BY
DR. MRUNALINI
S U B R A H M A N YA M
Case Control Study
Analytical Epidemiology
Introduction • Retrospective studies
• Type of Observational study
• Involves 2 populations :-
1. Cases
2. Controls
• Unit of Study :- INDIVIDUAL
Both outcome and exposure has
occurred before the start of the
study
Study proceeds backwards from
effect to cause
Uses a control or comparison group
Features to support or refute an inference
2×2 contingency table
Framework of a
case control study
Suspected or Cases/diseases Control/disease
risk factors present absent
Present a b
Absent c d
a+c b+d
Steps
01 02 03 04
Selection of Matching Measurement Analysis and
cases and of exposure interpretation
controls
Step 1
SELECTION OF CASES SOURCES OF CASES
• Definition of a case :- • Hospitals
Involves 2 specifications
• General population
Diagnostic criteria
Eligibility criteria
Controls
SELECTION OF CONTROLS SOURCES OF CONTROLS
• Free from disease under study • Hospital controls
• Relatives
• Neighborhood
• General population
1 2
Process of selecting Examples:
cases similar to Age
controls in pertinent Gender
variables occupation
Social class
Step 2 Matching
Step 3 Measurement
of Exposure
It is important to define cases and
controls Obtained by – Interviews
Questionnaires
Past hospital
records
Employment
records……
Analysis
Step 4 Analysis
Done to find out
Estimation of
Exposure rate
risk
Controls
Cases(with (without
lung
Total Example of case
lung cancer) Cancer)
control study of
Smokers
33 55 88 smoking and lung
(a) (b) (a+b) cancer
Non
smokers 2 27 29
(c) (d) (c+d)
Total n= a+b+c+d
35 82
(a+c) (b+d)
a. Exposure rates
- Cases= a/a+c = 33/35= 94.2%
- Controls = b/b+d= 55/82= 67%
frequency of lung cancer is high among smokers than
nonsmokers
b. Estimation of risk
• ODDS ratio - Measure of strength of association between risk factor and outcomes
• Estimation of disease risk associated with exposure
Odds ratio= ad/bc= 33×27/55×2 = 8.1
Smokers had risk of having lung cancer 8.1 times that Non smokers
1. Adenocarcinoma of vagina
2. Oral contraceptives and
thromboembolic diseases
3. THALIDOMIDE tragedy
Other examples of case control
studies
Bias due to confounding
memory or recall bias
Selection bias
Bias in case
control studies Interviewers bias
Easy to carry out
No risk to subjects,
few are required, Rapid and
rare disease can be inexpensive
find
Several etiological
and risk factors can No attrition
be identified
Disadvantages
Selection of controls is difficult
Cannot measure Incidence
Relies on memory and past records
Don't distinguish between causes and associated factors
Thank you