CASE-CONTROL STUDY
Exposure and outcome have occurred before the [Link] is a retrospective
study.(looking back in time). Uses a control group to support an inference outcome.
Also known as effect -to-cause. This study examines the relationship between
exposure and outcome, as well as disease and risk factors.
Case : A case is an individual who meets specific diagnostic or inclusion criteria
for the condition under investigation in a study within Epidemiology.
Control : A control is an individual selected from the same population as the cases,
who is free from the disease but at risk of developing it, used for comparison in a
study within Epidemiology.
Controls should come from the same population as cases
They must be comparable (age, sex, etc.)
They should represent the exposure distribution of the population
Proper selection is crucial to avoid selection bias
Advantages 1. Helps you determine the source
of an existing illness or epidemic.
2. Cheap and quick to conduct this
type of study. The health issue
has already occurred; you don’t
need a lab or special equipment.
3. Few ethical issues, as the patient
already has the health condition
4. Looks at multiple risk factors in
a patient’s life (environment,
work, diet).
Disadvantages 1. Patient recall about their history
can be inaccurate (recall bias).
2. Patients aware of certain risk
factors may focus on those and
ignore other exposures.
3. No randomisation is possible,
lowering internal validity of the
study.
4. Finding a control group that
matches the case group
appropriately can be difficult.
5. This study type does not prove a
clear causal relationship between
risk factors and illness; it only
calculates the odds.
Opportunities 1. Particularly useful for rare
conditions, it’s often the only
option as it’s not appropriate to
try to cause a disease using
RCTs.
2. Can be used to identify point of
outbreak, looking into the past to
see common exposure in the Case
histories.
3. Can look at many exposure
factors of the patient’s
environment, medical history,
diet. This is useful where no
single aspect of the patient’s life
has been narrowed down as a
potential cause.
Confounding 1. If the health issue causes
multiple common exposures, the
study should attempt to
differentiate between those
exposures to identify which is
more likely. For example, having
a chronic illness may mean you
spend more time in hospital
environments, have a lower
immune system, and suffer high
levels of stress – so any of these
could potentially be the cause of
that chronic illness.
2. The study must make an attempt
to eliminate some of these
exposures as causes if possible.
If the Control group is
overmatched and starts to
develop the health [Link] the
Control group is under-matched
and there aren’t enough
comparable factors to
differentiate what is specific to
the Case population. For example,
if you could not find female
Controls to match female Cases,
then you cannot assess whether
gender is or isn’t a contributing
risk factor.
Disease present Disease absent Total
Exposed 40 (a) 5 (b) 50
Non exposed 10 (c) 45(d) 50
Outcome measure
A) Exposure rates
Cases – a/a+c
Controls – b/b+d
B) Estimate of risk (Strength of association)
C) Odds ratio: The odds ratio is the main measure of association used in a case-
control study in Epidemiology. It tells us how strongly an exposure is related
to a disease.
Formula : axd /bxc
Suppose:
Smokers among cases = 40 (a)
Non-smokers among cases = 10 (b)
Smokers among controls = 20 (c)
Non-smokers among controls = 30 (d)
OR = 40 × 30 / 10 × 20 =6
Interpretation: Smokers have 6 times higher odds of developing the disease.