Clinical Epidemiology Evidence Based Medicine

Fundamental Principles of Clinical Reasoning

Research David L. Katz download

https://ebookname.com/product/clinical-epidemiology-evidence-
based-medicine-fundamental-principles-of-clinical-reasoning-
research-david-l-katz/

Get the full ebook with Bonus Features for a Better Reading Experience on ebookname.com
Instant digital products (PDF, ePub, MOBI) available
Download now and explore formats that suit you...

Principles and Practice of Clinical Research Second

Edition Principles Practice of Clinical Research
Principles Practice of Clinical Research John I. Gallin

https://ebookname.com/product/principles-and-practice-of-
clinical-research-second-edition-principles-practice-of-clinical-
research-principles-practice-of-clinical-research-john-i-gallin/

Family Violence Explanations and Evidence Based

Clinical Practice David M. Lawson

https://ebookname.com/product/family-violence-explanations-and-
evidence-based-clinical-practice-david-m-lawson/

Principles and Practice of Clinical Research Principles

Practice of Clinical Research 1st Edition John I.
Gallin

https://ebookname.com/product/principles-and-practice-of-
clinical-research-principles-practice-of-clinical-research-1st-
edition-john-i-gallin/

Visualising Small Traumas Contemporary Portuguese

Comics at the Intersection of Everyday Trauma 1st
Edition Pedro Moura

https://ebookname.com/product/visualising-small-traumas-
contemporary-portuguese-comics-at-the-intersection-of-everyday-
trauma-1st-edition-pedro-moura/
Evidence Based Physical Diagnosis Text with BONUS
PocketConsult Handheld Software 2nd Edition Steven
Mcgee

https://ebookname.com/product/evidence-based-physical-diagnosis-
text-with-bonus-pocketconsult-handheld-software-2nd-edition-
steven-mcgee/

The Emotional Organization Passions and Power 1st

Edition Stephen Fineman

https://ebookname.com/product/the-emotional-organization-
passions-and-power-1st-edition-stephen-fineman/

Collected Works of Poe Webster s Chinese Edgar Allan

Poe

https://ebookname.com/product/collected-works-of-poe-webster-s-
chinese-edgar-allan-poe/

Paper Time Machines 1st Edition Suckling

https://ebookname.com/product/paper-time-machines-1st-edition-
suckling/

Endings On Termination in Psychoanalysis 1st Edition

Fausta Ferraro

https://ebookname.com/product/endings-on-termination-in-
psychoanalysis-1st-edition-fausta-ferraro/
Whistler and Alfven Mode Cyclotron Masers in Space
Cambridge Atmospheric and Space Science Series 1st
Edition V. Y. Trakhtengerts

https://ebookname.com/product/whistler-and-alfven-mode-cyclotron-
masers-in-space-cambridge-atmospheric-and-space-science-
series-1st-edition-v-y-trakhtengerts/
 CLINICAL
EPIDEMOLOGY
&
EVIDENCE-BASED
MEDICINE
I dedicate this book to my wife, Catherine: the reason why.
 CLINICAL
EPIDEMIOLOGY
&

EVIDENCE-BASED
MEDICINE

Fundamental Principles of
Clinical Reasoning Si Research

DAVID L. KATZ
Yale University School of Medicine

® Sage Publications
International Educational and Professional Publisher
Thousand Oaks ■ London ■ New Delhi
Copyright © 2001 by Sage Publications, Inc.

All rights reserved. No part of this book may be reproduced or utilized in any form or by any
means, electronic or mechanical, including photocopying, recording, or by any information
storage and retrieval system, without permission in writing from the publisher.

For information:

Sage Publications, Inc.

2455 Teller Road
Thousand Oaks, California 91320
E-mail: [email protected]
Sage Publications Ltd.
1 Olivers Yard, 55 City Road
London EC 1Y ISP

SAGE Publications India Pvt Ltd

B-42 Panchsheel Enclave
PO Box 4109
NewDehli 110 017

Printed in the United States of America

Library of Congress Cataloging-in-Publication Data

Katz, David L.
Clinical epidemiology and evidence-based medicine: Fundamental
principles of clinical reasoning and research / by David L. Katz.
p. cm.
ISBN 0-7619-1938-4 (cloth: acid-free paper)
ISBN 0-7619-1939-2 (pbk.: acid-free paper)
1. Clinical epidemiology. 2. Evidence-based medicine. I. Title.
RA652.2.C55 K38 2001
614.4—dc21 2001001275

This book is printed on acid-free paper.

01 02 03 04 05 06 7 6 5 4 3 2 1

Acquiring Editor: C. Deborah Laughton

Editorial Assistant: Veronica Novak
Production Editor: Sanford Robinson
Editorial Assistant: Cindy Bear
Typesetter: Technical Typesetting, Inc.
Indexer: L. Pilar Wyman
Cover Designer: Ravi Balasuriya
 Contents

Preface ix
Acknowledgments xxi

Section I

Principles of Clinical Reasoning

1. Of Patients and Populations: Population-Based Data in

Clinical Practice 5
2. Test Performance: Disease Probability, Test
Interpretation and Diagnosis 13
Test Performance 13
Disease Probability 25
Test Interpretation 31
3. Quantitative Aspects of Clinical Thinking: Predictive
Values and Bayes' Theorem 45
Application 50
Alternative Applications 56
Odds and Probabilities: Bayes' Theorem
and Likelihood Ratios 57
Implications of Bayes' Theorem for Diagnostic Testing 60
Conceptual Factors Influencing Probability Estimates 62
Bayes' Theorem and the Sequence of Testing 64
4. Fundamentals of Screening: The Art and Science of
Looking for Trouble 69
Screening Defined 70
Screening Criteria 70
Statistical Considerations Pertinent to Screening 73
Sequential Testing 77
Statistics, Screening and Monetary Costs 79
Statistics, Screening and Human Costs 84
Screening Pros and Cons 87
5. Measuring and Conveying Risk 91
Measuring Risk to the Individual Patient 92
Risk Factors 96
Measuring Risk in Clinical Investigation 96
Measuring Risk Modification 97

Section II

Principles of Clinical Research

6. Hypothesis Testing 1: Principles 107
Association 110
Variation 112
Measuring Central Tendency: The Mean 113
Measuring Dispersion: Variance and Standard Deviation 114
Testing Hypotheses: The Signal to Noise Ratio 116
Types of Clinical Data 116
Characterizing Associations: Univariate, Bivariate,
and Multivariate Methods 119
Eliminating Alternative Explanations: The Threat
of Confounding and Bias 120
7. Hypothesis Testing 2: Mechanics 127
Parametric Methods 128
Nonparametric Methods 134
Odds Ratios and Risk Ratios 139
Other Methods of Hypothesis Testing 141
Hypothesis Testing and the Stipulation of Outcome 142
8. Study Design 147
Case-control Studies 151
Cohort Studies 155
Retrospective Cohort Studies 155
Prospective Cohort Studies 156
Randomized Clinical Trials 158
Meta-Analysis 160
Active Control Equivalence Studies (ACES) 163
Crossover Trials 164
Factorial Trial Designs 164
Other Study Designs 165
Sampling 166
Assessing Study Validity 167
The Strength of Evidence 169
Constructively Deconstructing the Medical Literature 170
9. Interpreting Statistics in the Medical Literature 181
Statistical Significance 182
One-tailed and Two-tailed Tests of Significance 185
Type I and Type II error 188
p -values 190
Sample Size 191
Confidence Intervals 194
Other Considerations 195

Section III

From Research to Reasoning: The Application

of Evidence in Clinical Practice

10. Decision Analysis 203

11. Diagnosis 211
12. Management 219
 Appendices

Appendix A: Getting at the Evidence 225

Appendix A. 1: Accessing the Medical Literature:
How to Get There from Here 225
Appendix A.2: A Walking Tour of Medline 241
Appendix A.3: Publication Bias: The Limits
of Accessible Evidence 245
Appendix B: Considering Cost In Clinical Practice:
The Constraint of Resource Limitations 249
Appendix C: Clinically Useful Measures Derived
from the 2 x 2 Contingency Table 257

Glossary 261
Tfext Sources 279
Books 279
Users' Guides to the Medical Literature 280
Other Articles 282
Epilogue 283
Index 285
About the Author 295
 Preface

E
vidence has securely claimed its place among the dominant con-
cepts in modern medical practice. To the extent possible, clinicians
are expected to base their decisions (or recommendations) on the
best available evidence. Physicians may see this as one possible silver lin-
ing in the dark cloud of managed care. Insurers competing for clientele
and revenue have increasingly made it a practice to include in benefits
packages only those items for which there is convincing proof of benefit.
Moreover, these items must provide their benefit at reasonable and ac-
ceptable cost. Thus, when applying evidence to the practice of medicine
the benefit of the evidence must be measurable and definable, the cost
must be measurable, and, perhaps the subtlest challenge of all, evidence
itself must be defined and measured.
Despite current efforts to bridge the gap between medicine and public
health through the Medicine-Public Health Initiative,1'2 the philosophical
divide between a discipline devoted to the concerns of populations and
one devoted to the advocacy of an individual seems impassable. However,
the consistent application of evidence to clinical decision making is the
bridge between the concerns of clinical practice and the goals of public
health.
Evidence-based practice is population-based practice. Evidence applied
clinically is derived from the medical literature, where the standards of ev-
idence, and therefore practice, continuously evolve. But what is reported
in the literature is not the experience of an individual patient (other than

IX
X ^ CLINICAL EPIDEMIOLOGY AND EVIDENCE-BASED MEDICINE

in case reports, a modest although time-honored and often important

source of evidence, or in rz-of-1 experiments), and certainly not the experi-
ence of our individual patient, but rather the experience of a population of
patients. Therefore the practice of evidence-based medicine requires the
application of population-based data to the care of an individual patient
whose experiences will be different in ways both discernible and not, from
the collective experience reported in the literature. All evidence-based de-
cisions made on behalf of (or preferably, with) individual patients are ex-
trapolation or interpolation from the prior experience of other patients.
Clinical medicine is evidence-based only if it is population-based.
This may or may not seem a provocative concept, but consider the al-
ternative. To base clinical decisions for an individual on the individual
alone, the outcome of an intervention would need to be known in ad-
vance. In other words, medicine would need to borrow from astrology or
some other system of predicting future events. The choice of an initial an-
tihypertensive drug for a hypertensive patient cannot be based, before the
drug is prescribed, on the response of the patient in question. Nor can the
benefits to the patient be known in advance. The drug is chosen based on
the published results of antihypertensive therapy in other patients. The
particular drug is selected based on how closely the characteristics of our
patient match those of others who have benefited from specific therapies.
Once the drug is selected, while the therapeutic effect on the surrogate
measure (e.g., blood pressure) is detectable, any outcome benefit to our
patient (e.g., stroke prevention) remains unknowable. We can never iden-
tify the stroke we have prevented in an individual. The strokes we prevent
by prescribing antihypertensives, the myocardial infarctions we prevent
by prescribing aspirin or statins, are statistical events. We know the rate
at which such conditions occur in particular populations, and research
demonstrates how these rates can be changed. By applying the interven-
tion to our patient, we expect the risk of the event to decline comparably.
But unless an intervention eliminates the risk of a clinical event entirely
(few, if any, do), our patient may suffer the event despite intervention. Al-
ternatively, our patient may have appeared to be at risk, but would have
never suffered the event even without intervention. We can never know.
We never base what we do for an individual on the outcomes particular
to that individual. We base what we do on the experience of populations,
and the probability that our patient will share that experience. Astute
medical care is predicated on the capacity to identify similarities between
a single patient and the particular population whose collective experience
is most likely to inform and anticipate the single patient's experience. In
 Preface % xi

his Poetics, Aristotle considers this "eye for resemblances/' or "intuitive

perception of the similarity in dissimilars," a mark of genius.3 If so, it is
a genius the clinician frequently has cause to invoke.
The science of applying the principles of population-based (epidemi-
ologic) evidence to the management of individual patients has come to
be known as clinical epidemiology. While epidemiology characterizes the
impact of health related conditions on populations, clinical epidemiology
applies such data to individual patient care. Clinicians are traditionally
uncomfortable with the notion of catering to populations rather than indi-
vidual patients. Clinical epidemiology asserts that the two are effectively
the same, or at least inextricably conjoined. Individual patient care is in-
formed by the interpretation of population-based data. When populations
are well served by the health care they receive, the individual members of
those populations are (generally) well served. When individuals are well
served by the clinical care they receive, in the aggregate, the pattern of
that care becomes (usually) the sound practice of health care delivery to
populations.
Implicit in the concept of evidence being the derivative of a popula-
tions experience is the need to relate that experience back to the indi-
vidual patient. The inapplicability of some evidence to some patients is
self-evident. Studies of prostate cancer are irrelevant to our female pa-
tients; studies of cervical cancer are irrelevant to our male patients. Yet
beyond the obvious exclusions is a vast sea of gray. If our patient is older
than, younger than, sicker than, healthier than, ethnically different from,
taller, shorter, simply different from the subjects of a study, do the results
pertain? As our individual patient will never be entirely like the subjects
in a study (unless they were a subject, and even then their individual ex-
perience might or might not reflect the collective experience), can the re-
sults of a study ever be truly pertinent? Clinical epidemiology is a sextant,
or in more modern but equally nautical terms, the geographical position-
ing system (GPS), on a vast sea of medical uncertainty. And, to extend the
metaphor, the skills of piloting can be acquired and increase the reliability
with which a particular destination (i.e., diagnosis, therapeutic outcome)
is achieved. Yet each crossing will be unique, often with previously unen-
countered challenges and hazards. No degree of evidence will fully chart
the expanse of idiosyncrasy in human health and disease. Thus, to work
skillfully with evidence is to acknowledge its limits. Judgment must be
prepared to cross those seas as yet uncharted by evidence.
It is expected that some of the material in this text, particularly the
more statistically involved, will diverge from what we would accept as
XII ^ CLINICAL EPIDEMIOLOGY AND EVIDENCE-BASED MEDICINE

intuitive. However, comfort can be taken from the fact that we are all
de facto clinical epidemiologists. As clinicians, we decide which infor-
mation pertains to a particular patient every day of practice: who does
and does not get prescribed an antibiotic (and if so, which one); who
does and does not get treated with insulin, metformin, a sulfonylurea,
a thiazolidinedione; who does and does not get advised to be x-rayed, in-
jected, phlebotomized, cannulated, or instrumented. Cognizant or not of
the subtleties as they play out, in each such decision we are comparing
our patient to others that have come before; others in our own practice,
relying as we tend to do (though we tell one another we should not) on the
compelling lessons of personal anecdote, or others whose experience has
been more formally conveyed, in the tables and graphs of a peer-reviewed
article. While the choices we ultimately make in clinical testing and man-
agement are a product of our interaction with patients, our shared and dis-
parate values, beliefs, and preferences, the decisions that delineate those
choices are largely the product of clinical epidemiology.
Because we all practice clinical epidemiology, an understanding of this
tool (or array of tools) we use is incumbent upon us all. If every clinical
decision derives in whole or in part (and it does) from the tacit compari-
son of our patient to a population of patients, then the skill with which
that comparison is made is fundamental to the skill with which medicine
is practiced. Integral to that comparison is the capacity to recognize the
defining characteristics of both patients and populations as the basis for
defining the bounds of similarity and dissimilarity. The physician's ca-
pacity to evaluate the context in which "evidence" was gathered is equally
important. The ability to evaluate the quality as well as the pertinence of
evidence is essential. Of course, finding the best available evidence when
one is uncertain about a clinical decision is prerequisite to its interpreta-
tion.
Viewed with the cool glare of reductionism, the practice of evidence-
based medicine requires a discrete and modest skill set. One must be
able to find the available evidence. One must be able to evaluate the rel-
evance and quality of evidence. And one must be able to interpret evi-
dence presented in terms pertinent to populations so that the same data
may inform patient care decisions. These skills, like any others, can be
learned and mastered. The various tools of our trade—stethoscopes and
sphygmomanometers—were handed to us along with the lessons that
made us competent in their use. While the tools of evidence-based prac-
tice have become enjoined among the more highly valued items in our
proverbial black bags, most of us have had no formal instruction in their
 Preface % xni

use. Consequently, many of us are likely using these tools less effectively
than we might.
While clinical choices (for both testing and treatment) are predicated
on, at a minimum, the knowledge, judgment, values, preconceived no-
tions, experiences, preferences and fears of both clinician and patient,
clinical decision-making is greatly influenced by three considerations:
probability, risk, and alternative. Probability is fundamental to such de-
cisions, as we evaluate and treat patients only for a given condition or
conditions it seems they might have. We do not order CT scans of ev-
ery patient's brain, yet we do order some. The distinction is derived from
our estimate of the probability of finding relevant pathology. A clinical
decision cannot be reached without a semiquantitative estimate of proba-
bility. A patient either seems likely enough, or not likely enough, to need
a particular test or treatment, to result in our recommending it. This is a
truism for any test applied only to some patients.
Some low probability diagnoses are pursued because they pose such
high risk. Here, too, the natural tendencies of our minds are in align-
ment with clinical epidemiology. We admit some patients to the hospital
to "rule out MI" even though we believe the probability of myocardial in-
farction (MI) to be low, because the risk associated with undetected MI is
high. We have all been taught to do a lumbar puncture (LP) whenever we
wonder "should I do an LP?" because of the devastating consequences of
missing meningitis.
Finally, we factor in alternatives: alternative treatments, alternative
tests, alternative diagnoses. When chest pain seems atypical for angina,
but no alternative explanation is at hand, we are more apt to treat the pain
as angina. When pneumonia is present to explain shortness of breath, we
will be less inclined to work up pulmonary embolism (PE), despite pleu-
ritic chest pain and tachycardia. When we have excluded the impossible
we are apt to focus on what remains, however improbable.4 By a process to
which we are, for the most part, comfortably incognizant, we make every
decision factoring in considerations of probability, risk, and alternatives.
But an unconscious process is a process that cannot be optimally regu-
lated. By knowing that our decisions are borne on our musings over prob-
ability, risk, and alternative, these parameters should become of sufficient
interest to us to warrant conscious monitoring. Each of these parameters
is population-based. There is no probability of genuine relevance to an
individual: there is the rate of occurrence in populations, and the degree
of concordance between individual and population characteristics. There
is no true individual risk; for an individual, an event occurs (100% risk)
XIV ^ CLINICAL EPIDEMIOLOGY AND EVIDENCE-BASED MEDICINE

or does not (0% risk). There, is however, the comparability of the patient
to groups in whom the event rate in question is higher or lower. The
alternatives available for an individual patient are those options and in-
terventions applied under similar circumstances to other patients, with
varying degrees of success.
Similar principles underlie the research that constitutes the evidence
base (or its greater portion) for clinical practice. As is detailed later in the
text, studies are constructed in an effort to establish the collective expe-
rience of a few (the study subjects) as representative of the many. While
the clinician looks for correspondence between patient and study partic-
ipants, the investigator must consider the relevance of the study to the
larger population of potential future patients. Just as the probability of
outcomes, good and bad, guides clinical management, the probabilities of
outcomes, good and bad, false and true, are estimated and factored into
the statistical stipulations and design of a study. The appropriateness of
a particular methodology depends as much on alternatives as does the
appropriateness of a clinical intervention. As is expressed by the conven-
tional safeguards against false-positive and false-negative error (see Chap-
ter 9) and the application of progressively stringent standards of human
subject protection,5'6'7'8 thorough consideration of risk is intrinsic to the
research process. Even less rigorous means of conveying evidence, such
as case reports and case series, depend for their interest and relevance
on probability, alternative, and risk. Such reports are meaningful only
when the putative association is apparent and convincing; the clinical
need nontrivial; the risks of application acceptable; alternative explana-
tions unlikely; and the pertinence to our patients probable. Probability,
alternative, and risk influence one another within the disciplines of clin-
ical practice and clinical research, and these disciplines in turn interact.
The needs, insights and frustrations of practice are an important source of
hypotheses and attendant methods in clinical research. The evidence gen-
erated by such studies naturally serves to inform clinical practice. These
interactions are displayed in Table 1.
Ultimately, then, while judicious practice depends on evidence, the
derivation of evidence depends on many of the same principles as judi-
cious practice. The thoughtful and diligent practitioner remains abreast
of the literature to apply the best practice an evolving evidence base sup-
ports. The diligent and thoughtful investigator exploits evolving method-
ologies to generate evidence most conducive to advances in science and
practice. The highest standards of evidence-based practice are achieved
not only when evidence is well applied, but also when that evidence is
 Preface % xv

TABLE 1 The Influence of Probability, Risk, and Alternatives on Clinical

Research and Clinical Reasoning, and the Salient Interactions

Factor
Discipline Interactions1,1 Factor Discipline Interactions

Probability: The estimated

probability of a meaningful
association between putative
Clinical cause (exposure, intervention)
Research and effect (measure of
outcome) is the basis for
generating and/or testing Needs for new
particular hypotheses. treatments and
technologies, and
Risk: Research design is con- insights derived
strained and/or influenced by from clinical anec-
intervention risks, the risks dote, are sources
of non-intervention (placebo) of hypotheses that
when treatment is available, serve as the basis
and the risks of false-positive or for clinical studies.
false-negative outcomes. Clinical experience
is often the basis for
Alternative: Hypotheses are initial probability
generated and tested when and risk estimates in
a particular causal association the generation and
is deemed more probable, testing of research
important, or testable than hypotheses.
alternatives. Among the condi-
tions for the establishment of
causality in research is the ex-
clusion (to the extent possible)
of alternative explanations of
the apparent association.

1
In research, higher risk is justified by greater probability of a particular outcome and the
relative lack of acceptable alternatives. The probability of demonstrating any particular
causal association w i l l vary inversely with the relative probability of alternative
explanations for the outcome of interest.
2
l n clinical practice, conditions that are highly probable w i l l be priority considerations
even if low risk. Lower probability considerations will become prioritized when they
represent high risk potential. The relative probability of any outcome or conclusion will
vary inversely with the plausibility, relative probability, and abundance of alternative
explanations.
Note: Dashed arrows indicate interactions among the factors within one discipline. Bold
arrows indicate interactions between disciplines.
XVI ^ CLINICAL EPIDEMIOLOGY AND EVIDENCE-BASED MEDICINE

TABLE 1 (continued.)

Factor
Discipline Interactions1 Factor Discipline Interactions

Probability: The estimated

probability of a particular
condition (or conditions) in
Clinical a given patient is an important
Reasoning basis for generating a differen-
tial diagnosis and conducting
and interpreting pertinent di-
agnostic tests. The estimated
probability of a response
(or range of responses) to a
given intervention (or to non-
intervention) is similarly an The published
important basis for the selec- results of studies be-
tion of management strategies. come the evidence
base (or a critical
Risk: Clinical practice is con- component of it)
strained and/or influenced by underlying clinical
intervention risks, the risks practice. Outcomes
of non-intervention when treat- and the patterns of
ment is available, and the disease and risk fac-
risks of false-positive or false- tors in populations
negative results of diagnostic become the basis
testing. for probability and
risk estimates in
Alternative: A particular di- practice.
agnosis is made when one
condition is deemed more
probable, important, or treat-
able than alternatives. Among
the conditions for the estab-
lishment of a diagnosis is the
exclusion (to the extent possi-
ble) of alternative explanations
for the patient's condition.

well produced. Part of the burden for the responsible cultivation of higher
standards and better outcomes in medicine falls, naturally, to researchers
and those that screen and publish their findings. But application is ulti-
mately the responsibility of the clinician, who is obligated to consider not
only the pertinence of particular evidence to his or her practice but the ad-
equacy and reliability of the evidence itself. At every step, from the design
of a study to clinical counseling, probability, alternative, and risk must be
 Preface %. XVII

addressed. For evidence to be well applied the correspondence of this one

patient to those that came before must be considered, the compatibility
of prior knowledge with current need revisited.
We cannot, therefore, practice clinical medicine and avoid population-
based principles. We cannot practice clinical medicine and avoid the prac-
tice of clinical epidemiology. But the discipline of evidence-based prac-
tice/clinical epidemiology (the terms might be used interchangeably) is
not one in which most of us have had any formal initiation. All of the art
and all of the science of medicine depend on how artfully and scientifi-
cally we as practitioners reach our decisions. The art of clinical decision-
making is judgment, an even more difficult concept to grapple with than
evidence. As the quality and scope of evidence to support clinical in-
terventions is, and will likely always remain, limited in comparison to
the demands of clinical practice, the practice of evidence-based medicine
requires an appreciation for the limits of evidence, and the arbiters of
practice at and beyond its perimeters. Judgment fortified by the highest
standards of decision-making science is a force to be reckoned with, en-
abling us each to extract the best possible results from a process to which
we are naturally inclined. Ultimately that is the validation of evidence-
based practice, or population-based practice, or clinical epidemiology—
the outcomes to which such concepts contribute. Rigorous reasoning is
the means, desirable outcomes the ends.

^ FOR WHOM I J T H I J BOOK INTENDED?

This book is about concepts, or rather the methodology of arriving at

robust clinical decisions that binds together an array of concepts. This
book is not about the facts, or current fund of medical knowledge, on
which such decisions rest. The life span of medical facts is short and
shortening further all of the time. Fortunately the methods for extracting
optimal performance from the prevailing facts of the day are enduring.
The intent here is to provide a basic mastery of such methods, that is, the
capacity to harness the power of our intrinsic heuristics (decision-making
pathways) and apply it to a constantly evolving body of knowledge. The
medical literature and clinical vignettes will be referenced as required to
demonstrate applications of the methods described. But the message is
in the methods rather than their application to any particular study or
article or case.
XVIII ^ CLINICAL EPIDEMIOLOGY AND EVIDENCE-BASED MEDICINE

The intended audience for this text is anyone who makes, or will
make, clinical decisions. Worth noting are a number of excellent texts
on the subjects of clinical epidemiology and evidence-based medicine al-
ready available, many of which I have used liberally as sources (see Text
Sources). Compared to most of these, this text is intended to be more
clinician-friendly and assumes less prior knowledge. Every effort has been
made to present material in a simple and uncluttered manner. Most tables
in the text, for example, should be interprétable at a glance.
One of the important distinctions I have made while writing this text
is to endeavor to teach less and clarify more. The contention on which
this text is based is that clinicians are intuitive clinical epidemiologists,
and therefore don't really need to learn to function as such. This text is
designed to help reveal the influence and application of this intuition.
By doing so, it should illuminate the processes of converting evidence
to decisions, research to practice. The more we understand the ways in
which we approach evidence and make decisions, the more reliably we
can control these processes, and their attendant outcomes.
While a fair amount of statistics is included, the use of a calculator in
clinical practice is certainly not intended. Rather, as quantitative princi-
ples already underlie clinical reasoning, one is well advised to have a basic
familiarity with those principles. Fundamentals of practice truly hang in
the balance. A positive or negative test result is at times highly reliable,
at other times highly unreliable. A bit of number crunching demonstrates
how different clinical conclusions can, and should, be under different cir-
cumstances. The numbers need not be recalled for the importance of the
concepts to be retained.
The consistent application of the basic principles of clinical epidemi-
ology infuses with the strengths of science the decision making that pre-
supposes all else in clinical practice, including its outcomes. That science
and evidence are limited and are dependent upon judgment for their ap-
plication is implicit in the text everywhere it is not explicit. Also implicit
throughout the text is that the medical decisions reached by clinicians
serve only to provide patients—the ultimate decision makers—with good
information upon which to base their decisions.
I am grateful to the many accomplished clinicians and clinical epi-
demiologists whose contributions I have drawn on so heavily, both in
the drafting of this text and in my own clinical and research efforts. I ac-
knowledge with appreciation and humility that in drafting this text I have
followed where many luminaries have led. That said, if I have wandered
off the trails blazed by the leaders of this field, I can blame no one but my-
 Preface %, xix

self. Any misstep—ambiguity, miscalculation, or distortion—is of course

my responsibility. While hoping that none is found, I apologize and offer
my sincere regret in advance on the chance that any is.
With a great reverence for the unique burdens and privileges of clinical
practice, I submit the principles of this text in the belief and hope that
they will enhance your ability to obtain the best possible outcomes for
your patients.

REFERENCE/

1. Reiser SJ. Medicine and public health. Pursuing a common destiny. JAMA.
1999;276:1429-1430.
2. Reiser SJ. Topics for our times: The medicine/public health initiative. Am J Public
Health. 1997;87:1098-1099.
3. Barnes J (ed). The Complete Works of Aristotle. Vol. 2. Princeton, NJ: Princeton Uni-
versity Press; 1984:2335.
4. Conan Doyle A. The sign of four. In: Conan Doyle A. The Complete Sherlock Holmes.
New York: Doubleday; 1930:87-138.
5. Amdur RJ. Improving the protection of human research subjects. Acad Med.
2000;75:718-720.
6. Bragadottir H. Children's rights in clinical research. / Nurs Scholarsh. 2000 ; 32:179-
184.
7. Beasley JW. Primary care research and protection for human subjects. JAMA.
1999;281:1697-1698.
8. High DM, Doole MM. Ethical and legal issues in conducting research involving elderly
subjects. BehavSciLaw. 1995;13:319-335.
 Acknowledgments

1
am grateful to Dr. Ralph Horwitz, chairman of medicine at the Yale
School of Medicine, for setting the standard so many of us strive
(without much hope of success) to meet.
I sincerely appreciate the vision of Dan Ruth, and the guidance and
support of C. Deborah Laughton, at Sage. The transition from idea to
book is a relay race, in which their laps were very well run indeed.
I acknowledge with thanks the contributions of my collaborators, Dr.
Laura Greci, a senior resident in preventive medicine/internal medicine
at Griffin Hospital in Derby, CT, and Dr. Haq Nawaz, associate director
of the same preventive medicine residency.
I am grateful to my parents, Dr. Donald Katz and Susan Katz, for
never (well, hardly ever...) discouraging me when, as a child, I incessantly
asked, "why?" To my Dad, I also add appreciation for the walks at Horse
Heaven; I do my best thinking there.
I am deeply indebted to Jennifer Ballard, administrator of the Yale Pre-
vention Research Center, who makes me wonder every day how I man-
aged before!
I am grateful to my children, Rebecca, Corinda, Valerie, Natalia, and
Gabriel, for their patience and unconditional love, seemingly unattenu-
ated by the countless times I have turned from them to the computer,
and turned down their invitations to play.
Above all, I am grateful to my wife, Catherine, my best editor as well
as my best friend, for the love and the coffee and the kind words as much
as for the ruthlessly honest (and always constructive) criticism.

XXI
Section I

PRINCIPLE/ OF
CLINICAL REA/ONING
T
his section is not an effort to encourage the use of statistics in clin-
ical decision-making. Rather, it is intended to disclose and char-
acterize the statistical aspects of decision-making as it is already
occurring. For if clinical decisions are already governed in part by quan-
titative principles, it follows that the process of reaching them is more
controllable, more predictable, and more reliable if the process is under-
stood.
The case that statistical principles influence our decision-making ef-
forts can be made with virtually any clinical scenario. Consider, for ex-
ample, the patient presenting with fever and headache. Not all such pa-
tients undergo lumbar puncture. Yet some patients do. How is the deci-
sion reached? First, by acquiring as much pertinent information as pos-
sible to support inferences regarding probability, alternative, and risk.
These three parameters influence (and perhaps even dominate) clinical
decisions.
In all patients with fever and headache, there is a finite probability of
meningitis. While none of us could comfortably commit to a precise nu-
merical estimate of probability in a given patient, we all manage to decide
whether meningitis (or any diagnosis) is probable enough to warrant in-
tervention (either diagnostic or therapeutic). Probable enough implies a
quantitative threshold. And while the placement of that threshold varies
widely among clinicians, each of us must place it somewhere; failure to
do so is failure to decide.
What makes a condition probable enough2. Until or unless a differ-
ential diagnosis list is shortened to a single item, the probability of any
diagnosis on the list is in part mediated by all of the competing probabil-
ities. In a patient with migraines, a headache mimicking prior migraines
makes meningitis less likely, but certainly not impossible. Automatically,
unthinkingly, we establish estimates of relative probability in deciding
whether or not a condition is probable enough. And in the unusual ad-
vent of a single-item differential, wisdom attributed to Sherlock Holmes
applies: when the impossible has been excluded, whatever is left, however
improbable, must be true. l At times, probable enough need not be very
probable at all.
This is particularly true when risk is high. But as all clinical decisions
carry risk, whether they dictate action or inaction, probability must be
estimated across the range of alternatives. Intervention for each diagnos-

1
Conan Doyle, A. The sign of four. In: Conan Doyle, A. The Complete Sherlock Holmes.
New York: Doubleday ; 1930:87-158.

3
4 ^ PRINCIPLES OF CLINICAL REASONING

tic consideration carries risk. Intervention for the wrong diagnosis carries
risk without the potential for benefit. Lack of intervention for the correct
diagnosis carries risk, perhaps compounded by the risk of inappropriate
intervention.
To establish the risk estimates for each consideration for a single un-
remarkable clinical encounter is a daunting prospect for even the most
mathematically inclined among us, yet we all do so. Decisions could
not otherwise be reached. To generate such risk estimates consciously
is unthinkable, but we unthinkingly do so with each patient encounter.
Somehow, the relative probabilities and risks of multiple permutations are
converted into decisions. Orders are written. Tests are done. Treatment is
administered. And often, because this ungoverned process serves us well,
patients respond.
But patients don't always respond. And while adverse outcomes de-
spite optimal decisions will always occur, certain adverse outcomes are
doubtless attributable to less than optimal decisions. To the extent that
the powerful heuristics of our decision-making operate of their own ac-
cord, we have little means to enhance the process or its products. An
understanding of the process is prerequisite to its improvement.
Clinicians all apply statistical principles in generating semiquantitative
estimates of probability and risk across a spectrum of alternatives. The
process is intrinsic and unavoidable. But to the extent that we subject
patients to the process without understanding it, we engage in an act
of faith rather than science. And we demand less understanding of the
performance characteristics of our minds than we do of far less potent
technologies.
Thus, the application of statistics to clinical decision-making is simply
an elucidation of the discrete principles already at work in the clinician's
mind. The content of this section characterizes the statistical principles
we unavoidably, but for the most part unintentionally and unthinkingly,
already apply. An understanding of these principles is essential to assess,
control, and refine the decision-making process on which clinical out-
comes depend.
 Of Patients
and Populations
Population-Based Data
in Clinical Practice

INTRODUCTION

The process of medical evaluation is unavoidably dependent on semiquan-

titative estimates of disease probability in individual patients. Such estimates
are in turn dependent on the prior experience of populations and our famil-
iarity and interpretation of this prior experience. This chapter demonstrates
both the need and the intrinsic capacity for clinicians to generate quantita-
tive estimates of disease probability and the dependence of these estimates
on population experience.

C
onsider that a patient presents to your office. How likely is it that
the patient has coronary disease? Naturally, you can't answer the
question. You have virtually no information about the patient. Yet
you have probably already begun the process of generating an estimate.
If you are a pediatrician, patients in your practice are unlikely to have
coronary disease. Therefore, this patient presenting to you is unlikely to
have coronary disease. Similarly, if you are a specialist (other than a car-
diologist) to whom patients are referred after passing through the filter
of primary care, it is also unlikely that the patient has coronary disease.

5
6 ^ PRINCIPLES OF CLINICAL REASONING

But if you are an internist or family practitioner, you may have already
started to consider the probability of coronary disease. If you practice in
the US, many of your adult patients will have coronary disease,- so, too,
might the patient in question. If you practice in certain other countries,
the probability of coronary disease may be so low that you need hardly
ever consider it; therefore, you would not consider it in this patient.
Of note, even at this very preliminary stage of evaluation, is the role of
bias or, more bluntly, prejudice in clinical decision making. We base our
decision on experience, either our own or that of others. Making infer-
ences about an individual based on the prior experience one has had with
others in the same population is the essence of prejudice, or prejudging.
This term is not meant to have negative connotations in clinical practice.
Prejudice—a tendency to judge the probability of a diagnosis in an indi-
vidual based on the probability of that condition in the population the
patient comes from—is appropriate and essential. It would be foolish to
consider coronary disease routinely in individual patients from a popula-
tion in which coronary disease almost never occurred. Almost never is not
never, so an individual patient might have coronary disease; it would just
be highly improbable in such a population. The prejudice borne of expe-
rience, and familiarity with the population in question, would influence
clinical judgment and decisions in an appropriate way.
So one immediately begins to formulate an impression of probability
based on the characteristics of other patients one has seen. But we want
more information. In this case, we would like to know whether or not
the patient has chest pain suggestive of angina. We would like to know
the patient's age and gender,- whether or not the patient is hypertensive
or diabetic,- whether the patient smokes, is sedentary, has hyperlipidemia;
whether the patient has a family history of heart disease; whether the
patient is obese. The way we conduct our histories and physicals is proof
that we want to know these things. But as we progress from one question
to another in what at times is an almost mechanical process, we often fail
to ask ourselves, "why?" Why do the answers to such questions matter?
How many such questions do we need to ask? How can we tell when we
have enough information from the history to progress to the physical?
How do we know when enough information has been gleaned from the
exam?
Our ability to reach conclusions is proof that we can find answers to
the above questions. But if we do so through a process to which we are
inattentive, the process may fail. A process so fundamental to our perfor-
 Of Patients and Populations % l

mance as clinicians is a process we should feel compelled to understand

and master.
The reason that we ask the questions we do is not the one that seems
most self-evident. We do not question our patients so that we can know
what condition they have. This statement will likely seem heretical, but
it is not intended to be. How often do we know with complete certainty
the explanation for symptoms in a patient? What we are after, at least
much of the time, is the establishment of a sufficiently familiar pattern
to invoke our prejudices with security, a sufficiently familiar narrative to
predict the conclusion with confidence. In attempting to determine how
probable coronary disease is in the patient in question, we ask questions
that sequentially allow us to place the patient in the context of popula-
tions in which coronary disease is more or less probable. If the patient
has chest pain typical of angina pectoris and happens to be a 70-year-old
male smoker with diabetes, hypertension, hyperlipidemia, and a family
history of heart disease, we can now answer the question with consid-
erable confidence; the probability of coronary disease is high.1 We have
not seen, and do not know, the patient in question. And the patient may
not, in fact, have coronary disease. But this patient is clearly very much
like others who, in what we have seen or been taught, do have coronary
disease. Thus, population-based practice is unavoidable. It infiltrates the
process of clinical decision making at every step. Our inferences about
an individual patient are derived from the historical experience of other
patients whom we believe to be much like our own.
If the process of relating past experience—the experience of populations
to the individual patients under our current care—happens of its own ac-
cord, why make a fuss about it? Is there intrinsic value in coming to know
a process that tends to proceed spontaneously? We naturally conduct our
histories so that we can characterize our patients and determine what
clinically meaningful conditions they might or might not have. Is there
value in transplanting the process to the purview of conscious oversight?
There is. The history and physical can be considered a process of se-
quential hypothesis testing. Each question asked tests hypotheses about
the population from which the patient might come. Once this is acknowl-
edged, there comes a point in the history when additional questions (and
answers) cannot dissuade us from a particular conclusion. In the case un-
der consideration, a point in the history would be reached when coronary
disease would seem sufficiently probable to warrant further investigation.
Even if the answers to subsequent questions were negative, lowering the
probability of coronary disease, our suspicion, based on both probability
8 ^ PRINCIPLES OF CLINICAL REASONING

and risk, might be great enough to warrant commitment to a workup.

Recognizing this semiquantitative element in our decision making is es-
sential to manage the result. For example, if the patient seemed very likely
to have coronary disease, would we abandon that belief if the ECG were
normal? Probably not. What if a stress test were normal? Would we pro-
ceed from a routine to a nuclear stress test or refer the patient for cardiac
catheterization? The answer would depend on how robust the clinical
suspicion of coronary disease was, compared with the negative results of
any testing done. In such a scenario, would the clinical impression or the
results of testing rule the day?
To decide, we would need to be able to compare the reliability of the
clinical test results with that of the judgment that resulted in the testing
in the first place. Of course, we do this all the time. But we do it quite
variably. You will probably agree that permutations of the above scenario
presented to 100 different physicians might well result in 100 slightly
different plans. And while some of that variability might be legitimized
in the allowance for medicine's art, some would simply be the result of
deficient science, of failing to apply rigorous methods to the decision-
making process. If a negative stress test makes angina pectoris less likely,
we need to know how much less likely in order to decide whether or
not further testing is required. But we also need to know how much less
likely relative to what7. What was the probability with which we started?
This vaguely characterized concept, the clinical impression, is prerequi-
site to every clinical decision, whether in pursuit of diagnosis, prognosis,
or treatment. Yet in routine practice, it defies definition. While we expect
to know how reliably a stress test performs in identifying angina pectoris,
do we demand the same of the medical evaluation on which the decision
to order the stress test rests? In general, we do not. And we most certainly
should.
Our questioning generally cannot lead us to certain conclusions about
the individual patients under our care. In the process of sequentially test-
ing hypotheses about the patient, we are in essence endeavoring to define
as narrowly as possible the population of which the patient is representa-
tive. Once that goal is achieved, epidemiology can offer us a fairly stable
estimate of the probability of the particular condition under considera-
tion. That estimate is the prevalence of the condition in the population
on which we've settled. Prevalence, the proportion of a specified popula-
tion with a particular condition at a particular point in time, is related
to the probability of disease in an individual member of that population.
Incidence, the number of new cases of a particular condition in a defined
 Of Patients and Populations '% 9

population during a given period of time (typically a year) is related to the

risk of that condition in an individual member of that population.
Clinical epidemiology allows us, then, to convert the population data of
the epidemiologist into a concept of practical utility for patient care. The
analogue of prevalence for the individual patient is the prior probability,
the probability of the condition in question prior to any subsequent testing
that might be indicated to further evaluate our impression(s). In essence,
there is a discrete probability of a condition prior to every question posed
during the history that is modified by each answer to become an estimate
of the posterior probability, the probability resulting from, or following,
a test. Each such posterior probability becomes the prior probability esti-
mate in advance of the next question. The questions should be tailored
to the continuously revised probability estimate, so that the pertinent hy-
potheses are tested. There is, of course, no such thing as a comprehensive
H&JP; we cannot ask all questions of possible clinical significance in the
course of an interview. The same pertains to the physical; a comprehen-
sive physical exam is as unachievable as a comprehensive history. The
exam is tailored to test the hypotheses generated through the history.
Each relevant aspect of the exam completed serves in turn to modify the
probability of whatever condition or conditions are under consideration.
Thus the physical exam, like the history, is sequential hypothesis test-
ing and sequential conversion of prior probability to posterior probability,
then back to prior probability in advance of the next maneuver.
This construct may seem artificial, yet it is the mere application of
principles we already accept as fundamental to the medical workup. We
could not legitimately order an echocardiogram, stress test, MRI, or V/Q
scan with no advance knowledge of how reliably each performs. We
should know in advance of any test we order how secure we will be with
the result if positive or negative and what the implications for clinical
management will be of each possible outcome. Yet by far the most potent
"technology" to which our patients are subject is our decision making, the
antecedent to all other medical technology. No other technology will be
applied unless so guided by the "results" of our clinical evaluation. How
reasonable is it to ignore the performance characteristics of the process
that underlies all use of technology, then require that the performance
characteristics of the secondary technologies be known? We may know
how reliably a CAT scan performs, yet we generally do not know in any
systematic way how reliably we determine the need for the CAT scan.
And as will be demonstrated in Chapters 2 and 3, the performance of the
CAT scan, or any other technology, is dependent on the context in which
10 ^ PRINCIPLES OF CLINICAL REASONING

it is obtained. Just as a 5mm PPD reaction is positive for TB in a high-

risk patient but negative in a low-risk patient,2'3 a positive test when the
probability of the condition is very low cannot mean the same thing as it
does when the probability is high.
This can be demonstrated by use of an extreme, and therefore absurd,
example. If a pregnancy test is accurate almost all of the time (e.g., 99%
accurate), a presupposition rarely met by tests we use routinely, and it
returns positive when ordered for a male patient, we know it must be
wrong. Not because the test has suddenly lost our trust but because the
prior probability of the condition (pregnancy) in the patient (a male) is 0.
If a prior probability of 0 invalidates the positive result of a highly accurate
test, a prior probability of 1 (100%) similarly invalidates the results of a
negative test, no matter how reliably the test generally performs. These
extremes help clarify that the range of prior probabilities between 0 and 1
must also influence how we interpret a test. If a prior probability is very
close to 0, we should require a much higher standard of technological
evidence before we conclude with confidence that a condition has been
ruled in. If a prior probability is very close to 1, we will similarly require
a high standard of evidence to conclude that disease can be ruled out.
But the implications of this construct run deeper still. If a prior prob-
ability between 0 and 1 can be established and the performance charac-
teristics of the relevant technologies are known, then the ability of each
diagnostic test to change our probability enough to alter clinical manage-
ment can be determined in advance. Certainly, a test not reliable enough
to alter a very high or low prior probability estimate is of little clinical util-
ity. But worse, if the principles of clinical epidemiology are overlooked,
the test is potentially harmful. A positive test, if interpreted identically
in all patients regardless of whether their probability of disease is high
or low, will provide misinformation as often as information. Dissociated
from the principles of clinical epidemiology, medical technology, whether
invasive or noninvasive, poses very real threats to patient care.

SUMMARY/TAKE-AWAY MESSAGES

The care of the individual patient is directed toward the best achievable
outcome for that individual. But the future outcome for an individual pa-
tient is unknown and cannot be ascertained. Rather, the probability of
both diagnosis and treatment response is based on the prior experience of
 Of Patients and Populations % 11

similar patients. The reliability of the prior probability estimate is based

on the confidence with which disease prevalence in the population is
known, the appropriateness of the particular diagnosis under considera-
tion, and the degree of correspondence between the patient and the pop-
ulation. The results of diagnostic testing must be interpreted in light of the
prior probability estimate and not used to replace it. When test results are
incompatible with patient characteristics, even highly accurate tests may
be more apt to be wrong than right. The reliability of the diagnostic pro-
cess is substantially dependent on the prior probability estimate as well
as on the judicious selection and performance characteristics of the diag-
nostic tests themselves. Measures of test performance, their quantitative
interaction with estimates of disease probability, and the implications for
diagnosis are considered in Chapter 2.

^ ^ »

^ REFERENCE/

1. Cannon CP. Diagnosis and management of patients with unstable angina. Curr Probl
Cardiol 1999;24:681-744.
2. Mackin LA. Screening for tuberculosis in the primary care setting. Uppincotts Prim
CarePract. 1998;2:599-610.
3. Starkey RD. Tuberculin testing: Placement and interpretation. AAOHNf. 1995;43:371-
375.
 *m
Test Performance
Disease Probability, Test
Interpretation and Diagnosis

INTRODUCTION

Few, if any, diagnostic tests perform perfectly. Instead, tests are accurate or
inaccurate, reliable or unreliable, to varying degrees, under varying clinical
circumstances. Tests yield positive or negative, correct or incorrect results
in patterns that are dependent on patient characteristics, and substantially
predictable. The reliability of the diagnostic process is dependent in part on
an appreciation for the quantitative impact of test performance on disease
probability. The probability of a particular diagnosis depends on underlying
estimates of disease probability prior to testing, the performance of selected
tests, and the assumptions imposed to facilitate test interpretation. This chap-
ter explores the quantitative aspects of the diagnostic process.

TEXT PERFORMANCE

T
he performance of tests used to titrate the probability of any par-
ticular diagnosis can be cast in terms familiar to most of us. Sen-
sitivity is the ability of a test to detect disease when it is present.
Specificity is the ability of a test to exclude disease when it is absent.

13
14 ^ PRINCIPLES OF CLINICAL REASONING

While ostensibly simple, these concepts underlie much of the reasoning

in clinical epidemiology and therefore merit more rigorous treatment.
Each term is used in medicine much as it is used in the vernacular.
Consider the use of sensitivity in the vernacular, such as sensitive skin.
Skin that is sensitive will feel the lightest touch of the wind. If the wind
blows, sensitive skin will feel it. But we know nothing about how such
skin will feel when the wind does not blow.
The denominator for sensitivity is the presence of the condition in
question (e.g., disease). Sensitivity can tell us nothing about the test's
performance in patients who are condition (disease) free. In terms of a
population, the denominator for sensitivity is the prevalence, the propor-
tion of the population with the condition. Of those with the condition,
some will test positive (true positives), and some will test negative (false
negatives). Sensitivity is the proportion of disease positives in whom the
test is positive. If a test is negative in a patient with disease, it is a false-
negative result. Thus, sensitivity (the true-positive rate) and the rate of
false-negative error are complementary and add to 1.
Specificity pertains to the proportion of the population that is free of
disease. In comparable terms, the denominator for specificity, all disease-
free individuals, is 1 -prevalence. A test is specific when it reacts only to
the singular condition under investigation. The proportion of those free of
disease identified as disease free is the specificity. Specificity pertains only
to the denominator of those who are negative for the condition. Those
who are disease free but have a positive test result are false positives. The
specificity (the rate of true negatives) plus the false-positive error rate are
complementary and add to 1. Note that among those with disease there
are true positives and false negatives,- sensitivity defines the rate at which
true positives are identified. Among those free of disease, there are true
negatives and false positives. Specificity defines the rate at which true
negatives are identified. In both cases, recalling the denominator is help-
ful. Specificity is relative to those free of disease, and sensitivity to those
with disease. Only those with disease are subject to being false negatives;
only those who are disease free are subject to being false positives.
Just as the P wave orients one to the basic interpretation of an electro-
cardiogram, the denominator orients one to the fundamental interpreta-
tion of epidemiologic (clinical or otherwise) principles. To consider how
sensitivity and specificity measure the performance of a test, the denom-
inator to which each pertains is the first item to recall. In clinical epi-
 Test Performance % 15

demiology, start with the denominator, the portion of the statement that
indicates "in those with..." some particular characteristic. The numerator,
the occurrence of events within the denominator population, will tend to
follow quite readily (see 2 x 2 table, Appendix C).
The relationships between sensitivity, specificity, false-negative error,
false-positive error, and prevalence can best be demonstrated with use of
a 2 x 2 contingency table. By convention, such tables set true disease
status across the top and test status along the side. The cells are labeled a
through d, beginning in the upper left and proceeding left to right through
each row. Such a table is shown in Box 2.1.
As depicted in the 2 x 2 table, the entire study population, or n, is
a +b + c +d. The population with disease is [a + c). The population that
is disease free is (b + d). Sensitivity is the proportion of those with disease
(the denominator,- a + c), in whom the test is positive (the numerator,- a).
Therefore, sensitivity is a I [a + c). Specificity is the proportion of those
without disease (the denominator; b + d), in whom the test is negative
(the numerator; d). False-negative error can occur only when disease is
truly present. It is the proportion of those with disease [a + c) in whom
test results are incorrectly negative (c). False-positive error can occur only
when disease is truly absent. It is the proportion of those without disease
[b + d) in whom test results are incorrectly positive [b).
Sensitivity and specificity are performance characteristics pertinent to
every diagnostic test. To interpret the results of a test, its performance
characteristics, as well as the context in which it is operating, must be
known.
Consider a common clinical scenario, such as the need to evaluate a
patient with chest pain for evidence of ischemia. Prior to ordering any
test, we will have asked a series of questions and performed a physical
exam. If the pain sounds atypical, we may begin thinking that the prob-
ability of angina is low. But if we find the patient to have multiple car-
diac risk factors, we are likely to revise our estimate of the probability of
angina upward. If the physical exam fails to reveal any support for an al-
ternative diagnosis, we will again revise upward our impression of angina.
Conversely, if we find the pain to be reproducible with palpation, we will
revise our estimate downward. The knowledge gained from the history
and physical alters the context in which we select and interpret our sub-
sequent diagnostic tests and should therefore influence our confidence in
the results obtained.
16 ^ PRINCIPLES OF CLINICAL REASONING

Box '2.1
I The 2 x 2 contingency table andks applications to the performance |
J of diagnostic studies*
I Disease
I -f
I 4* Û h
I Ttest
1 ~~ c d

Cells:
I a «s true positives* disease present, test positive
1 b » false positives; disease absent, test positive
1 c - false negatives; disease present, test negative
I d * true negatives* disease absent, test negative
1 Rows;
{a + h} » all test positives
(c + d] « all test negatives
1 Columns:
{a + c} *» all disease positives
(jb * df| a* all disease negatives
Pertinent formulas:
Disease free population ( 1. - prevalence) — b^d
Disease positive population (prevalence) — a+c I
Prevalence — {a 4* c) 1
I Prior probability — ''prevalence rate" - ia + c}/n I
Sample population — n « aH-fo-fc-frf I
Sensitivity — Ö/)Ö + C) 1
Specificity — df{h + d) I

Altering Test Performance

Consider that we have evaluated a patient with chest pain and con-
cluded, on the basis of the history and physical exam, that the probability
of ischemia is moderate, say, 20%. While a specific numerical probabil-
 Test Performance % 17

ity may seem contrived and uncomfortable, the notion that disease is
improbable, probable, or highly probable is not. Applying a numerical es-
timate to such impressions facilitates demonstration of how important
test performance is in the diagnostic process. In this scenario, we will
want to conduct diagnostic tests to confirm or exclude a diagnosis of is-
chémie heart disease. The performance characteristics of the test or tests
we choose will influence the reliability with which we can make a final
diagnosis and develop an appropriate plan of management.

Scenario 2.1a

Consider that we test for ischemia by use of an ECG with a sensitivity

of 70% and a specificity of 60%.l We can determine the influence of test-
ing results on the probability of disease by using a 2 x 2 table in which
the prior probability estimate becomes the prevalence in a hypothetical
population of n patients. Box 2.2 is such a table, in which n — 100.
In a population of 100, a prior probability of 20% translates into a
prevalence of 20. Therefore the disease positive cells, a and c, must add
up to 20. The remaining patients, (1 -prevalence), are all in the disease-
free cells, b and d. Therefore, b and d add up to 80. Sensitivity is the

Box 2.2
The results of performing an £CG with a sensitivity of 70% and a .
specificity of 60% when thé prior probability of angina is 20%,
A population of fOO is assumed.

Disease

1 ~i- 14 32
Tfest
- 6 m
n =* 100

( d * c } ~ 20 \b HK:=d) ** 80
18 ^ PRINCIPLES OF CLINICAL REASONING

capacity of a test to detect disease when it is truly present. An ECG with

a sensitivity of 70% will detect 70% of those with disease. Thus, cell a
will be 14 (70% of the 20 patients with disease). Cell c is those patients
with disease in whom test results are negative, or the false negatives. In
this case, there are 6 such cases.
In this example, 80 patients are disease free. Given a specificity of 60%,
48 of the 80 patients without disease will be correctly identified by the
cardiogram as being free of ischemia. Thus, cell d is 48. The remainder of
the 80 disease-free subjects will be incorrectly identified as having disease.
Cell b, the false positives, is 32.
What do we know after this exercise? We don't know for sure whether
or not our patient has ischemia. We are still engaged in the process of
narrowing the population characteristics we can apply to our patient. We
have decided our patient is from a population in which the probability
of ischemia is moderate. If the cardiogram is negative in such a patient,
the patient then comes from an even more narrowly defined population:
those in whom clinical evidence suggests moderate probability of coro-
nary ischemia and in whom the ECG is negative for ischemia. Clearly, in
such patients, the probability of ischemia is lower. How much lower? The
contingency table allows us to answer.
The population we are now interested in is those patients with a prior
probability of ischemia of 20%, who subsequently have a nonischemic
cardiogram. In Box 2.2, this group is represented by cells c and d. Again,
it is helpful to begin with the denominator. To answer the question, how
probable is ischemia if the cardiogram is negative, the denominator is all
patients in whom the cardiogram is negative, whether correctly or incor-
rectly. In Box 2.2, there is a total of 54 patients in whom the cardiogram
is negative. Only 6 of these have disease. Thus, the probability of disease
in this group, the posterior probability, is 6/54, or 11%. Conversely, the
probability that ischemia is truly absent given a negative cardiogram in
this population is 48/54, or 89%. This measure, true negatives (cell d)
over all negative test results (the sum of cells c and d) is the negative pre-
dictive value (NPV). The negative predictive value reveals the probability
of being disease free given a negative test result. The remaining negative
test results are false negatives (cell c) and reveal the probability of disease
being present despite a negative test result. The proportion of mislead-
ing negative tests is ( 1 -NPV). In this case, the proportion (or percentage)
of misleading negative tests is 11%. Thus, the probability of ischemia is
 Test Performance %, 19

11% with a negative cardiogram, down from 20% before the cardiogram.
(The term proportion misleading negatives (PMN) is clinically useful, but
not currently in use. The term is coined here, and its use recommended.
See 2 x 2 table in Appendix C.) In other words, the negative cardiogram
has reduced the prior probability of 20% to a posterior probability of 11%.
One final way of evaluating the negative cardiogram is by use of what
may be termed the false-negative index (FNI), the ratio of false- to true-
negative test results. The FNI for Box 2.2 is the false negatives (c = 6) over
the true negatives (d = 48). The FNI for Box 2.2 is therefore 0.125. (The
false-negative index is a newly coined term; see 2 x 2 table in Appendix
C.) For every true-negative test result, there will be 0.125 false negatives,
or 1 false negative for 8 true negatives. A FNI below 1 indicates that there
will be more true than false negatives; conversely, a FNI above 1 indicates
that more negative test results will be false than true.
What if the cardiogram is positive? We again want to know the proba-
bility that ischemia is present or absent. Now, the denominator is those
from our population in whom the cardiogram is positive. In Box 2.2, test
results are positive in cell a (true positives; 14) and in cell b (false pos-
itives; 32). Therefore, the denominator is [a + b), or 46. The probabil-
ity of ischemia in this group is a I [a 4- b), which is 14/46, or 30%. This
is the probability of disease in those with a positive test result, or the
positive predictive value (PPV). The remainder of those with a positive
test result are false positives. The proportion misleading positives (PMP)
is (1-specificity), or bl[b + d). In Box 2.2, the PMP is 70%. (The term
proportion misleading positives (PMP) is coined here, and its use is rec-
ommended. See 2 x 2 table in Appendix C.) In other words, 70% of those
from the population with a 20% prior probability of ischemia in whom the
cardiogram is positive are nonetheless disease free. Another way to look at
this is the false-positive index (FPI), the ratio of false positives [b = 32) to
true positives (a = 14). (The false-positive index is a newly coined term.
See 2 x 2 table in Appendix C.) In this case the FPI is approximately 2.3;
for every 1 true-positive test, there will be 2.3 false positives. A FPI below
1 indicates that there will be more true than false positives,- conversely,
a FPI above 1 indicates that more positive test results will be false than
true. Thus, a positive cardiogram in scenario 2.1a, does not allow us to
diagnose ischemia with confidence. Ischemia remains improbable despite
a positive cardiogram.
20 ^ PRINCIPLES OF CLINICAL REASONING

Scenario 2.1b
Consider that we again test for ischemia by use of an ECG in a pa-
tient with a 20% prior probability of disease. Assume now, however, that
the sensitivity of the cardiogram is reduced to 40%, perhaps because of
abnormalities in the patient's resting cardiogram, and that the specificity
of the cardiogram is reduced to 30%. This scenario reveals an important
aspect of diagnostic tests often overlooked: the performance characteris-
tics of almost any test are dependent to varying degrees on the population
to which they are applied.2 While the ECG may be a fairly good test for
acute cardiac ischemia, the test may perform much less reliably in a pop-
ulation with left bundle branch block. An x-ray may be accurate in the
diagnosis of pneumonia, but not in a patient with lung cancer. There-
fore, patient/population characteristics not only influence the probability
of disease (prior probability) but also the accuracy with which diagnos-
tic tests perform. In this scenario, the prior probability of angina is held
constant, while the accuracy of the cardiogram is altered.
In a population of 100, a prior probability of 20% translates into a
prevalence of 20; therefore the disease-positive cells, a and c, must add to
20 as shown in Box 2.3. The remaining patients, (1 -prevalence), are all
in the disease-free cells, b and d. Therefore, b and d add to 80. Sensitivity

Box 2.3
The results of performing an ECG with a sensitivity of 40% and a
specificity of 30% when the prior probability of angina is 20%.
A population of 10Û is assumed,

Disease
4T -

4- 8 56
Test
12 24
n « 100
 Test Performance % 21

is the capacity of a test to detect disease when it is truly present. An ECG

with a sensitivity of 40% will detect 40% of those with disease. Thus, cell
a will be 8 (40% of the 20 patients with disease). Cell c is the number
of patients with disease in whom test results are negative, or the false
negatives. In this case, there are 12 such cases.
In this example, 80 patients are disease free. Given a specificity of 30%,
24 of the 80 patients without disease will be correctly identified by the
cardiogram as being free of ischemia. Thus, cell d is 24. The remainder of
the 80 disease-free subjects will be incorrectly identified as having disease.
Cell b, the false-positives, is 56.
We are still engaged in the process of narrowing the population charac-
teristics we can apply to our patient. We have decided our patient is from a
population in which the probability of ischemia is moderate. If the cardio-
gram is negative in such a patient, the patient then comes from an even
more narrowly defined population: those in whom clinical evidence sug-
gests moderate probability of coronary ischemia and in whom the ECG
is negative for ischemia. Clearly, in such patients the posterior probabil-
ity of ischemia should be lower. How much lower? The contingency table
allows us to answer. The population we are now interested in is those
patients with a prior probability of ischemia of 20% who subsequently
have a nonischemic cardiogram. In Box 2.3, this is cells c and d. Again,
it is helpful to begin with the denominator. To answer the question "How
probable is ischemia if the cardiogram is negative/' the denominator is all
patients in whom the cardiogram is negative, whether correctly or incor-
rectly. In Box 2.3, there is a total of 36 patients in whom the cardiogram
is negative. Of these, 12 have disease. Thus, the probability of disease
in this group, the posterior probability, is 12/36, or 33%. Conversely, the
probability that ischemia is truly absent given a negative cardiogram in
this population is 24/36, or 67%. This measure, true negatives (cell d)
over all negative test results (the sum of cells c and d) is the negative
predictive value (NPV). The remaining negative test results are false neg-
atives (cell c) and reveal the probability of disease being present despite
a negative test result (proportion misleading negatives). In this case, the
proportion misleading negatives is 33%. Thus, the probability of ischemia
is 33% despite a negative cardiogram, up from 20% before the cardiogram.
In other words, the negative cardiogram has raised the prior probability
of 20% to a posterior probability of 33%! The poor performance charac-
teristics of the test in this patient (i.e., this population of patients) result
in a paradox: the probability of disease is actually higher after a negative
test than it was before any testing was done. This is because under the
22 ^ PRINCIPLES OF CLINICAL REASONING

conditions and assumptions imposed, the test is wrong more often than
it is right (see below). The FNI for Box 2.3 is the false negatives (c = 12)
over the true negatives (d = 24). The FNI for Box 2.3 is therefore 0.5. For
every true-negative test result, there will be 0.5 false negatives.
What if the cardiogram is positive? We again want to know the proba-
bility that ischemia is present or absent. Now, the denominator is those
from our population in whom the cardiogram is positive. In Box 2.3 test
results are positive in cell a (true positives; 8) and in cell b (false posi-
tives; 56). Therefore, the denominator is [a + b), or 64. The probability of
ischemia in this group is a I [a + b), which is 8/64, or 12.5%. This is the
probability of disease in those with a positive test result, or the positive
predictive value (PPV). Thus, a positive cardiogram in scenario 2.1b ac-
tually lowers the posterior probability of disease. The remainder of those
with a positive test result are false positives. The proportion misleading
positives, 1 -PPV, is 87.5%. In other words, 87.5% of those from the pop-
ulation with a 20% prior probability of ischemia in whom the cardiogram
is positive are nonetheless disease free. In this case, the false-positive in-
dex (FPI), the ratio of false positives [b = 56) to true positives [a = 8), is 7;
for every 1 true-positive test, there will be 7 false positives. An FPI below
1 indicates that there will be more true than false positives. An FPI above
1 indicates the converse, that more positive test results will be false than
true.
Scenario 2.1b demonstrates that a test with poor performance char-
acteristics will not only fail to help reach an accurate diagnosis but is
actually apt to do harm. An ischémie cardiogram in this patient would
tend to make us think that the probability of ischemia had gone up when
in fact it would have gone down. A nonischemic cardiogram would nat-
urally make us believe that ischemia could be more securely excluded,
but the probability of ischemia in this scenario is actually higher after a
negative cardiogram than before testing. The performance characteristics
of tests are therefore of vital importance to the diagnostic process. Natu-
rally, a test performing this poorly would not be used. Yet many diagnos-
tic tests are employed with little consideration given to their performance
characteristics. This scenario, extreme though it may be, points out the
potential hazards.

Scenario 2.1c
Consider that we again test for ischemia using an ECG in a patient
with a 20% prior probability of disease. Assume now, however, that the
 Test Performance %, 23

sensitivity of the cardiogram is raised to 90%, perhaps because the patient

has a perfectly normal baseline cardiogram, and that the specificity of the
cardiogram is also raised to 90%. In this scenario, the prior probability of
ischemia is held constant, while the accuracy of the cardiogram is altered.
In a population of 100, a prior probability of 20% translates into a
prevalence of 20. Therefore the disease-positive cells, a and c, must add
to 20 as shown in Box 2.4. The remaining patients, (1 -prevalence), are
all in the disease-free cells, b and d. Therefore, b and d add to 80. An ECG
with a sensitivity of 90% will detect 90% of those with disease. Thus, cell
a will be 18 (90% of the 20 patients with disease). Cell c is the number
of patients with disease in whom test results are negative, or the false
negatives. In this case, there are 2 such cases.
In this example, 80 patients are disease free. Given a specificity of 90%,
72 of the 80 patients without disease will be correctly identified by the
cardiogram as being free of ischemia. Thus, cell d is 72. The remainder of
the 80 disease-free subjects will be incorrectly identified as having disease.
Cell b, the false-positives, is 8.
If the cardiogram is negative, the population we are now interested in
is those patients with a prior probability of ischemia of 20% who subse-
quently have a nonischemic cardiogram. In Box 2.4, this is cells c and
d. Again, it is helpful to begin with the denominator. In Box 2.4, there

1 Box 2.4

The results of performing an ECG with a sensitivity of 90% and a

specificity of 90%when the prior probability of angina is 20%.
A population of 100 is assumed* 1

Disease
*f - 1
18 8
Test
2 72
n - 100

{a + c) « 20 (fr + tf)-80
24 ^ PRINCIPLES OF CLINICAL REASONING

is a total of 74 patients in whom the cardiogram is negative. Of these, 2

have disease. Thus, the probability of disease in this group, the posterior
probability, is 2/74, or 2.7%. Conversely, the probability that ischemia is
truly absent given a negative cardiogram, the NPy in this population is
72/74, or 97.3%. In this case, the PMN is 2.7%. Thus, the probability
of ischemia is 2.7% with a negative cardiogram, down from 20% before
the cardiogram. In other words, the negative cardiogram has reduced the
prior probability of 20% to a posterior probability of 2.7%. The FNI for
Box 2.4 is 0.028. For every true-negative test result, there will be 0.028
false negatives.
What if the cardiogram is positive? The denominator is those from our
population in whom the cardiogram is positive. In Box 2.4 test results
are positive in cell a (true positives; 18) and in cell b (false positives; 8).
Therefore, the denominator is [a + b), or 26. The positive predictive value
(PPV), the probability of ischemia in this group, is a I [a + b), which from
Box 2.4 is 18/26, or 69%. Thus, a positive cardiogram in scenario 2.1c
considerably raises the posterior probability of disease. In Box 2.4, the
PMP is 31%. In other words, 31% of those from the population with a
20% prior probability of ischemia in whom the cardiogram is positive are
nonetheless disease free. In this case, the (FPI), the ratio of false positives
(b = 8) to true positives [a = 18), is 0.44; for every 1 true-positive test,
there will be 0.44 false positives.
Scenario 2.1c demonstrates that a test with good performance char-
acteristics can substantially alter the accuracy with which a disease is
diagnosed or excluded. In this scenario, an ischémie cardiogram raised
the probability of disease from 20% to almost 70%. A negative cardio-
gram lowered the probability of ischemia from 20% to 2.7%. When there
is considerable diagnostic uncertainty a test that performs well can be
tremendously helpful.

Summary: Altering Tfest Performance

Accurate tests are more helpful in reaching correct diagnoses than in-
accurate tests. This concept is hardly earth-shattering, but it conceals
surprising subtleties. The performance characteristics of any diagnostic
test, its sensitivity (ability to detect disease when disease is present) and
specificity (ability to exclude disease when disease is absent) are based on
the characteristics of the populations in which the test itself was tested.
To the extent that an individual patient differs from the test population
 Test Performance % 25

used to validate the test, test performance may vary. Even when patient
characteristics are not thought to threaten test performance, the stan-
dards for testing should vary based on the degree of clinical uncertainty
Test performance is uninfluenced by disease prevalence, but diagnostic
accuracy (predictive value) is influenced by disease prevalence. Sensitivity
and specificity of a test should be good enough to enhance meaningfully
one's certainty that disease is present or absent before that test is or-
dered. As shown above, a test that performs poorly, either in general or in
a particular patient, may actually increase, rather than diminish, diagnos-
tic uncertainty. Worse, if the test is simply believed it may result in the
inappropriate exclusion of disease or in an inappropriate diagnosis. The
implications of this section are that test performance should be known
with some confidence before a test is ordered, as should the prior prob-
ability of disease. To enhance or refute a very strong suspicion of disease
(a high prior probability), a test will need to be highly accurate, especially
with regard to negative results. To reduce or reverse a very low suspicion
of disease (a low prior probability), a test will similarly need to be highly
accurate but this time especially with regard to positive results. Tests with
operating characteristics common to clinical experience—sensitivity and
specificity in the 75%-90% range—are best utilized when a considerable
degree of diagnostic uncertainty (i.e., prior probability close to 50%) exists.

^ DI/EA/E PROBABILITY

What if it is the degree of diagnostic uncertainty (the prior probability)

that varies, while the performance characteristics of the test remain con-
stant? This situation also has marked effects on disease probability and
the reliability of testing, and is often overlooked in the diagnostic process.

Scenario 2.2a

On the basis of history and physical we conclude that the probability

of angina is low. For purposes of demonstration, a "low" probability will
be set at 10%.
An ECG will be ordered in such a patient and will either reveal evi-
dence of ischemia or not. To know how these results can be interpreted,
the reliability of the cardiogram must be known. Consider that the sen-
sitivity of a resting cardiogram is approximately 70% and the specificity
26 ^ PRINCIPLES OF CLINICAL REASONING

approximately 60%.3 Box 2.5 demonstrates the results of performing an

ECG when the prior probability of angina is 10%. If a population of 100
is assumed, a prior probability of 10% translates into a prevalence of 10.
Therefore the disease-positive cells, a and c, must add to 10. The re-
maining patients, (1 -prevalence), are all in the disease-free cells, b and
d. Therefore, b and d add to 90. An ECG with a sensitivity of 70% will
detect 70% of those with disease. Thus, cell a will be 7 (70% of the 10
patients with disease). Cell c is those patients with disease in whom test
results are negative, or the false negatives. In this case, there are 3 such
cases.
In this example, 90 patients are disease free. Given a specificity of 60%,
54 of the 90 patients without disease will be correctly identified by the
cardiogram as being free of ischemia. Thus, cell d is 54. The remainder of
the 90 disease-free subjects will be incorrectly identified as having disease.
Cell b, the false-positives, is 36.
The population we are now interested in is those patients with a prior
probability of ischemia of 10% who subsequently have a nonischemic car-
diogram. In Box 2.5, this is cells c and d. Again, it is helpful to begin
with the denominator. To answer the question, how probable is ischemia
if the cardiogram is negative, the denominator is all patients in whom the
cardiogram is negative, whether correctly or incorrectly. In Box 2.5, there

Box 2.5
The results of performing an ECG with a sensitivity of 70% and a
specificity of 60% when the prior probability of angina h 10%.
A population of 100 is assumed.
Disease

+ 7 36
Test
~~ 3 54
n « 100

(ö + c ) » i 0 [h + d)^90
 Test Performance % 27

is a total of 57 patients in whom the cardiogram is negative. Only 3 of

these have disease. Thus, the probability of disease in this group, the pos-
terior probability, is 3/57, or 5%. Conversely, the probability that ischemia
is truly absent given a negative cardiogram, the NPy in this population
is 54/57, or 95%. The remaining negative test results are false negatives
(cell c). In this case, the PMN is 5% (i.e., of 100 negative test results, on
average 95 will be correct, while 5 will be incorrect). Thus, the probability
of ischemia is 5% with a negative cardiogram, down from 10% before the
cardiogram. In other words, the negative cardiogram has reduced the prior
probability of 10% to a posterior probability of 5%. The FNI for Box 2.5 is
the false negatives (c = 3) over the true negatives (d = 54), therefore 0.06.
For every true-negative test result, there will be 0.06 false negatives.
What if the cardiogram is positive? In Box 2.5, test results are positive
in cell a (true positives; 7) and in cell b (false positives; 36). Therefore, the
denominator is (a + b), or 43. The probability of ischemia in this group,
the PPy is a I [a + b), which from Box 2.5 is 7/43, or 16%. The remainder
of those with a positive test result are false positives. In Box 2.5, the PMP
is 36/43, or 84%. In other words, 84% of those from the population with a
10% prior probability of ischemia in whom the cardiogram is positive are
nonetheless disease free. The FPI is approximately 5, so for every 1 true-
positive test there will be 5 false positives. Thus, a positive cardiogram
in scenario 2.2a does not allow us to diagnose ischemia with confidence.
Ischemia remains improbable despite a positive cardiogram.

Scenario 2.2b

Consider another patient in whom, after history and physical exam,

the probability of ischemia is estimated at 80%. A cardiogram, with the
same performance characteristics already noted (sensitivity of 70%, speci-
ficity of 60%) is obtained. The results are shown in Box 2.6.
If a population of 100 is assumed, a prior probability of 80% translates
into a prevalence of 80. Therefore the disease-positive cells, a and c, must
add to 80. The remaining patients, (1 -prevalence), are all in the disease-
free cells, b and d. Therefore b and d add to 20. An ECG with a sensitivity
of 70% will detect 70% of those with disease. Thus, cell a will be 56 (70%
of the 80 patients with disease). Cell c is those patients with disease in
whom test results are negative, or the false negatives. In this scenario
there are 24 such cases.
Twenty patients are disease free in this example. Given a specificity of
60%, 12 of the 20 patients without disease will be correctly identified by
Exploring the Variety of Random
Documents with Different Content
prec] fa'bliau (-16), n. (pi. -x, pr. -z). Metrical tale of early French
poetry. [F] fa'brie, n. Thing put together; edifice, building ; frame,
structure, (lit. & fig.) ; (often textile f.) woven material; construction,
texture, tissue, [f. F fabrique f. Lfabrica (faber artificer)] fa'brieate,
v.t. Construct, manufacture, (rare) ; invent (lie, &c), forge
(document). So fabriCA #tion, fa'brieatOR 2, nn. [f. L fabricare as
prec, -ate3] fa'bulist, n. Composer of fables or apologues ; liar. [f. F
fabidiste (fable1, -ist)] fa'bulous, a. Given to legend (/. historians);
celebrated in fable ; unhistorical, legendary ; incredible, absurd,
exaggerated. Hence or cogn. fabulS'siTY, fa'bulousNESS, nn.,
fa'buIously 2 adv. [f. L fabidosus (fable 1, -ous)} faea'de (-sahd), n.
Face of building towards street or open space. [F (foil., -ade a)]
face1, n. Front of head from forehead to chin (look one in the /.,
confront him steadily ; show one's/., appear ; /. tof., confronted ; /.
to f. with, confronting ; set one's /. against, oppose; with wind, sun,
in one's /.. straight against one ; fly in thef. of, openly disobey ; in f.
of, opposite to ; in the f. of, or in f. of, despite ; to one's/., openly in
his sight or hearing ; in the f. of day, openly) ; expression of
countenance ; grimace (make, pull, a f. or- ff.); composure,
coolness, effrontery, (have the /., be shameless enough ; save
one's/., forbear from or evade shaming him or oneself openly) ;
outward show, aspect, (on the f. of it, to judge by appearance ; put
a new /. on, alter aspect of ; pxd a good f. on matter, make it look
well, show courage in facing it) ; surface (from the f. of the earth) ;
front, facade, right side, obverse, dial-plate of clock &c, working
surface of implement &c ; f.-ache, neuralgia ; /. value, nominal value
as stated on coin, note, &c Hence -facED 2 a. [F, f. pop. L facia-
facies (facere make, or fa- shine)] face 2, v.t. & L Meet confidently
or defiantly (/. matter out, carry it through ; /. opponent down,
browbeat him), not shrink from, stand
 PACER 291 FAHRENHEIT fronting, (/. the music, not quail
at moment of trial) ; present itself to (the problem that faces its) ;
turn (card) f. upwards ; (of persons &c.) look, (of things) be
situated, in a certain direction (on, to, or North, Eastwards, &c.) ;
front towards, be opposite to, (to f. page 20) ; (Golf) strike (ball) full
with middle of club-f. in driving from tee; (Mil.) turn in certain
direction on one's ground {left, about, f. ; also trans., he faced his
men about); supply (garment) with facings; cover (surface) with
layer of other material; dress surface of; coat (tea) with colouring
matter, [f. prec] fa*eer, n. Blow in the face ; great & sudden
difficulty. [FACEi-f-ER1] fa'cet, n. One side of a many-sided body,
esp. •f a cut gem ; one segment of a compound eye. Hence
fa*cetED2a. [f. Ffacette (face l, -ette)] faee'tiae (-shie), n. pi.
Pleasantries, witticisms; (book catalogues) books of humorous or
obscene character. [L (facetus urbane)] facetious (-shus), a.
Addicted to or marked by pleasantry, waggish. Hence face'tiousLV2
adv., faee'tiousxESS n. [f. F facetieux (facetie f. L facet ia sing, of
prec.)] fa'cia (-sha), n. Plate over shop-front with occupier's name
&c. [var. of fascia] fa'cial (-shl), a. Of the face (esp. in Anat., as /.
artery) ; f. angle, that formed by two lines from nostril to (1) ear &
(2) forehead. [F, f. med.L facialis (face1, -al)] -facient (-shnt), suf .
forming adj j. representing L -facient- (facere make, -ent) added to
infin. in -e, as calefacere, liquefacere, w. sense producing the action
of the vb. E forms, on strict anal. w. L, absorbefacient Sec, &,
loosely. aborti facient, calori facient, &c. , where L would have vbs in
-jicare, adjj. in -Jicus -fic. fa'cile, a. Easily done or won ; working
easily, ready, fluent; of easy temper, gentle, flexible, yielding. [F, f. L
facilis (facere do)] fa'cile pri'neeps, pred. a. Easily first. [L] facilitate,
v.t. Make easy, promote, help forward, (action or result). Hence
faei'litA'tion n. [f. Ffaciliter as prec] faei'lity, n. Being easy, absence
of difficulty, unimpeded opportunity (giveff. for, of doing); ease or
readiness of speech &c, aptitude, dexterity, fluency ; pliancy, [f.
Ffacilite f . hfacilitatem (facile, -ty)] fa'eingr, n. In vbl senses of face
2 ; esp. : (pi.) cuffs, collar, &c, of soldier's jacket, differently coloured
from rest ; coating of different material, esp. of stone &c. on wall, [-
ing1] facsimile, n., & v.t. Exact copy, esp. of writing, printing,
picture, &c. (reproduced in /., exactly) ; (vb) make f. of. [L fac
imperat. of facere make + neut. of similis like] fact, n. Perpetration
of act, occurrence of event, (now only in before, after, thef, confess
the /.) ; thing certainly known to have occurred or be true, datum of
experience, (often with explanatory clause or phrase, as the f that
fire burns, of my having seen him) ; thing assumed as basis for
inference {his ff. are dispxdable) ; (sing, without a) the true or
existent, reality, (so matter of /., independent of inference ; matter
1-of-f. ; inf. ; in point off. ; thef. of the matter is), [f. L factum neut.
p.p. of facere do] fa'etion, n. Self-interested, turbulent, or
unscrupulous party, esp. in politics ; prevalence of party spirit. Hence
or cogn. fa'etionAL, fa'etious, aa. , fa'etiousLY2 adv. , fa'ctiousness n.
[F, f . L f actionem (facere fact- do, -ion) way of making (fashion),
class, clique] -faction, suf. repr. L -f actio, forming nn. of action
related to vv. in -FY, prop, only when -fy represents L -facere, F -
faire, as in satisfaction, but also used when -fy represents L -Jicare,
F -Jier, as in petrifaction. factitious, a. Designedly got up, not natural,
artificial. Hence facti'tiousLY - adv.. faeti'tiousNESS n. [f. L facticius
(facere fact- make) + -ous] fa'etitive, a. (gram.). F. verb, one with
sense make, call, or think, that takes obj. & compl. (he thought her
mad). Hence fa'etitiveLY 2 adv. [irreg. f. L facere fact- make, -ive]
fa'etop, n. Agent, deputy ; merchant buying & selling on
commission, whence fa*etoPAGE(4) n. ; (Sc.) land agent, steward ;
(Math.) one of the components that make up a number or
expression by multiplication ; circumstance, fact, or influence,
contributing to a result, [f. Ffacteitr f. L factor (prec, -or-)] factorial,
a. & n. (math.). Product of series of factors in arithmetical
progression ; product of an integer & all lower integers (adj., /. U,
symbol |fc=4 x 3 x 2 x 1). [-ial] fa'etopy, n. Merchant company's
foreign trading station ; manufactory, workshop, (F. Acts, regulating
management in interest of the hands), [ult. f. med.L factoria (factor,
-y1)] faeto*tum, n. Man of all work; servant managing his master's
affairs. [med.L, as facsimile + neut. of L tot us whole] fa'ctual, a.
Concerned with, of the nature of, fact. Hence fa'ctualLY 2 adv. [f .
fact on false anal, of actual] fa'ctum, n. Statement of facts or points
in controversy, memorial. [L, see fact] fa'cula, n. (astron. ; pi. -ae).
Bright spot or streak on sun. Hence fa'eulAR 1, fa'culous, aa. [L,
dim. of fax fac- torch] fa'eultative, a. Permissive ; optional ;
contingent ; of a faculty. [F (-if, -ive) ; foil., -ive] fa'eulty, n. Aptitude
for any special kind of action ; executive ability (chiefly U.S.) ; power
inherent in the body or an organ ; a mental power, e. g. the will,
reason ; branch of art or science, department of University teaching
(the four ff.. Theology, Law, Medicine, Arts), Masters & Doctors in
any of these (pop., The F., members of medical profession) ; liberty
of doing something given by law or a superior, authorization, licence,
(esp. eccl.). ff. Ffaculte f. ljfacultatem (facilis easy)] fad, n. Pet
notion or rule of action, craze, piece of fancied enlightenment.
Hence fa*ddiSH1, fa'ddY2, aa., fa'ddiNESS, fa'ddish' ness,
fa*ddiSM(3), fa*ddiST(2), nn. [?] fade, v.i. & t. Droop, wither, lose
freshness & vigour; (of colour &c) grow dim or pale; cause to lose
colour ; disappear gradually. Hence fa'deLESS a., fa'delessLY 2 adv.
[f. OF fader (fade dull, insipid, perh. f. L vapidus)) faeces (fe-), n. pi.
Sediment ; excrement of the bowels. Hence fae'CAL a. [L, pi. oifaex}
Fa'epie, -y, n. & a. Fairyland, the fairies, esp. as represented by
Spenser ; (attrib.) vision* ary, fancied, [var. of fairy] fag, v.i. & t., &
n. Toil painfully ; (of occupation) tire, make weary ; (at schools, of
seniors) use the service of (juniors), (of juniors) do ser« vice for
seniors ; (Cricket) /. out, field ; f.-end^ inferior or useless remnant.
(N. ) drudgery, unwelcome task (what af. !), exhaustion (brainf.) ;
(at schools) junior who has to f. [perh. corrupt, of flag v.] fa'g-grot,
fa'g-ot, n., & v.t. & i. Bundle of sticks or twigs bound together as fuel
; bundle of steel rods ; f.-vote, manufactured by transferring
sufficient property to unqualified person, so /. -voter; (vb) bind in ff,
make ff. [F (fagot), etym. dub.] Fahrenheit (-it), a. (abbr. F.). F.
th&rmo*
 FAIENCE 292 FALCONER meter, with 32° & 212° for
freezing & boiling points (used esp. in giving temperatures, as5(r F.).
(Prussian inventor d. 1736J faience (F), n. Earthenware & porcelain
of all kinds, [f. F faience f. Faenza Italian town] f ail K n. Witho ut f. ,
for certain, irrespective of hindrances, (emphasizing injunction or
promise), [f. OF faile (faillir fail2)] fail2, v.L & t. (strictly, intr. with
ind. obj.). Be missing (see failing2) or insufficient, not suffice for
needs of (person), run short, {time would f. me to tell ; his heart
failed him) ; become extinct, die away ; flag, break down ; prove
misleading, disappoint hopes of, (the prophecy failed ; the xcind
failed us) ; be insufficiently equipped in, not succeed in the
attainment of; not succeed (in doing or to do) ; miscarry, come to
nothing ; suspend payment, go bankrupt ; be rejected as candidate,
[f. OF faillir f. pop. L *fallire = hfallere deceive] fai'lingr1, n. In vbl
senses; also, foible, shortcoming, weakness, [-ing1] fai'lingr2, prep.
In default of (/. this, if this does not happen ; whomf. or/, whom in
proxy appointments). [-ING2] fai'lupe (-yer), n. Non-occurrence,
non-performance ; running short, breaking down ; ill success;
unsuccessful person, thing, or attempt ; insolvency, [earlier failer for
F faillir fail 2, cf. -er4, -ure] fain1, pred. a., & adv. "Willing under the
circumstances to ; left with no alternative but to ; (adv.) would /.,
would be glad to. [OE fxgen cogn. w. OHG (gi)fehan rejoice] fain2,
fains, fen(s), child's formula (usu. fains /as v.t.) stipulating for
exemption from unwelcome office &c. (/. 1 wicket-keeping !). [?]
faineant (F), n. & a. Idle(r), inactive (official). [F, perversion on / aire
do, niant nothing, of OF faignant sluggard (faindre skulk)] faint K a.
Sluggish ; timid (f. -heart, coward ; so faint-heaptED2a., faint-
hear'tedi.v 2 adv., faint-heap*tedNESs n.) ; feeble (a f. show of
resistance) ; dim, indistinct, pale (/. or feint lines, ruled f or feint, of
paper with lines to guide writing ; af. idea, inadequate) ; giddy or
languid with fear, hunger, &c, inclined to swoon ; (of air, scents, &c.)
sickly, oppressive. Hence fal,ntisH1(2) a., fai'ntLY2 adv., fai*ntNESS
n. [OF, p.p. of faindre feign] faint2, v.i., & n. Lose courage, give way,
(archaic) ; swoon (v. & n. ; fainted away ; in a deadf, utterly
insensible), [f. prec] faints, n. pi. Impure spirit coming over at
beginning & end of distillation, [f. faint1] fair1, n. Periodical
gathering for sale of goods, often with shows & entertainments, at
place & time fixed by charter, statute, or custom (a day after thef.
too late) ; fancy-/. If. OF feire (now foire) f. L/eria holiday] faiP2, a.
& n. Beautiful (the f. sex, the /., women; also archaic as n., a f. =a
woman); satisfactory, abundant, (af. heritage) ; specious If.
speeches) ; blond, not dark, (af. man, complexion,hair, whence faip-
haiPED2a.) ; clean, clear, unblemished, (/. water-, /. copy1; /. fame)
; just, unbiased, equitable, legitimate, tf. & square a. & adv., without
finesse, aboveboard ; /. trade, principle that reciprocity should be
thecondition of free trade; af. field1 & no favour ; all \s /. in love &
war ; /. play, equal conditions for all) ; of moderate quality, not bad,
pretty good, whence faip*iSH1(2) a. ; favourable, promising, gentle,
unobstructed. If. or foul weather ; /. -weather friends, not good at
need ; in a f. way to succeed ; by f. means, without violence or fraud
; fairway, navigable channel) ;/. -maid, - fumade; February Fair-
maids, snowdrops. Hence faip*NES8 n. [com.-Teut. ; OE fxger cf.
OHG fagar] faiP3, adv. Speak one /., address him courteously ;f. -
spoken (of person), courteous, bland ; write outf, as fair2 copy ; hit,
fight, /., according to the rules ; bid 1/. ; (with strike, fall, &c.)
straight, plump, clean. [OEfxgre (prec.)] faii»*ingr, n. Present bought
at a fair. 1-ing1] faip'ly, adv. In adj. senses ; (also) utterly,
completely, (/. beside himself ; there is sometimes doubt between
this sense & that of rather, tolerably, as in/, good), [-ly2] faip'y, n. &
a. Small supernatural being with magical powers ; Fairyland, home
of ff., enchanted region ; f.-ring, circular band of darker grass caused
by fungi & attributed to f. dancing ; f.-tale, about fF., also account of
strange incident, coincidence, marvellous profress, &c. ; hence
faip'yDOM, faip*yHOOD, aip'yiSM, nn. (Adj.) of ff. ; imaginary,
fictitious; f.-like, beautiful & delicate or small, whence faip'iLY2 adv.
[f. OF faerie (now f eerie) f. OF/ae fay] fait accompli (F), n. Thing
done & no longer worth arguing against. faith, n. Reliance, trust, in ;
belief founded on authority (pin one's/, to or upon, believe implicitly)
; (Theol. ) belief in religious doctrines, esp. such as affects character
& conduct, spiritual apprehension of divine truth apart from proof ;
system of religious belief (the Christian. Jewish, f. ; defender of the
F. ; the f. , the true religion) ; things (to be) believed ; warrant (on
thef. of); promise, engagement, (give, pledge, plight, keep, break,
violate, one's/.); loyalty, fidelity, (goodf, honesty of intention ; badf.,
intent to deceive ; Punic f. treachery) ;f.-cure, -curer, -healing, -
healer, acting by prayer, not drugs &c. [f. OF feid f. L fides] farthful,
a. Loyal, constant, (to person, one's word), conscientious ;
trustwortby ; true to fact, the original, &c, accurate; thef. (pi.), true
believers, esp. Mohammedans (Father of the /., Caliph). Hence
fai'thfulNESS n. [-ful] fai'thfully, adv. In adj. senses ; esp. : yours /.,
formula of rather distant tone for closing letter ; dealf. with, speak
home truths to or of ; promise f, emphatically (colloq. ). [-ly2]
fai'thless, a. Unbelieving ; perfidious, false to promises ; unreliable.
Hence fai'thlessLYadv., fai'thlessNESS n. [-less] fake1, v.t., & n.
(naut.). Coil (rope) ; (n.) one round of a coil. [?] fake2, v.t., & n.
(slang). Do up, make presentable or specious, contrive out of poor
material ; (n. ) piece of faking, thing faked up, dodge, cooked
report, [perh. f. Gfegen sweep] fakir* (-er), n. Mohammedan (or
Hindu) religious mendicant, devotee, [f. Arab, faqir poor man]
fa'lbala, n. Flounce, trimming. [?] fa*lcate,a. (anat.,bot.,zool.).
Hooked, sickleshaped, [f. \j falcatus i. falx sickle, -ate 2(2)] fa'lcated,
a. (astron.). = prec. (of moon &c). [as prec, see -ate2] falchion
(fawltshn), n. Broad curved convex-edged sword, [f. OF fauchon f.
pop. L *falcionem nom. -o (hfalx sickle)] fa'leifopm, a. (anat). Sickle-
shaped, [f. L falx -cis sickle -f -form] falcon (fawkn, fawlkn), n. Small
diurnal bird of prev, esp. as trained to hawk for sport (in falconry the
female only, cf. tiercel). So fal*conRY(2, 5) n. [f. OF faucon f. LL
falconem perh. f. hfalx sickle] fal'conep (fawk-), n. Keeper and
trainer of hawks ; one who hunts with hawks, [f. OF faulconnier, see
prec, -er2(2)J
 FALCONET 293 FAMILIARIZE fal'conet (fawk-), n. (Hist.)
light cannon ; species of shrike, [first sense f. It. falconetto dim. of
falcone falcon ; last f. falcon + -et *] faldera'l, n. Gewgaw, trifle,
[earlier as meaningless refrain in songsj fa'ldstool (faw-), n. Bishop's
armless chair ; movable desk for kneeling at ; desk for litany to be
said from. [f. med.L faldistolium f. OHG faldstuol (faldan to fold,
stool)] Faler'nian, n. A famous wine of ancient Campania, [f. L
(vinum) Falernum Falernian (wine) + -1AN] fall ! (fawl), v.i. (fell ;
fallen often conjugated with be, see -ed](2), & used as adj.).
Descend freely {falling star, meteor), drop, (the remark fell from him
; lambs f., are born), come down, lose high position (statesmen /.),
swoop (vengeance fell) ; become detached, hang down ; sink to
lower level (barometer, prices, f.), decline, slope ; disembogue into ;
subside, ebb, abate ; show dismay (faces f), droop (eyesf.) ; cease
to stand (falling sickness archaic, epilepsy), become prostrate, come
to ground, sin, be overthrown, perish, (/. prostrate, flat ; plans f. to
the ground, are abandoned, fail; /. on one's sword, in suicide ;
wicket falls, batsman is out ; fortress falls, is taken ; woman falls,
loses chastity ; many fell, were killed in battle ; seven lions fell to his
rifle ; /. a prey or sacrifice to ; f. into error ; houses f., tumble in
fragments ;/. to pieces, intwo, asunder) ; take such adirection (his
eye fell upon me), have such a place (accent falls on first syllable),
alight, come by chance *&c, (the lot fell upon me ; cost falls to you ;
it fell in my way; f. among thieves, uponacorrupt age ; subject falls
into three divisions) ; pass into such a state (fell into a rage, in love),
become so-&-so (/. dumb, due) ; lapse, revert, (revenues f. to the
Croum) ; occur, have date, (Easter falls early), find place (what now
falls to be described). With prepp. : /. a — ing, begin ; /. behind, be
passed by ;/. into, (line) take one's place in the ranks, combine with
others, (conversation with) begin talking to, (habit &c) adopt it ; /.
(up)on, assault, come across, (one's feet or legs) get well out of
difficulty ; /. to — ing, take to, begin, (also/, to work) ;f. under, be
classed among, be subjected to (observation &c.);/. within, be
included in. With advv. : /. astern, (of ship) drop behind ;/. away,
desert, revolt, apostatize, decay, vanish ; /. back, retreat ; /. back
upon, have recourse to ;/. behind, lag ; f.fotdof, come into collision
with, quarrel with, attack ;/. in, (Mil.) take or cause to take places in
line, (of buildings &c.) give way inwards, (of debt &c.) become due,
(of land &c.) become available, (of lease) run out ; /. in with,
happen to meet, accede to (views), agree with (person), coincide
with, humour ; /. off, withdraw, decrease, degenerate (so falling off,
n.), (of ship) refuse to answer helm, (of subjects) revolt ; /. on, join
battle, begin feeding ; /. out, quarrel, come to pass, result well &c,
(Mil.) leave the ranks ; /. out of, give up (habit &c.) ; /. short,
become insufficient, (of missile) not go far enough;/, short of, fail to
obtain ;/. through, miscarry, fail ; /. to, begin eating or fighting.
[com.-Teut. ; OE feallan cf. G fallen ; also L fallere deceive] fall 2, n.
Act of falling (see prec. ) ; also or esp. : amount of rain &c. that
falls; (now chiefly U.S.; also/, of the year or leaf) autumn ; number
of lambs born ; cataract, cascade, (of ten pi. ) ; downward trend,
amount of descent ; wrestling-bout, throw in this, (try af, lit. & fig.) ;
amount of timber cut down ; succumbing to temptation (theF.,f. of
man, Adam's sin and its results); kind of woman's veil. [f. prec]
fallacy, n. Misleading argument, sophism, (Log.) flaw that vitiates
syllogism, one of the types of such flaws ; delusion, error, (pathetic
/.); unsoundness, delusiveness, disappointing character, (of
arguments or beliefs). So fallA •cious a., falla'ciousLYSadv.,
falla'ciousness n. [f. L fallacia (fallax deceiving f. fallere deceive) see
-acy] fal-la*l, n. Piece of finery. Hence falla*lery(5) n.
[contemptuous reduplication, cf. gewgaw, perh. f. falbala] fa'llible, a.
Liable to err or be erroneous. Hence falliBrLlTY n. [f. lAjfallibilis
(fallere deceive, -ble)J fa-llow i, n., a., & v.t. (Ground) ploughed and
harrowed but left uncropped for a year ; uncultivated (land) ; (vb)
break up (land) for sowing or to destroy weeds. [ME f alive ploughed
land, cf. OE fealga harrows] fa'llow2, a. Of pale brownish or reddish
yellow (now only inf. -deer, species smaller than red deer). [OE falu
cf . Gfahl, prob. cogn. w. L pallidus pale & Gk polios grey] false
(fawls), a. & adv. Erroneous, wrong, incorrect, (/. idea, verdict ;/.
concord, breach of agreement rules in grammar ; /. quantity,
incorrect length of*Vowel in verse or pronunciation ; /. note in music
;/. drawing ;/. imprisonment, illegal ; /. weights &c. ; f. pride,
shame, based on wrong notions ; /. position, one that forces person
to act against his principles ; /. step, stumble, transgression ; /.
start, wrong start in racing) ; lying, deceitful, treacherous, unfaithful
to ; deceptive (/. mirror, medium) ; spurious, sham, artificial, (/.
coin, god, prophet, hair, teeth ;/. colours, flag one has no right to.
lit. & fig.); improperly so called, pseudo-, (/. acacia ; /. bottom,
horizontal partition in vessel ; /. keel) ; hence or cogn. fa'lseLY2 adv.,
fa'lseNESS, fa'lsiTY, nn. ( Ad v. ) play person /., cheat, betray.
[OE/aZs f. lifalsus p.p. of fallere deceive] falsehood, n. Falsity ;
something untrue, contrariety to fact ; lying, lie(s). [-hood] falsetto
(fawl-), n. Forced shrill voioe above one's natural range (inf., af.
tone, &c, often of sham indignation). [It., dim. off also false] fa'lsify
(fawl-), v.t. Fraudulently alter (document); misrepresent; make
wrong, pervert; disappoint (hope, fear, &c). So falsincA"TiON n. [f. F
falsifier f. ISLfalsificare (false, -fy)] fa'ltep (fawl-), v.i. & t. Stumble,
stagger, go unsteadily ; stammer, speak hesitatingly, (/. out, utter,
say, thus) ; waver, lose courage, flinch. Hence fa'ltening'LY2 adv. [?]
fame1, n. Public report, rumour; reputation (house of illf., bawdy-
house), good reputation ; renown, celebrity. [F, f. hfama = Gk
pheme (fa- speak)] fame 2, v.t. (Pass. ) be currently reported as, for,
to be or do; (p.p.) famous, much spoken of, (for valour &c). [f. OF
famer (prec.)] fami'liar (-jar), a. & n. Of one's family (archaic for
family attrib.) ; intimate (with), in close friendship (/. spirit, or /. as
n., demon attending & obeying witch &c); closely acquainted with
(some subject) ; well-known, no longer novel, (to) ; common,
current, usual ; unceremonious, free, over-free. (N.): (R.-C. Ch.)
person rendering certain services in Pope's or bishop's household ;
intimate friend or associate; f. spirit. Hence fami*liaPLY2adv. [f. OF
familieri. hfamiliaris (family, -ar1)] fami'lia'Pity, n. Close intercourse,
intimacy with person or some subject ; unceremoniousness,
treatingof inferiors or superiors as equals, (/. breeds contempt), [f. F
familiarite f. L familiaritatem (prec, -ty)] fami'liarize (-ya-), v.t. Make
(thing) well
 FAMILY 294 PAR known ; make (person, person's mind &c,
oneself) well acquainted or at home with. Hence fami'liapiZATiON n.
[ize] fainily, n. Members of a household, parents, children, servants,
&c. (happy /., animals of different kinds in one cage) ; set of parents
& children, or of relations, living together or not (Holy F„ The Virgin,
Jesus, St Joseph, & often SI John Baptist & St Elizabeth, as grouped
in pictures) ; person's children ; all descendants of common
ancestor, house, lineage, (of f., nobly born) ; race, group of peoples
from common stock ; brotherhood of persons or nations united by
political or religious ties ; group of objects distinguished by common
features ; group of allied genera, usu. subdivision of order ; /.
butcher &c, supplying ff. as opp. to the army &c. ; /. hotel, with
special terms for ff. ; in af. way, without ceremony ; in the f. way,
with child ; /. Bible, large Bible with fly-leaves for registering births
&c. ; /. coach, large closed carriage, a game of forfeits; Fr Compact,
in 18th c. between Bourbons of France, Spain, & Two Sicilies, esp.
against England & Austria ; /. likeness, that between relations,
vague resemblance ; /. living, benefice in gift of head of f. ; /. man,
one with f., domestic person ; /. tree, genealogical chart, [f. L familia
household (famulus servant, -ia j)] fa'mine, n. Extreme scarcity of
food in a district &c. ; dearth of something specified, as water f. (f.
prices, raised by scarcitv) ; hunger, starvation, (die of /.). [F, f. LL
*famina f. L fames hunger, -ixe 4] fa'mish, v.t. & i. Reduce, be
reduced, to extreme hunger; (colloq.) be famishing, feel hungry,
[obs. fame v. f. L fames hunger, -ish2] fa'mous, a. Celebrated, well
known ; (colloq.) capital, excellent, whence fa'mousLY2 adv. [f. OF
fameus f. hfaynosus (fame, -ose1)] fa'mulus, n. (pi. -li). Attendant
on magician. [L,= servant] fan1, n. Winnowing-machine ;
instrument, usu. folding & sector-shaped when spread out, on
radiating ribs, for agitating air to cool face; anything so spread out,
as bird s tail, wing, leaf, kind of ornamental vaulting (/. tracery) ;
rotating apparatus giving current of air for ventilation &c; (Naut.)
(blade of) screw, propeller ; (in windmill) small sail for keeping head
towards wind ; /. -light, f. -shaped Avindow over door ;/.tail, f. -
shaped tail or end, kind of pigeon, coalheaver's hat or sou'- wester.
[OE fann f. L vannus winnowing-basket] fan 2, v.t. & i. Winnow
(corn), whence fa*nnERJ(2) n.; winnow away (chaff), sweep away
(as) by wind from f.; move (air) with f.; drive current of air (as) with
f. upon, to cool (face &c.) or to kindle (flame ; /. the flame, increase
excitement &c.) ; (of breeze) blow gently on, cool ; spread out (t. &
i.) in f. shape, [f. prec] fana'tic, a. & n. (Person) filled with excessive
& mistaken enthusiasm, esp. in religion. Hence fana'tieAL a.,
fana'tiealLY 2 adv., fana'tieiSM n., fana*tieiZE(2, 3) v.i. & t. [f.
\jfanaticus (fanum temple, -atic)] fa'neiep, n. Connoisseur in some
article or animal (of which the name is usu. prefixed, as dog, rose, -
f). [fancy 2, -er l] fa'neiful, a. Indulging in fancies, whimsical,
capricious; fantastically designed, ornamented, &c, odd-looking;
imaginary, unreal. Hence fa'neifalLv 2 adv., fa'neifulNESs n. [ful]
fa'ney \ n. & a. Delusion, unfounded belief ; faculty of calling up
things not present, of inventing imagery; mental image; arbitrary
supposition ; caprice, a whim ; individual taste, inclination, (take a f.
to, for; catch thef. of, please) ; thef, those who have a certain hobby,
= -fanciers, esp. the patrons of boxing ; art of breeding animals with
certain points of excellence ; f.-free, not in love. (Adj.; not pred.)
ornamental, not plain, (/. bread ; /. dress, masquerade costume, so
f.-d. or/, ball ; /. work, ornamental sewing &c. ; /. fair, bazaar for
sale of f. goods) ; (of flowers &c.) particoloured ; capricious,
whimsical, extravagant, (at af. price ;/. franchise, based on
complicated or arbitrary qualifications ; /. dog, pigeon, &c, bred for
particular points of beauty &c.) ; based on imagination, not fact (/.
picture) ; f. man, sweetheart, (slang) man living on earnings of a
prostitute, [contraction of fantasy] fa'ney2, v.t. Picture to oneself,
conceive, imagine, (/. oneself dead ; /. a blue dahlia ; /. him to be
here, that he is here ; imperat. as excl. of surprise, fancy !,f. his
believing it!); be inclined to suppose, rather think ; (colloq.) have
good conceit of (oneself, one's game &c.) ; take a f. to, like ; breed,
grow, (animals, plants) with attention to certain points, [f. prec]
fanda'ngle, n. Fantastic ornament, tomfoolery, [perh. f. foil.]
fanda'ngro, n. Lively Spanish dance ; tune for this. [Sp.] fane, n.
(poet.). Temple, [f. hfanum] fa'nfape (or as F), n. Flourish of
trumpets, bugles, &c. [F] fanfapona'de (or as F), n. Arrogant talk,
brag ; = prec. [f. F fanfaronnade f. fanfaron (prec, -oon), -ade]
fang"1, n. Canine tooth, esp. of dogs & wolves; serpent's venom-
tooth ; spike of tool held in the stock ; (prong of) root of tooth.
Hence (-)fang*ed2, fa*ng"LESS, aa. [OE, cf. Gfangen seize] fang 2,
v. t. Prime (pump) by pouring in water to start it. [f. prec] fantasia (-
aze'a, -ah'zla), n. Musical composition in which form is subservient to
fancy. [It.,= fantasy] fa'ntast, ph-, n. Visionary, dreamer, [f. med.L f.
Gk phantastes (phantazomai make a show f. phaxno show)]
fanta'stic, a. Fancied (rare) ; extravagantly fanciful, capricious,
eccentric ; grotesque or quaint in design &c Hence (thr. obs. -
i'ca?)fantastiea'IiTV, fanta'sticalNESS, fanta'stieisivi, nn.,
fanta'stiealLY 2 adv. [f. med.L fantasticus f. LLf. Gk phantastikos
(prec, -ic)] fa'ntasy, ph-, n. Image-making faculty, esp. when
extravagant or visionary ; mental image ; fantastic design ; =
fantasia ; whimsical speculation, [f. OF fantasie f. L f. Gk phantasia
(see fantast)] fantocci'ni (-tshene), n. pi. Mechanically worked
puppets ; marionette show. [It.] faquir. See fakir. fapi.adv. (farther, -
thest, further, -thest), & n. At a great distance, a long way off, (often
with away, off, out ; also fig., as /., so f, from doi?ig, f. from it ; f. be
it from me to, I would on no account) ; to a great distance or
advanced point (driven f. into the ground ; f. gone, advanced in
some progress ; he will gof, do much ; gof to effect &c, nearly do so
; by a great interval, by much, (/. different, better, the best ; also/. &
away) ;sof, to such a distance, (also) up to now ; howf, to what
extent; asf. as, right to, not short of, (place) ; as or sof. as, in sof as,
to whatever extent ;f.-aicay, remote, long-past, (of look &c) absent,
dreamy ;/.-6etween, infrequent ;/. -famed, widely known ; /.fetched,
of simile, illustration &c) studiously sought out, strained ; /. forth ; /.
off, remote; f.-reaching, widely applicable, carrying many
consequences ; f. -seeing, -sighted, prescient,
 PAR 295 FAST prudent, (of vision) seeing distant things
more clearly than near ones. (N.) a distance (do you come from/. ?)
; large amount (by/., -with comEar. & superl., prefer, surpass, &c).
[OE/eor(r) OTeut./er-f. Aryan per- cf. Glapcran beyond] far2, a.
(farther, -est, further, -est). Distant, remote, (af. cry 2). [OE feorr f.
prec] fa'rad, n. (electr.). Electro-magnetic unit of capacitj*. [f.
Faraday, electrician, d. 1867] faracia'ie, a. (electr.). Inductive,
induced, (of current), [as prec, -icj farce \ n. Dramatic work merely
to excite laughter; this species of drama ; absurdly futile
px*oceeding, pretence, mockery. Hence far*eical a., fap'eicalLY2
adv., fareiea'liTY n. [F, orig. = stuffing, f. L farcire to stuff, used
metaph. of interludes &c] faree2, v.t. (archaic). Season, spice, stuff,
(in cookery, & fig. of literary compositions), [f. OF farsir f. L as prec]
farceur' (-ser), n. Joker, wag. [F] far'ey, n. Disease, esp. of horses,
allied to glanders ; /. bud. button, small tumour in this, [f. F farcin f.
Lifarciminum {farcire stuff)] far'del (-dl), n. (archaic). Bundle,
burden. [OF, dim. of farde burden perh. f . Arab, fardah ] fare a, n.
1. Cost of passenger's conveyance, passage-money ; passenger in
hired vehicle. 2. Food provided (usu. good, bad, plentiful, &c, /*.;
bill 4 off). [OE/ser (sense 1), Scfaru (sense 2), f. st. of foil.] fare2,
v.i. Journey, go, travel, (poet.; so/. forth, start) ; happen, turn out,
(how fares it?) ; get on well, ill, &c, have such luck; be entertained,
be fed or feed oneself, well &c [com.Teut.; OE & OHG faran f. Aryan
por- pass through cf. Gk poros ford, L portare carry] farewell, int. &
n. Good-bye !, Adieu !, (/. to, no more of) ; (n.) leave-taking, parting
good wishes, [imperat. of prec. + well] fappna, n. Flour or meal of
corn, nuts, or starchy roots; powdery substance ; (Bot.) pollen ;
(Chem.) starch. Hence farlnA'CEOUS a. [L (far corn, -ine4)] fa'Pinose
(-s), a. Mealy, sprinkled with powder, [prec, -ose1] farl, n.(Sc). Thin
cake,orig. quadrant-shaped, of oatmeal or flour, [for obs. fardel
quarter (FOURTH, DEAL), Cf. FARTHING] farm \ n. Tract of land used
under one management for cultivation (orig. only of leased land ;
home-f, reserved & worked by owner of estate containing other ff.) ;
(also f. -house) dwelling-place attached to f. ; place where children
are farmed (see foil.) ; f. -stead, f. with buildings on it ; f.-yard,
enclosure attached to f. -house, [f. F ferme f. med.L/irma fixed
payment ( firmare fix f. firmms)] farm 2, v.t. & i. Take proceeds of
(tax, office, &c.) on payment of fixed sum ; (also/, out) let out
proceeds of (tax &c) to person for fixed sum ; let the labour of
(persons) for hire ; contract to maintain and care for (persons, esp.
children) for fixed sum ; cultivate, till ; till the soil, be a farmer.
Heuce far'mER1, far'mingJ, nn. [f. prec] far*o, n. Gambling card-
game. [f. Pharaoh (significance doubtful)] farou'ehe (-oosh), a.
Sullen, shy. [F] farra'go (-ah-, -a-), n. Medley, hotchpotch. Hence
farraginous a. [L (genit. -inls),= mixed fodder (far corn)] fa'rrier, n.
Shoeing-smith ; horse-doctor ; N.C.O. in charge of cavalry
regiment's horses. Hence fa*rriERY(2) n. [f. OF ferrier f. L ferrarius f.
ferritin iron, -er2(2)] fa'ppow, n., & v.t. & i. Giving birth to, litter of,
pigs (20 at one f.) ; (vb) produce (pigs), produce pigs, [vb f. n., OE
fearh f. OTeut. farhoa cf. L porcus] fart, n., & v.i. (indecent).
Emission of, emit, wind from the anus. [Aryan, cf. Gk perdomai]
far'ther (-dh-), adv. & a. (used as comp. of far1. 2, see etym.), & v.
t. To or at a more advanced point or greater extent or distance (I'll
see youf. or further /ms£) ; in addition, also, besides, moreover,
(now usu. further). (Adj.) more extended, additional, more ; niore
distant or advanced, whence far'therMOST a. (Vb, rare) = further,
[var. of further ; both used as comp. of far, but with tendency to
restrict/, to lit. & further to secondary senses] farthest (-dh-), a. &
adv. Most distant (at the, at,f, at the greatest distance, at latest, at
most); (adv.) to or at the greatest distance, [var., now more usu., of
furthest] far'thing (-dh-), n. Quarter of a penny ; least possible
amount (doesn't matter a /.). [OE feorthing (feortha fourth, -ing3)]
far'thing-ale, n. (hist.). Hooped petticoat, [f. OF verdugale f. Sp. -ado
(verdugo rod, -ado)] fa'sees, n. pi. (Rom. hist.). Bundle of rods with
axe in the middle carried by lictor before high magistrate ; ensigns of
authority. [L (pi. olfascis bundle)] fascia (fa'shia), n. (Arch.) long flat
surface of wood or stone ; (Anat.) thin sheath of fibrous tissue ;
stripe, band, fillet, belt. [LJ fa'sciated (fashl-), a. (Bot.; of contiguous
parts) compressed, growing, into one (so fa*seiATiON n.); striped,
[f. obs. fasciate f. L fasciare (prec), -ate2] fa'scicle, -ieule, -i'eulus,
(fasi), n. (Bot. &c) bunch, bundle, whence fa'seielED2, fasei'eulAR i,
fasei'eulATE 2, -a ted, aa., faseicuIa-tion n.; one part of book
published by instalments, [f. L fasciculus (fasces, -cule)] fa'scinate, v.
t. Deprive (victim) of power of escape or resistance by one's look or
presence (esp. of serpents) ; attract irresistibly, enchant, charm,
whence fa*seinatiNG2a., fa'scinatingLY2adv. Hence or cogn.
fa'seinA'TioN, fa'seinatOR2 (esp., = opera-hood), nn. [f. L fascinare
(fascinum spell), -ate3] fasci'ne (-sen), n. Long faggot used for
engineering purposes & esp. in war for filling trenches, making
batteries, &c. ;/. dwelling, prehistoric lake dwelling supported by
cross layers of sticks sunk below surface. [F, f. hfascina (fascis
bundle, -ine4)] fash, v.t., & n. (Sc). Bother, trouble, inconvenience,
[n. f. vb, f. OF fascher (now fdcher)] fa'shion (shn), n.,&v.t. Make,
shape, style, pattern, manner, (after the f. of, like ; so -/. = -wise, as
walk crab-f); after, in, a f., not satisfactorily, but somehow or other;
prevailing custom, esp. in dress (f. -plate, picture showing style of
dress) ; conventional usages of upperclass society, (thef., whatever
is in accord with these for the time being ; set thef, give the example
in changing them; the /., also, admired & discussed person or thing;
in, out of,f. or thef, agreeing or not with current usage ; man &c off,
of social standing, moving in & conforming with upper-class society)
; hence -fashionED 2 a. ( Vb) give shape to, form, mould, (into, to,
or abs.). [f. OF/aco?i, OXF fachon, f. L f actionem (facerefact- make,
-ion)] fa'shionable, a. & n. Following, suited to, the fashion ;
characteristic of, treating of, or patronized by, persons of fashion.
Hence fa#shionableNESsn., fa'shionabLY2 adv. (N.) f. person, [prec.
n., -able] fast !, v.i. 1. Abstain from all or some kinds of food as
religious observance or in sign of mourning (fasting-day, =
FAST^cfoy). 2. Go
 FAST 296 PATTY without food. [com. -Teut. ; OBfaestan cf .
Goth. fastan = orig. keep, observe, f. OTeut. fasteja] fast2,n. Act of
fasting (prec, 1); season or (also/. -day, fasting-day) day appointed
for fasting ; going without food (break one's/. = breakfast v.). [prob.
f. ONfasta f. OTeut. as prec] fast3, a. Firmly fixed or attached
(stake/, in the ground ; /. friend or friendship, steady, close ; ship f.
aground ; f. asleep ; af. prisoner ; /. colour, unfading, not washing
out; make/., fasten ; play f. & loose, ignore obligations, be
unreliable; door is f, locked &c; takef. hold of, tight ; /. with gout,
confined) ; rapid, quickmoving, producing quick motion, (/. train ; f.
cricket-ground,racquet-court,billiard-table,on which ball bounds
smartly ; watch is/., shows too advanced time ; /. person, dissipated,
see foil.). Hencefa-stiSH1(2)a. [com. -Teut. ; OE fxst cf . G fest cogn.
w. Goth, fastan fast i] fast-*, adv. (-er,-est). Firmly, fixedly, tightly,
securely, (stand, sit, stick, /.; /. bind, /. find, lock up" what you
would not lose ; eyes f. shut ; sleep f., soundly); (poet. & archaic)
close beside, by, upon, &c; quickly, in quick succession ; live /., live
in a dissipated way, expend much energy in short time. [OEfxste
(prec.)] fa'sten (-sn), v.t. & i. Make fast, attach, fix, secure by some
tie or bond, (to, tipon, on adv. or prep., together, up, in adv. or prep.
; or abs. ; f. parcel, garment, door, &c, or string, bolt, &c. ; f. off
thread &c, secure with knot or otherwise), whence fa'steniNGMl) (-
sn-) n. ; direct (look, thoughts, &c.) keenly (up)on; fix (nickname,
imputation, &c.) (up)on;f. quarrel upon, pick quarrel with ; become
fast (door will not f) ;/. (up)on, lav hold of, single out for attack,
seize \ipon (pretext). Hence fa*stenER1(2) n. [OE fxstnian cf.
Gfestnen (fast3, -en6)] fa'sti, n. pi. Chronological register of events,
annals. [L, = calendar] fasti'dious, a. Easily disgusted, squeamish,
'hard to please. Hence fasti'diousLY2 adv., fasti'diousN'ESS n. [f. L
fastidiosus (fastidium loathing, -ose1)] fa'stness, n. In adj. senses
(fast 3) ; also, stronghold, fortress, [-ness] fat, a., n., & v.t. & i. Fed
up for slaxighter, fatted; well-fed, plump, (cut upf., leave much
money), corpulent ; thick, substantial, (esp. of printing- type) ;
greasy, oily, unctuous, (cut itfi, make a display); (of coal) bituminous
; (of clay &c.) sticky ; fertile, rich, yielding abundantly, (/. lands,
benefice, job) ; slow-witted, indolent, (f.-head, dolt ;f.-witted, stupid)
; f.-guts, corpulent person ; f.-hen, kinds of goose-/oo£ ; hence
fa,ttiSH1(2) a., fa'tXESS n. (N.) the f. part of anything (live on the f.
of the land, have the best of everything) ; oily substance composing
f. parts of animal bodies (the f. is in the fire, there will be an
explosion); (Theatr.) part of r61e that enables actor to show off;
(Chem.) orfanic compound of glycerin with acid ; hence a'tLESS a.
(Vb) = fatten ; kill the fatted calf for, receive (returned prodigal) with
joy. [OE faett(ian) cf. Du. vet f. OTeut. faitido- p.p. of faitjan fatten
(faito- adj. fat)] fa'tal, a. Like fate, inevitable, necessary ; of,
appointed by, destiny (/. sisters, the Fates ; /. thread, allotted length
of life; /. shears, death), fateful, important, decisive ; destructive,
ruinous, ending in death, (to) ; deadly, sure to kill; (by exag?.)
mischievous, ill-advised. Hence fa*talLV2 adv. [f. hfatalis (fate, -al)]
fa'talism, n. Belief that all events are predetermined by arbitrary
decree ; submission to all that happens as inevitable. So fa*taliST(2)
n. &a.,fatali*stica.,fatali,stiCALLYadv. [-ism] fata'lity, n. Subjection to,
supremacy of, fate ; predestined liability to disaster ; fatal influence ;
misfortune, calamity ; death by accident, in war, &c. [f. FfataliU f.
hhfatalitatem (as fatal, see -ity)] fa'talize, v.i. & t. Incline to fatalism
; subject to government by fate, [-ize] fa'ta morga'na (fah-, -gah-),
n. Kind of mirage seen esp. in strait of Messina. [It. (fata fay,
Morgana female name)] fate, n., & v.t. Power predetermining events
unalterably from eternity; (Mythol.) goddess, one of the three Greek
goddesses, of destiny ; what is destined to happen ; appointed lot of
person &c; person's ultimate condition (decide, fix, seal, one's/.);
death, destruction. (Vb, usu, pass. ) preordain (he was fated to&o or
be ; it was fated that) ; (p.p.) doomed to destruction, [f. L fatum
neut. p.p. oifari speak] fa'teful, a. Prophetic ; fraught with destiny,
important, decisive ; controlled by, showing power of, fate. Hence
fa'tefulLY2 adv. [-ful] fa'ther l (fahdh-), n. Male parent (also fig.; the
wish is f. to the thought, one believes because one wishes to ; the
child is f. to the man, lays down the lines of his development; =f.-
inlaw ; = step-f. ; (also adoptive /.), one who has adopted a child ;
progenitor, forefather ; originator, designer, early leader, (/. of
English poetry ; F. of lies, the devil ; F. of the faithful; Ff. of the
ChurchorFf., Christian writers of first five centuries) ; one who
deserves filial reverence (/. of his country) ; religious teacher ; God ;
First Person of the Trinity ; confessor : priest belonging to religious
order, superior of monastic house ; Right, Most, Reverend F. in God,
title of bishop, archbishop ; The Holy F., the Pope ; = priest as
prefixed title ; venerable person, god (F. Thames) ; oldestmember,
doyen, (F. of House of Commons, member with longest continuous
service); (pi.) leading men, elders, (Ff. of the City, Conscript Ff,
Roman senators) ; f. -in-law, father of one's wife or husband ;
fatherland, native country (the FA., Germany). Hence
fa*therHOOD,fa*thepsHiP, nn., fa'tnePLESS a., fa'thePLiKE,
fa'thePLV1'2, aa. & advv., fa'thepliNESS n. [Aryan; OE fxder cf. G
rater, L pater, Gk pater] fa'ther2, v.t. Beget ; be the f. of ; originate
(statement &c); pass as, confess oneself, thef., author, of (child,
book) ; govern paternally ; fix paternity of (child, book) upon. [I.
prec] fa'thom1 (-dh-), n. (pi., with numbers, often fathom). Measure
of six feet, chiefly used in soundings; quantity of wood 6ft square in
section, whatever the length. [OE fsethm the outstretched arms, cf.
Gfaden 6ft cogn. w. Gk petannumi spread] fa'thom2, v.t. Encircle
with the arms (archaic) ; measure with f.-line, sound, (depth of
water), (fig.) get to the bottom of, comprehend, whence
fa'thoiriLESs a., fa'thomlessLY * adv. [OE faethmian (prec. )]
fati'dieal, a. Gifted with prophetic power, [f. L fatidicus (fate, -dicus -
saying) + -al] fati'g'ue (-eg), n., & v.t. Weariness after exertion ;
weakness in metals after repeated blows or long strain ; task &c.
that wearies ; iier's non-military duty (f. -party or/., party told off for
this ; so f -dress). ( Vb) tire, exhaust, whence fati'grueLESS, fati'g-
iiiNG2, aa. ; weaken (metal ; see above), [f. Ffatigue(r) I. h fatigare
prob. cogn. w. fatiscere gape] fa*tlingr» n. Young fatted animal, [-
ling1] fa'tten, v.t. & i. Make fat (esp. animals for slaughter) ; grow
fat ; enrich (soil), [-en6] fa'tty J,a. Like fat, unctuous, greasy ;
consisting of fat, adipose ; with morbid deposition of fat (/.
degeneration of heart or kidney). [- y 2]
 FATTY 297 FEATHER fa'tty 2, n. Fat child &c. (usu. voc. ). [-
Y 3] fatuous, a. Vacantly silly, purposeless, idiotic. Hence or cogn.
fatu'iTV n., fa'tuously2 adv., fa'tuousNESS n. [f. Lfatuus + -ousj
faubourg (fo'boorg, or as F), n. Suburb, esp. of Paris. [F] fau'cal, a.
& n. (phonol. ). Of the throat, deeplv guttural (sound). If. L fauces
throat + -al] faireet, n. (dial. & U.S.). Tap for barrel. If. F fa usset
vent-peg etyni. dub.J faug-h (faw), int. of disgust. fault, n., & v.
(geol.) i. & t. Defect, imperfection, blemish, of character or of
structure, appearance, &c. (generous &c. to af, excessively ; with all
JT., at buyer's risk) ; transgression, offence, thing wrongly done,
(Racquets &c.) ball wrongly served ; find f. {with), complain (of),
whence fau'ltfindER l n., fau'ltfindiNG1'2 n. & a. ; responsibility for
something wrong (the f. was mine ; it will be our ownf.), defect that
causes something (the f. is in the patient) ; (Hunt.) loss of the scent,
check so caused, (be at /., also fig. = be puzzled, not know what to
do) ; (Geol.) break in continuity of strata or vein (vb, break
continuity of, show such break) ; (Telegr.) imperfect insulation,
leakage. Hence fau'ltless a., fau\ltlessLY2adv., fau'ltlessNESS n.,
fau'ltY2 a., fau'ltiLY 2 adv., fau'ltiNESS n. [ME & OF faut(e) f. pop. L
*fallita fern, p.p. of fallere fail2] faun, n. One of a class of Latin rural
deities with horns & tail. [f. L Faunus Latin god identified w. Gk Pan]
fau'na, n. (pi. -ae). The animals of a region or epoch ; treatise upon
these. Hence fau'nAL a., fau'niST(3) n., fauni'stic(AL) aa. |mod.L, f.
name of Goddess sister of Faunus see prec] fauteuil (F), n. Arm-
chair; theatre stall. faux pas (lb pah), n. Act that compromises one's,
esp. a woman's, reputation. [F] fa'voup \ n. Friendly regard,
goodwill, (find f. in the eyes of, be liked by ; curr y '2f. ), appro
val(look withf. on), good graces (be, stand high &c, in person's /.) ;
kindness beyond what is due (should esteem it af. ; by f. of — ,
written on letter conveyed by friend ;domethef. of—ing\have
received your f. of yesterday, letter ; woman bestows herff. on lover,
yields) ; leave, pardon, (archaic ; byyourf. ; under f., if one may
venture to say so) ; partiality, over-lenient or over-generous
treatment; aid, furtherance, (under f. of night) ; inf. of, on behalf or
in support of, on the side of, to the advantage or account of, (am in
f. of woman's suffrage ; cheques to be drawn in f. of the treasurer);
thing given or worn -as mark of f., knot of ribbons, rosette, cockade,
badge ; (archaic) looks, countenance, whence well, ill, hard, Sec, -
favouPED2 a. [OF, f. L favorem (J'avere show kindness to, -OR1)]
fa'voup2, v.t. Look kindly upon, approve ; treat kindly, countenance ;
oblige with ; treat with partiality, be unjust on behalf of; aid, support
; serve as confirmation of (theory &c.) ; prove advantageous to
(person), facilitate (process &c), whence fa*voupiNG2a. ; resemble
in features (/. one's father) ; (p.p.) having unusual advantages;
favoured by' (of letter), by f. of. [f. OF javorer f. med.L J avorare as
prec] t(^ , fa'voupafole, a. Well disposed, propitious ^j
commendatory, approving; giving consent (J\ answer) ; promising,
auspicious, (/. aspect) ; helpful, suitable, (to). Hence
fa'voupableness n., fa'vourabLY 2 ad v. [f. ¥ favorable f. ~Lfavorabilis
(favour1, -able)] fa'vounite, n. & a. (Person) preferred above others
(thef. of, af. with or of) ; (Racing) thef., competitor generally
expected to win ; person chosen as intimate by king or superior &
unduly favoured, whence fa*voupitiSM(3) n. [f. OF favoHt, -ri, p.p.
of favorir favour] fawn1, n., a., & v.i. & t. Young fallow deer, buck or
doe of first year (inf., pregnant) ; f. or f.-colour(ed), (of) light
yellowish brown ; (vb ; of deer) bring forth (young, or abs.). [f. OF
faon I. med.L Jetonem nom. -o (foetus)] fawn2, v.i. (Of animals,
esp. dog) show affection by tail-wagging, grovelling, &c. (/. on,
upon, lavish caresses on) ; (of persons) behave servilely, cringe
(upon patron, or abs. ), whence fawniNG2 a., fawningxY2 adv. [OE
fahnian cogn. w. fain1] fay, n. Fairy, [f. OF fae f. Rom. fata sing, f.
hjata pi. the fates] fe'alty, n. Feudal tenant's or vassal's
(acknowledgment of obligation of) fidelity to his lord (do, make,
receive, swear, J.). [I. OF './ eaulte f. L fidelitatem (fidelis f. fides
faith, -ty)] feap1, n. Painful emotion caused by impending danger or
evil, state of alarm (was in /.), dread of, that, or lest; forf. of, (that),
lest, in order that so-&-so may not occur ; dread & reverence (thef.
of God) ; anxiety for the safety of (inf. of his life) ; no /., it is not
likely. Hence feap'LESs a., feap'lessLY 2 adv., feap'lessness n. [OEfxr
cf. G gefahr danger] fear2, v.i. & t. Be afraid (also as archaic refl. in
parenthesis, //. me ; never J., there is no danger of that) ; be afraid
of ; hesitate to do, shrink from doing ; revere (God) ; apprehend,
have uneasy anticipation of ; be afraid that (or with that omitted ;
also need not &c. J. but or but that). [OKfxran f. prec] feap'ful, a.
Terrible, awful; (by exagg.) annoying &c (in af. mess); frightened,
timid; apprehensive of, lest, (that) ; wanting resolution to;
reverential. Hence feap'fulLY2 adv., feap'fulNESS n. [-ful] fear'nought
(-awt), n. Stout woollen cloth used at sea for clothing & for
protecting portholes &c [fear 2, nought] feap'some, a. Appalling,
esp. in appearance (usu. joe). Hence feap'someLY 2 adv.,
feap'someNESS n. [-some] fea'sible (-z-), a. Practicable, possible ;
(loosely) manageable, convenient, serviceable, plausible. Hence
feasiBixiTY n. [OF (fais- imperf. st. off aire f. L facere do -f -ible)]
feast, n., & v.i. & t. Joyful religious anniversary (movable,
immovable, /., recurring on different, same, date) ; annual village
festival ; sumptuous meal, esp. one given to number of guests and
of public nature, (fig.) gratification to the senses or mind (/. oj
reason, intellectual talk). (Vb) partake of f., fare sumptuously,
whence fea'stER l n. ; pass (night &c. ) away in feasting ; regale
(guests), [f. OF Jeste(r) f. L fsta neut. pi. of Jestus festal] feat1, n.
Noteworthy act, esp. deed of valour (often/, of arms) ; action
showing dexterity or strength, surprising trick, [f. OF fait fact] fe t,2
a. (ai-chaic). Adroit, smart, dextrous, neat, fence fea'tLY 2 adv. [f.
OF fait made f. hfn us p.p. of facere make] fea'thep l (fedh-), n. One
of the appendages growing from bird's skin, consisting of quill, shaft,
& two vanes of barbs (show the ivhitef., betray cowardice — white f.
in game-bird's tail being mark of bad breeding — ; crop one's ff.,
humiliate him) ; (collect.) plumage (in high &c. /., in good spirits &c;
birds of a f, people of one sort) ; feathered game (fur &f, game
beasts & birds) ; piece(s) of f. attached to arrow ; plume worn in hat
&c. (af. in one's cap, something one may be proud of) ; very light
object (could have knocked me down xoith af.); ridge of upright hair
; f.-like flaw in gem ; (Rowing) action of 10*
 FEATHER 298 FEELING feathering (see foil.); f.-bed,
mattress stuffed with ff.; /.-edge, (n.) fine edge of wedge-shaped
board, (v.t.) bring (board) to this; f.-head(ed), -brain(ed), -pate(d),
silly (person );/. -stitch, ornamental zigzag sewing; /.-weight, very
light thing or person, esp. jockey not over 4 st. 7 lb., boxer 9 st.
Hence (-)feathePED 2, fea'therless, fea'thepv2, aa., fea'thepiNESS,
fea*thePLET, nn. [com.-Teut. ; OE f ether cf. G feder, also Gk pteron
wing, petomai fly] fea'theP2, v.t. & i. Furnish, adorn, line, coat, with
ff. (/. an arroxo ; /. one's nest, enrich oneself; TXR&f.) ; form f.-like
covering or adornment for; float, move, or wave, like ff.; turn (oar),
turn oar, so as to pass through the air edgeways; (Shoot.) knock fF.
from (bird) without killing ; (Hunt. ; of hound) make quivering
motion of body & tail while seeking scent. [OE gefithrian f. prec]
feathering, n. In vbl senses; esp.: plumage; feathers of arrow ;
feathery structure in animal's coat; (Arch.) cusps in tracery;
featherlike marking in flower, [-ing 2] feature (fetsher), n., & v.t.
(Usu. pi.) part(s) of the face, esp. with regard to shape & visible
effect; distinctive or characteristic part of a thing, part that arrests
attention. ( Vb) stand as distinctive mark upon ; portray, sketch the
prominent points of. Hence -featuPED 2, fea*tUPGLESS, aa. [f. OF
faiture f. L factura (facere fact- make, -ure)] fg'bpifuge, n. Medicine
to reduce fever, cooling drink. Hence febPi'fugAL a. [f. F febrifuge
(hfebris fever, fitga re drive away)] fe'bpile, a. Of fever, feverish. [F,
f. hfebrilis (prec, -il] Fe'bpuapy, n. Second month of year (F. filldike,
name referring to its rain & snow), [f. L Februarius (februa
purification)] fe'citf fecer"unt, (abbr. fee), v. sing. & pi. 3rd pers. (So-
&-so) made this picture &c. (used with artist's signature).
[L,perf.of/aceremake] fe'ckless, a. Feeble, futile, inefficient. Hence
fe'eklessLY2adv., fe'cklessxESS n. [So. feck perh. for effect + -less]
fe'eulent, a. Turbid, fetid. So fe'eulEXCE n. JF (fe-) f. L faeculentus
(faeces, -ulent)] fe'eund, a. Prolific, fertile ; fertilizing. So fepu'ndiTY
n. [f. Yfecond f. Jjfecundus] fe*eundate,v.t. Make fruitful ;
impregnate. Hence feeundA"TiONn. [L FECUNDare, -ate3] fed. See
feed K fe'deral, a. (Theol.) based on doctrine of Covenants; (Pol.) of
the polity in which several States form a unity but remain
independent in internal affairs, concerning this whole & not the
separate parts; (U.S. Hist.) favouring central government, of
Northern party in Civil War. Hence fe*depaliSM(3), fe*depaliST(2),
nn., f e*depalrzE(3) v.t. , f ederaliz a tio nt n. , fed*epalLY 2 adv. [f.
F federal (L foedus -eris covenant, cogn. w. fides faith, + -al)]
fe'depate, v.t. & i. Band together (t, & i.) in league for some
common object ; organize (t. & i. of States) on a federal basis. So
fe*dePATE2 (-at), fe'dePATivE, aa,, fe'depativeLY2 adv. [Iifoederare
(prec), -ate3] f edera'tion, n. Federating, whence fedepa'tioniST(2)
n. ; federated society, esp. federal empire or group of States ;
imperial/, [f. F fe'de'ration f. nfoederationem (prec, -ation)] fee, n., &
v.t. (feed). Fief, feudal benefice, (hist.); inherited estate (f. -simple,
without limitation to particular class of heirs ; f.-tail, with such
limitation ; hold in f. -simple or/., have as absolute property); sum
payable to public officer for performing his function ; remuneration
of lawyer, physician, or any professional man (reTAiNi'ngr/.);
entrance money for examination, society, &c; terminal school-monev
; gratuity ; (vb) pay f. to, engage for a f. [f. OF ft, fief, fiii, med.L
feodum, feudum, etym. dub.] fee'ble, a. & n. Weak, infirm ; deficient
in character or intelligence ; wanting in energy, force, or effect; dim,
indistinct; hence fee'bleness n., fee*bliSH](2) a., fee'bLY2 adv. (N.,
Fenc) = FoiBLE. [f. OF feble, foible, (now faible) f. L flebilis
lamentable (fiere weep)] feed1, v.t. & i. (fed). Supply with food ; put
food into mouth of (cannot f. himself); graze (cattle); gratify (vanity
Sec, also eyes Sec), comfort (person) with hope &c; take food, eat,
(at the high table ; well, high, Sec; often/, on, consume) ; serve as
food for ; nourish, make grow, (/. up, fatten, also satiate) ; keep
(reservoir, fire, &c.) supplied ; supply (machine) with material ; use
(land) as pasture (often /. down, close); deal out (fodder) to
animals; supply (material) (in)to machine ; (of cattle) eat, eat down,
(pasture); feeding-bottle, for hand-fed infants. [OEfe'dan cf. OHG
fuotan, & see food] feed 2, n. Act of feeding, giving of food, (two
biscuits at one f. ; out at/., turned out to graze ; ozone's/., with no
appetite; on the f. of fish, feeding or looking out for food) ;
pasturage, green crops ; horse's allowance of oats &c ; fodder ;
(colloq.) meal, feast; feeding of machine, material supplied, charge
of gun ; /-tank, -tro ugh, holding water for locomotive, [f. prec] feed
3. See fee. fee'dep, n. In vbl senses ; esp.: large, quick, gross, &c /.,
one who eats much &c; child's feeding-bottle; tributary stream (also
fig.); (Rounders &c ) player who tosses ball to striker; hopper or
feeding apparatus in machine, [-er j] feeding-, a. In vbl senses ;
also,/, storm, one that constantly increases, [-ing2] fee-faw-fum, int.
& n. Ogreish exclamation ; (n.) nonsense fit only to terrify child, [in
Jack the Giant-Killer] feel, v.t. & i. (felt), & n. Explore by touch (/.
the pulse o/, lit., & fig. = cautiously ascertain sentiments of ; /.one's
way, grope it out, proceed carefully) ; search (about) with hand
a/ter, /or ; try to ascertain by touch whether, if, hoiv ; (Mil.)
reconnoitre (ground, enemy); perceive by touch [/ a hard substance,
heat, pain, a blow ; I /elt him, move, moving, that he was cold; /
one's legs, realize one's power of walking, also fig., be at ease) ;
have sensation of touch ; be conscious of (sensation, emotion,
conviction) ; be consciously (/. well, warm, angry ; /. quite oneself,
be fit, self-possessed, &c; /. wtoxoork Sic; / like doing, have
inclination to do) ; experience, undergo, (he shall /. my vengeance ;
felt the stor7n severely), be affected by, behave as if conscious of,
(ship feels her helm) ; be emotionally affected by, have sympathy
with or compassion for, (/. the censure keenly) ; have vague or
emotional conviction (that) ; (quasipass.) be realized as, seem,
produce impression of being, (air /eels chilly; feels like velvet). (N.)
sense of touch (firm to the/) ; testingby touch ; sensation
characterizing something, [com.WG ; OE Man, cf. G /uhlen, f. Aryan
palwhence Gk palame, L palma, palm of hand] f ee'lep, n. In vbl
senses ; esp. : organ in certain animals for testing things by touch or
searching for food ; (Mil.) scout ; tentative proposal or hint, ballon
d'essai, [-er !] fee'ling" J, n. In vbl senses; esp.: sense of touch ;
physical sensation ; emotion (often 0/ hope, fear, &c); (pi.)
susceptibilities, sympathies, (hurts my ff., offends me) ; readiness to
feel, tenderness for others' sufferings, (good/, avoidance of
unkindness &c); consciousness
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebookname.com