Case-Control Study: Dr. Sourab Kumar Das

CASE-CONTROL
STUDY
DR. SOURAB KUMAR DAS
3 DISTINCT FEATURE OF CASE
CONTROL STUDY:
a) Both exposure and out come (disease) have occurred
before the start of the study.
b) The study proceeds backwards from effect to cause,
and
c) It uses a control or comparison group to support or
refute an inference.
By definition a case control study involves two population

case and control. In case control studies, the unit is the
individual rather than the group. The focus is on a
disease or some other health problem that has already
developed.
Design of a Case-Control Study
TIME
Direction of inquiry
Start with :
Exposed CASES
(people with disease)
Non Exposed
POPULATIO
N
Exposed CONTROLS
(people without disease)
Non Exposed
BASIC STEP IN CASE-CONTROL STUDY
• SELECTION OF CASES
• SELECTION OF CONTROLS
• MATHCING
• MEASURMENT OF EXPOSURE
• ANALYSIS AND INTERPRETATION.
SELECTION OF CASES
• 1)Selection by definition of case

• 2) Selection by source of case
SELECTION BY
DEFINITION
• a)Diagnostic criteria: The definition must be specified
before the study is undertaken.
• Supposing, we are investigating a case of cancer, we should
be quite clear about a group( case) histologically the same.
Once diagnostic criteria established they should not change
or altered till the study is over.
• B) Eligibility criteria: Only newly diagnosed (incident)
cases within a specified period of time are eligible than old
cases or cases in advanced stage of disease (Prevalent case).
WHY INCIDENCE CASES??
• PATIENT RECALL BETTER THAN PREVALENT CASES

• AVAILABILITY OF RECORDS
SOURCES OF CASES
1. HOSPITAL BASED
-all persons with the disease seen at a particular
medical care facility or group of facility
-more common, easy, and inexpensive to carry

out
2.COMMUNITY BASED
-all persons with the disease detected in a more general

population, city or county in a period time
-more difficult, involves special efforts to locate and obtain
necessary data from all affected individuals as well as medical
records
-advantage-computation of rates of disease in total population
related to etiologic factors
SELECTION OF CONTROLS
• SHOULD BE FROM THE ”SOURCE-POPULATION”,

i.e., the study base that gives rise to the cases.
• SHOULD BE FREE OF THE STUDY DISEASE
• SHOULD BE SIMILAR TO THE CASES IN REGARD TO PAST
POTENTIAL EXPOSURE
• MOST DIFFICULT AND CONTROVERSIAL ASPECTS OF THIS
STUDY DESIGN
• SOURCES AND METHOD(S) OF DATA COLLECTION SHOULD
BE SIMILAR FOR CASES AND CONTROLS
SOURCES OF CONTROLS
1.POPULATION OF AN ADMINISTRATIVE
AREA
• appropriate when the cases represent all, or the majority of, cases
occurring in the same area
• expensive, time consuming
• non-response is high
• concern on different information quality is raised, if the situations of the
interview are different.
2. HOSPITAL PATIENTS
• more practical
• patients with other diseases are subjected to the same selective
factors that influence the cases to come to the particular facility
• controls are ill persons, thus likely to be unrepresentative of the
population as a whole with respect to factors, e.g., smoking,
socioeconomic status
• should be of variety of conditions so that no
excessive representation in the control group
• note that the factors leading a patient to come to
a particular hospital are not the same for all
diseases
3. ASSOCIATES OF THE CASES
• attendance at the same school, similar type of employment, friends,
or residence in the same neighborhood
• these controls would be fairly similar matched to cases on
socioeconomic status and environmental exposures
• multiple call-backs are necessary, great efforts needed
• relatively expensive
• tendency to overmatched(cases and friends).
4. RELATIVES OF THE CASES
 spouses or siblings since they are generally
similar in ethnic and social background to the
cases
 not suitable choice if the study exposure is one
of which close relatives are likely to be similar,
e.g., diets, genetics background
EXAMPLES OF CASE-CONTROL STUDY
• The relationship between thalidomide and unusual limb

defects in babies born in the Federal Republic of Germany
in 1959 and 1960. In 1961, the study was conducted to
compare the affected children with the normal children. Of
46 mothers whose babies had typical malformations, 41 had
taken thalidomide between the 4th and the 9th weeks of
pregnancy, whereas none of the 300 control, whose children
were normal, had taken the medications during the period
CALCULATION OF ODDS RATIO
ODDS RATIO = odds that an exposed person

develops disease/odds that a non-exposed
person develops the disease
a/c=ad
b/d bc
Cases Controls
Exposed a b
Not Exposed c d
Odds Odds that the cases were exposed

Ratio = Odds that controls were exposed
a/c ad
= b/d = bc
MATCHING
• THE PROCESS OF SELECTING CONTROLS SO THAT
THEY ARE SIMILAR TO THE CASES IN CERTAIN
CHARACTERISTICS, SUCH AS AGE, RACE, SEX,
SOCIOECONOMIC STATUS, AND OCCUPATION
• 2 TYPES: 1. GROUP MATCHING
2. INDIVIDUAL MATCHING
GROUP MATCHING
• ALSO KNOWN AS FREQUENCY MATCHING
• CONSISTS OF SELECTING THE CONTROLS IN SUCH
A MANNER THAT THE PROPORTION OF CONTROLS
WITH A CERTAIN CHARACTERISTIC IS IDENTICAL
TO THE PROPORTION OF CASES WITH THE SAME
CHARACTERISTIC
• EG. IF 25% CASES ARE MARRIED, CONTROLS ARE
SELECTED SO THAT 25% IS ALSO MARRIED
INDIVIDUAL MATCHING
• ALSO KNOWN AS MATCHED PAIRS
• HERE, FOR EACH CASE SELECTED FOR THE
STUDY, A CONTROL IS SELECTED WHO IS SIMILAR
TO THE CASE IN TERMS OF SPECIFIC VARIABLE
OR VARIABLES OF CONCERN
• EACH CASE IS INDIVIDUALLY MATCHED TO A
CONTROL
• OFTEN USED WHEN IT’S HOSPITAL CONTROLS
USE OF MULTIPLE
CONTROLS
• CONTROLS OF THE SAME TYPE
• IN CONDITIONS WHEN DISEASES RARE, OR CASES ARE
DIFFICULT TO OBTAIN
• INCREASES THE POWER OF THE STUDY (IF RATIO IS 1:4)
USE OF MULTIPLE
CONTROLS
• MULTIPLE CONTROLS OF DIFFERENT
TYPES
• TO GENERALISE THE RESULT TO GENERAL
POPULATION
• FOR EXAMPLE: USING THE HOSPITAL AND THE
NEIGHBORHOOD CONTROLS
• INVESTIGATOR WILL HAVE TO DECIDE WHICH
COMPARISON GROUP WILL BE USED AS “GOLD
STANDARD”
FACTORS TO KEEP IN MIND
WHEN MATCHING
ANALYSIS
• The final step is analysis, to find out
a)Exposure rates among case and controls to
suspected factor
b)Estimation of disease risk associated with
exposure(odds ratio).
EXAMPLE: A CASE CONTOL STUDY OF SMOKING
AND LUNG CANCER
Case(with lung Control( witho

cancer) ut lung
cancer)
Smokers (less than 5 33 55
cigarettes a day) (a) (b)
Non -smokers 2 27
(c) (d)
Total 35(a+c) 82(b+d)
Exposure rate
a.Case=a/a+c=33/35=94.2%
b.controls=b/b+d=55/82=67.0%
P<001
The frequency rate of lung cancer was definitely higher
among smoker than among non-smokers.
Next step will be to ascertain whether there is a statistical
association between exposure status and occurrence of lung
cancer . This question can be resolve by calculating the P
value.
• Relative risk
• =Incidence among exposed group/ incidence
among non expose group.
• =a/(a+b)/c/(c+d)
• Odds ratio
• =ad/bc
• 33×27/55×2=8.1
• In the above example, smokers of less than 5
cigarettes per day showed a risk of having lung
cancer 8.1 times than that of non-smokers.
• Odds ratio is the key parameter in analysis of case
control studies.
LIMITATIONS OF CASE CONTROL STUDY
• PROBLEM OF BIAS : RECALL BIAS, INFORMATION

BIAS, CONFOUNDING, SELECTION,
INTERVIEWERS’
• DIFFICULTY IN SELECTION OF APPROPRIATE
CONTROL GROUP
• CANNOT MEASURE INCIDENCE
• DO NOT DISTINGUISH BETWEEN CAUSES AND
ASSOCIATED FACTORS
• NOT SUITABLE FOR EVALUATION OF TREATMENT
OR PROPHYLAXIS
ADVANTAGES
• INEXPENSIVE
• RAPID
• NO RISK TO SUBJECTS
• ALLOWS STUDY OF SEVERAL DIFFERENT
ETIOLOGICAL FACTORS IN ONE STUDY
• SUITABLE FOR INVESTIGATION OF RARE DISEASES

Case-Control Study: Dr. Sourab Kumar Das

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Case-Control Study: Dr. Sourab Kumar Das

Uploaded by

Copyright:

Available Formats

CASE-CONTROL

By definition a case control study involves two population

• 1)Selection by definition of case

• PATIENT RECALL BETTER THAN PREVALENT CASES

-more common, easy, and inexpensive to carry

-all persons with the disease detected in a more general

• SHOULD BE FROM THE ”SOURCE-POPULATION”,

• The relationship between thalidomide and unusual limb

ODDS RATIO = odds that an exposed person

Odds Odds that the cases were exposed

Case(with lung Control( witho

• PROBLEM OF BIAS : RECALL BIAS, INFORMATION

You might also like