Transcript Epidemiology

Department of Epidemiology, Medical University of Silesia
EPIDEMIOLOGY
AND ITS CONTRIBUTION
TO MEDICAL RESEARCH
(selected slides)
Jan E. Zejda
TOPICS
Roots of modern epidemiology
Epidemiology among health sciences
Definition of epidemiology
Epidemiological paradigms
Application of epidemiology in medical research
Clinical epidemiology
TOPICS
Roots of modern epidemiology
Epidemiology among health sciences
Definition of epidemiology
Epidemiological paradigms
Application of epidemiology in medical research
Clinical epidemiology
CHOLERA IN EUROPE
1829 in Orenburg (Russia)
1830 in Moscow (Russia)
1831
↓
Poland, Hungary, Austria, Prussia
1832
↓
England, France, Scandinavia,
Spain …
CHOLERA IN EUROPE
Second Wave of Epidemia
1848 in England: 53000 †††
Dr WILLIAM FARR
1807 – 1883
Physician
Founder of Medical Statistics
(measurement of health events:
birth, mortality, etc.)
CHOLERA CARRIED BY AIR POLLUTED BY MIASMA
(miasma = rotting organic matter)
Farr’s hypothesis („miasma theory”)
Deaths due to cholera versus elevation of the residence place in London
140
Deaths
Zgony/10000
120
100
80
60
40
20
0
<20
20-40
40-60
60-80
80-100
100-120
340-360
Stopy nad poziomem morza
Feet above sea level
Źródło: Gordis L.: Epidemiology. W.B. Sounders Company, Philadelphia 1996)
MIASMA THEORY
(miasma in Greek = pollution)
...
Thickness of miasma layer
340-360
100-120
80-100
60-80
40-60
Stopy nad poziomem morza
Feet above sea level
20-40
<20
DR JOHN SNOW
1813 – 1858
One of the first physicians to test the
use of ether and chloroform
as anaesthetics.
Designer of devices to safely
administer ether and chloroform.
C2H5OC2H5
CHCl3
↓
Kidney
Liver
Heart
JOHN SNOW (1813-1858)
on causes of cholera epidemics in London
JOHN SNOW (1813-1858)
on causes of cholera epidemics in London
Snow’s statistical argument
regarding water-borne disease
Mortality distribution according to the place of residence / source of drinking water
Water Supply Company
Number of
houses
Number of
deaths
Deaths rate
(n/10000 houses)
Southwark&Vauxhall
40 046
1263
315
Lambeth
26 107
98
37
Source: Fox J.P. et al.: Epidemiology. Man and Disease. The Macmillan Company, London 1970)
JOHN SNOW (1813-1858)
on causes of cholera epidemics in London
SNOW’S DISTINCT METHODS
• Symptoms of acute intestinal infection (medical
background)
• Location (plotting the cases)
• Measurement of mortality rates (comparisons)
observation → hypothesis → measurement
knowledge → action
IMPORTANT MESSAGE
Cholera put under control long before
the causative agent was discovered
•
•
•
•
response
1848 - Establishment of General Health Department
(London)
1851 - International Health Conference (Paris)
1875 - Public Health Act (Great Britain): water supply,
sewage transportation, etc.
1907 - Establishment of International Health Organization
(Paris)
ROBERT KOCH (1843 – 1910)
[Vibrio cholerae – 1884]
Koch – 1884 !
FILIPPO PACINI (1812 – 1883)
[Vibrio Cholerae – 1854]
EPIDEMIOLOGY TODAY
(from infectious to non-infectious diseases)
Non-infectious diseases
Ch. Niezakaźne
Ch. Zakaźne
Infectious diseases
↑ YESTERDAY
TODAY
↑
√ ADVANCES IN MEDICINE
√ DEMOGRAPHY
√ QUALITY OF LIFE …
TOPICS
Roots of modern epidemiology
Epidemiology among health sciences
Definition of epidemiology
Epidemiological paradigms
Application of epidemiology in medical research
Clinical epidemiology
HEALTH SCIENCES
RESEARCH FIELD
Health
PRIMARY FOCUS
Health care systems
Services
Populations
Epidemiology
Individual patients
Clinical Sciences
Biologic Sciences
Animal models, cells &
transmitters, molecules, genes,
drugs development
Source: Fletcher R.W., Fletcher S.W.: Clinical Epidemiology. The essentials .Lippincott Williams & Wilkins 2005
HEALTH SCIENCES
RESEARCH FIELD
Health
PRIMARY FOCUS
Health care systems
Services
Populations
Epidemiology
Individual patients
Clinical Sciences
Biologic Sciences
Human & animal models, cells &
transmitters, molecules, genes,
drugs development
Source: Fletcher R.W., Fletcher S.W.: Clinical Epidemiology. The essentials .Lippincott Williams & Wilkins 2005
TOPICS
Roots of modern epidemiology
Epidemiology among health sciences
Definition of epidemiology
Epidemiological paradigms
Application of epidemiology in medical research
Clinical epidemiology
DEFINITION OF EPIDEMIOLOGY
The study of the distribution
and determinants of health related states
or events in specified populations,
and the application of this study
to control of health problems
(Last, 2001)
!
EPIDEMIOLOGY
The study of the distribution and determinants of health
related states or events in specified populations, and the
application of this study to control of health problems
STUDY
surveillance, observation, hypothesis
testing, analytic research,
and experiments.
EPIDEMIOLOGY
The study of the distribution and determinants of health
related states or events in specified populations, and the
application of this study to control of health problems
DISTRIBUTION
analysis by person, place and time
„who” – „where” – „when”
?
EPIDEMIOLOGY
The study of the distribution and determinants of health
related states or events in specified populations, and the
application of this study to control of health problems
DETERMINANTS (function)
risk factors, protective factors,
modyfying factors …
DETERMINANTS (source)
HOST
FACTORS
SOCIOECONOMIC
FACTORS
ENVIRONMENTAL
FACTORS
LIFE STYLE
FACTORS
OTHER
FACTORS
- Genes
- Gender
- Age
- Poverty
- Employment
- Isolation
- Air quality
- Water
quality
- Occupational
environment
- Home
environment
- Social
environment
- Nutrition
- Physical
activity
- Tobacco
- Alcohol
- Drugs
- Sexual
activity
-Education
-- Health care
system
- Transport.
- Recreation
EPIDEMIOLOGY
The study of the distribution and determinants of health
related states or events in specified populations, and the
application of this study to control of health problems
HEALTH RELATED STATES
OR EVENTS
diseases, causes of death, behavior,
reactions to preventive regimens,
provision and use of health services …
HEALTH-RELATED PHENOMENA
EPIDEMIOLOGY
The study of the distribution and determinants of health
related states or events in specified populations, and the
application of this study to control of health problems
APPLICATION TO CONTROL
to assess the public health importance of
diseases, identify the population at risk,
identify the causes of disease, describe the
natural history of disease, and evaluate the
prevention and control of disease
EPIDEMIOLOGY
The study of the distribution and determinants of health
related states or events in specified populations, and the
application of this study to control of health problems
APPLICATION TO CONTROL
health promotion
preventive measures
diagnostic standards
therapeutic standards
EPIDEMIOLOGY versus CLINICAL MEDICINE
PRINCIPAL TASKS
CLINICAL MEDICINE
EPIDEMIOLOGY
Diagnosis
Epidemiological study
Treatment
Prevention, health
promotion
Prognosis
Risk assessment
EPIDEMIOLOGY
FROM KNOWLEDGE
Descriptive epidemiology (what, who, where ..?)
Analytical epidemiology (why ?)
TO APPLICATION
Prevention
New diagnostic / therapeutic procedures
Refined provision of health services
…
TOPICS
Roots of modern epidemiology
Epidemiology among health sciences
Definition of epidemiology
Epidemiological paradigms
Application of epidemiology in medical research
Clinical epidemiology
EPIDEMIOLOGIC PARADIGMS
Non-random occurrence of diseases
Between-subject variability in health and
exposures* leading to co-variability
• - a very broad concept (smoking=exposure; gene variant=exposure; etc..)
& Reliability of the results
DISEASES DO NOT OCCUR BY RANDOM
?








?
?
?
WHY ME ?
NON-RANDOM OCCURRENCE IMPLIES CAUSATION
EXPOSURE TO WHAT ? (HARMFUL AGENTS ? POOR GENES ? … ?)
Q: IS KNOWN HEALTH DISORDER
RELATED TO KNOWN EXPOSURE ?
A:
COVARIABILITY – Y/N
(1) Between-subject variability in health status
(2) Between-subject variability in potential risk factors
(1) follows (2) = covariability
Reliable Measurement of Health and Exposure
Evidence of Covariability
SIR DAVID ROXBEE COX
FOUR CHALLENGES:
MEASUREMENT
STUDY DESIGN
DATA ANALYSIS
INTERPRETATION
David R. Cox: Some challenges for medical statistics (symposium „Modern Statistical Methods
in Medical Research” - Nobel Foundation Series „Frontiers in Medicine”
< European Journal of Epidemiology 2005;30:5-9 >
LORD KELVIN (1824-1907)
When you can
measure what you
are speaking about,
and express it in
numbers, you know
something about it.
But when you
cannot your
knowledge is of
a meager and
unsatisfactory kind
(Lord Kelvin)
MEASUREMENT
- Quantitative scale: blood glucose level …
- Qulitative scale: round opacity on chest x-ray …
SOURCES
Interview, standardized questionnaire, laboratory tests,
composed indices … case definition …
MEASUREMENT - ERRORS
TYPE I: above, above, below, above, below, below, above, below, above, above
TYPE II: either
or
above, above, above, above, above, above, above
below, below, below, below, below, below, below
Repeat measurements and you are
Repeat measurements and you
close to the true value (average)
are always one way away from
the true value
RANDOM ERROR
SYSTEMATIC ERROR
(BIAS)
GOOD EPIDEMIOLOGY PRACTICE
EPIDEMIOLOGICAL STUDY – A MEASURING TOOL
Prevalence of diabetes
Role of sleep deprivation in high blood pressure
Protective significance of face mask in COPD
etc., etc.
MEASUREMENT IS SUBJECT TO ERROR
RANDOM ERROR
Divergence – due to chance alone – of an observation
(measurement) on a sample from the true population value
(estimate ≠ parameter). Three sources:
• Biological variation;
• Sampling error;
• Measurement error.
Given biological variation and limited accuracy
of the measurement random error can be reduced
by increasing the size of the sample
( estimate obtained in a sample composed of all candidates = parameter)
SYSTEMATIC ERROR = BIAS
Systematic divergence (either ‘in plus’ or ‘in minus’) of
an observation (measurement) on a sample from the
true population value (estimate ≠ parameter). Three
principal classes of bias:
•Selection bias;
•Information bias;
•Confounding.
SELECTION BIAS
Distortion resulting from the manner in which subjects
are selected (from population to sample)
• Berkson’s bias (hospital admission bias): affects studies done
in hospitalized populations – the relation found in hospital may
not reflect the one in the population;
• Nonresponse bias: „refusers” are more likely to engage in
unhealthy behaviors
• Healthy worker effect: seriously ill people are unable to join and
remain in the workforce
• …
Way to avoid (minimize) = representative, probability sampling
INFORMATION BIAS
Causes: measurement device defects, inappropriate
questionnaires, inaccurate diagnostic procedures,
Result: placement of a subject in a wrong category
(misclassification)
•
Recall bias: sick individuals are more likely to recall hazardous
exposures, events, etc. (mothers of leukemia children remember all,
mothers of healthy children „do not care”)
•
Diagnostic suspicion bias: increased diagnostic attention toward
exposed (endometrial cancer versus hormone replacement therapy –
estrogens)
•
…
CONFOUNDING
Distortion in an association between a study exposure and
disease brought about by the influence of „a third factor”
Confounder:
Is associated with the exposure in question;
Is an independent risk factor for the disease
Confounder
↨
Exposure
→→
Disease
→?
Is not intermediate in the causal pathway: exposure - disease
RANDOM ERROR AND SYSTEMATIC ERROR
True Value of the Studied Event = bull’s eye
Result of the Study = dot
Random Error (Imprecision)
Low
Low
Systemtic
Error (Bias)
High
High
VALIDITY
• Internal validity (state-of-art measurements on
sample)
• External validity (generalizability)

GOOD EPIDEMIOLOGY PRACTICE !
GOOD EPIDEMIOLOGY PRACTICE
Identify potential sources of errors
Avoid errors
Control errors
Adjust for errors
TOPICS
Roots of modern epidemiology
Epidemiology among health sciences
Definition of epidemiology
Epidemiological paradigms
Application of epidemiology in medical research
Clinical epidemiology
Core Contributions
1,2,3,4
EPI CONTRIBUTION TO MEDICAL RESEARCH
I – INVESTIGATION INTO NATURAL HISTORY OF DISEASES
Good
Health

Subclinical
Changes

Clinical
Disease

Death
or
Recovery
EPI CONTRIBUTION TO MEDICAL RESEARCH
II – DESCRIPTION OF HEALTH STATUS OF POPULATION
Occurrence
of diseases
(health events)
Incidence



According to sex
According to age
According to …
Prevalence



According to sex
According to age
According to …

EPI CONTRIBUTION TO MEDICAL RESEARCH
III – EXPLORATION OF CAUSATION
Lifestyle
factors

Genetic
factors
Genetic
factors


Good
Health

Onset
of
disease

Ill
Health


Environmental
factors
Environmental
factors

Lifestyle
factors
EPI CONTRIBUTION TO MEDICAL RESEARCH
IV – EVALUATION OF INTERVENTION
Treatment

Good
Health
← ← ←
Ill
Health


Health promotion
Preventive measures
Public health services
Medical Care

Diagnostic
Procedures
TOPICS
Roots of modern epidemiology
Epidemiology among health sciences
Definition of epidemiology
Epidemiological paradigms
Application of epidemiology in medical research
Clinical epidemiology
EPIDEMIOLOGY and CLINICAL MEDICINE
CLINICAL MEDICINE
EPIDEMIOLOGY
Patient
Population/Group
Treatment
Prevention, health
HOWEVER
promotion
Prognosis
Risk
assessment
Patient
A responds to the treatment T, patient
B does
not, and patient C …
(what is the true effect of the treatment T ?)
Low level of enzyme E has a prognostic value in patient A and not in patient B
(what is the true diagnostic value of enzyme E ?)
EPIDEMIOLOGY and CLINICAL MEDICINE
CLINICAL MEDICINE
Treatment
CLINICAL
Prognosis
EPIDEMIOLOGY
Prevention, health
promotion
EPIDEMIOLOGY
Risk assessment
CLINICAL EPIDEMIOLOGY
The science of making predictions about individual
patients by counting clinical events in groups of
similar patients and using strong scientific methods
to ensure that the predictions are accurate.
The purpose of clinical epidemiology is to develop
and apply methods of clinical observation that will
lead to valid conclusions by avoiding being misled
by systamatic error and the play of chance
Source: Fletcher R.W., Fletcher S.W.: Clinical Epidemiology. The essentials .Lippincott Williams & Wilkins 2005
CLINICAL EPIDEMIOLOGY
„Clinical” because it seeks to answer questions and
to guide clinical decisions making with the best
available evidence
„Epidemiology” because many of the methods
used to answer the questions have been developed
by epidemiologists (patient as a member of the
population of patients – disease specific)
Source: Fletcher R.W., Fletcher S.W.: Clinical Epidemiology. The essentials .Lippincott Williams & Wilkins 2005
CLINICAL EPIDEMIOLOGY ISSUES
Abnormality
Diagnosis
Frequency
Risk
Prognosis
Treatment
Prevention
Cause
Source: Fletcher R.W., Fletcher S.W.: Clinical Epidemiology. The essentials .Lippincott Williams & Wilkins 2005
FROM CO-EXISTENCE TO SYNERGISM
EPIDEMIOLOGY vs CLINICAL MEDICINE
EPIDEMIOLOGY ↔ CLINICAL MEDICINE