COHORT STUDY

Epidemiologic Study Designs

 WHAT IS COHORT
• Group of persons with a
common characteristic and
experiences
• Examples
– birth cohort
– Marriage cohort
– occupational cohort
– injured
 Cohort studies
l longitudinal
l Prospective studies
l Forward looking study
l Incidence study
l startswith people free of disease
l assesses exposure at “baseline”
l assesses disease status at “follow-up”
Design of Cohort study
 Design of Cohort study
Time
Disease

Study Population
Outcome
(exposed)

Non-
Disease
Population

Disease

Control Population
Outcome
(Non-exposed)

Non-
Disease
 INDICATION OF A COHORT STUDY

• When there is good evidence of exposure and

disease.
• When exposure is rare but incidence of
disease is higher among exposed
• When follow-up is easy, cohort is stable
• When ample funds are available
 Framework of Cohort study

Cohort group Development of Disease Total

Yes No

Study cohort(exposed to a b a+b

suspected factor)

Control cohort( not c d C+d

exposed to suspected
factor)
2 x 2 contingency table for Calculation of Measures
of Association

Outcome

Exposure Present Absent TOTAL

Present a b a+b

Absent c d c+d
 Types of Cohort Study
• Prospective cohort study
• Retrospective (historical) cohort study
• Combination of Retrospective and Prospective
cohort study.
 Steps in cohort study
• Selection of study cohorts (exposure cohorts)
• Selection of control cohorts (un-exposure
cohorts)
• Obtaining data on exposure
• Selection of comparison group
• Follow up
• Analysis
 Selection of study subjects
• General population
– Whole population in an area
– A representative sample
• Special group of population
– Select group
• occupation group / professional group (Dolls study )
– Exposure groups
• Person having exposure to some physical, chemical or biological
agent
– e.g. X-ray exposure
 Obtaining data on exposure
• Personal interviews / mailed questionnaire
• Reviews of records
– Dose of drug, radiation, type of surgery etc
• Medical examination or special test
– Blood pressure, serum cholesterol, ECG
• Environmental survey

• By obtaining the data of exposure we can classify

cohorts as
– Exposed and non exposed and
– By degree exposure we can sub classify cohorts
 Selection of comparison group

• The study and control cohorts should be similar in

demographics

• Control cohorts should not be exposed to risk factors

 Follow-up
• To obtain data about outcome to be determined
(morbidity or death)
– Mailed questionnaire, telephone calls, personal interviews
– Periodic medical examination
– Reviewing records
– Surveillance of death records
Follow up is the most critical part of the study

• Some loss to follow up is inevitable due to death

change of address, migration, change of occupation.
• Loss to follow-up is one of the draw-back of the
cohort study.
 ANALYSIS

• Calculation of incidence rates among exposed

and non exposed groups

• Estimation of risk
 Framework of Cohort study

Cohort group Development of Disease Total

Yes No

Study cohort(exposed to a b a+b

suspected factor)

Control cohort( not c d C+d

exposed to suspected
factor)
 Incidence rate
• Incidence rate in study(or exposed) cohorts(IE) =
a
a+b 1000
• Incidence rate in control(or non-exposed) cohorts(IO) =
c
c+d 1000
If IE is greater than IO it suggests the existence of the
association between the suspected factors(e.g smoking)
and the disease(Lung cancer)
To study the strength of the association between the
suspected factors and the disease, risk of getting the
disease in both the groups following indicators are used
• Relative risk
• Risk difference
• Attributable risk
• Absolute risk
• Population attributable risk
2000 smokers and 1000 matched non smokers were followed-up. Lung
cancer developed in 120 smokers and 20 nonsmokers. Estimate the
strength of the association between smoking and lung cancer by the
appropriate indicators

Framework
Smoking Developed lung Not developed lung Total
cancer cancer
Present(exposed) 120(a) 1880(b) 2000

Absent(not 20(c) 980(d) 1000

exposed)
 Estimation of risk
• Relative Risk(risk ratio):
incidence of disease among exposed
RR = ______________________________
Incidence of disease among non-exposed

a/a+b X1000
= _________
c/c+d x1000

RR= 120/2000 X1000

20/1000 X1000
= 3( smokers are relatively 3 times at a greater risk of developing lung cancer than
non-smokers)
 Estimation of risk
• Risk difference: it is the difference between
incidence of disease among exposed and
unexposed
RD= IE-IO
= 60-20
= 40 per 1000 population.
 Estimation of Risk
• Attributable Risk
Incidence of disease among exposed(IE) – incidence of disease
among non exposed(IO)
AR = ____________________________
Incidence of disease among exposed(IE)

a/a+b x1000 – c/c+d X1000

AR = _______________ X100
a/a+b x1000

AR= 60-20 X100 =66.6%

60

(66.6% of lung cancer among the smokers is attrubuted to smoking)

• Absolute Risk:
The risk of developing the disease, irrespective
of exposure to the risk factor and is expressed
as percentage.

a+c
AR = ____________
a+b+c+d 100
AR= 140/3000 x100= 4.67%
Smoking Lung cancer Total

YES NO

YES 70 6930 7000

NO 3 2997 3000

73 9927 10000

Find out RR and AR for above data

• Incidence of lung cancer among smokers
70/7000 = 10 per 1000
• Incidence of lung cancer among non-smokers
3/3000 = 1 per 1000
RR = 10 / 1 = 10
(lung cancer is 10 times more common among
smokers than non smokers)
AR = 10 – 1 / 10 X 100
= 90 %
(90% of the cases of lung cancer among smokers are
attributed to their habit of smoking)
 Population attributable risk
• Is the difference between the incidence of the disease or
(death) in the total population(IP) minus the incidence of the
disease or (death) among those who were not exposed to the
suspected causal factor(IO)
• PAR= IP-IO
IP = a+c X100
a+b+c+d
IO= c/c+d X1000
PAR= 26.67%
 Advantages of Cohort Studies
• Can establish population-based incidence
• Accurate relative risk, AR (risk ratio) estimation
• Can examine rare exposures
• Temporal relationship can be inferred (prospective design)
• Time-to-event analysis is possible
• Dose response ratios can be calculated
• Magnitude of a risk factor’s effect can be quantified
• Selection and information biases are decreased
• Multiple outcomes can be studied
(smoking > lung cancer, COPD, CHD, peptic ulcer)
 Disadvantages of Cohort Studies
• Lengthy (takes long time to complete and expensive)
• May require very large samples
• Not suitable for rare diseases, uncommon disease and low
incidence in population
• The study itself may alter people’s behaviour
• Unexpected environmental changes may influence the association
• Non-response, migration and loss-to-follow-up biases
• Attrition problem
• Ethical problem
 Cohort studies
Strengths
Weaknesses
• Find out incidence rate
and risk • losses to follow-up
• More than one disease • often requires large
related to single sample
exposure • ineffective for rare
• can establish cause - diseases
effect • long time to complete
• good when exposure is
rare • expensive
• minimizes selection and • Ethical issues
information bias
 Main differences
Case control Cohort
• Proceeds from effect to cause • Proceeds from cause to effect
• Starts with disease • Starts with people exposed to
RF or suspected cause
• First approach to testing the • Reserved for testing of
hypothesis precisely formulated
hypothesis
• Involves fewer number of • Involves large number of
subjects subjects
• Long follow up and delay for
• Yeild relatively quick result result
 Cont…
Case control Cohort
• Suitable for the study of rare • Inappropriate when the
disease disease under investigation is
rare
• Generally yields only estimate • Yields incidence rate, RR as
of RR ( Odd’s ratio) well as AR

• Cannot yield information about • Can yield information about

disease other than that more than disease outcome
selected for study
• Results are more reliable
• Results are less reliable
• Relatively inexpensive • Expensive
• Techniques are crude • Techniques are refined
THANK
YOU