DEPARTEMTN OF COMMUNITY HEALTH

FACULTY OF MEDICINE
ADDIS ABABA UNIVERSTY

COMH 603
Epidemiology I
Principles of Epidemiology
(3 credits)

LECTURE NOTE

By
YEMANE BERHANE

1997
 TABLE OF CONTENTS (PAGE)

I. INTRODUCTION TO EPIDEMIOLOGY.....................................................................2
II. COMMUNICABLE DISEASE EPIDEMIOLOGY.....................................................2
III. TYPES OF EPIDEMIOLOGIC STUDIES................................................................2
IV. MEASUREMENT IN EPIDEMIOLOGY...................................................................2
V. EPIDEMIOLOGICAL DESIGN STRATEGIES.........................................................2
VI. EVALUATION OF EVIDENCE.................................................................................2
VII. PRESENTATION OF EPIDEMIOLOGIC...............................................................2
INFORMATION.................................................................................................................2
VIII. OUTBREAK INVESTIGATION AND.....................................................................2
MANAGEMENT................................................................................................................2
IX. EPIDEMIOLOGIC SURVEILLANCE.......................................................................2
X. SCREENING..................................................................................................................2
XI. ETHICS OF EPIDEMIOLOGIC RESEARCH..........................................................2
DCH/AAU: Epidemiology Note 1

I. INTRODUCTION TO EPIDEMIOLOGY
A. Definition

Epidemiology is the study of the frequency, distribution, and determinants of health-

related states or events in specified population, and the application of this study to the
control of health problems.

The definition emphasizes that epidemiologists are concerned with the collective health
of the people in communities; unlike the clinicians who are concerned with the health of
the individual.

B. Basic Epidemiologic Assumptions

1. Human disease does not occur at random.e.g HIV sexually active,IVDU

2. Human disease has causal and preventive factors that can be identified through
systematic investigation of different populations or subgroups of individuals within a
population in different places or at different times.

C. Scope/use of epidemiology

Epidemiology has been used in several ways in the planning and evaluation of health
intervention in an effort to improving the health status of the population. Four uses are
mentioned here:

1. Elucidation of the natural history of disease.

2. Description of the health status of the population.e.g IMR,MMR
3. Establishing causation of disease.
4. Evaluation of intervention

D. Major categories of epidemiology

1. Descriptive Epidemiology – Defines the amount and distribution of health problems.

It answers the questions: Who, what and where.
2. Analytic Epidemiology – Analyses the causes or determinants of health and disease.
Answers the questions why and how.

E. History of Epidemiology

Although epidemiologic thinking has been traced to the time of Hippocrates, the
discipline did not flourish until the 1940s. Some key dates and contributions to the
development of epidemiologic thinking and methods include:

1
DCH/AAU: Epidemiology Note 2

1662- John Graunt published Natural and Political Observations on the Bills of
Mortality. He was the first to quantify patterns of birth, death and disease
occurrence, noting male- female disparities, high infant mortality, urban-rural
differences, and seasonal variations.

1747- Lind used an “experimental” approach to prove the cause of scurvy by

showing it could be treated effectively with fresh fruit.

1839- William Farr took responsibility for medical statistics in the Office of the
registrar General for England and Wales. He extended the epidemiologic analysis
of morbidity and mortality data, looking at effects of marital status, occupation,
and altitude.

1854 – John Snow demonstrated that the risk of mortality due to cholera was
related to the drinking water provided by a particular supplier in London. He used
a “natural experiment” to test his hypothesis. In another study conducted by Snow
in 1854, he linked an epidemic of cholera to a specific pump, the “Broad Street
Pump”. According to literatures, Snow removed the handle of that pump and
aborted the cholera epidemic.

Originally epidemiology was concerned with epidemics of communicable disease.

Lately, epidemiology was extended to endemic communicable diseases and non-
communicable infectious diseases. More recently, epidemiologic methods have been
applied to chronic diseases, injuries, birth defects, maternal and child health, occupational
health, and environmental health. Now, even health behaviours, such as care-seeking,
safety practices, violence, and hygienic practices are valid subjects for epidemiologic
investigation.

II. COMMUNICABLE DISEASE EPIDEMIOLOGY

A. Introduction

Despite the great scientific advances that have reduced morbidity and mortality from
communicable diseases over the past decades, communicable diseases continue to
account for a major proportion of acute illness, even in technologically advanced
countries, though the types of diseases vary from place to place. Some important aspects
of infectious diseases are discussed below.

B. Natural History of Diseases

The natural history of disease refers to the progress of a disease

process in an individual over time, in the absence of
intervention.

2
DCH/AAU: Epidemiology Note 3

The process begins with exposure to the causative agent capable of causing disease.
Without medical intervention, the process ends with recovery, disability, or death. Most
disease has a characteristic natural history, although the time frame and specific
manifestations of disease may vary from individual to individual. The usual course of a
disease may be halted at any point in the progression by preventive and therapeutic
measures, host factors, and other influences. The stages in the natural history of disease
are shown in Figure 1.

(What happen after exposure?)

Figure 1
NATURAL HISTORY OF DISEASE

USUAL TIME
OF DIAGNOSIS

PATHOLOGIC ONSET OF
EXPOSURE CHANGES SYMPTOMS

STAGE OF STAGE OF STAGE OF STAGE OF RECOVERY

SUSCEPTIBILITY SUBCLINICAL DISEASE CLINICAL DISEASE DISABILITY OR DEATH

(1) (2) (3) (4)

C. Components of Infectious Disease Process

Infectious diseases result from the interaction of infectious agent, susceptible

host/reservoir and environment that brings the host and the agent together.

Agent: refers to an Infectious micro-organism – virus, bacteria, parasite, or other

microbe.

Host: host factors influence individual’s exposure, susceptibility or response to a

causative agent. For example – age, sex, race, socioeconomic status, and
behaviours (smoking, drug abuse, lifestyle, sexual practices and
contraception, eating habits) affect exposure.

Environment: environmental factors are extrinsic factors which affect the agent and the
opportunity for exposure. Physical factors such as geology, climate, and
physical surrounding (e.g., maternal waiting home, hospital); biologic
factors such as insects that transmit the agent; and socioeconomic factors
such as crowding, sanitation, and the availability of health services.

3
DCH/AAU: Epidemiology Note 4

D. Causal Concepts of Disease

Not all associations between exposure and disease are causal. A cause of a disease can be
defined as a factor (characteristic, behaviour, event, etc.) that influences the occurrence of
disease. If disease does not develop without the factor being present, than we term the
causative factor “necessary”. If the disease always results from the factor, then we term
the causative factor “sufficient”.
Example: Tubercle bacilli is a necessary factor for tuberculosis.
Rabies virus is sufficient for developing clinical rabies.

The epidemiologic triad or triangle is the traditional model of infectious disease

causation. It has three components: an external agent, a susceptible host, and an
environment that brings the host and agent together, as shown in the two diagrams in
figure 2:

Gigure 2
EPIDEMIOLOGIC TRIANGLE AND TRIAD (BALANCE BEAM)

Agent

Agent Host

Host Environment

Environme

In recognition of the multifactorial nature of most diseases, several other models have
been proposed. Causal pie model is one of the models, which takes into account multiple
factors, which are important in causation of disease. In the causal pie model., the factors
are represented by pieces of the pie called component causes, as shown in figure 3:

Figure 3
ROTHMAN’S XAUSAL PIES: CONCEPTUAL SCHEME FOR DISEASE CAUSATION

SUFFICIENT SUFFICIENT SUFFICIENT

CAUSE CAUSE CAUSE
I II III
D
E H G J I
C
A A F A F
B B
C

4
DCH/AAU: Epidemiology Note 5

E. Chain of Infection

Infection implies that the agent has achieved entry and

begun to develop or multiply, whether or not the process
leads to disease. This is sometimes called the chain of
infection, or transmission cycle, which can be illustrated as
follows:

Components of Chain of Infection

1. Causative agent
2. Reservoir host
3. Portal of exit
4. Mode of transmission
5. Portal of entry
6. Susceptible host.

Figure 4
CHAIN OF INFECTION

The reservoir of an agent is the habitat in which an

infectious agent normally lives, grows, and multiplies.

5
DCH/AAU: Epidemiology Note 6

Agents with a human reservoir include measles, mumps,

and most respiratory pathogens. Human reservoirs may be
persons with symptomatic illness, or carriers. A carriers is a
person without apparent disease who is nonetheless capable
of transmitting the agent to others. The importance of
carriers in the transmission of disease depends on their: 1)
number, 2) detectability, 3) mobility, and 4) chronicity.

Carriers may be:

Asymptomatic carriers (transmitting infection without
ever showing signs of the
disease),
incubatory carriers (transmitting infection by shedding
the agent before the onset
of clinical manifestations), or
convalescent carriers (transmitting infection after the
time of recovery from the
disease).
Chronic carriers shed the agent for a long period of
time, or even indefinitely.

The chain of infection may be interrupted if the agent does

not find a susceptible host. This may occur if a high
proportion of individuals in a population is resistant to the
agent. Through such herd immunity, immune persons limit
the spread to the relatively few who are susceptible by
reducing the probability of contact between infected and
susceptible persons. Herd immunity operates best when
there is: 1) a single reservoir, 2) direct transmission, 3) total
immunity, 4) no shedding of the agent by immune hosts, 5)
a uni for distribution of immunes, and 6) no overcrowding.

6
 DCH/AAU: Epidemiology Note 7

Figure 5
TIME COURSE OF A DISEASE
IN RELATION TO ITS CLINICAL EXPRESSION AND
COMMUNICABILITY
Symptomatic

Clinical case
Clinical expression

Convalescent carrier
Incubatory carrier

Clinical
Threshold
Asymptotic

Asymptomatic carrier
Chronic carrier

Time

Time of agent first recovery agent

relapse
Infection starts manifestations stops
(biological being of disease being
onset) shed (clinical onset) shed

incubation period latent period

prepatent communicable period

period

E. Modes of Transmission of infectious Agents

The mechanism by which the agent escapes from a reservoir host and enter
into a susceptible host is referred as mode of transmission. There are two
major modes:

7
DCH/AAU: Epidemiology Note 8

1. Direct Transmission - immediate transfer of the agent from a reservoir to

a susceptible host by direct contact or droplet
spread.
Touching
Kissing
Sexual intercourse
Blood transfusion
Transplacental (vertical) from mother – child

2. Indirect Transmission- an agent is carried from reservoir to a

susceptible host by suspended air particles or by
animate (vector-mosquitoes, fleas, ticks …) or
inanimate (vehicle-food, water, biologic products,
fomites) intermediaries.
Vehicle-born: food, water, towels, …
Vector-borne: insect animals, …
Airborne: dust, droplets
Parenteral injections

F. Levels of Disease Prevention

Disease prevention means to interrupt or slow the progression of disease.
Therefore, the aim is to push back the level of detection and intervention to
the precursors and risk factors of disease. Fluctuation in patterns of
morbidity and mortality over time in countries, and the observation that
migrants slowly develop the patterns of disease are preventable. Hence,
epidemiology plays a central role in disease prevention by identifying those
modifiable causes.

Table 1. Levels of prevention in relation to the stage of disease.

Level of Stage of disease Target
prevention
Existence of underlying condition
leading to causation.
The aim is to avoid the emergence Total population
Primordial and establishment of the social. Or/ and selected
Economic and cultural patterns of groups
living that are known to contribute

8
DCH/AAU: Epidemiology Note 9

to an elevated risk of disease.

Example: smoking, environmental

pollution

Specific causal factors exist

The causative agent exists but the
Primary aim is to Total population,
prevent the development of disease. selected groups
and health,
Example: immunization individuals
Measles, polio

Early stage of disease

The aim is to cure patients and
Secondary prevent the Patients
development of advanced disease.

Example: Early detection & treatment

of cases of tuberculosis & STD
Late stage of disease (treatment &
rehabilitation)
Tertiary The aim is to prevent severe Patients
disability and
death.
Example: Leprosy

H. Levels of Disease Occurrence

Diseases occur in a community at different levels at a particular point in

time. Some diseases are usually present in a community at a
certain predictable level, this is called the expected level,
but at times disease may occur in excess of what is
expected.

1. Expected levels

9
DCH/AAU: Epidemiology Note 10

a) Endemic: a persistent level of low to moderate

occurrence
b) Hyperendemic: a persistently high level of
occurrence
c) Sporadic: occasional cases occurring at irregular
intervals

2. Excess of what is expected

a) Epidemic: occurrence of disease in excess of what
is expected in a limited period.
b) Outbreak: same as epidemic, often used by public
health officials because it is less provocative to the
public.
c) Pandemic: an epidemic spread over several
countries or continents, affecting a large number of
people.

I. Disease Classification

Disease is often classified according to: 1) its time course,

or 2) its cause. The time course classifies disease as acute
(characterized by a rapid onset and short duration) or
chronic (characterized by a prolonged duration). A chronic
disease may have both acute and chronic manifestations.
The cause of a disease may be classified as infectious
(caused by living organisms which are transmissible)or
non-infectious.

The outcome of exposure to an infectious agent are referred

as:

10
DCH/AAU: Epidemiology Note 11

Infectivity: the proportion of exposed persons who

become infected.

Pathogenicity: the proportion of infected persons who

develop clinical disease.

Virulence: the proportion of persons with clinical

disease who become severely ill or die.
Figure 6

OUTCOMES AT EACH STAGE OF INFECTION

EXPOSURE Infection disease Disease outcome

Infectiousness pathogenicity
virulence
(infection rate)(clinical-to- (case-
fatality rate,
subclinical hospitalization
rate)
ratio)
J. Variation in Severity of Illness

The infectious process is a wide spectrum of clinical

effects, which ranges from in apparent infection to sever
clinical illness or death. The effect depends on the nature of
the infectious agent and host susceptibility. Case fatality
rate (CFR) is the measure of severity of illness.

* CFR = Number of deaths from a disease

Number of clinical cases of that disease

11
DCH/AAU: Epidemiology Note 12

Recognizing in apparent infections require the use of

laboratory tests on seemingly healthy individuals.
Information thus obtained are useful in planning public
health interventions. A good example could be HIV testing
to determine the potentials for the spread of the disease and
to plan appropriate control strategies.

Figure 7
The Spectrum of Illness from Communicable Disease

INAPPARENT MILD SEVERE

DEATH
INFECTION DISEASE DISEASE

No signs or Clinical illness with signs and

symptoms

K. Spread of Disease through Person to Person

transmission

Person to person transmission of an infectious agent is one

of the main methods of disease spread in a community and
is dependent on:

1. Generation time: This refers to the period between

exposure/infection and the maximum

12
DCH/AAU: Epidemiology Note 13

communicability of the exposed host

regardless of whether the disease is apparent
or inapparent. In case of apparent infections
generation time may be equivalent to
incubation period.
2. Herd immunity: This refers to a community
resistance to spread of an infectious agent as
a result of immunity gained by high
proportion of individual members of the
community. Though it may not be important
to achieve 100% immunity, successful
breakage of the chain of infection can be
achieved if the immunity is close to 100%.
3. Secondary attack rate: This is an important measure of
spread of disease among contacts of an
index case. It has great use in epidemic
situations.

Secondary AR = New cases among contacts of index cases

during the period
Total number of contacts with the index
cases

 The index cases are excluded from both numerator and

denominator.
 Index case: The case that brings a household or any other
group (community) to the attention of the public health
personnel.

Attack Rate (AR)

13
DCH/AAU: Epidemiology Note 14

An attack rate is a variant of an incidence rate,

applied to a narrowly defined
population observed for a limited time, such as
during an epidemic. It is usually
expressed as a percent.

AR= New cases among the population during the

specified period
Population at risk at the beginning of the
period

III. TYPES OF EPIDEMIOLOGIC STUDIES

A. Descriptive Epidemiology

Descriptive epidemiology is a way of organizing data

related to health and health related events by person (Who),
place (Where) and time (When) in a population.
Information organized as such is easy to communicate and
provides information about:

1) the magnitude of the problem,

2) the populations at greatest risk of acquiring a
particular disease, and
3) the possible cause (s) of the disease.

Time – Information organized by time easily shows the

trend of the disease over time and establishes the usual

14
DCH/AAU: Epidemiology Note 15

occurrence of the disease in the population which is

essential in identifying excess occurrence (epidemics). It
can also be used to predict seasonal and secular (long-time)
trends.

Place – This provides information on geographic

distribution of the disease. Such information provides clue
in identifying factors influencing the occurrence of the
disease either in the host or environment.

Person – Describing disease occurrence by personal

characteristics is important to identify some modifiable
factors in order to prevent or control the disease. Person
data include: the inherent characteristics of people (age,
ethnic group, gender), their acquired characteristics
(educational marital, immune, or nutritional status), their
activities (occupation, leisure activities, use of alcohol,
tobacco, or medications), or the conditions in which they
live (socioeconomic status, access to health care).
Figure 1. Purpose of Epidemiological Studies

Descriptive Analytic
Characterize disease Concerned with the
occurrence by time, search for causes and
place and person. effects.

Generate testable Test hypothesis about

Hypothesis as to the association between
cause of disease. exposure and outcome.

B. Analytic Epidemiology

Analytic epidemiology uses Comparison groups to determine whether the

characteristics of those with a given health condition are alike or different from that
expected. When persons with a particular characteristic are more likely than those

15
DCH/AAU: Epidemiology Note 16

without the characteristic to develop a certain health problem, we say that the
characteristic is associated with that health problem. Thus analytic epidemiology is
concerned with the search for causes and effects, or the why and how. We use analytic
epidemiology to quantify the association between exposures and outcomes and to test
hypotheses about causal relationships.

The principles of analytic epidemiology are applicable to all types of disease, whether
acute or chronic, infectious or non-infectious (even to states of health, behaviours,
phenomena). They are most often associated, however, with studies of chronic disease.
This reflects the difference between chronic and acute disease in terms of: 1) duration of
latency (short for actue, long or variable (or obscure) for chronic disease, 2) magnitude of
incidence (uaually high for acute disease, low for chronic), and complexity of causation
(single causes common for acute disease, multifactorial etiologies for chronic). Because
of these characteristics, chronic diseases more often require relatively elaborate studies to
establish causes, while acute disease can often be linked to its cause through simple
descriptive or short-term cross-sectional studies.

Analytic epidemiology uses two categories of studies to understand causes and effects: 1)
experimental studies and 2) observational studies. In an experimental study, we
determine the exposure status for each individual (clinical trial) or community
(community trial); we then follow the individuals or communities to detect the effects of
the exposure. In an observational study, which is more common, we simply observe the
exposure and outcome status of each study participant.

Two types of observational studies are the cohort study and the case-control study. A
cohort study is similar in concept to the experimental study, except that we observe the
exposure status rather than determining it. Cohort studies categorize subjects on the basis
or their exposure and observe the frequency of disease occurrence. Case-control studies
enroll a group of people with disease (“cases”) and a group without disease (“controls”)
and compare their patterns of previous exposures to risk factors.

IV. MEASUREMENT IN EPIDEMIOLOGY

Measurement in Epidemiology

1. Measures of disease occurrence

Prevalence
Incidence

2. Measures of association
Rate ratio
Etiologic fraction

A. Measures of Disease Occurrence

16
DCH/AAU: Epidemiology Note 17

The number of cases in a given community can give more epidemiologic sense if they are
related to the size of the population. Such tie of the number of cases with the population
size can be determined by calculating ratios, Proportions, and rates. These measures
provide useful information about the probability of occurrence of health events,
population at a higher risk of acquiring the disease. They are also important in designing
an appropriate public health interventions.

Ratio: the value of x and y may be completely independent, or x may be included

in y.

Example: Male: Female (male to female ratio)

Proportion: is a ratio (expressed as a percent) in which x is included in y.

Example: Male/Both sexes (proportion of male in a community)

Rate: measures the occurrence of an event in a population over time. The time
component is important in the definition. Rates are often proportions. Rates must: 1)
include persons in the denominator who reflect the population from which the cases in
the numerator arose: 2) include counts in the numerator which are for the same time
period as those from the denominator; and 3) include only persons in the denominator
who are “at risk” for the event.

Example: Measles cases in under five in 1995

Under five children in 1995

When we call a measure a ratio we usually mean a nonproportional ratio. When we call a
measure a Proportion, we usually mean a proportional ratio that doesn’t measure an
event over time. When we use the term rate, we frequently refer to a proportional ratio
that does measure an event in a population over time. The following table depicts the
common uses of the three measures

Table 1. Common uses of measures of disease occurrence.

As measure of As measure of As measure of Impact

Disease occurrence comparision Of intervention
Ratio - Rate ratio (OR,RR) Rate Ratio
Proportion Prevalence - Etiologic fraction
Rate Incidence - -

Table 2. Measures Described by Type of Event

EVENT RATIO PROPORTIONS RATES

Morbidity Relative risk Attributable Proportion Incidence Rate

17
DCH/AAU: Epidemiology Note 18

(Disease) Odds Ratio Point Prevalence Attack Rate

Period Prevalence
Crude Mortality Rate
Death-to-case Ratio Cause-Specific Mortality
Mortality Maternal Mortality Rate Proportionate Mortality Age-Specific Mortality
(Death) Proportionate Mortality Ratio Case-Fatality Rate Sex-Specific Mortality
Postneonatal Mortality Rate Race-Specific Mortality
Age-Adjusted Mortality
Neonatal Mortality
Infant Mortality
Years of Potential Life Lost

Natality Crude Birth Rate

(Birth ) Low Birth Weight Crude Fertility Rate
Rate of Natural Increase

B. Common Measures of Disease Frequency

The frequency of health related events are measured by risk, prevalence and incidence
rate.

Risk (cumulative incidence): the likelihood that an individual will contract a

disease.

the proportion of unaffected individuals who, on

average, will contract the disease of interest over a
specified period of time.

New cases occurring during a given time period

Risk = Population at risk during the same period

Prevalence: the amount of disease that is present already in a population.

indicates the number of existing cases in a population.

All new and pre-existing cases during a given time period

Prevalence = population during the same time period

Incidence: measures the rapidity with which newly diagnosed patients develop over
time.

Most common way of measuring and comparing the frequency of disease

in populations.

the period of time for the rate must be specifies.

Number of new cases during observation period

Incidence Rate = Person – time observed

18
DCH/AAU: Epidemiology Note 19

Crude versus Adjusted Rates

Crude rates apply to the total population of a given area. Specific rates apply to specific
subgroups in the population (such as by age, sex, or occupation) or specific diseases.
Adjusted rates and age-specific rates are often used to permit comparison of mortality
rates in populations which differ in age structure. Mortality rates computed with
adjustment techniques are called age-adjusted or age-standardized mortality rates.

C. Measures of Association

A. Rate Ratio

Measures of association between risk factors and disease are often calculated from data
presented in a two by two tables:

Table 3. TWO-BY-TWO TABLE SHOWING ASSOCIATION

EXPOSURE DISEASE
YES (+) NO(-)
YES (+) A B
NO (-) C D

By convention, the capital letters (A,B,C,D) designate study populations (e.g., in cohort
studies) defined by risk exposure and disease occurrence. The small letters (a,b,c,d)
represent samples of populations (e.g., in case-control studies), usually of unknown and
different sampling frequencies.

The relative risk or risk ratio compares the risk of some health-related event (often
disease or death) in two groups, typically in persons exposed to the disease to those not
exposed:

_A_ _C_
A+B C+D

An odds ratio, or cross-product ratio, is another measure of association which quantifies

the relationship between an exposure and health outcome from a comparative study. The
formula can be derived (to be the same as that for the relative risk) by dividing the odds
that a case will have been exposed to the risk actor (a/c) by the odds that a control will
have been exposed (b/d):

_a_
Odds Ratio = _c_ = _ad_

19
DCH/AAU: Epidemiology Note 20

_b_ bc
d

When the health outcome is uncommon, the odds ratio provides a good approximation of
the relative risk or risk ratio. The odds ratio is also useful in analysis of data from case-
control studies, since the size of the control group is arbitrary and the true size of the
population from which the cases come is usually not known. Under these circumstances,
we cannot calculate rates or the relative risk. The relative risk can, nonetheless, be
approximated by calculating the odds ratio, using only the data in the cells of a two-by-
two table.

Since cases of disease in most chronic disease studies represent only a small fraction of
exposed and unexposed populations, B is about equal to A+B and D to C+D. The formula
can, under these circumstances, be simplified as follows.

_A_ _A_
A+B = _B_ = _AD_
_C_ _C_ BC
C+D D

A different procedure is used to calculate relative risk in case-control studies when

controls are selected by matching. If matching is used in selecting controls, the matching
(or pairing) should be retained for analytic purposes as shown in the following table of
case-control pairs:

Table 4. TWO-BY-TWO TABLE FOR MATCHED CASE-CONTROL STUDY

Control Exposed Case Exposed to Risk Factor

To Risk Factor YES NO
YES e f
NO g h

In calculating relative risk, one need only consider discordant case-control pairs,
represented by g (case exposed, control not exposed) and by f (control exposed, case not
exposed):

Relative Risk = _g_

_f_

g = ___case exposed___
control not exposed

f= __control exposed__
case not exposed

20
DCH/AAU: Epidemiology Note 21

B. Etiologic Fraction

The attributable risk is the difference between the disease rate in exposed persons (or in
the total population) and the rate in non-exposed:

_A_ - _C_
A+B C+D

The attributable proportion, also known as the attributable risk percent, is a measure of
the public health impact of a causative factor. In calculating this measure, we assume that
the occurrence of disease in a group not exposed to the factor under study represents the
baseling or expected risk for that disease. Thus, we attribute any risk above that level in
the exposed group to their exposure. It represents the expected reduction in disease if the
exposure could be eliminated. The calculation is shown in the summary table.

Summary of Measures of Association

Attributable risk (AR) or Risk difference (RD) indicate how much of the risk is due to (or attributable to)
the exposure. Quantify the excess risk in the exposed that can be attributable to the exposure by removing
the risk of disease that could have occurred anyway due to other causes.

AR = Risk in exposed – Risk in non-exposed

Relative risk (RR): estimates the magnitude of the association between exposure and disease and indicates
the likelihood of developing the disease in the exposed group relative to those who are not exposed.

RR = Risk in exposed
Risk in unexposed

Odds of disease: is a simple ratio, not a proportion. Indicates odds of diseased relative to the exposure
status.

Odds of disease in exposed = a/b or a:b

Odds of disease in unexposed = c/d or c:d

Odds Ratio (OR): is the odds in the exposed over the odds in the unexposed. Some people call it cross
product.

21
DCH/AAU: Epidemiology Note 22

OR = a/b  c/d = ad/bc

Attributable Risk Percent (AR%) among exposed: estimate the proportion of disease among the exposed
that is attributable to the exposure, or the proportion of the disease that could be prevented by eliminating
the exposure.

AR% = Risk in the exposed – Risk in unexposed

Risk in exposed

= OR – 1 X 100
OR

Population Attributable Risk (PAR) is the risk in total population minus risk in the non-exposed. Estimate
the excess rate of disease in the total study population that is attributable to the exposure.

PAR = Risk in population – Risk in unexposed

Population Attributable risk Percent (PAR%) Estimate the proportion of disease in the study population
that is attributable to the exposure and thus could be eliminated if the exposure were eliminated.

PAR% = Risk in population – Risk in unexposed X 100

Risk in population

Possible Outcomes in studying the relationship between disease and exposure

1. No association between exposure and disease

Attributable risk = 0
Relative risk =1

2. Positive association between the exposure and the disease (i.e., more exposure, more
disease)
Attributable risk > 0
Relative Risk >1

3. Negative association between the exposure and the disease (i.e., more exposure, less
disease)
Attributable risk < 0 (negative)
Relative risk < 1 (a fraction)

 Association is dependent on your definition of exposure.

Example: Exposure sex of CHA

22
DCH/AAU: Epidemiology Note 23

CHA Female Male

AR >0 <0
RR >1 <1

Positive association with female CHA and negative association with male CHA.

V. EPIDEMIOLOGICAL DESIGN STRATEGIES

Epidemiology is primarily concerned with the distribution and determinants of disease in
human populations. The basic design strategies in epidemiologic research are categorized
into two according to their focus of investigation. Descriptive studies focus on the
distribution of disease and analytic studies focus in elucidating the determinants of
disease.

Table 4. Types of Epidemiologic Design Strategies

DESCRIPTIVE ANALYTIC
Dealing with population Observational studies
 Correlational or ecological  Case-control
 Cohort
Dealing with individuals
 Case report or series Intervention studies
 Cross sectional survey

A. Descriptive studies

 Mainly concerned with the distribution of diseases with respect to time, place
and person.
 Useful for health managers to allocate resource and to plan effective
prevention programmes.
 Useful to generate epidemiological hypothesis, an important first step in the
search for disease determinant or risk factors.
 Can use information collected routinely which are readily available in many
places. So generally descriptive studies are less expensive and less time-
consuming than analytic studies.
 It is the most common type of epidemiological design strategy in medical
literature.
 There are three main types:
 Correlational
 Case report or case series
 Cross-sectional

A.1. Correlational or Ecological

23
DCH/AAU: Epidemiology Note 24

 Uses data from entire population to compare disease frequencies – between

different groups during the same period of time, or in the same population at
different points in time.
 Does not provide individual data, rather presents average exposure level in the
community.
 Cause could not be ascertained.
 Correlation coefficient ® is the measure of association in correlational studies.
It is important to note that positive association does not necessarily imply a
valid statistical association.

e.g. Hypertension rates and average per capita salt consumption compared between
two communities.

Average per capita fat consumption and breast cancer rates compared between
two communities.

Comparing incidence of dental cares in relation to fluoride content of the water

among towns in the rift valley.

Mortality from CHD in relation to per capita cigarette sales among the regions of
Ethiopia.

Strength: Can be done quickly and inexpensively, often using available data.

Limitation:
i. Inability to link exposure with disease. Data on exposure and outcome are
not linked at the individual level. For example, in a society with high fat
intake, perhaps it is the individual women with low intake that get breast
cancer. Again in the association with the reduced mortality from cervical
cancer and PAP smear screening, it is difficult to know whether the
reduction is really in those women who were screened by smear or
otherwise.

ii. Lack of ability to control for effects of potential confounding factors.

There may be other things that at the true cause. For example, often people
with high fat consumption also have high meat consumption. Perhaps it is
the meat that is actually responsible for the breast cancer – or may be
because of reduced vegetable intake or merely a reflection of
socioeconomic status. Another example is the correlation found between
the high per capita color TV and mortality from CHD, again here it is
obvious that color TV owning is not a good reason for increased mortality
from CHD.

24
DCH/AAU: Epidemiology Note 25

iii. It may mask a non-linear relationship between exposure and disease. For
example alcohol consumption and mortality from CHD have a non-linear
relationship (the curve is “J” shaped), but this type of relationship is
impossible to demonstrate in correlational studies.

A.2. Case Report and Case Series

Describes the experience of a single or a group of patients with similar diagnosis. Has
limited value, but occasionally revolutionary.

E.g. 5 young homosexual men with PCP seen between Oct. 1980 and May
1981 in Los Angeles arose concern among physicians. Later, with further
follow-up and thorough investigation of the strange occurrence of the
disease the diagnosis of AIDS was established for the first time.

One case of pulmonary embolism observed 5 weeks after oral

contraceptive usage was the first clue to the association, later established,
between oral contraception and increased risk of venous
thromboembolism.

Strength: very useful for hypothesis generation.

Limitations: Report is based on single or few patients, which could happen just by
coincidence. Lack of an appropriate comparison group.

A.3. Cross Sectional Studies (Survey)

Information about the status of an individual with respect to the presence or absence of
exposure and disease is assessed at the same point in time. Easy to do-many surveys are
like this.

For factors that remain unaltered overtime, such as sex, race or blood group, the cross-
sectional survey can provide evidence of a valid statistical association.

Useful for raising the question of the presence of an association rather than for testing a
hypothesis.

Limitation: “chicken or egg” dilemma – difficult to know which occurred first, the
determinant/ exposure or the outcome. Therefore, difficult to distinguish whether the
exposure preceded the development of the disease or whether presence of the disease
affected the individual’s level of exposure.

E.g. In the study of knowledge of modern contraceptive, and use of

contraception, you may show that women who know about modern

25
DCH/AAU: Epidemiology Note 26

contraception are more likely to use it. So you may want to educate
women about it, believing that this will lead to higher rate of use. The
problem is, did the women know about it and then start to use it, or did
they learn about it because they were using it?

Another example is community health agent activity and health station

supervision. Are the CHAs that they know are doing something. There are
factors which clearly come before the outcome of interest in time. For
example, if more women CHAs are active than men CHAs, one can be
sure that their sex came before their activity as a CHA in time, and thus it
is their sex that causes them to be active, not their activity which cause
them to be female.

B. ANALYTIC STUDIES

Focuses on the determinants of a disease by testing the hypothesis formulated from

descriptive studies, with the ultimate goal of judging whether a particular exposure
causes or prevents disease. Broadly classified into two – observational and interventional
studies. Both types use “controls”. The use of controls is the main distinguishing feature
of analytic studies.

B.1. Observational studies

Information are obtained by observation of events. No intervention is done. Cohort and
case-control are in this category.

i. Cohort
Subjects are selected by exposure, or determinants of interest, and followed to see
If they develop the disease or outcome interest.

E.g. Take Awarjas with trained manager and untrained manager and untrained
managers and follow them to see which group will do better to increase
coverage.

Follow 100 children who received BCG vaccination and another 100 who
didn’t get BCG vaccination and see how many of them get tuberculosis.

ii. Case Control

Subjects are selected with respect to presence or absence of disease, or outcome of

interest, and then inquiries are made about past exposure to the factor(s) of interest.

E.g. Take people with and without TB, ask them if they ever had BCG vaccination.

Take Awrajas with high and low EPI rates, ask them if their Awraja health

26
DCH/AAU: Epidemiology Note 27

managers were trained.

B.2. Interventional / Experimental

 The researcher does something about the disease or exposure and observe the
changes.
 Investigator has control over who gets exposure and who don’t. The key is
that the investigator assign into either group, whether it is done randomly or
not.
 Always prospective.

E.g. Assign children randomly to get chloroquine or not, and see how many develop
symptomatic malaria.

CASE-CONTROL STUDIES

An epidemiologic research method in which the two study groups are selected on their
disease status. This is a design strategy developed in response to the difficulty of studying
diseases with very long latency period. The design is capable of evaluating the
association of a disease to exposure many years after the actual exposure. Because of this
and its efficiency in time and cost case-control studies have became the most common
analytic design encountered in medical literature. The prototype study on lung cancer and
smoking was done in 1950’s.

Design and conduct of case-control studies

In the design of the study always seek for the comparability between cases and controls,
this is the basis for valid conclusion.

Defining Cases:

Establish a clear operational definition or use standard definition of disease (outcome) of

interest in order to have a clear understanding of exposure-disease association.

E.g. “uterine Ca”, before 1940 include Ca of the body of uterus and Ca of the cervix.
“congenital malformation” and drug use-specify malformation
John’s criteria to diagnose rheumatic heart disease is a good example.

If you are not certain about the diagnosis, and if the information collected is
adequate perform analysis separately for cases classified as definite, probable or
possible.

Selection of Cases:

27
DCH/AAU: Epidemiology Note 28

Do not always go for random representation of cases, rather it is better to restrict your
self to cases on which you can get complete and reliable information. Select controls
which are comparable to the cases entered into the study, do not try to represent the
population of all non-diseased persons.

Hospital-based Vs population-based cases

Hospital-based: easy and inexpensive to conduct but it is prone for selection bias.

Population-based: avoids selection bias, allows the description of a disease in the

entire population and the direct computation of rates of disease in
exposed and non-exposed persons.

Incident Vs Prevalent cases

Prevalent cases:
- Increase sample size available for rare disease.
- Difficult to establish temporal sequence between exposure and outcome. E.g.
Coffee consumption and peptic ulcer disease.

- Use is unavoidable in certain situations, like in studying congenital

malformations which are rare to find.

Incident cases:

- Helpful to establish temporal relationship between exposure and outcome.

So, it is better to limit cases to those newly diagnosed within a specified
period of time if the aim is also to establish temporal relationship.

- Records are easily obtainable and recall is not a serious problem.

Selection of controls:

There is no control group that is optimal for all situations. Controls are made for a
particular group of cases, do not try to represent the entire non-diseased population rather
try to achieve comparability between the cases and controls. Selection of controls should
consider besides comparability, practicability and economic impact.

The control series is intended to provide an estimate of the exposure rate that would be
expected to occur in the cases if there were no association between the study disease and
exposure.

Sources of controls.

28
DCH/AAU: Epidemiology Note 29

1. Hospital Controls

Advantages:
- Easily identified and readily available in sufficient number with reduced cost
than population controls.
- More likely than healthy individuals to be aware of antecedent exposures or
events minimize recall bias.
- Controls are also likely to have been subject to the same intangible selection
factors that influence cases to come to this particular physician or hospital
minimize selection bias.
- More likely to be cooperative because they anticipate benefit from their
involvement or might think that its related with their illness reduce bias
due to non-response.

Disadvantages:
- Because they are ill they are different from health individuals in many ways.
Several studies in the West have demonstrated that hospitalized patients are
more likely to smoke cigarette, use oral contraceptive, and be heavy drinkers
of alcohol than non-hospitalized individuals.
- There is danger of altering the direction of association or masking a grue
association between exposure and outcome of interest. Patients with diseases
known to be associated either positively or negatively, with the exposure of
interest, should be excluded from the control series. For example, in studying
the association of cigarette smoking and lung Cancer, individuals with other
respiratory illnesses could not be taken as controls, since smoking is also
known to have some association with other respiratory illnesses.

2. General population controls

Advantages:
- Generalizability is possible
- Good when cases are selected to represent affected individuals in a defined
population. For example, if cases to that particular hospital are coming from
a geographically defined area selection of controls from the entire population
could be possible.

Disadvantages:

- Costly and time-consuming

- Recall bias – controls may not recall exposures with the same level of
accuracy since they may not be seriously concerned about their illness.
- People might be less motivated to participate for the same reason given
above increase non-response rate selection bias.

29
DCH/AAU: Epidemiology Note 30

3. Special controls

Special controls are individuals, which are related to the cases in some way. These are
friends, household members (siblings,…), neighbours,…

Advantages:
- They are healthy
- More likely to be cooperative than members of the general population,
because of their interest in the cases.
- Offer a degree of control over some confounding factors, such as ethnicity,
socioeconomic status, or environment.

Disadvantage:
- If the study factor is likely to be similar to the cases, an underestimate of the
true effect of the exposure of interest may result. E.g. if the study factor is
diet, it will be similar for both cases and controls, if controls are siblings.

Number of control groups and case-control ratio

A single control group is optimal in most of the times. Add more control groups only
when you are not confident with the control group or when you see a clear deficiency in
your control group or when there is a clear advantage by adding another control.

Conditions for multiple controls:

- When the control is not considered appropriate.

- When the selected group has a specific deficiency that could be overcome by
inclusion of another control group.
Control-case ratio

The optimal control-case ratio is 4:1. As the number of controls per case increases, the
power of the study also increases. But, beyond 4:1, there is only a small increase in
statistical power, which can not justify the expenditure of additional resources.

Ascertainment of disease and exposure status

Potential source of information must be carefully considered in terms of its ability to

provide accurate as well as comparable information for all study groups. Procedures used
to obtain information must be similar for cases and controls:

- Place and circumstances of interview must be the same.

- Blind interviewers or record reviewers, if possible.
- Data collectors should be unaware of the specific hypotheses being tested
to reduce observation bias.

30
DCH/AAU: Epidemiology Note 31

- The ability to obtain exposure information from records completed before the
occurrence of outcome events is especially valuable. E.g. record of X-ray
during pregnancy in studying its effect on the child (congenital
malformation).
- Ascertainment of exposure should involve defining the part of a person’s
exposure history that could be relevant to the etiology of the disease under
study.

E.g. Smoking & lung Ca – duration of smoking is important than the amount
currently smoked.
Smoking & Myocardial infarction – current smoking is most important.

Collect information in such a way that it allows you to identify the most
appropriate time window for the evaluation of the possible harmful effects
of an exposure – try to avoid collecting information over too wide a
period, such as “ever use” in order to avoid the inclusion of some period in
time that cannot be causally related to the disease.

Issues in analysis

Comparison is made primarily by estimating the relative risk as computed by the

odds ratio. If Case Control study is population based, or if estimates of disease
incidence are available from an outside source, rates of disease for the exposed
and non-exposed can be computed and compared directly.

COHORT STUDIES

A design in which the two groups are defined according to their exposure status to a
suspected risk factor for a disease. the two groups should be free of the study outcome.

Types: There are two types of cohort studies, prospective and retrospective, depending on
the temporal relationship between the initiation of the study and the occurrence of
the disease.

1. Prospective- At the beginning of the study the outcome has not yet occurred.
Regarded as more reliable than the retrospective, if the sample size is
large and follow-up complete.

 The outcome has not occurred at the beginning of the study

2. Retrospective- Both exposure and outcome status have occurred at the beginning of
the study. Efficient in cost and time. Often uses of data collected for
other purposes, so information obtained might be incomplete and non-
comparable for all subjects.

 Both exposure and outcome have occurred before the beginning of the study

31
DCH/AAU: Epidemiology Note 32

Exposed
Retrospective Prospective
Outcome Unexposed Outcome

Selection of Exposed Group

Selection of exposed group should consider scientific and feasibility issues which
include:

- The frequency of the exposure of interest in the study population.

- The need to obtain complete and accurate exposure and outcome information
on all study subjects. Example use of physicians or nurses.
- The ability of obtaining sufficient exposed individuals in a reasonable period
of time-identify high-risk population (special group) to the exposure of
interest.

Selection of high risk group also allows the evaluation of a rare disease.
Although cohort studies are in general not optimal for the evaluation of rare
diseases, if the outcome of interest is relatively common among those
exposed; i.e., if the attributable risk percent is high the design can be used
efficiently.

- The ease to collect relevant information and to follow-up.

Selection of controls

Always attempt to select a control group which is comparable to the characteristics of the
exposed population. There is no single optimal control group that can be used for any
circumstance.

Source of data

The major consideration should be the availability of accurate and complete information
on exposure and outcome of interest in the study groups in a way that is comparable to
both.

Exposure ascertainment:

1. Using Pre-existing records: from hospital, employers record.

Advantages:
- Can make available information for high proportion of cohort.
- Relatively inexpensive to obtain.
- Allow objective and unbiased classification of exposure status.

32
DCH/AAU: Epidemiology Note 33

Disadvantages:
- Information on exposure level may be insufficient.
- May not contain adequate information on potential confounders.

2. By conducting Interview and filling questionnaire

Advantages:
- enables to record exposure information that are not routinely recorded,
particularly lifestyle factors.

Disadvantages:
- Potential for information bias, particularly recall. In such situations, where
objective sources can not be used, it is important that information is obtained
in a comparable manner for all participants.

Outcome ascertainment:

With adequate consideration to the resources available for the study, the aim is to obtain
complete, comparable and unbiased information on the subsequent health experience of
every study subject. One or a combination of the following sources could be used: routine
surveillance, death certificate, periodic health examination, autopsy records, hospital
records surveillance, death certificate, periodic health examination, autopsy records,
hospital records, etc.

Always try to have a firm outcome criteria and standard diagnostic procedure which are
equally applied for exposed and non-exposed individuals. Do not do any diagnostic
examination only for one group, because the difference which might be observed could
be just due to the greater opportunity offered to be diagnosed.

Follow-up

This the major challenge in cohort studies, as well as the major cost in terms of time.
Unless complete or nearly complete information could be obtained the results might be
un-interpretable. If the loss to follow-up is not comparable between the two exposed
groups, this will also be a source for bias. Therefore, if there is a need for long follow-up
period, the mechanism to achieve complete follow-up should be thought carefully in the
planning of the study.

Analysis

The basic analysis in cohort studies are:

- Calculation and comparison of rates of the incidence of the outcome for
exposed and non-exposed.
- Comparison of the two groups with baseline characteristic to ensure
similarity.

33
DCH/AAU: Epidemiology Note 34

Issues in interpretation of cohort studies

Role of bias:

Misclassification bias to some extent might be unavoidable. So always attempt must be

done to avoid the introduction of any systematic misclassification.

Random misclassification or error unrelated to the outcomes of interest may not affect
comparability, rather it dilute or underestimate any true association that may exist
between the exposure and outcome. As a result, the observed RR estimate will always be
biased towards the null value of 1. On the other hand differential misclassification can
result in a biased risk estimate that is either an underestimate, an over estimate, or, by
chance, the same as the true measure of association.

Effects of losses to follow-up:

If the probability of loss is related to exposure, outcome or to both. Or if the proportion is

large the estimate of exposure-disease association may be biased.

Because of the difficulty to know which factors are related to loss, the best way to
eliminate bias is by reducing loss to follow-up to an absolute minimum.

For losses:
- try to get at least mortality status from other sources.
- Examine previously collected data to determine whether there are systematic
differences between the losses and follow-ups.
- Indirectly calculate exposure-disease association, assuming the two extreme
outcomes. One assuming all those who were lost to follow-up developed the
outcome of interest and the other assuming that none developed the outcome
–this provides a range within which the true association will lie. If losses to
follow-up are large, the observed range will be so wide as to provide little
useful information.

Effect of non-participation

This does not affect validity unless non-response is related to both the exposure and other
risk factors for the outcome under study. The effect of the difference is mainly on
generalizability of the study results.

The possible effect of non-response on either generalizability or validity can be assessed

by comparing basic social and demographic characteristics of those who do and do not
participate in a study.

34
DCH/AAU: Epidemiology Note 35

Table 4.2 Advantages and Limitations of Cohort and Case-Control Study Designs.

Case-Control Cohort
Advantages
 Optimal for the evaluation of  Valuable when the exposure is rare
RARE disease
 Can examine multiple effects of a
 Can examine multiple etiologic single exposure
factors for a single disease
 Can elucidate temporal relationship
 Quick and inexpensive
 Allows direct measurement of risk
 Relatively simple to carry out
 Minimize bias in ascertainment of
 Guarantee the number of persons exposure
with cases.
Limitations
 Inefficient for the evaluation of rare  Inefficient in evaluation of rare diseases
exposure
 Expensive
 Can not directly compute risk
 Time consuming

35
DCH/AAU: Epidemiology Note 36

 Difficult to establish temporal

relationship  Loss to follow-up create problem

 Determining exposure will often

relay on memory

 Persons who die as a result of

disease caused by the determinant
may not be known to the study

Measures of Association: case-control studies

Odds of disease: is a simple ratio, not a proportion.

Odds of disease in exposed = a/b

Odds of disease in unexposed = c/d

Relative Odds or Odds Ratio (OR)

OR = ad
bc

Attributable Risk Percent (AR%) among exposed:

AR% = Risk in the exposed – Risk in unexposed

Risk in exposed

= OR-1 X 100
OR

Population Attributable Risk Percent (PAR%):

36
DCH/AAU: Epidemiology Note 37

PAR% = Risk in population – Risk in unexposed

Risk in population

*Attributable risk (AR) or Risk difference (RD)

AR = Risk in exposed – Risk in unexposed

= a/a + c – c/c+d

* Population attributable risk (PAR)

PAR = R total population – Risk in unexposed

= AR X proportion of exposed individuals in population

* if the study is population based or if incidence rates can be estimated.

Measures of association: cohort studies

Attributable risk (AR) or risk difference (RD)

AR = Risk in exposed – Risk in non-exposed

= a/a + b – c/c +d

Relative risk (RR) or Risk Ratio

RR = Risk in Exposed
Risk in unexposed

Attributable Risk Percent (AR%) among exposed:

AR% = Risk in the exposed – Risk in unexposed

Risk in exposed

= AR – 1 X 100
Ie

Population attributable risk (PAR)

37
DCH/AAU: Epidemiology Note 38

PAR = Risk total population – Risk in unexposed

= AR X proportion of exposed individuals in population.

Population Attributable Risk Percent (PAR%):

PAR% = Risk in population – Risk in unexposed

Risk in population

= ______________PAR_____________ X 100
Incidence rate of disease in population

Interpretation of Measures of Association

Rate Ratio: measures the strength of association between an exposure and disease and
provide information that can be used to judge whether a valid observed
association is likely to be causal.

Attributable Risk: measures the public health impact of an exposure, assuming that the
association is one of cause and effect.

Relative and attributable risks of mortality from lung cancer and coronary heart disease
among cigarette smokers in a cohort of British male physicians

Annual mortality rate per 100,00

Lung cancer CHD

Cigarette smokers 140 669

Nonsmokers 10 413

38
DCH/AAU: Epidemiology Note 39

Relative risk 14.0 1.6

Attributable risk 130/10s/year 256/s/year

The above study demonstrated a 14-fold increased death rate from lung cancer among
smokers compared with nonsmokers. The relative risk of CHD mortality among current
smokers compared with nonsmokers was 1.6. Thus, cigarette smoking is a much stronger
risk factor for mortality from lung cancer than coronary heart disease. However, if
smoking is causally related to both diseases, the elimination of cigarettes would prevent
far more deaths among smokers from coronary heart disease than from lung cancer, as
shown by the attributable risks of 256/100,000 and 130/100,000, respectively. The
explanation for this is that while death from lung cancer is a relatively rare occurrence,
accounting for only 10 deaths/100,000 population each year among nonsmokers, the
annual death rate of coronary heart disease in that same group is 413/100,000.
Consequently, even a 60% increased risk of CHD mortality associated with cigarette
smoking will affect a much larger number of people than a 14-fold increased risk of death
from lung cancer. Thus, the potential public health impact of smoking cessation on
mortality will be far greater for coronary heart disease than for lung cancer.

INTERVENTION STUDIES

This is an epidemiologic design that closely resembles the controlled experiment used in
basic science researches, and can produce high quality data if done properly. The main
distinction from other types of analytic studies is that individuals are allocated into
experiment or control group by the investigators.

Classification

1. Based on population

A. Clinical trial - usually performed in clinical setting and the subjects are patients.
B. Field trial - used in testing medicine for preventive purpose and the subjects
are healthy people. E.g. vaccine trial
C. Community trial - unit of the study is group of people/community. E.g. fluoridation
of water to prevent dental caries.

2. Based on design

A. Uncontrolled trial – no control group. Control will be past experience (history).

B. Non-randomized controlled - there is control group but allocation into either group is
not randomized.

39
DCH/AAU: Epidemiology Note 40

C. Randomized controlled - there is control group and allocation into either group is
randomized.

3. Based on objective

A. Phase I - trail on small subjects to test a new drug with small dosage to determine
the toxic effect.
B. Phase II - trial on small group to determine the therapeutic effect.
C. Phase III - study on large population – usually a randomized control trial.

Problems Related to Intervention Studies

1. Ethical considerations prevent evaluation of many treatments or procedures using an

intervention design strategy.

Some of the ethical issues are:

 Practices or substances already known to be harmful should not be used in this study.
 Therapies known to be beneficial should not be withheld from any affected
individuals in the study population.
 Investigators have to have a complete knowledge of the subject under study..
 The researcher must have at least informed consent from each study participant and
subjects should be left free to withdraw from the study at anytime.
 A written research protocol is a must.

2. Feasibility/ practical issues

 Subject recruitment, getting adequate individuals to enrol into a study is not easy.
 Conducting trial on a widespread practice poses difficulty in getting sufficiently large
population who are willing to undergo through a new treatment or practice believed
to be more beneficial than the old treatment or practice for the duration of the entire
study period; i.e., its difficult to achieve satisfactory compliance from all study
subjects for a long time, particularly if study period is quite long. Getting an
appropriate control group is also difficult sometimes. For example, if you want to see
the association of chat chewing and dental caries and the prevalence of chat chewing
is 70%, it will be difficult for you get an adequate control.

3. Cost-experimental studies are very expensive.

Issues in the design and conduct of clinical trials

Intervention studies to represent the “gold standard” for epidemiologic research should
consider the following:

1. Selection of a study population

Reference population: The general group to whom investigators expect the results of the

40
DCH/AAU: Epidemiology Note 41

particular trial to be applicable. Represents the scope of the public

health impact of the intervention. And, it is related to the issue of
generalizability.

Experimental population: The actual group in which the trial is conducted. It is preferable
that if this group is not difference population for the sake of
generalizability, but this should not be a concern.

Considerations in choosing the experimental group:

 Check weather the proposed experimental population is sufficiently
large to achieve the required sample size for the trial.
 Choose population that will experience a sufficient number of
endpoints to permit meaningful comparison.
 Likelihood of obtaining complete and accurate follow-up information
for the period of trial.

2. Allocation of study groups

Allocation into either group must be done after determining eligibility and getting
consent. It is always advantageous to do the allocation at random.

Randomization: can be done using random-number table or using small computers,

which are capable of generating random numbers. If the sampling frame is small a lottery
method can be applied.
Advantages:
. Treatment groups will not be known by the researcher.
. “On average” the study group will be comparable; i.e., known and unknown potential
confounders will be equally distributed between the two groups.
. Randomization can provide a degree of assurance about the comparability of the study
groups that is simply not possible in any observational design.
. the impression it poses on the readers (consumers) – less proof if needed to show that
the observed result is due to a selection bias or confounder effects.

Maintenance and assessment of compliance

Subjects may decline from the treatment protocol for various reasons after randomization,
and this related to the length of time that subjects are expected to adhere to the
intervention, as well as to the complexity of the study protocol. It is always important to
obtain as complete follow-up information as possible since they will be included in the
primary analysis.

Methods to enhance compliance:

. select population who are both interested and reliable.
. arrange frequent contacts with individuals
. use incentives, such as providing medical information

41
DCH/AAU: Epidemiology Note 42

Assessment of compliance:

Noncompliance will decrease the statistical power of a trial to detect any true effect of the
study intervention. Therefore, to see its effect compliance levels in any study must be
measured. Measuring compliance is not easy, and all the measures available have
inherent limitations. Some of the measures are.
. Self-report, the simplest and the only way to assess behavioural modification and
exercise programs.
. Pills count – ask participants to bring unused pills to each clinic visit, this may
eliminate inaccuracies due to poor memory, it assumes that all the unreturned
pills has been ingested.
. Biochemical tests
. used to validate self-report
. objective but expensive and logistically difficult – Riboflavin is a safe
biochemical marker that has been used added in the treatment. But, can
only reflect the ingestion of the pills the preceding day or two and thus
can not be used as reliable measure for long-term compliance.
It is inevitable that some portion of participants in a trial will become noncompliant
despite all reasonable efforts. In any case, it is important to obtain as complete follow-up
information as possible since they will be included in the primary analysis.

Ascertainment of outcome

Use uniform ascertainment of outcome for complete follow-up period for all study
subjects. To eliminate a possible bias, maintain a high level of follow-up and reduce the
proportion of outcomes that are not ascertained to the minimum and comparable between
the two groups. Follow-up is short in assessing the effect of acute disease and long in
assessment of chronic disease outcomes. The difficulty in maintaining complete
ascertainment of outcome increases with increasing length of follow-up.

Potential for observation bias in ascertainment of outcome can exist in an intervention

study in that knowledge of a participant’s treatment status might, consciously or not,
influence the identification or reporting of relevant events. This can be overcome by the
use of placebo and blinding.

Placebo- an inert agent indistinguishable from the active treatment. Use of placebo
minimize bias in the ascertainment of both subjective disease outcomes and
side effects.

Placebo effect: tendency for individuals to report favourable response to any therapy
regardless of the physiologic efficacy of what they received.

42
DCH/AAU: Epidemiology Note 43

The use of placebo ensures that all aspects of the intervention offered to participants are
identical except for the actual experimental treatment. With no placebo, it is impossible
to tell wether subjective outcomes are due to the actual trial treatments, to the extra
attention participants receive, or merely to their belief that the treatment will help.

The primary strength of a double-blind design is to eliminate the potential for observation
bias. Of course, a concomitant limitation is that such trials are usually more complex and
difficult to conduct. Circumstances in which double-blinding is not possible are
evaluation of programs involving substantial changes in life-style, such as exercise,
cigarette smoking or diet, surgical procedures, or drugs with characteristics side effects.

Problems associated with unblinded trails:

- subjects who are not on the new or experimental program may become
dissatisfied and dropout of the trial, thus resulting in differential compliance
or loss to follow-up.
- Knowledge of the intervention to which group the participant has been
assigned might raise the potential for observation bias in the reporting of side
effects or assessment of outcome.

The quality of old standard” in intervention studies can be achieved through:

Randomization
Use of placebo
Double Blinding

Stopping Rules: Decision for early termination of a trial

To assure the welfare of the participants is protected, interim results should be monitored
by a group that is independent of the investigators conducting the trial. Consider
termination if the interim results indicate a clear and extreme benefit on the primary end
point due to intervention, or if one treatment is clearly harmful. It would also be unethical
to stop a trial prematurely based solely on emerging trends from a small number of
patients – the aim must be to achieve an equitable balance between, on the one hand,
protection of randomized participants against real harm and, on the other, minimizing the
risk of mistakenly modifying or stopping the trial prematurely.

Requirements for modification or termination of an ongoing trial:

43
DCH/AAU: Epidemiology Note 44

1st – Observation of a sustained statistical association that is so extreme, and, therefore,

so highly significant, that it is virtually impossible to arise by chance alone.

2nd – Consider the observed association in the context of totality of evidence:

. Is there known or postulated biologic mechanisms that might explain the
observed effect?
. Is it in-line with other randomized trials, or those from observational studies?
. How does the observed association (effect) affect the risk-to-benefit ratio of the
intervention?

POWER OF THE STUDY

The statistical power of a trial to detect a postulated difference between treatment groups,
if one truly exists, is dependent on:

1. Sample Size

Trials with inadequate sample size might have a great potential for scientific harm –
could be as a result of misinterpretation. Always its advisable to take sample large
enough to detect small to moderate (10-20%) benefit or differences that resulted from the
intervention.

2. Accumulation of adequate end points

There are at least two major strategies to obtain adequate numbers of end points:
a. Selection of a high-risk population

The collection of baseline data can be planned to allow the identification of

particular subgroups who might experience the effects of an intervention more
than others; i.e., those at a higher risk of developing the outcome of
intervention.

b. Length of follow-up period

It is always better to consider that the actual rate of occurrence of end points
will be less than the projected level, which could be due to the low incidence
of the outcome of interest in the volunteer study population, this is referred as
“healthy volunteer effect” – the only way to compensate for this deficit is to
extend the length of follow-up to get more events.

Secular changes in disease rates during the course of the trial might be
sometimes as great as that due to the intervention. E.G. During the decade in
which MRFIT trail was conducted, the entire U.S population including all
MRFIT participants, experienced a marked 25 to 30% decline in Coronary
Heart Disease (CHD) mortality. As a result, the expected numbers of deaths in

44
DCH/AAU: Epidemiology Note 45

the trial was less by two-third, so the follow-up was extended to increase the
number of end-points (the outcome).

Consider the postulated mechanism by which the study agent (the exposure)
exerts its effect in deciding the length of follow-up period. That is, how long
will it take for the study agent to exert its effect on the end result.

Every effort should be made to incorporate an adequate length of follow-up

during the planning phase of the trial. IF, for any reason, there is a need to
alter the follow-up period, the decision should be made as early in the trail as
possible to maintain the scientific credibility of the study and avoid the
implication that the change in study design was based on last-minute efforts to
achieve statistical significance.

3. Effect of Compliance

Compliance must be assessed in all study participants, regardless of their particular

treatment assignment. The effect of noncompliance in any participant is to make the
intervention and comparison groups more alike, as a result decreases the ability of the
trial to detect any true difference between the groups.

One strategy to increase compliance is to use “Run-in or Wash out” period prior to the
actual randomization- all participants receive either the active treatment or the placebo
for a number of weeks or months before formal randomization to a treatment group. The
only limitation to this strategy is the limitation to generalize study into reference or
general population, but the primary goal of atrial must be to attain a valid result.

Issues in Analysis and Interpretation of Intervention studies

Basically the same with analysis of cohort studies. The fundamental comparison to
estimate the true benefit of the intervention program should be obtained through
analysing the data by intention to treat “once randomized, always analyzed” – so always
maintain high level of compliance, keep losses to follow-up at a minimum, and collect
information of all randomized subjects.

Reasons:

1. Noncompliance may be related to factors that also affect the risk of the outcome
under the study, and failure to analyze data on all randomized participants could
introduce bias. In most studies, perfect compliers represent only a fraction of the total
study population.

2. Analysis of compliers data does not address the actual research question posed in an
intervention study. First, it is only the entire groups allocated by randomization that
are truly comparable – so preserve the power of randomization by analysing the entire

45
DCH/AAU: Epidemiology Note 46

population. Secondly, if a particular regimen is so difficult and uncomfortable that it

is likely to be accepted and used by only a small proportion of the reference
population, it may not be practical to recommend its use, no matter how effective the
actual treatment may be.

3. Rule out other possible alternative explanations for the observed findings. Alternative
explanations for the observed result in any analytic epidemiological study include:

3.1. Chance
. obtaining adequate sample size for the study could reduce the likelihood of
chance as a possible explanation.
. Statistically significant finding leave little room for chance.

3.2. Bias
. Selection bias is best eliminated by randomization
. information bias can be eliminated by:
. using blinding procedures
. using standard and comparable exposure and outcome ascertainment in both
groups.

3.3. Confounding
. ways to control for confounding include:
. use appropriate analytic tools to control known confounding factors –
multivariate analysis
. control for known and unknown cofounders can be best achieved by

randomization
. matching if properly applied, is another method used for control of known
cofounders
. compare basic socio-demographic characteristics to assure that balance was
achieved.

VI. EVALUATION OF EVIDENCE

Figure 1. Judging Observed Association

OBSERVED ASSOCIATION

Could it be due to
Selection or
Measurement bias?

46
DCH/AAU: Epidemiology Note 47

NO

Could it be due
To confounding?

NO

Could it be a
Result of chance?

PROBABLY NOT

Could it be causal?

Apply guidelines
And make judgement
Accuracy of Measurement

Accuracy = Validity + Precision

Validity is the extent to which data collected actually reflect the truth. The concepts of
sensitivity (ability to detect true positive) and specificity (ability to detect true negatives)
can be used to characterize the validity of a measure (“measurement validity”). Study
results are also described as “valid” when there is no systematic misrepresentation of
effect or “bias” (“validity in the estimation of effect”). Validity is often described as
internal or external.

Internal validity concerns the validity of inferences that do not proceed beyond
the target population for the study. Internal validity is threatened when the
investigator does not have sufficient data to control or rule out competing
explanations for the results.

47
DCH/AAU: Epidemiology Note 48

External validity, on the other hand, concerns generalizeability, or inferences to

populations beyond the study’s restricted interest. External validity is threatened,
for example, when the investigator attempts to apply the findings of the study to a
population, which is not comparable to the population in which the research was
completed. Internal validity should be the primary objective in study design,
however, since efforts to ensure the generalizeability of results often introduce
problems of bias or confounding.

Precision, on the other hand, describes the extent to which random error (i.e., sampling
variation and the statistical characteristics of the estimator) alters the measurement of
effects. Misclassification may result in problems with either validity (due to systematic
misclassification bias attributable to methodolgical aspects of study design or analysis) or
precision (due to random misclassification error attributable to sampling variation).
Random misclassification errors always bias measures of relative risk toward one.
Systematic misclassification bias can either increase or decrease the strength of the
measured association.

A. Bias

Bias may be defined as any systematic erro in

an epidemiologic study that results in an
incorrect estimate of the association between
exposure and risk of disease.

Bias may result from systematic error (or difference between exposed and
unexposed populations or between cases and controls) in the collection, recording,
analysis, or interpretation of data. Evaluating the role of bias as an alternative
explanation for an observed association is a necessary step in interpreting any
study result. Unlike chance (including lack of precision) and confounding, which
can be evaluated quantitatively, the effects of bias are far more difficult to
evaluate and may even be impossible to take into account in the analysis. For this
reason, it is important to design and conduct studies in such a way that every
possibility for introducing bias has been taken into account and to take steps to

48
DCH/AAU: Epidemiology Note 49

minimize chances of bias. In evaluation of study results, it is important to estimate

the magnitude and direction of any suspected bias.

Types of bias may be grouped into two categories:

1) Selection bias refers to any error that arises in the process of identifying the study
populations. Selection bias can occur whenever the identification of individual
subjects for inclusion in the study on the basis of either exposure (cohort) or disease
(case-control) status depends in some way on the other axis of interest.

Examples of selection bias include:

1) Berkson’s bias- Case-control studies carried out exclusively in hospital settings are
subject to selection bias attributable to the fact that risks of
hospitalization can combine in patients who have more than one
condition.

2) Ascertainment bias – Differential surveillance or diagnosis of individuals make those

exposed or those diseased systematically more or less likely to
be enrolled in a study.

3) Non-response bias- Rates of response to surveys and questionnaires in many studies

may also be related to exposure status, so that bias is a
reasonable alternative explanation for an observed association
between exposure and disease.

4) Loss to follow-up- This is a major source of bias in cohort studies. Persons lost to
follow-up may differ from with respect to both exposure and
outcome, biasing any observed association.

5) Volunteer / Compliance bias– In studies comparing disease outcome in persons who

volunteer or comply with medical treatment to those who
do not, better results might be expected among those
persons who volunteer or comply than among those who
do not.

6) Cohort bias – Refers to the biased view of the natural history of disease presented in
survival cohorts, since only the prevalent cases (those with less lethal
disease) are available for study in the latter part of the period of
observation.

2) Observation or information bias includes any systematic error in the measurement of

information on exposure or outcome.

Examples of Information bias include:

49
DCH/AAU: Epidemiology Note 50

1) Interviewer bias This can occur if the interviewer or examiner is aware of the
disease status (in a case-control study) or the exposure status (in
cohort and experimental studies). This kind of bias may affect
every kind of epidemiologic study.

2) Recall bias May result because affected persons may be more (or less) likely
to recall an exposure that healthy subjects, or exposed persons
more (or less) likely to report disease. This source of bias is more
problematic in retrospective cohort or case-control studies.

3) Social desirability bias- Occurs because subjects are systematically more likely to
provide a socially acceptable response.

4) Hawthorn effect Refers to the changes in the dependent variable which may be due
to the process of measurement or observation itself.

5) Placebo effect - In experimental studies which are not placebo-controlled observed

changes may be ascribed to the positive effect of the subject’s
belief that the intervention will be beneficial.

6) Regression to the mean- Refers to the statistical phenomenon that extreme values will
tend to “regress” to more average values. Thus a change
from a very high or very low values in the dependent
variable may be attributable to simple random variation,
rather than to changes in the independent variable.

7) Healthy worker bias - Refers to the bias in occupational health studies which
tends to underestimate the risk associated with an
occupation due to the fact that employed people tend to be
healthier than the general population.

8) Lead-time bias - Results in overestimation of the effectiveness of a screening

program for a condition which is actually caused by the early
detection of a condition. It is more exaggerated in conditions with
a long “lead-time” (such as cervical carcinoma). The detection of a
condition before the person shows clinical signs and symptoms
(the “lead time”) is the cause of the measurement of prolonged
survival in persons who participate in screening programs rather
than a real prolongation of a real survival. Those individual who
are diagnosed early may actually gained more “disease time”.

9) Length/time bias- Occurs in studies of screening tests for cancer. This occurs due to
the fact that screening tests for cancer tend to detect more slow-
growing tumors with a better prognosis (since faster growing
tumors are more often detected because they cause symptoms). As
a result, the mortality rate of cancers found on screening will

50
DCH/AAU: Epidemiology Note 51

appear better than that of tumors not found on screening (though

the effect is not due to the screening itself)

Some recommendations to minimize bias at the time of study design are:

1) Choose study design carefully. If ethical and feasible, a randomized double blind trial
has the least potential for bias. If loss to follow-up will not be substantial, a
prospective cohort study may have less bias than a case-control study. Controls for
case-control studies should be maximally comparable to cases except for the variable
under study.
2) Choose “hard” (i.e., objective) rather than subjective outcomes.
3) “blind” interviewers or examiners wherever possible.
4) Use well-defined criteria for identifying a “case” and use closed-ended questions
whenever possible.
5) Collect data on variables you do not expect to differ between the two groups. If such
a “dummy” variable regarding exposure, for example, in a case-control study shows
an unexpected difference, it may alert you to recall bias.

C. Confounding

Confounding refers to the mixing of the effect

of an extraneous variable with the effects of the
exposure and disease of interest.
Confounding arises when some cause other than the exposure under study is more or less,
prevalent in the exposed group than in the unexposed. Such variable is defined as an
extraneous (third) variable which is associated with the exposure and, independent of that
exposure, be a risk factor for the disease.

Characteristic of a confounding variable

1. Associated with the disease of interest in the absence of exposure

1.a. Risk factor for the study outcome among exposed group
1.b. Risk factor for the study outcome among non-exposed

2. Associated with the study exposure but not as a consequence of the exposure.

Effect of Confounding

Without prior knowledge of the effect of the variable on the outcome and exposure it is
very difficult to predict the direction of effect of a suspected confounding variable.
However, the effect could be categorized into three:

1. Totally or partially accounts for the apparent effect

51
DCH/AAU: Epidemiology Note 52

2. Mask an underlying true association

3. Reverse the actual direction of the association

Control for Confounding Variables

The list of potential confounders in a study is limited to established risk factors for the
disease of interest, though still some other variables may play a confounding role in the
association it might be difficult to identify them and explain their effects.

In the design confounding could be minimized by:

Randomization
Restriction
Matching

Evaluation of confounding in the analysis by:

Standardization
Stratification/pooling
Multivariate analysis

C. Chance

One of the alternative explanation to the observed association between an exposure and a
disease is chance. Since the general aim of epidemiological studies is to make
generalization about a larger group of individuals on the basis of a sample population it is
always important to evaluate the role of chance or sampling variability in any study
which tries to elucidate association. Evaluation of the role of chance is mainly the domain
of statistics and it involves:

1. Hypothesis Testing (Test of Statistical Significance)

Test of statistical significance quantifies the degree to which sampling variability may
account for the observed results. The “P value” is used to indicate the probability or
likelihood of obtaining a result at least as extreme as that observed in a study by chance
alone, assuming that there is truly no association between exposure and outcome under
consideration (i.e., H0 is true). For medical research, the P value < 0.05 is set
conventionally to indicate statistical significant.

P value is a function of:

 The magnitude of the difference between the groups

52
DCH/AAU: Epidemiology Note 53

 Sample size

The fact implies that even a very small difference may be statistically significant is the
sample size is sufficiently large, and a large difference may not achieve statistical
significance if variability is substantial due to a small sample size. Hence, one cannot
make a definite decision about the role of a factor based only on the P value.

Steps in testing for statistical significance

1. Assume that the exposure is not related to disease – state the null hypotheses.
2. Compute a measure of association – relative risk or odd ratio.
3. Calculate chi-square statistical test of significance.
4. For the value of chi-square calculated, look up its corresponding P-value in the
table of chi-squares.

 A very small P-value means that you are very unlikely to observe such an
association if the null hypotheses is true.

2. Estimation of Confidence Interval

The confidence interval represents the range within which the true magnitude of effect
lies within a certain degree of assurance. It is more informative than just P value because
it reflects on both the size of the sample and the magnitude of the effect.
D. Establishing a Causal Association

Sequence in establishing a causal association between an exposure and outcome

1. Incidental observation of possible causal association between an exposure and

outcome.
2. Descriptive epidemiologic analysis establishing the association on a population level.
3. Analytic epidemiologic studies establishing the association on an individual level.
4. Experimental reproduction of the outcome by the exposure and/or looking for
biologic explanation.
5. Observation that removal of the exposure (or modification of the host response to it)
decreases the occurrence of the outcome.

Our primary objective in epidemiology is to judge whether an association between

exposure and disease is, in fact, causal. Scientific proof of a cause-effect relationship is
often difficult to obtain, since experimental studies may be either not feasible or not
ethical. Since associations documented by other kinds of epidemiologic studies do not
constitute proof of causation, one must assess the validity of individual studies and

53
DCH/AAU: Epidemiology Note 54

examine the totality of evidence from all available studies and make a judgement about
the likelihood of a cause-effect relationship.

Judgements of causality must first consider whether, for any individual study, the
observed association is valid (i.e., Whether the findings reflect the true relationship
between exposure and disease or may be explained by chance, bias, or confounding) and,
second, whether the accumulated evidence supports a cause-effect relationship. The
validity of an observed association is established by eliminating alternative explanations
of that association. Associations can be:

1. Artifactual (spurious) associations, which may be:

a) the result of chance variation (i.e., type 1 error). Statistical tests and confidence
intervals can help to evaluate the likelihood of this as an explanation for an
association.
b) The result of bias, or systematic error in the design or conduct of the study.
Examples of how bias can lead to atrifactual (i.e., not real) associations are
presented below.

2. Noncausal (indirect) associations, which may occur when:

a) the associated factor is itself an effect, rather than a cause (reverse causation), or
both a cause and an effect (reciprocal causation). For example, in the association
between vitamin A deficiency and diarrhea, vitamin A deficiency could be a
cause or an effect of diarrhea, or both. Vitamin a deficiency results in
abnormalities in epithelial surfaces and, thus could impair resistance to the
infectious agents of diarrhea. On the other hand, diarrhea leads to reduced food
intake through loss of appetite and to impaired absorption of nutrients, both of
which could result in vitamin A deficiency after repeated episodes of diarrhea.
b) The association is due to a confounding effect by a third variable. A
confounding variable is one independently associated with both the exposure and
the disease. To confound, a variable must fulfil each of the following two criteria:
1) it must be related to both the frequency of disease exposure and the frequency
of disease recognition, and 2) it must occur with differing frequencies in groups
being compared (cohorts or cases and controls). For example, the association
between anaemia and illiteracy is likely non-causal, due rather to the confounding
effect of socioeconomic status, which is independently related to both anaemia
(because of poor diet) and illiteracy (because of reduced access to educational
opportunities).

3. Causal associations, which can be established only when other potential explanations
of the association can be ruled out.

In observational studies, there are many potential confounders and sources of bias, some
of which may remain undetected. The results of one observational study rarely provide
adequate support for concluding that there is a cause-and-effect relationship between an

54
DCH/AAU: Epidemiology Note 55

exposure and a disease. Properly conducted experimental trials do provide direct proof of
causality, yet are often impossible because of ethical considerations.

In the absence of experimental evidence, the following criteria (called the Bradford-Hill
criteria) are used to assess the strength of evidence for a cause-and –effect relationship.
The criteria are listed in descending order of importance:

1. Strength of the Association – The stronger the association, the more likely
that it is causal.
2. Consistency of the Relationship – The same association should be
demonstrable in studies with different methods, conducted by different
investigators, and in different populations.
3. Specificity of the Association – The association is more likely causal if a
single exposure is linked to a single disease.
4. Temporal Relationship – The exposure to the factor must precede the onset
of the disease.
5. Dose-response relationship – The risk of disease often increases with
increasing exposure to a causal agent.
6. Biological plausibility – The hypothesis for causation should be coherent
with what is known about the biology and the descriptive epidemiology of the
disease.

VII. PRESENTATION OF EPIDEMIOLOGIC

INFORMATION
The ultimate goal of epidemiologic studies is to generate information which are useful for
planning of health services and promotive, preventive and control activities. Usually
these studies generate an enormous data which are difficult to comprehend in the form
they are collected. Therefore, it is mandatory to reduce the data in the form easily
understandable by everybody. To that end epidemiologist use several data
reduction/summary methods to display their data as simple as possible. Some of the
methods are discussed below.

1. Table

Table summarize a set of data arranged in rows and columns. Tables are useful for
demonstrating patterns, exceptions, differences or other relationships. Tables may also
serve as the basis for preparing more visual displays of data, such as graphs and charts,
where some of the detail may be lost. Tables designed to present data should be as simple
as possible. Two or three small tables, each focusing on a different aspect of the data, are

55
DCH/AAU: Epidemiology Note 56

easier to understand than a single large table that contains many details or variables. To
create a table that is self-explanatory, use the following guidelines:

 Use a clear and concise title that describes the what, where, and when of the
data in the table. Precede the title with a table number.
 Label each row and each column clearly and concisely and include the units
of measurement for the data.
 Show totals for rows and columns. If you show percents, also give their total
(always 100).
 Explain any codes, abbreviations, or symbols in a footnote.
 Note any exclusions in a footnote.
 Note the source of the data in a footnote if the data are not original.

Types of tables

A. One-variable table (frequency distribution table)

This display; the values or categories of one variable and the number and percentage of
people falling into that category.

Table 1: Distribution of students by sex.

Category Number Percent

Female 80 40
Male 120 60

Total 200 100

B. Two-variable table (contingency table)

Data displayed in contingency tables is frequently used to calculate measures of

association and tests for statistical significance. The cells of a table can just as easily
contain means, rates, years of potential life lost, relative risks, and other statistical
measures. A common type of contingency table is the two-by-two table, in which each of
the two variables has only two categories.

Table 2: A two-by-two table of measles and vaccination.

Measles
Yes No
Yes 10 90
Vaccination No 70 30

C. Three-variable table

56
DCH/AAU: Epidemiology Note 57

Though its use is not much recommended sometimes three variables can be displayed in a
table. At this point it is important to remember that elegant tables are simple and easy to
understand.

C. Table shells or dummy tables

Dummy tables are prepared as part of the analysis plan to show how the data will be
organised and displayed once the data is collected. Table shells are complete except for
the data, showing titles, headings and categories. In developing table shells which include
continuous variables such as age, we create more categories than we may later use, in
order to disclose any interesting patterns.

Dummy Table 3: Distribution of children in village X by age.

Category (in year) Number Percent

<1
1-4
5-9
10-14

Total

Ordinal variables are presented according to their intrinsic natural categories. For
continuous variables an artificial categories must be created based on the purpose of the
study, however, it is advisable to create more categories (narrow intervals) in order not to
miss any interesting patterns.

In creating categories, or class intervals, for continuous variable of epidemiologic data,

remember the following guidelines:
 Create class intervals that are mutually exclusive and that include all of the
data.
 Use a relatively large number of narrow class intervals for your initial
analysis. You can always combine intervals later. In general, you will have 4
to 8 intervals when analysis is complete.
 Use natural or biologically meaningful intervals when possible. Try to use age
groupings that are standard or used most frequently in the particular field of
study. If rates are to be calculated, the intervals for the numerator must be the
same as the intervals used for the available population data.
 Create a category for unknowns.

Table 4 below lists some standard class intervals used for age (age-groupings) used for
data presentation and analysis.

Table 4: SOME STANDARD AGE GROUPINGS FOR EPIDEMIOLOGIC

REPORTING

57
DCH/AAU: Epidemiology Note 58

Notifiable Pneumonia and Final Mortality HIV/AIDS

Diseases Influenza Mortality Statistics
< 1 YEAR < 28 days < 1 year < 5 years
1-4 28 days-< 1 year 1-4 5-12
5-9 1-14 5-14 13-19
10-14 15-24 15-24 20-24
15-19 25-44 25-34 25-29
20-24 45-64 35-44 30-34
25-29 65-74 45-54 35-39
30-39 > 85 55-64 40-44
40-49 unknown 65-74 45-49
50-59 75-84 50-54
> 60 > 85 55-59
age not stated not stated 60-64
> 65

If no natural or standard class intervals are available, several strategies can be used for
creating intervals. These include:
 Divide the data into groups of similar size. To apply this strategy, divide the
total number of observations by the number of intervals you wish to create
(usually 4, but you might start with 8). Next, develop a cumulative frequency
column of a rank-ordered distribution of your data to find where each interval
break would fall.
 Base intervals on mean and standard deviation. With this strategy you can
create 3,4 or 6 class intervals.
 Divide the range into equal class intervals. This method is most common and
simplest, and is most readily adapted to graphs.

2. Graph

A graph is a way to show quantitative data visually, using a system of coordinates. It is a

kind of statistical snapshot that helps us see patterns, trends, aberrations, similarities, and
differences in the data. We usually use the horizontal axis (or x-axis) to show the values
of the independent (or x) variable. We use the vertical axis (or y-axis) to show the
dependent (or y) variable, which is usually a frequency measure such as number of cases
or rates of disease. Each axis should be labeled (with both the name of the variable and
the units in which it is measured) and a scale of measurement marked along the line.

2.1 Histogram

A histogram is a graph of the frequency distribution of a continuous variable. It

uses adjoining columns to represent the number of observations for each class
interval in the distribution. The area of each column is proportional to the number
of observations in that interval. Histograms with unequal class intervals are,

58
DCH/AAU: Epidemiology Note 59

therefore, difficult to construct and are not recommended. A second variable may
be displayed using a histogram by shading each column into the component
categories of the second variable. Epidemic curves (wich are not really “curves”
at all) are frequently displayed as histograms.

2.2 Frequency polygon

A frequency polygon, like a histogram, is the graph of a frequency distribution. In

a frequency polygon, we mark the number of observations within an interval with
a single point placed at the mid-point of that interval, and then connect the points
with a straight line. A frequency polygon of a set of data must enclose the same
area as a histogram of the data. (Note that for each area of histogram that the
polygon leaves out, it includes another area of equal size.) Frequency polygons
are often used to compare two or more distributions on the same axis. Other
commonly used graphic displays of epidemilolgic data include cumulative
frequency curves, survival curves, and scatter diagrams (or scatter grams).

3. Chart

Charts are methods of illustrating statistical information using only one coordinate. They
are most appropriate for comparing data with discrete categories other than “place”.
Variables shown in bar charts are either discrete and non-continuous (e.g., race or sex) or
are treated as though they were discrete and non-continuous (e.g., age groups rather than
age intervals). The length or height (bar charts can be presented either horizontally or
vertically) of each bar is proportional to the frequency of the event in that category (and,
therefore, scale breads should not be used). The simplest bar chart is that used to display
data from a one-variable table. This presentation makes it very easy to see the relative
importance of different variables.

3.1 Grouped bar chart

A grouped bar chart can be used to illustrate data from two-variable or three-
variable tables, when an outcome has only two separate categories. Bars within a
group are usually adjoining, should be no more than three, and must be illustrated
distinctively and described in a legend.

3.2 Stacked bar chart

Stacked bar charts can be used to show categories of a second variable as

components of the bars that represent the first variable.

3.3 Deviation bar chart

Deviation bar charts can be used to show deviations in a variable, both positive
and negative, from a variable.

59
DCH/AAU: Epidemiology Note 60

3.4 100% component bar chart

100% component bar charts are useful for comparing the contribution of different
components to each of the categories of the main variable. This is a variation of
the stacked bar chart in which we make all the bars the same height (or length)
and show the components as percents of the total rather than as actual values.

3.5 Pie chart

Pie charts are simple, easily understood charts in which the size of the “slices”
show the proportional contribution of each component part. Pie charts are useful
for showing the component parts of a single group or variable. Conventionally,
we begin at 12 O’clock and arrange the component slices from largest to smallest.

3.6 Geographic coordinate charts

Geographic coordinate charts (maps) are used to show the location of events or
attributes. Spot maps used dots or other symbols to show where an event occurred
or a condition exists. Although it can show the geographic distribution of an
event, a spot map does not show risk because it does not take the size of the
population into account. Area maps can overcome this problem by using shaded
or coded areas to show either the incidence of an event in sub areas, or the
distribution of some condition over a geographic area. Area maps can show either
numbers or rates.

To construct a bar chart, observe the following guidelines:

 Arrange the categories that define the bars, or groups of bars, in a natural
order, such as alphabetical or by increasing age, or in an order that will
produce increasing or decreasing bar lengths.
 Position the bars either vertically or horizontally, except for deviation bar
charts, in which the bars are usually positioned horizontally.
 Make all of the bars the same width.
 Make the length of bars in proportion to the frequency of the event. Do not
use scale breaks.
 Show no more than three bars within a group of bars.
 Leave a space between adjacent groups of bars, but not between bars within a
group.
 Code different variables by differences in bar color, shading, cross-hatching,
etc. and include a legend that interprets your code.

Table 5: Guide to Selecting a Graph or Chart to Illustrate Epidemiologic Data

Type of Graph or When to Use

Chart

60
DCH/AAU: Epidemiology Note 61

Arithmetic-Scale Line Trends in numbers or rates over time

Graph
Semilogarithmic-scale line 1) Emphasize rate of change over time
graph 2) Display values ranging over more than 2 orders of
magnitude
Histogram 1) Frequency distribution of continuous variable
2) Number of cases during epidemic (epidemic curve)
over time
Frequency Polygon Frequency distribution of continuous variable, especially
to show components
Cumulative Frequency Cumulative frequency for continuous variable
Scatter Diagram Plot association between two variables
Simple Bar Chart Compare size or frequency of different categories of a
single variable
Grouped Bar Chart Compare size or frequency of different categories of 2-4
series of data
Stacked Bar Chart Compare totals and illustrate component parts of the total
among different groups
Deviation Bar Chart Illustrate differences, both positive and negative, from
baseline
100% Component Bar Chart Compare how components contribute to the whole in
different groups
Pie Chart Show components of a whole
Spot Map Show location of cases or events
Area Map Display cases/events or rates geographically
VIII. OUTBREAK INVESTIGATION AND
MANAGEMENT
PATTERNS OF EPIDEMICS

Two principal types are well recognized. These are the common source and
propagated/progressive. The two types can be distinguished by plotting an epidemic
curve. An epidemic which shows the features of both types is referred as mixed.

1. Common source epidemics

Caused by exposure of a group of people to a common noxious influence, such as an

infectious agent or a toxin. If the exposure is brief and simultaneous all exposed will
develop the disease within one incubation period – referred as point,or point source
epidemic/outbreak. A rapid rise and fall of an epidemic curve suggests a point source
epidemic – this is called log-normal distribution. If the source of an outbreak remains for
a longer time, days, weeks or longer either continuously or intermittently, there will be
multiple exposures with variable incubation period, this will make an epidemic curve
with no clear peak and the duration of the outbreak will be prolonged. Continuous

61
DCH/AAU: Epidemiology Note 62

common source – makes wide peak in the epidemic curve, because of the range of
exposures and range of incubation periods. Intermittent common source – results in an
irregular pattern of the epidemic curve that reflects the intermittent nature of the
exposure.

2. Propagated or progressive epidemics

Outbreak of this type can occur through direct person-to-person transmission or the
transmission could pass through a vector from infected to healthy person.

The epidemic curve would have a successive series of peaks reflecting increasing
numbers of cases in each generation. The epidemic usually wanes after a few generations,
either because the number of susceptible falls below some critical level, or because
intervention measures become effective. In reality, few propagated outbreaks provide a
classic pattern. Diseases with short incubation period and are highly infectious, can create
a rapidly rising and falling epidemic curve similar to that of a point source epidemic.

* When one can not distinguish the two by the epidemic curve, studying the geographic
distribution will help to differentiate them. The propagated epidemics tend to show
geographic spread with successive generations of cases.

3. Mixed Epidemics

Epidemics having the features of both common source and propagated epidemics are
referred as mixed epidemics. For example a common source outbreak may be followed
by secondary person-to-person spread.
Steps of an Epidemic Investigation

There is no rigid step to follow during investigation of an outbreak. Several activities

could be accomplished simultaneously. The steps to follow are set by the individual
investigator depending on the suspected cause of the outbreak. Verification of the
diagnosis and establishment of the existence of an epidemic are commonly among the
first steps.

1. Prepare for field work

Before leaving for the field an investigator must be well prepared to under take the
investigation. Preparations can be categorized into three:

A) Investigation related: Investigator must have the appropriate scientific

knowledge, supplies, and equipment to carry out the
investigation. Discuss the situation with knowledgeable
people, review applicable literature, and collect sample
questionnaire.

B) Administration related: No matter there is urgency in handling the situation, it is

62
DCH/AAU: Epidemiology Note 63

useful to observe all administrative procedures. This

include arrangement of transportation and organising
personnel matters.

C) Consulation: Clarify your and your team role in the field. Identify local contacts
at the site where the outbreak is reported and arrange where and
when to meet them.

2. Verify the existence of an epidemic

Compare the current number of cases (or incidence) with the past levels of disease in that
community, considering the seasonal variation in the occurrence of the disease, to
determine whether an excessive number of cases have occurred, i.e., compare the
observed number of cases (reported as outbreak) with the expected number of cases in the
area.

Be careful, excess may not always indicate an outbreak. The excess may be due to
changes in local reporting procedures, change in case definition increased interest
because of local or national awareness, or improvements in diagnostic procedures. In area
with sudden changes in population size such as resort areas, college towns, and migrant
farming areas, changes in the numerator (number of reported cases) may simply reflect
changes in the denominator (size of the population) – absolute numbers (without
proportion or rates) should be carefully analyzed.

Outbreak/epidemic: is the occurrence of more cases of disease than expected in agiven

area or among a specific group of people over a particular period of time.
Cluster: is an aggregation of cases in a given area over a particular period without regard
to whether the number of cases is more than expected.

3. Verify the diagnosis

Review clinical and laboratory findings to establish diagnosis. This is to ensure that the
problem has been properly diagnosed and to rule out laboratory error as the basis for the
increase in diagnosis. If you have any doubt about the laboratory findings review the
laboratory techniques being used with the qualified laboratorian or send specimen for
confirmation to reference laboratory.

Summarize the clinical findings with frequency distribution. They are useful in
characterizing the spectrum of the illness, verifying the diagnosis, and developing case
definitions. Visit as much patients as you can. Conversation with patients are very helpful
in generating hypothesis about disease etiology and spread. Depending on the type of the
problem under investigation establish criteria for labeling persons as “cases”.

3.1. Case definition

63
DCH/AAU: Epidemiology Note 64

Case definition is a standard set of criteria for deciding whether an individual

should be classified as having the health condition of interest.

A Case definition includes clinical criteria, particularly in an outbreak

investigation, restricted by time, place, and person. Set simple and objective
measures, and apply them consistently and without bias to all persons under
investigation, do not include an exposure or risk factor, which is going to be
tested in the case definition. For example, if one of the goal of the investigation is
to determine whether work place is associated with the illness, the case definition
should not be restricted with regard to working place. Use “loose” case definition
early in the investigation to identify the extent of the problem and the population
affected. But, during testing the hypothesis generated from this process using
analytic epidemiology, specific or “tight” case definitions must be used.

Keeping in mind the uncertainty of some diagnosis, it is advisable to classify

cases as confirmed, probable, or possible. This may help to keep track of a case if
the diagnosis is not confirmed or if there is a decision not to order laboratory test
required to confirm the diagnosis because the test is expensive, difficult to obtain,
or unnecessary. It is also customary that investigators usually confirm the
diagnosis of a few cases and rely on clinical features to identify the rest of the
cases.

Confirmed/definite: a case with laboratory verification.

Probable: a case with typical clinical features of the disease without laboratory
confirmation.
Possible: a case presented with fewer of the typical clinical features.

3.2. Surveillance – identifying and counting cases.

Often the case which creates the concern are small and non representative fraction
of the total number of cases. Therefore, epidemic investigators should “cast the
net wide “to determine the geographic extent of the problem and the population
affected by it. In order to do that one must adopt an appropriate methods, for the
setting and disease in question, to identify cases. The two types of surveillance
commonly utilized in an outbreak investigation are:

1. Stimulated or enhanced passive surveillance, includes:

 Sending out a letter describing the situation and asking for reports.
 Alerting the public directly, usually through local media, to see a
physician if they have symptoms compatible with the disease in
question.
 Asking case-patients if they know anyone else with the same condition.

2. Active surveillance:

64
DCH/AAU: Epidemiology Note 65

 Making telephone call or visit the facilities to collect information on

cases.
 Conducting a survey of the entire population.

Regardless of the disease under investigation collect the following types of information
about every case:

 Identifying information – allows you to contact patients and to map the

geographic extent of the problem.
 Demographic information – Provides the “person” characteristics of the
population at risk.
 Clinical information – allows verification of the case definition and charting
the time course of the outbreak, and helps to describe the spectrum of the
illness.
 Risk factor information – inquire specifically exposure to the suspected cause.
 Reporter information – to help you inquire additional information if there is a
need or to report back the results of your investigation.

4. Describe the epidemic with respect to time, place, person.

Collect relevant information related to the investigation. Information could be obtained

from the already existing records or you can obtain using a case investigation form
specifically designed for a particular situation under investigation. Information must be
collected carefully, so that, at the end they will enable the investigator to characterize the
outbreak with respect to time, place and person.

By using a well established descriptive epidemiological tools, like epidemic curve and
spot mapping, an outbreak can be characterized by time, place and person.

Epidemic curve – plots the cases by the time of onset and provides a time frame for
the outbreak investigation.

Spot map – plots the cases by location and shows the geographic spread of
cases.

Attack rates - Calculate rates of illness in population at risk by exposure to

specific suspected items and other relevant attributes. The
identification of “relevant” attributes may be a crucial step in the
solution of the problem.

5. Formulate and Test Hypotheses

Formulate the hypotheses based on your characterization of the epidemic by time, place,
and person. The hypotheses should address the source of the agent, the mode of
transmission, and the exposures that caused the disease. Determine the type of epidemic-
common source Vs propagated. Based on characteristics of the epidemic define the

65
DCH/AAU: Epidemiology Note 66

population at the highest risk and consider the possible source(s) of the disease
(infection). The hypotheses should be testable.

In an outbreak investigation, evaluation of hypotheses can be done in two ways: either by

comparing the hypotheses with the established fact, or by using analytic epidemiologyto
quantify relationships and explore the role of chance.

 Analytic approach:

The analytic technique utilizes the cohort and the case-control approach to identify
possible source of an outbreak. Cohort approach identifies the comparison group based
on exposure status. The case-control method identifies the comparison groups on the
basis of their disease status.

 Compute Odds ratio to find association between cases and controls (non-ill)
with regard to exposure to the suspected case-case-control.
 Calculate relative risk (attack rate) to determine whether there is association
between exposed and non-exposed – cohort.
 Compute statistical tests to determine how likely it is that the investigation
results could have occurred by chance alone, if exposure was not actually
related to disease.

In both analytic approaches a test of significance has to be worked out to determine

whether the differences observed between the two groups (cases Vs control, or exposed
Vs non-exposed) are not due to chance. Statistically significant difference only provides
supportive evidence on the possible source of an outbreak. Causation can only be
established after careful assessment of the whole situation and requires laboratory proof,
which is not always easy.

Incidence (attack) rate among exposed

Relative Risk = Incidence (attack) rate among unexposed

= a/a + b
c/c + d

Odds Ratio = ____Proportion of exposed in diseased (cases)____

Proportion of exposed in non-diseased (controls)

= a/a + c
b/b + d if a & b are small relative to c & d

= a/c = ad
b/d bc

66
DCH/AAU: Epidemiology Note 67

6. Search for additional cases: Locate unrecognized or unreported cases.

Passive: inquire physicians or hospitals or both whether they have seen similar
cases.

Active: do intensive investigation in the community on asymptomatic persons or

contacts of the cases. For example, doing liver function test in an
investigation of hepatitis A outbreak.

7. Analyze the Data

. Assemble all results.
. Interpret findings.

8. Make a decision on the hypotheses tested.

. All the findings must be consistent with one, and only one, hypotheses.

9. Intervention and follow-up.

Although it is discussed late, intervention must start as soon as possible depending on the
specific circumstances. Aim control measures at the weak link or links in the chain of
infection. One might aim control measures at the specific agent, source, or reservoir. For
example, an outbreak might be controlled by destroying contaminated foods. Sterilizing
contaminated water, or destroying mosquito breeding sites, or an infectious food handler
could be removed from the job and treated. (see discussion on epidemic management).

10. Report of the investigation

At the end prepare a comprehensive report and submit to the appropriate/concerned

agency (or agencies). The report should follow the usual scientific format: introduction,
background, methods, results, discussion, and recommendations.

The report should discuss in detail:

- factors leading to the epidemic.
- evaluation of measures used for the control of the epidemic.
- recommendations for the prevention of similar episodes in the future.

Managing Outbreak/epidemics

Management of epidemics require an urgent and intelligent use of appropriate measures

against the spread of the disease. Action to be taken is dependent on the type of the

67
DCH/AAU: Epidemiology Note 68

disease as well as the source of the outbreak. However, the action can be generally
categorized as presented below to facilitate easy understanding of the strategies.

1. measures Directed Against the Reservoir

Understanding the nature of the reservoir is necessary in the selection of an appropriate

control methods and their likelihood of success. The following are examples of control
measures against disease with varies reservoir:

Domestic animals as reservoir:

. Immunization
. Testing of herds
. Destruction of infected animals
Example: brucellosis and bovine tuberculosis.

Wild animals as reservoir:

. post-exposure prophylaxis
Example: rabies

Humans as reservoir
 removal of the focus of infection – e.g., cholecystectomy in a chronic typhoid
carrier.
 Isolation of infected persons. This is separation of infected persons from non-
infected for the period of communicability. Not suitable in the control of
diseases in which a large proportion are inapparent infection or in which
maximal infectivity precedes overt illness.
 Treatment to make them noninfectious – e.g., tuberculosis.
 Disinfection of contaminated objects.

 Quarantine- is the limitation of freedom of movement of apparently healthy persons

or animals who have been exposed to a case of infectious disease. Usually imposed
for the duration of the usual maximal incubation period of the disease.

* Cholera, Plaque, and yellow fever are the three internationally quarantable
diseases by international agreement.
* Now quarantine is replaced in some countries by active surveillance of the
individuals – maintaining close supervision over possible contacts of ill persons
to detect infection or illness promptly; their freedom of movement is not
restricted.

2. Measures that interrupt the transmission of organisms

* Action to prevent transmission of disease by ingestion:
. purification of water
. pasteurization of milk
. inspection procedures designed to ensure safe food supply

68
DCH/AAU: Epidemiology Note 69

. improve housing conditions.

* Attempts to reduce transmission of respiratory infections

. chemical disinfection of air and use of ultraviolet light.
. work on ventilation patterns, like unidirectional (“laminar”) air flow to
reduce the transmission of organisms in hospitals.

* Action to interrupt transmission of disease whose cycles involve an intermediate

host
. clearing irrigation farms from snails to control schistosomiasis.

3. Measures that reduce host susceptibility

* Active immunization, when either the altered organism or its product is given to a
person to induce production of antibodies – EPI.

* Passive immunization, has lesser role in the control of communicable diseases than
active immunization:

 Transfer of maternal antibodies to the fetus through the placenta.

 Prophylaxis administration of immune serum globulin (ISG). E.g., TAT for

unimmunized persons who receive penetrating wounds, antitoxin against
Clostridium botulinum and antiserum against rabies.

 Chemoprophylaxis:

 Use of antibiotics for known contacts of cases – for example, in

tuberculosis, gonorrhoea, and syphilis.

 Use of chlorquine to persons travelling to malaria endemic areas.

Uncovering outbreaks

Outbreaks are detected in one of the following ways:

. Through timely analysis of routine surveillance data, this may reveal an increase in
reported cases or unusual clustering of cases.

. Report from clinician

. Report from the community, either from the affected group or concerned citizen.

Why Investigate Possible Outbreaks

1. To institute control and prevention measures

69
DCH/AAU: Epidemiology Note 70

In order to design and implement appropriate control measures assessment of the extent
of the outbreak and the size and the characteristics of the population at risk needs to be
done.

2. Opportunity for research

Outbreaks are natural experiment waiting to be analyzed and exploited. It gives a unique
opportunity to study the natural history of diseases. May also help to assess the impact of
control measures and the usefulness of new epidemiology and laboratory techniques.

3. Training opportunity

Investigating an outbreak requires a combination of diplomacy, logical thinking,

problem-solving ability, quantitative skills, epidemiologic know-how, and judgement.
These skills improve with practice and experience. Therefore, an outbreak may provide a
good opportunity for an epidemiologist in-training to learn these skills by working with
experienced epidemiologist.

4. Opportunity for program evaluation

An outbreak of a disease targeted by a public health program, such as EPI, tuberculosis or

STDs, may reveal a weak point in that program and provide the opportunity to change or
strengthen the program’s effort.

5. Public, political, or legal concerns

Public, political, or legal concerns some override scientific concerns in the decision to
conduct an investigation. The call from these parties usually have no scientific basis and
such investigations mostly do not identify a causal link between exposure and disease.
Nevertheless, health departments have to be responsive to public concerns, because it at
least provides an opportunity to educate the public.

IX. EPIDEMIOLOGIC SURVEILLANCE

Epidemiologic Surveillance is the systematic collection, analysis, interpretation and
dissemination of health data in an ongoing basis. Surveillance provides “information for
action” which can be used to investigate, prevent, and control disease in communities. Its
purpose is to provide a factual basis for setting priorities, planning programs, and taking
action to promote and protect community health. Surveillance can be conducted globally
(as in the AIDS surveillance system managed by WHO), regional (as in the polio
surveillance in Latin America), national, or institutional (as in the surveillance for
hospital acquired or nosocomial, infections or for potential causes of epidemics in
refugee camps)

70
DCH/AAU: Epidemiology Note 71

We do not limit surveillance to diseases for which we have effective control measures.
Surveillance can be justified for two additional purposes: 1) to learn more about the
natural history, clinical spectrum, and epidemiology of a disease, and 2) to obtain
baseline data which we can use to assess the effectiveness of prevention and control
measures when they are developed and implemented.

Surveillance is a system of close observation of all aspects of the occurrence and

distribution of a given disease through systematic collection, tabulation, analysis,
and dissemination of all relevant data pertaining to that disease.

We monitor health events for the following purposes:

 To detect sudden changes in disease occurrence and distribution (determines

the need for epidemic investigation and control) and to ensure that effective
action to control the disease is being done.
 To follow secular (long-term) trends and patterns of disease (alerts decision
makers of the need to reallocated resources or shift policy)
 To identify changes in agents and host factors (helps to assess the potential for
future disease occurrence)
 To detect changes in health care practices (points up the need for changes in
preventive measures)

Interpretation of surveillance data may also provide the basis for generating hypotheses
and stimulating community health research, test hypotheses regarding the impact of
exposures on disease occurrence. Archival surveillance data have also been used to
develop statistical models of diseases, such as to predict the feasibility of proposed
programs to eradicate measles and polio.

The following are some key sources of surveillance data, not all of which are available in
every country:
 Census data
 Mortality reports (birth and death certificates, autopsy reports)
 Morbidity reports (notifiable disease reports)
 Hospital data (discharge diagnoses, surgical logs, hospital infection reports)
 Absenteeism records (school, workplace, compensation claims)
 Epidemic reports
 Laboratory test utilization and result reports
 Drug utilization records
 Adverse drug reaction reports
 Special surveys (e.g., research data, serologic surveys)
 Police records (especially for injury, alcohol-related crime)
 Information on animal reservoirs and vectors (e.g., for rabies, plague, Lyme
disease)

71
DCH/AAU: Epidemiology Note 72

 Environmental data (hazard surveillance, water and food testing)

 Special surveillance systems (e.g., for injury and occupational illness)

Types of Surveillance

Passive surveillance is that in which health care providers send reports based on a known
set of rules and regulations.

Active surveillance is that in which public health officials contact providers to solicit
reports of events or diseases. Such active surveillance is usually limited to specific
diseases over a limited period of time, such as after a community exposure or during an
epidemic. Incomplete reporting, especially in passive surveillance systems, is very
common.

Sentinel surveillance uses a pre-arranged sample of reporting sources to report all cases
of one or more conditions. Usually the sample sources are selected to be those most likely
to see cases. Particularly in developing countries, sentinel surveillance provides a
practical alternative to population-based surveillance. Under this strategy, health officials
define homogenous population subgroups and the regions to be sampled. They then
identify institutions that serve the population subgroups of interest, and that can and will
obtain data regarding the condition of interest.

Surveillance systems based on secondary data analysis can make productive use of data
sets collected for other purposes. Data collected for marketing surveys, patient
management records, police records, and other information sources can be exploited as
sources of surveillance data. Such data may be of lesser quality and timeliness than data
collected through systems designed specifically for surveillance.

As with all descriptive epidemiologic data, surveillance data is first analyzed in terms of
time, place, and person. Data are analyzed as rates rather than simply the numbers of
cases reported. When delays occur between diagnosis and reporting, we analyze data by
the date of onset, rather than the date of the report. A critical step before calculating rates
is the identification of the appropriate denominator. Simple tabular and graphic
techniques are used initially to display the data, although sophisticated techniques such as
cluster and time series analysis and computer mapping may also be used.

Surveillance data may be assessed for changes over time by comparing the number of
cases for the current period with the number reported for the same period in each of the
last three years. Secular trends, or long-term trends, are usually analyzed by graphing the
occurrence of disease by year. Any key events, such as initiation or cessation of a control
program. Should be noted on the graph. Changes in the surveillance system (such
changes in diagnostic criteria, reporting requirements, screening programs, or publicity
about the condition) which may influence the appearance of long-term trends should also
be indicated on the graph.

72
DCH/AAU: Epidemiology Note 73

The surveillance data should also be analyzed by place. Even when the secular trend
reveals no increases in overall incidence, analysis by place may reveal a geographic
cluster of cases, which deserves investigation. Analysing surveillance data by the
characteristics by person variables (age, sex, and behavioral risk factors) may also reveals
patterns or clues.

There is no single “threshold” above which disease patterns are different enough from the
expected to warrant further investigation. The excess necessary to trigger action may
depend on the priority assigned to the disease and the interests, capabilities and resources
of the ministry or agency. Public, political, or media attention and pressure, however, can
sometimes make it necessary to investigate minor variations in disease occurrence, which
might no otherwise be pursued.

Apparent increases should be treated as real until proven otherwise. However other
causes of apparent increases should also be considered, including an increase in the
denominator population, improved detection, “batch” reporting, or other changes in the
system itself. Surveillance data should be disseminated to those who provide reports, and
those who need to know for administrative, program-planning, and decision-making
purposes. Newsletters and other reports of surveillance data can also help to maintain the
quality of a surveillance system by providing motivation for continued reporting by
health care providers. Like other epidemiologic data, surveillance data should be
“information for action”, collected only if it is functionally linked with community health
programs.

In the design or evaluation of a surveillance system, proposals or recommendations for

modifications should be made with an understanding of the trade-offs (e.g., the impact
that efforts to improve representativeness may have on cost). In justifying, designing or
evaluating a surveillance system, the following aspects of the system should be assessed:

1) The importance to the public health of the health event under surveillance
a) incidence and prevalence
b) severity (case-fatality or death-to-case ratio)
c) mortality (overall and age-specific mortality rates, years of potential
life lost)
d) health care costs
e) potential for spread
f) preventability

2) The objectives and operation of the system

a) the case definition of the health event
b) the population under surveillance
c) the time period for data collection (weekly, monthly, annually)
d) what information is collected (is it what programs need?)
e) the reporting sources
f) how data are handled (transfers, delays, confidentiality)
g) how data are analyzed (by whom? Frequency, thoroughness)

73
DCH/AAU: Epidemiology Note 74

h) how data are disseminated

3) The system’s usefulness

a) action taken to date as a result of the information
b) future or potential uses

4) Attributes or qualities of the surveillance system

a) simplicity
b) flexibility (with changes in case definition or funding, to add new
diseases)
c) acceptability (often judged by proportion who report, completeness of
forms)
d) sensitivity (ability to detect events it is intended to detect)
e) predictive value positive (proportion of reported cases which truly are
cases, or of epidemics which are actual epidemics)
f) representativeness (extent to which one can generalize or draw
conclusions from surveillance data, such as for calculating rates)
g) timeliness

5) Cost or resource requirements for system operation

Surveillance systems are never perfect. Understanding the limitations of surveillance data
is important to ensure correct interpretation. The most common limitations of
surveillance systems include:
1) Under reporting (such as due to lack of knowledge of reporting requirements,
negative attitudes toward reporting)
2) Lack of representativeness of reported cases (such as due to a bias toward
reporting severe cases, or increased likelihood of reporting after publicity)
3) Lack of timeliness
4) Inconsistency of case-definitions

These limitations suggest specific steps, which may be taken to improve a surveillance
system. Most commonly, surveillance systems are strengthened by improving awareness
of practitioners, simplification of the process of reporting, frequent feedback to those
reporting, widening the “net” (for example, obtaining reports from laboratories or
schools, rather than relying on physicians), and using active (rather than passive)
surveillance. Remember to “share the data, share the responsibility, share the credit”.

Important Points
Factors related with the selection of disease for surveillance:
 Magnitude of the disease
 Feasibility of control measures
 Need for monitoring and evaluating the performance of a control program
 Resource availability

74
DCH/AAU: Epidemiology Note 75

Activities in surveillance:
 Data collection and recording
 Reporting and notification
 Compilation, data analysis, and interpretation
 Dissemination of findings for action
Conditions in which active surveillance is appropriate:
 For periodic evaluation of ongoing programs
E.g. HIV/AIDS, EPI…
 For programs which have time limit of operation
E.g. Small pox
 With the occurrence of unusual situations:
 When a new disease/event discovered
 When investigating a new mode of transmission
 When a high-risk period is recognized
 When a disease appears in a new geographic area or found to affect a
new subgroup of the population
 When previously eradicated disease reappear or low incidence
disease occur at a higher level of endemicity
Features of good surveillance system
 Uses a combination of passive and active mechanisms to collect data.
. Emphasize the collection of minimum data in s simplest possible
way.
. To assure quality and enhance compliance make sure that the data
collected is useful for the workers who collect the data.
 Timely reporting.
 Timely and comprehensive action.
Action must be targeted towards both case detection and treatment and as well as to the
control of the disease.
 Strong laboratory services for accurate diagnosis.

X. SCREENING
Although the principles of interpretation of diagnostic and screening tests are classically
applied to tests done in a laboratory, these principles apply equally well to information
obtained from other clinical assessments (such as history or physical examination) as well
as any indicator or indirect measure used in science. Most ordinal scales for variables are
used to simplify the interpretation and use of data. Data are also often expressed as
simple dichotomies (e.g., exposed/ unexposed, ill/well), although realities are rarely so
simple.

Test results are likewise often reported as simply normal or abnormal, although some
may be more or less abnormal (or normal) than others. The following table summarizes
the four possible relationships between a diagnostic test and the actual presence of
disease:

75
DCH/AAU: Epidemiology Note 76

DISEASE STATUS
TEST RESULT
PRESENT ABSENT

POSITIVE True Positive (a) False Positive (b)

NEGATIVE False Negative (c) True Negative (d)

Such assessments of what is “true” or “false” depend on the selection of a “gold

standard”. Although the truth or falsehood of measures by the “gold standard” method
may, themselves, be questioned, these are usually the best available information, which is
the basis for evaluation of the performance of a second diagnostic tast which is usually
cheaper, easier, or safer.

Sensitivity is defined as the proportion of people with a disease who have a positive test
for the disease (a/a +c).

Specificity is defined as the proportion of people with a disease who have a positive test
(d/b +d).

A sensitive test is preferable when there is an important penalty for failing to detect a
disease (e.g., when trying to detect a dangerous but treatable condition). Sensitive tests
are also used when the probability of disease is relatively low and the purpose of the
testis to discover possible cases. A sensitive test is; therefore, most helpful when the test
result is negative. Specific tests, on the other hand, are most useful when the test result is
positive, and are often used to confirm a diagnosis, which has been suggested by other
data. A highly specific is preferable when false positive results might have negative
(physical, emotional, or financial) consequences.

Case definitions used in epidemiology may also be characterized by their sensitivity and
specificity. For rare but potentially severe communicable diseases, where it is important
to identify every possible case, health officials use a sensitive or “loose” case definition.
On the other hand, investigators of the causes of a disease outbreak want to be certain
that any person included in the investigation really had the disease. In this case, the
investigator prefers a specific or “strict” case definition.

In theory, the sensitivity and specificity of a test are independent of the prevalence of the
condition being detected. In practice, however, several characteristics of cases (such as
the stage and severity of the disease) may be related to both the sensitivity and specificity
of the test and to the prevalence of the disease, since different kinds of cases are found in
high-and low-prevalence situations. Tests are often assessed to be more valuable than
they actually are, since a positive test result may prompt the health care provider to
continue pursuing a diagnosis, while a negative result may cause a clinician to abandon
further testing.

76
DCH/AAU: Epidemiology Note 77

When assessing the implications of a positive or negative test, the sensitivity and
specificity (which are more useful in deciding whether to perform the test) are no longer
of primary importance.

Positive predictive value (or predicative value positive) (-PV = a/b +b) is the probability
of disease in a person with a positive (abnormal) test result.

Negative predictive value (or predictive value negative) –PV = d/c +b) is the probability
of not having the disease when the test result is negative (normal).Predictive value is
sometimes called posterior or post-test probability.

The predictive value of a test is not a property of the test alone. It is determined by the
sensitivity and specificity of the test and the prevalence of the disease in the population
being tested. Positive results, even for a very specific test, when applied to a population
with a low likelihood of disease, will be largely false positives. Similarly, negative
results, even for a very sensitive test, will be largely false negatives when the test is
performed in a population with a high chance of having the disease.

The criteria for a successful screening program were first summarized in a WHO
publication in 1968. They can be broadened to screening for problems other than human
disease:
1. The problem to be detected should be important enough to be worth detecting.
2. There should be an acceptable intervention, which is effective.
3. The intervention should be feasible and available.
4. There should be a recognizable latent or early “asymptomatic” stage.
5. There should be a suitable test.
6. The test should be acceptable to the population to be tested.
7. The natural history of the condition should be adequately understood.
8. There should be an agreed policy regarding when the intervention is
appropriate.
9. The cost of detecting the problem and its remedy should be reasonable.
10. The screening program should be ongoing, and not a “one-time” effort.

XI. ETHICS OF EPIDEMIOLOGIC RESEARCH

Epidemiologic research, with its continually expanding potential for collection, storage
and use of data on individuals and communities, encounters inevitable conflicts between
the rights and freedoms of the individual and the needs of society. Past ethical abuses,
particularly in clinical research, have underlined the need for clear guidelines for the
ethical conduct of both clinical and epidemiologic research. The Proposed International
Guidelines for Biomedical Research Involving Human Subjects were adopted by the
World Medical Association in 1964 to guide application of the ethical principles for
clinical research specified in the Declaration of Helsinki. The additional need for special

77
DCH/AAU: Epidemiology Note 78

ethical guidelines for epidemiologic studies has also recently been accentuated by the
complex issues raised by research regarding HIV infections and AIDS

In 1991, the Council for International Organizations of Medical Sciences (CIOMS), in

collaboration with the World Health Organization (WHO), published the “International
Guidelines for the Ethical Review of Epidemiological Studies”. The following discussion
of epidemiologic research ethics draws largely on these guidelines. Although such ethical
guidelines cannot resolve all the moral ambiguities that are encountered in everyday
epidemiologic research and practice, they can draw attention to the ethical implications of
professional action and thereby improve ethical standards.

Ethical issues often arise as a result of conflict among competing sets of values. Many
situations require careful discussion and informed judgements on the part of
investigators, ethical review committees, administrators, health care practitioners, policy-
makers, and community representatives. Externally sponsored epidemiological studies in
developing countries merit special attention in ethical review.

The purpose of ethical review is to consider the features of a proposed study in light of
ethical principles, so as to ensure that investigators have anticipated and satisfactorily
resolved possible ethical objections, and to assess their response to ethical issues raised
by the study. Not all ethical principles weigh equally. A study may be assessed as ethical
even if a usual ethical expectation, such as confidentiality of data, has not been
comprehensively met, provided the potential benefits clearly outweigh the risks and the
investigators give assurances of minimizing risks. It may even be unethical to reject such
a study, if its rejection would deny a community the benefits it offers. The challenge of
ethical review is to take into account potential risks and benefits, and to reach decisions
which best reflect the consensus of the review committee. Different conclusions may
result from different ethical reviews of the same issue or proposal, and each conclusion
may be ethically reached, given varying circumstances of place and time; a conclusion is
ethical not merely because of what has been decided, but also owing to the process of
conscientious reflection and assessment by which it has been reached.

General Ethical Principles

General ethical principles may be applied at the individual and community levels. At the
level of the individual (micro ethics), ethics governs how one person should relate to
another and the moral claims of each member of a community. At the level of the
community, ethics applies to how one community relates to another, and to how a
community treats each of its members (including prospective members) and members of
other groups with different cultural values (macro ethics). Procedures that are unethical at
one level cannot be justified merely because they ear considered ethically acceptable at
the other.

All research involving human subjects should be conducted in accordance with four basic
ethical principles: 1) respect for persons. 2) beneficence. 3) non-maleficent. And 4)
justice.

78
DCH/AAU: Epidemiology Note 79

Respect for persons incorporates at least two other fundamental ethical principles,
namely:

a) autonomy, which requires that those who are capable of deliberation about
their personal goals should be treated with respect for their capacity for self-
determination: and,
b) Protection of persons with impaired or diminished autonomy, which requires
that those who are dependent or vulnerable be afforded security against harm
or abuse.

Beneficence is the ethical obligation to maximize possible benefits and to minimize

possible harms and wrongs. This principle gives rise to norms requiring that the risks of
research be reasonable in the light of expected benefits, that the research design be sound,
and that the investigators be competent both to conduct the research and to assure the
well-being of the research subjects.

Non-maleficence (“Do no harm”) holds a central position in the tradition of medical

ethics, and guards against avoidable harm to research subjects.

Justice requires that cases considered to be alike be treated alike, and that cases
considered to be different be treated in ways that acknowledge the difference. When the
principle of justices is applied to dependent or vulnerable subjects, its main concern is
with the rules of distributive justice. Studies should be designed to obtain knowledge that
benefits the class of persons of which the subjects are representative. The class of persons
bearing the burden should receive an appropriate benefit, and the class primarily intended
to benefit should bear a fair proportion of the risks and burdens of the study.

Ethical Principles Applied to Epidemiology

1. Informed Consent

When individuals are the subject of epidemiologic studies, their individual informed
consent will usually be sought. Consent is informed when it is given by a person who
understands the purpose and nature of the study, what participation in the study requires
the person to don and to risk, and what benefits are intended to result from the study. An
investigator who proposes not to seek informed consent has the obligation to explain how
the study would be ethical in its absence (such as because informed subjects might alter
the behaviour under study or feel needlessly anxious, or because it is public knowledge
that personal data is made available for epidemiologic studies). Consent is not required
for use of publicly available information, although countries and communities differ with
regard to the definition of what information about citizens is regarded as public.
Investigators must provide assurances that strict safeguards will be maintained to protect
confidentiality by minimizing disclosure of personally sensitive information.

79
DCH/AAU: Epidemiology Note 80

When it is not possible to obtain informed consent from every individual to be studied,
community agreement through a representative of a community or group may be sought,
but the representative should be chosen according to the nature, traditions and political
philosophy of the community or group. Approval given by a community representative
should be consistent with general ethical principles. Even if a leader expresses agreement
on behalf of a community, the refusal of individuals to participate has to be respected.
Representatives of a community or group may sometimes be invited to participate in the
design of a study and in its ethical assessment.

Selective disclosure may be used in epidemiologic research, provided that it does not
induce subjects to do what they would not otherwise consent to do. For certain
epidemiologic studies, such non-disclosure is permissible, even essential, so as not to
influence the spontaneous conduct under investigation, and to avoid obtaining responses
that the respondent might give in order to please the questioner.

Prospective subjects may not feel free to refuse requests from those who have power or
undue influence over them. It is ethically questionable whether subjects should be
recruited from among groups that are unduly influenced by persons in authority if the
study can be conducted with subjects who are not in this category.

Individuals or communities should not be pressured to participate in a study. However, it

can be hard to draw the line between exerting pressure or offering inappropriate
inducements to participate and creating legitimate motivation. The benefits of a study,
such as improved knowledge or health, are appropriate inducements. However, when
people or communities lack basic health services or money, the prospect of being
rewarded by goods, services or cash payments can induce participation. It is acceptable to
repay incurred expenses, such as for travel.

2. Maximizing Benefit

Part of the benefit that communities, groups, and individuals may reasonable expect from
participating in studies is that they will be told of findings that pertain to their health. A
strategy for communication of study results to policy-makers and to participating
individuals and communities (with due consideration of levels of literacy and
comprehension) should be included in the study protocol. When findings indicated a need
for health care, those concerned should be appropriately advised and arrangements should
be made for treatment or referral. Health professionals have an obligation to advocate
release of study results that is in the public interest. Training of local health personnel in
skills and techniques that can be used to improve health services or research may also be
an important way of ensuring that communities will benefit from the proposed research.

3. Minimizing Harm

Epidemiologic studies must consider all harm or disadvantage to individuals and

communities which may be incurred due to the research. For example, diversion of scarce
health personnel from their routine duties to serve the needs of a study or alteration of

80
DCH/AAU: Epidemiology Note 81

health care priorities may constitute harm. Ethical review should also assess the risk of
subjects or groups suffering stigmatization, prejudice, loss of prestige or self-esteem (e.g.
due to being identified as HIV-positive), or economic loss as a result of taking part in a
study. Investigators must be able to demonstrate that benefits outweigh the risks for both
individuals and groups. When a health person is a member of a population or sub-group
at increased risk and engages in high-risk activities, it is unethical not to propose
measures for protecting the population or sub-group.

Disruption of social mores is usually regarded as harmful. Although investigators must

respect social mores, it may be the specific aim of an epidemiologic study to stimulate
change in certain customs or behaviours to improve health. Investigators must respect the
ethical standards of their own country or culture as well as the cultural expectations of the
societies in which epidemiological investigations are under taken.

4. Confidentiality

Research may involve collecting and storing data relating to individuals and groups, and
such data, if disclosed to third parties, may cause harm or distress. Consequently,
investigators should make arrangements for protecting the confidentiality of such data by,
for example, omitting information that might lead to identification of individual subjects,
or limiting access to the data, or by other means. When personal identifiers remain on
records used for a study, investigators should explain why this is necessary and how
confidentiality will be protected.

Unlinked information is that which cannot be linked, associated or connected with the
person to whom it refers. As this person is not known to the investigator, confidentiality
is not at stake and the question of consent does not arise. Linked information may be 1)
anonymous (when the information cannot be linked to the person to whom it refers
except by a code or other means known only to that person, and the investigator con not
know the identity of the person), 2) non-nominal (when the information can be linked to
the person by a code which is not a personal identifier and which is known to the person
and the investigator), and 3) nominal or nominative (when the information is linked to the
person by means of personal identification, usually the name).

5. Conflict of Interest

It is an ethical rule that investigators should have no undisclosed conflict of interest with
their study collaborators, sponsors, or subjects. Conflict can arise when a commercial or
other sponsor may wish to use study results to promote a product or service, or when it
may not be politically convenient to disclose findings. Honesty and impartiality are
essential in designing and conducting studies, and presenting and interpreting findings.
Data must not be withheld misrepresented or manipulated.

81