this null hypothesis. In Fisher’s time, before the advent
of electronic computers, this simplicity encouraged a focus on the
null hypothesis—a focus that remains, inappropriately, to the
present day in most statistics texts and analyses.50
Borrowing the terms “statistical significance” and “significance
test” from earlier statistical writers—but not their meaning of the
term—Fisher computed what he called a “level of significance” for
the experimental outcome in relation to the hypothesis tested, which
in more neutral terms is known as a P-value. For simple experiments
designed to “test” the null, this P-value was the frequency
(expressed as a probability) with which one should expect the
experimental design to produce an effect estimate as large as
observed or larger if in fact there was no effect of the treatment.
There are many subtleties and qualifications that should be a ached
to the P-value and this description, which we will review at length in
Chapter 15. For now, the most important one, which Fisher also
emphasized, is that the P-value as Fisher defined it is not the
probability of the tested hypothesis. Unfortunately, it is often
misinterpreted as if it were that probability.
A P-value may, however, be viewed as measuring, under ideal
conditions, the compatibility or consistency between the
experimental outcome and the tested hypothesis, with P = 1
indicating that the outcome is exactly what the hypothesis predicted,
and P = 0 indicating that the outcome is impossible under the
hypothesis. In this usage, it is crucial to note that P = 1 does not mean
the hypothesis is correct, for there may be many other hypotheses
that are also highly compatible with the outcome. In addition, the
results may only reflect departures from the ideal se ing assumed
by the computation of the P-value; the result may have arisen from
invalid assumptions, from computing errors, and in the extreme, a
clever analyst can manipulate the data or methods to produce a large
or small P-value.
Another random document with
no related content on Scribd:
only in undertones, and make no sound, save as their clogs clatter
on stones and gravel.
It is impossible to carry away more than a general and bewildered
impression of the splendid walled and lanterned courts, the superb
gate-ways, and the temples themselves, but certain details, upon
which the guides insist, remain strangely clear in memory. Over the
doors of the stable where the sacred white pony is kept are colored
carvings representing groups of monkeys with eyes, or ears, or
mouth covered with their paws—the signification being that one
should neither see, hear, nor speak any evil. In one superbly-carved
gate-way is a little medallion of two tigers, so cunningly studied and
worked out that the curving grain and knots of the wood give all the
softly-shaded stripes of their velvet coats and an effect of thick fur.
One section of a carved column in this gate is purposely placed
upsidedown, the builder fearing to complete so perfect and
marvellous a piece of workmanship. Above another gate-way curls a
comfortable sleeping cat, which is declared to wink when rain is
coming, and this white cat has as great a fame as anything along the
Daiyagawa.
The strangest hierophant in Nikko is the priestess who dances at
the temple of Iyeyasu. She looks her three-score years of age, and is
allowed a small temple to herself, where she sits, posed like an altar
image, with a big money-box on the sacred red steps before her, into
which the pious and the curious toss their offerings. Then the
priestess rises and solemnly walks a few steps this way, a few steps
that way, poses before each change, shakes an elaborate sort of
baby’s rattle with the right-hand, and gesticulates with an open fan in
the left-hand. The sedate walk to and fro, the movements of the
rattle and fan constitute the dance, after which this aged Miriam sits
down, bows her head to the mats, and resumes her statuesque
pose. She wears a nun-like head-dress of white muslin, and a loose
white garment without obi, over a red petticoat, the regular costume
of the Shinto priestesses. She seems always amiable and ready to
respond to a conciliatory coin, but the visitor wonders that the cool
and shaded sanctuary in which she sits, with nearly the whole front
wall making an open door, does not stiffen her aged joints with
rheumatism and end her dancing days.
At some of the great mineral springs there are now separate pools
for men and women, in deference to foreign prejudice; but more than
one generation will pass before promiscuous bathing is done away
with.
At all medicinal springs the baths are owned and managed by the
Government and are free to the people. Here at Yumoto, men,
women, and children walk into the one large room containing the
pools, undress, lay their clothing in a little heap on the raised bench
or platform running around the edge of the room, and step into the
water; and, as has been said, no one sees any impropriety in this
custom. Women sit or kneel on the edges of the pool, scouring
themselves with bags of rice-bran, and chattering with their friends in
or out of the water. People stop at the open doors, or breast-high
windows, to talk to the bathers, and conduct is as decorous, as
reserved, and as modest as in a drawing-room. The approach of a
foreigner sends all the grown bathers deep into the water, simply out
of respect to his artificial and incomprehensible way of looking at
natural things. They know, though they cannot understand, that the
European finds something objectionable, and even wrong, in so
insignificant a trifle as being seen without clothes.
At our tea-house in Yumoto our three rooms in the upper story
were thrown into one during the daytime, making an apartment open
to the gallery on three sides. Hibachis, or braziers, with mounds of
glowing charcoal, tempered the morning and evening air, and all day
we could sit on piles of futons, and enjoy the superb picture of
mountains and lake before us. We were poled over the placid water
in a queer ark of a boat, and the mountain-paths were always
alluring, the roughest trail often passing under torii, or leading past
some shrine, just when it seemed that no foot had ever preceded
ours. At night, when the chilling air presses the sulphur fumes closer
to earth, Yumoto streets resound with the wailing whistle of the blind
shampooer, or amah. These amah are found everywhere—in the
largest cities and in the smallest mountain villages—and, whether
men or women, are never young, or even middle-aged. Theirs is an
indefinite, unscientific system of massage, and their manipulations
often leave their charges with more lame and aching muscles than
before. But the amah are an institution of the country, and Yumoto
streets would ring with their dreary music, and our screens would be
slipped aside by many an ill-favored crone, as soon as it was time for
the usual evening baths to be prepared at the tea-houses.
Upon another visit to Nikko and Chiuzenji in late October there
was a more splendid autumnal pageant than the most gorgeous hill-
sides of America had ever shown me. Frost had done its most
wonderful work, and the air was exhilarating to intoxication. The
clear and brilliant weather moved the coolies to frisk, play, and chant
like children—even that dignified little man, Ito, relaxing his gravity to
frolic like a boy, and to pry bowlders over the edges of precipices to
hear them crash and fall far below. Chiuzenji looked a vast, flawless
sapphire, and Nantaisan was a mosaic of richest Byzantine coloring.
Kegon-no-taki, the fall of three hundred feet by which the waters of
Chiuzenji drop to the valley in their race to the Daiyagawa, seemed a
column of snow in its little amphitheatre hung with autumn vines and
branches. But we dared not remain, for already Yumoto was closed
and boarded up for the season, and on any day the first of the
blockading snows of winter might shut the door of the one tea-house
left open at Chiuzenji, and end the travel from the Ashiwo copper-
mines.
CHAPTER XVII
THE ASCENT OF FUJIYAMA