Analytical Studies(Analytical

epidemiology
Analytical Studies(Analytical epidemiology)

• Observational study
• Testing the etiological hypothesis formulated
by descriptive epidemiology
• Types

Case-control study Cohort study

 From each of these studies one can
show:
• Whether any association exists between the
suspected factor and the disease of the
hypothesis
• If so, what is the strength of the association
between the suspected factor and the disease
under study.
Case Control Studies
 Introduction
• Imagine you are a clinician
• You have seen few patients with certain type
of cancer
• Almost all of them have been exposed to a
particular chemical
• You hypothesize that their exposure is related
to their risk of developing this type of cancer
• How will you go about confirming or refuting
your hypothesis?
 Example 1
• In the early 1940s, Alton Ochsner, a surgeon in New
Orleans, observed that virtually all of the patients on
whom he was operating for lung cancer gave a history
of cigarette smoking
• Although this relationship is accepted and well
recognized today, it was relatively new and
controversial at the time that Ochsner made his
observation.
• He hypothesized that cigarette smoking was linked to
lung cancer.
• Based only on his observations in cases of lung cancer,
was this conclusion valid?
 Example 2
• Again in the 1940s, Sir Norman Gregg, an Australian
ophthalmologist, observed a number of infants and young children
in his ophthalmology practice who presented with an unusual form
of cataract.
• Gregg noted that these children had been in utero during the time
of a rubella (German measles) outbreak. He suggested that there
was an association between prenatal rubella exposure and the
development of the unusual cataracts.
• Keep in mind that at that time there was no knowledge that a virus
could be teratogenic.
• Thus, he proposed his hypothesis solely on the basis of observational
data, the equivalent of data from ambulatory or bedside practice
today.
Classification of research methods

Research
methods

Observational Experimental

Uncontrolled,
Descriptive Analytical Controlled
Non-random

Ecological Cross- Case control Cohort

Case series, sectional
case reports,
CS, cohort
Position in the evidence hierarchy

9
Design of a case-control study

Hallmark of Case Control Study:

from cases and controls and searches for exposure.
 Introduction:
• Most frequently used study design
• Participants selected on the basis of whether
or not they are DISEASED

• Those who are diseased are called CASES.

• Those who are not diseased are called
CONTROLS
 Definition
• A case control study is defined as an
epidemiological approach in which the researcher
starts by picking up ‘cases’ who have already
developed particular disease or ‘outcome’ of
interest and a comparison group (controls) who
have not developed the disease but are similar to
cases.
• Than he/she tries to find out the presence of
particular exposure which he/she thinks is a risk
factor and compares the two groups in regards to
presence of history of exposure.
• Case : A person in the population or study
group identified as having the particular
disease, health disorder or condition under
investigation. (Dictionary of Epidemiology: 3 ed; John M Last. 2000)
rd

• Control: Person or persons in a comparison

group that differs, in disease experience (or
other health related outcome) in not having
the outcome being studied. (Dictionary of Epidemiology: 3 rd

ed; John M Last. 2000)

Design of Case-control study
 Features of case controls study
• Both exposure and outcome has happened
before the start of the study.

• The study proceeds backwards from effect to

cause.

• It uses a comparison group to support or

refute an inference.
 Framework of case control study
( The 2 2 contingency table)

Suspected or Cases Control

risk factors ( Disease ( Disease
present) absent)
Present a b
Absent c d
a+c b+d
Design of case control study
 Basic steps
• Selection of cases and controls

• Matching

• Measurement of exposure

• Analysis and interpretation

1.Selection of cases and controls
 Selection of cases
a. definition of the case:
• disease or outcome of interest
1. diagnostic criteria – Strict establishment of assurance of
homogeneous

2. Eligibility criteria: New cases are employed for the study

 Cont…
b) sources of cases:
1) hospitals
2) general population
Approaches:
• First approach: Hospital based case control
study
• Second approach: Population-based case
control study
 Selection of controls
• Selection of comparison group – perhaps the
most difficult and critical issue

Sources:
• Hospital controls
• Relatives
• Neighborhood control
• General population
 Selection of controls
• Controls must be free from disease under study
• Must be similar to the cases except for the
disease under study
• Selection of controls is the most difficult
• Sources of controls
– Hospitals
– Relatives
– Neighborhood
– General population
Source Advantage Disadvantage

Hospital based Easily identified. Not typical of general population.

Available for interview. Possess more risk factors for disease.
More willing to cooperate. Some diseases may share risk factors
Tend to give complete and with disease under study. (whom to
accurate information exclude???)
(recall bias). Berkesonian bias

Population based Most representative of the Time, money, energy.

(registry cases) general population. Opportunity of exposure may not be
Generally healthy. same as that of cases.

Neighbourhood Controls and cases similar in Non cooperation.

controls/ Telephone residence. Security issues.
exchange random Easier than sampling the Not representative of general
dialing population. population.

Best friend control/ Accessible, Cooperative. Overmatching.

Sibling control Similar to cases in most
aspects.
 How many controls are needed?
• 1:1 for larger studies
• 1:2
• 1:3
• 1:4
• Multiple control

• Failure to select appropriate/comparable

control group results in bias.
2.Matching
 Matching
• Matching is a process in which we select
controls in such a way that they are similar to
cases with regard to certain pertinent
variables (eg. age) which are known to
influence the outcome of disease and which if
not adequately matched for comparability
could distort or confound the results.
 MATCHING

• Matching is defined as the process of selection of

the controls in such a way that they resemble
cases in all the attributes and not only the free
from the disease under study but also free from
confounding factors, such as age, race, sex,
socioeconomic status, and occupation.
• Matching may be of two types:
– group matching
– individual matching.
Confounding factor:
• It is the factor that is independently associated
with the disease as well as risk factor and is
distributed unequally in both the groups and
independently change the risk of outcome.
Cont…
 What is a confounding factor?

Alcohol Esophageal cancer

Smoking
 Matching types

• Group matching (Frequency matching)

• Pair matching ( Individual matching)

3.Measurement of exposure
Measurement of exposure in both
the groups
• Consists of collection of data among both

cases and control groups about their exposure

to the suspected cause with reference to

duration and frequency of exposure.

Measurement of exposure in both
the groups
Such information can be obtained by
• Interviews

• Questionnaires

• Past records
– Hospital records
– Employment records
4. Analysis
 Analysis
• Find out

– Exposure rates among cases and controls to

suspected factor

– Estimation of disease risk associated with

exposure ( Odds Ratio)
 Exposure rates
A case control study between smoking and lung cancer
Cases Controls ( without Total
(with Ca Lung) Ca lung)

Smokers 33 55 88
( >5/day) (a) (b) (a+b)

Non-smokers 2 27 29
(c) (d) (c+d)

Total 35 82 117
(a+c) (b+d) (a+b+c+d)
 Exposure rates
A case control study between smoking and lung cancer
Cases Controls ( without Total
(with Ca Lung) Ca lung)

Smokers 33 55 88
( >5/day) (a) (b) (a+b)

Non-smokers 2 27 29
(c) (d) (c+d)

Total 35 82 117
(a+c) (b+d) (a+b+c+d)
 Exposure rates
• Cases= a/(a+c)= 33/35= 94.2 %
• Controls= b/(b+d)= 55/82= 67 %

• So frequency of smoking was definitely higher

among lung cancer patients than those
without cancer
• Odds and odds ratio

– Odds: Odds of an event is defined as the ratio of the

number of ways an event can occur to the number of
ways an event cannot occur. (Epidemiology; Leon Gordis. 2004)

• If the probability of event X occurring is P, then odds of it

occurring is = P/ 1-P.

– Odds ratio: Ratio of the odds that the cases were

exposed to the odds that the controls were exposed.
 Outcomes of Case Control Study
• Odds ratio:

Diseased/ Cases Not diseased/

Controls
Exposed a b
Not exposed c d
Odds that case was exposed
Odds ratio =
Odds that control was exposed

= (a/c)/ (b/d) = ad / bc
 Estimation of risk
• Odds Ratio (Cross-product ratio)

• Odds that cases were exposed= a/c

• Odds that controls were exposed=

b/d

• Odds ratio= (a/c)/(b/d)= ad/bc= 8.1

 Interpretation
• The odds of smoking more than 5 cigarettes
per day was 8.1 times more in the lung cancer
patient than those without lung cancer.

OR
• Smoking (>5/day) was found be associated
8.1 times more in patients with lung cancer
than those without lung cancer.
On interrogation, 40 out of 50 cases of lung cancer and
60 out of 150 controls gave the history that they were
smoking cigarettes. Does smoking predispose to lung
cancer?
Exposure Cases Controls

Yes 40(a) 60(b)

No 10(c) 90(d)

Total 50(a+c) 150(b+d)

Exposure rate among cases=a/a+c X100
= 40/50 X100
= 80%
Exposure rate among controls= b/b+d X100
= 60/150 X100
= 40%
Since the exposure rate is more among cases than among
controls, the relation between smoking and lung cancer
exists.
Odds ratio= ad/bc
=6
That means smokers are 6 times at a greater risk of getting
lung cancer( This indicates the strength of the
association.
 Bias in case control studies
• Bias is an error, often occurring in research work,
resulting in contradictory results and thus leading
to wrong conclusion.
• Different types of bias in case control study are:
1. Memory or recall bias: The person who is
diseased is more likely to remember about the
exposure than the non – diseased.
X-ray exposure and congenital malformation
Unprotected sexual intercourse and HIV
2. Selection bias: Selection of inappropriate
control group. Basic dictum is controls should
be derived from the same source population
from which the cases have come and that the
controls should be equally at risk.
3. Confounding bias: since the confounding
factor itself independently can result in the
disease, care must be taken while selecting
the control that they must be free from the
confounding factors also. This means there
must be proper matching between the cases
and control.
 4. Berkesonian bias

• Hospital selective admissions

• Eg. Cigarette smoking causes peptic ulcer????

• Eg. High fever and febrile seizures.

5. Interviewer bias:
This occurs when the interviewer knows who
is in the study group and who is in the control
group.
• This can be overcome by double blind study,
wherein neither the interviewer nor the
participants knows their group allocation.
 Advantage of case-control study
• Easy to carry out
• Rapid and inexpensive
• Rare disease investigation
• No risk to subjects
• Allows study of several etiological factors
• Rational prevention and control measures
• No attrition
• Minimal ethical problems
 Disadvantages
• Bias
• Control selection is difficult
• Incidence cannot be measured
• Cannot differentiate between causes and
associated factors
• Not suitable for evaluation of Rx.
THANK YOU