Sie sind auf Seite 1von 27

Population Health

for
Health Professionals

1
Module 2

Epidemiology
The Basic Science
of Public Health

Epidemiology
The Basic Science of Public Health

The purpose of this module is to provide an overview of the fundamentals of


epidemiology as the basis for the science of public health. The first two
hours are in the same format as the other modules. The third hour
incorporates an interactive epidemiology exercise from the University of
Illinois, Chicago School of Public Health web site. The program takes
students through the investigation of an outbreak of disease in a school
setting, and can be completed as an in-class group exercise or by students
individually as an out-of-class homework assignment.
The module does not include details on basic statistical terms (e.g., mean,
median) or basic statistical calculations (e.g. chi-square). If your students
have not been introduced to these methods, it will be necessary to insert
them into the module at appropriate points. On the other hand, if your
students have had extensive courses in statistical methods, you may want to
abbreviate this module.
It is also advisable that you identify current issues or cases most relevant to
your students area of practice for use during this module.

2
OBJECTIVES

z epidemiology and role as foundation for


public health
z common measures of disease frequency
z strengths and weaknesses of study
designs
z applications of epidemiologic method

At the conclusion of this module, the learner will be able to:

define epidemiology and explain its role as the foundation for public health
describe three common measures of disease frequency
discuss the strengths and weaknesses of three major epidemiologic study
designs
identify at least two applications of the epidemiologic method

3
Part I

Key Terms and Measures

4
Epidemiology

z Study of distribution of determinants


and antecedents of health and disease
in human populations
z Application of results to control of health
problems

Epidemiology is the study of the distribution of determinants and


antecedents of health and disease in human populations, and the application
of study results to the control of health problems.

5
From Hippocrates
to John Graunt
z Fifth century BCE, Hippocrates pointed
to the need to understand the
environment and the risks it posed to
understand the experience of disease
z 1662, John Graunt analyzed weekly
reports of births and deaths in London,
quantifying patterns of disease in the
population

Hippocrates: Whoever wished to investigate medicine properly should


proceed thus: in the first place to consider the seasons of the year, and what
effects each of them produces. Then the winds, the hot and the cold,
especially such as are common to all countries, and then such as are
peculiar to each locality. In the same manner, when one comes into a city to
which he is a stranger, he should consider its situation, how it lies as to the
winds and the rising of the sun; for its influence is not the same whether it
lies to the north or the south, to the rising or to the setting sun. One should
consider most attentively the waters which the inhabitants use, whether they
be marshy and soft, or hard and running from elevated and rocky situations,
and then if saltish and unfit for cooking; and the ground, whether it be naked
and deficient in water, or wooded and well watered, and whether it lies in a
hollow, confined situation, or is elevated and cold; and the mode in which the
inhabitants live, and what are their pursuits, whether they are fond of
drinking and eating to excess, and given to indolence, or are fond of exercise
and labor.
Graunt: Noted excess of men compared with women in births and deaths,
high infant mortality rate, and seasonal variations alluded to by Hippocrates.
Also attempted a numerical assessment of the impact of the plague on the
population, and the characteristics of years in which such outbreaks
occurred.

6
From William Farr
to John Snow
z 200 years later, Dr. William Farr was
made responsible for medical statistics
in the Office of the Registrar General for
England and Wales
z A mere 20 years later, John Snow
completed his study of cholera

Farrs contribution is the tradition of using vital statistical data to study health
problems. His studies compared married to single persons, workers in
different occupations (metal mines and earthenware industry), elevations
above sea level, and the effects of incarceration.

If your class did not complete Module 1a, you might want to include at this
point the slides from that module on John Snow and cholera in London,
which provide a dramatic illustration of epidemiologic analysis.

7
Modern Experiences

z Evaluation of risk factors for chronic


diseases using case controls
z Long term population studies using
cohorts
z Design of clinical trials to evaluate
interventions

Evaluation of risk factors for chronic diseases using case controls


Case control studies compare a group of people with a disease or condition
to another group of people without it. The Doll and Hill (1950) study of
cigarette smoking and cancer in Britain is a classic example, and is credited
with starting our current series of efforts to control tobacco use.
Long term population studies using cohorts
In cohort studies, subjects are categorized on a predetermined basis and
followed over time for the development of health conditions. One well-known
example is the Framingham Heart Study in which 5200 residents were
followed over 35 years. Findings of this study have been used to develop
improved cardiovascular disease prevention methods.
Design of clinical trials to evaluate interventions
Clinical trials in humans are conducted to determine whether methods found
effective in laboratory conditions can be safely applied to a large population
under normal conditions to demonstrate its application to the control of
disease.

8
Three Essential Components

z Disease distribution
z Disease determinants
z Disease frequency

Expected level
Endemic
Sporadic
Epidemic
Pandemic

Disease distribution: how are cases of the condition of interest spread


across a population differently by gender, age, geographic location, socio-
economic status, other features?
Disease determinants: what risk factors or antecedent events are
associated with the appearance of a disease or condition?
Disease frequency: how many cases of the condition occur over a given
time period?
Expected level - baseline level of observed occurrence of a particular
disease
Endemic - persistent occurrence at a low to moderate level,
sometimes referred to as a high background rate
Sporadic - irregular pattern with occasional cases occurring at
irregular levels
Epidemic - occurrence of a disease within an area exceeds expected
level for a given time. Also called an outbreak.
Note that these mean basically the same thing, but public
perspective is that epidemic is much more serious than an
outbreak.
Pandemic - epidemic that has spread over several countries or
continents, affecting large numbers of people

9
Epidemiologic Studies

z Descriptive
z Analytic

Descriptive Epidemiology provides the Who, What, When, and Where of


health-related events in a population.

Analytic Epidemiology attempts to provide the Why and How of health-


related events in a population.

10
Descriptive Studies

z Frequency of occurrence of particular


condition
z Patterns of occurrence according to
person, place and time

In Descriptive Studies, the epidemiologist is concerned with collecting


information that characterizes and summarizes a health event or health
problem.
Routinely collected data from such sources as death certificates, hospital
discharge records, health surveys (e.g., cross-sectional surveys) and
disease surveillance programs are used for most descriptive studies.
Characteristics related to person may include age, gender, race,
ethnicity,marital status, socioeconomic class and occupation.
Descriptive studies on occurrence of conditions according to place might
involve examining their frequency within or between natural or political
boundaries, urban versus rural localities, or latitude.
Examination of time relationships can both identify and evaluate possible
causes for changes in health conditions.

11
Analytic Studies

z Observational studies
case-control studies
cohort studies
prospective
retrospective

z Experimental studies

Case control studies: Case-control studies are those in which persons


with a specified condition (the cases) and pesons without the condition (the
controls) are selected for study. The proportion of cases and controls with
certain characteristics or exposure is then measured and compared. For
example, knowing that there are 10 school children with purple spots in
grade 3, a set of other third grade children from the same school but without
purple spots would be identified as controls, and analysis done to see what
different exposures the purple-spotted children had than the non-spotted.

Cohort studies: groups of individuals with some common feature (age and
geography, for example) are identified for study over time to learn about
differing health and illness experiences. For example, one might enroll in a
study all third graders in a school and follow them until graduation,
attempting to identify the differences in experiences of those who maintained
a body weight close to recommended and those who did not.

The next slides illustrate some of the differences in timing in these types of
studies: case control, prospective cohort, and retrospective cohort.

12
Case Control Study

Exposure Disease
?

Key
Basis for selection of group for study

= present = absent

KEY

Investigator at beginning of study

? To be determined

At the beginning of the case control study, the investigator knows that there
are some people with a disease; they are matched with similar individuals
(controls) who do not have the disease. The investigator looks backward to
identify what different exposures the two groups might have had. For
example, when some individuals attending a picnic become ill, they could be
matched with controls who also attended the picnic but did not become ill,
and all interviewed about what was eaten, to identify a possible source of
food-borne illness.

13
Prospective Cohort Study

Exposure Disease
?

Key

Basis for selection of group for study

= present = absent

KEY

Investigator at beginning of study

? To be determined

At the beginning of a prospective cohort study, the investigator is aware of a


group of individuals, some of whom have been exposed to a hazard. All
members of the cohort will be followed over time to see if those exposed and
those unexposed have different disease experiences.

For example, a public health department may be informed of the exposure of


a portion of a school class to an individual with an active case of a
communicable disease in the course of a field trip. The entire class (the
cohort) would be observed over time to identify any cases of disease that
arise, and any difference in disease rate between the two groups.

14
Retrospective Cohort Study

Exposure Disease
?

Key

Basis of selection of group for study

= present = absent

KEY

Investigator at beginning of study

? To be determined

At the beginning of a retrospective cohort study, the investigator is aware of


an exposure to a hazard that occurred at some time in the past, sufficiently
long ago that if disease were to have occurred, it should by now be evident.
A cohort that includes the exposed individuals is identified, and the health
histories of all members explored to identify the presence or absence of
disease in all individuals, and the difference in rate between exposed and
non-exposed.

Many of the studies of association between environmental exposures and


disease have been of this type. A cohort of individuals who lived or worked
in an area but had different experiences of exposure/non-exposure to a
chemical will be identified. Their health histories in the intervening years will
be examined to identify differences, if any, in their rates of disease.

15
Analytic Studies

z Observational studies
z Experimental studies
Intervention studies
Clinical trials

Experimental studies generally provide the strongest evidence that a given


response is the cause of a disease or that a specific intervention is effective.
The controlled clinical trial, the gold standard of studies of efficacy of
disease interventions or prevention methods, is conducted by identifying an
eligible population, dividing it into control and test groups (usually randomly),
applying the intervention, and watching to see what the effect is compared to
the control group.
Experimental studies typically are expensive and challenging to conduct.
The methodology for these studies is extensively described in a variety of
resources, and is not a major focus of this module.

16
Basic Presentation of Results

Disease Disease
TOTAL
Yes No
Exposed
Yes A B A+B
Exposed
C D C+D
No
TOTAL A+C B+D A+B+C+D
All rates and ratios discussed can be
calculated from this.

Chi-square is the basic statistic used to test for the significance of any differences
noted in distribution of disease or risk in a table such as this.

A= the number of individuals who are exposed and have the disease
B= the number of individuals who are exposed and do not have the disease
C= the number of individuals who are not exposed and have the disease
D= the number of individuals who are both non-exposed and non-diseased

A+B+C+D= the total number of individuals in the population being studied

A+B= the total number of individuals exposed


C+D= the total number of individuals non-exposed
A+C= the total number of individuals with the disease
B+D= the total number of individuals without the disease

Any rates needed for epidemiologic analysis can be calculated from this basic
table. Unless your students are extremely familiar with statistics, you should insert
an example that will be of interest, and take some time to discuss it.

17
Smoking
and Carcinoma of the Lung
Disease status # of smokers # of nonsmokers P-value
Males
Lung cancer 647 (99.7%) 2 (0.3%)
Controls 622 (95.8%) 27 (4.2%) 0.00000064
Females
Lung cancer 41 (68.3%) 19 (31.7%)
Controls 28 (46.7%) 32 (53.3%) 0.016

Doll R, Hill AB 1950 Smoking and carcinoma of the lung: preliminary report. British Medical
Journal 2:739.

This table shows the association between smoking and lung cancer, taken
from the historic study by Doll and Hill.

Men and women with and without lung cancer were categorized as smokers
or non-smokers. Thus, the exposure was smoking; the disease outcome of
interest was lung cancer.

Lung cancer among exposed individuals is far in excess of lung cancer


among non-exposed individuals.

18
Interpreting Results:
Measurement Errors
z Bias
information
selection
z Confounding
extraneous factors
z Effect modification
statistical interaction

Bias is the tendency of a measurement to deviate from the true value.


Information bias derives from measuring the outcome or exposure such that
information is more accurate or more complete in one group than another; this
includes interviewer or observer bias, recall bias and reporting bias. Selection
bias occurs because of differences between those who are not selected for a
study or a group within a study; this includes ascertainment bias, detection bias,
and response bias.

Confounding can occur when a variable related to the condition under study is
associated with, but not a consequence of, the exposure under investigation.
While it is never possible to control all factors in an epidemiologic study, findings
will be easier to interpret if extraneous factors are eliminated or made identical
for cases and controls.

Effect modification occurs when the magnitude of the association between one
variable and another differs according to the level of a third variable. Detecting
effect modification is an important aspect of data analysis.

19
Interpreting Results: Cause-
Effect Relationship
z Strength of the association
z Consistency
z Temporality
z Plausibility
z Biological gradient

In general, there are five criteria that must be met to establish a cause and
effect relationship. These are:

Strength of association: Power and sample size must be sufficient to


demonstrate a statistically significant difference.

Consistency: Observation of the association must be repeatable in different


populations at different times.

Temporality: The cause must precede the effect.

Plausibility: The explanation must be biologically plausible.

Biological gradient: The presence of a dose-response relationship.

20
Measures of Disease
Frequency
z Prevalence
z Incidence

Prevalence and incidence are commonly confused. They are similar, but
differ in the number of cases included in the numerator:

Prevalence includes all cases during a given time period

Incidence includes only the number of new cases during a given time period.

21
Prevalence

z Prevalence= number of existing cases


divided by total population

Visual examination survey


310 X 100 = 12.5%
2477

Prevalence is the proportion of persons in a population who have a


particular disease or attribute at a specified point in time or over a specified
period of time.

The numerator for prevalence includes all persons during a specified interval
or point in time, regardless of when the condition began.

For example, a visual examination survey of 2477 persons between the ages
of 52 and 85 years showed that 310 had cataracts. The prevalence of the
condition was

310 X 100 = 12.5%


2477

22
Types of Prevalence Measures

R a te N u m erato r D en o m in ato r
Disease rate at Number of cases of disease Number of persons
autopsy autopsied

Birth defect rate Number of babies with a given Number of live births
abnormality

Smoking rate Number of people who smoke Total population

This chart shows several types of prevalence rates. All have in common a
numerator that includes all cases (new and old) of the condition under study.

Note: The term period prevalence is sometimes used to denote the


number of existing cases plus new cases diagnosed during a given time
period, divided by the total population.

23
Incidence

z Incidence = number of new cases in a


given period of time divided by the total
population at risk
Bacteremia among contraceptive users

27/483 X 100 = 5.6%

Incidence is the proportion of persons in a population who develop or


demonstrate a particular disease or attribute during a specified period of
time.

The numerator for incidence includes only those persons who develop the
condition during the specified time period.

For example, in a study of 2390 women between 16 and 49 years of age, it


was found that 482 used oral contraceptives. Twenty-seven of the oral
contraceptive users developed bacteremia. The incidence was therefore:
27/482 X 100 = 5.6%

24
Types of Incidence Measures

Rate Numerator Denominator


New cases of non- Total population at risk
Morbidity
fatal disease
Number of deaths Total population
Mortality
from a disease
Number of deaths Number of cases of that
Case-fatality
from a disease disease
Number of cases of a Total population at risk
Attack
disease during a specified period

This chart shows several types of incidence rates. All have in common a
numerator that includes only new cases of the condition under study.

Note: When the denominator is the size of the population at the start of the
time period, the measure is sometimes called cumulative incidence. It is a
measure of the probability or risk of the disease or condition, i.e., what
proportion of the population will develop an illness or condition during the
specified time period. In contrast, the incidence rate indicates how quickly
people become ill measured in people per year.

25
Rates Commonly Used in
Epidemiology
z Crude
z Category specific
z Age adjusted

Crude rates are those calculated for an entire population, such as the
annual cancer mortality rate.

Category specific rates are based on the number of persons in the


category and the number of cases occurring in that group, such as the age-
specific cancer death rate.

Age adjustment allows for more appropriate comparisons when differences


in age distribution in the two populations may mask real differences in the
condition of interest.

Rates imply a change over time. For disease incidence rates, the change is
from a healthy state to disease for a specified period of time.

26
Module Author

Fran C. Wheeler, Ph.D

School of Public Health


University of South Carolina

Dr. Wheeler is the Director of the Office of Public Health Practice at the School of Public
Health, University of South Carolina. Prior to joining the faculty at USC, she worked for
over twenty years in chronic disease prevention and health promotion programs at the
South Carolina Department of Health and Environmental Control. As a long-time
consultant with the Centers for Disease Control and Prevention, she also has experience
in public health policy development at the national level. Dr. Wheeler has published
research on the epidemiology of chronic diseases and related risk factors, as well as
community-based public health interventions to address those problems in South
Carolina. Her professional interests include policy and environmental interventions for
chronic disease prevention and health promotion, translating research into public health
practice, and minority health.

Fran C. Wheeler, Ph.D.


Director, Office of Public Health Practice
School of Public Health
University of South Carolina
Columbia, SC 29208
Phone: 803.777.5054
Fax: 803.777.4845
E-mail: fran.wheeler@sc.edu

27