Case Control Study
V.B.Malleswara Rao
Definition
CASE CONTROL STUDY IS AN
ANALYTICAL AND COMPARATIVE
METHOD OF OBSERVATIONAL
NATURE TO TEST THE CAUSAL
HYPOTHESIS.
To
establish causal
association & its strength
Purpose
Case Control Design
Types of Case Control Study
Unmatched Case Control Study
Matched Case Control Study
Nested Case Control Study
Basic Steps
Selection of Cases and Controls
Matching
Measurement of exposure
Analysis and Interpretation
Selection of Cases
The Specifications
Diagnostic Criteria
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 cases].
Sources of Cases
Hospitals
General population
The cases should be
representative of all cases in the
community.
1. The prevalence of exposure
among controls should reflect
the prevalence of exposure in
the source population.
2. Controls should come from
the same source population as
cases (e.g. would have been
cases if diagnosed with the
disease).
Selection of Controls
3. The time during which a
subject is eligible to be
a control should be the
time in which the
individual is also
eligible to be a case.
If #1, #2, or #3 are not
met = Selection Bias
Selection of Controls
Sources of Controls
Hospital Controls
Relatives
Neighborhood Controls
General population
Random digit dialing
AXIOM:
The benefit of increased sample size is
not as relevant past the 1:4 ratio (e.g
increase in statistical power).
Matching
Matching is a comparative
technique of neutralising
all other variables present
in cases and controls,
except the variable
[Disease] under study, to
eliminate the systematic
errors [biases] while
conducting the study.
Group matching
Matching in pairs
Matching
Advantages Disadvantages
May increase the
precision of case-control
comparisons and thus
allow a smaller study
May be time-consuming
and expensive to perform
The sampling process is
easy to understand and
explain
Some potential cases and
controls may be excluded
because matches cannot
be made
If analyzed correctly,
provides reassurance that
matched variables cannot
explain case-control
differences in the risk
factor of interest
The matched variables
cannot be evaluated as
risk factors in the study
population
Confounding is a
distortion of results
that occurs when the
apparent effects of the
exposure of interest are
attributable entirely or
in part to the effects of
an extraneous variable.
Example.age
Confounding factors
Example
Cigarette Smoking
Benzene
Exposure
(Confounder)
Lung Cancer
Measurement of exposure
By
Interviews
Questionnaires
Past records
Hospital records
Employment records etc.
Analysis
Unmatched design
Matched design
Nested
Unmatched Case Control
Data is expressed in a four-
fold table, and an odds ratio is
calculated
Cases Controls
Exposed a b
Unexposed c d
Odds Ratio = ad/bc
Odds Ratio
Odds of exposure in cases:
A / C
Odds of exposure in controls:
B / D
Odds ratio:
[A / C] / [B/ D)]
Odds ratio:
AD / BC
Case Control
E E
E E
E N
E N
E N
E N
N E
N N
N N
Case Control
Exposed
Not exposed
6 3
4 7
Odds ratio = 6/4 =3.5
3/7
Odds ratio in unmatched case-control
Odds Ratio in Matched Case
Control
Concordant pairs
Both case and control were
exposed
Neither case nor control were
exposed
Discordant pairs
Case exposed but control not
exposed
Case not exposed but control
exposed
Cont..
Odds Ratio = Ratio of Discordant pair
= b/c
Exposed Not
exposed
Exposed a b
Not
exposed
c d
Control
Case
Matched Case-Control Study of
Association Between Use of Oral
Conjugated Estrogens and Cervical
Cancer
Estrogen
Use
Present Absent Total
Present 12 43 55
Absent 7 121 128
Total 19 164 183
• OR=43/7=6.14
Interpretation
OR > 1 indicates a positive
association between the factor
and the disease.
OR < 1 indicates the factor is
protective
OR = 1 indicates no
association
Bias
Confounding bias
Memory bias
Selection bias
Berkesonian bias
Interviewer's bias
Advantages
Quick and Inexpensive
Optimal for rare diseases
Useful for diseases of long
latency from exposure to
disease development
Can evaluate multiple risk
factors
Disadvantages
Cannot calculate incidence
rate
Not suited for rare
exposures
May not elucidate temporal
sequence of exposure and
disease
Susceptible to bias
(selection & recall)
Examples of Case Control Study
Doll’s 1951 study of smoking and
lung cancer.
Doll
Male lung
cancer patients
Male patients
other diseases
Smokers 647
(99.7%)
622
(95.8%)
Nonsmokers 2 27
Total 649 649
OR = 647 x 27 = 14.04
622 x 2
Doll and Hill’s Data
Thalidomide tragedy
A retrospective study of 46 mothers
who delivered deformed babies were
found to have Thalidomide during
their early pregnancy.This was
compared with a control of 300
mothers who delivered normal babies.
Cont….
1950’s
Cigarette smoking and lung
cancer
1970’s
Diethyl stilbestrol and
vaginal adenocarcinoma
Post-menopausal estrogens
and endometrial cancer
Cont….
1980’s
 Aspirin and Reyes syndrome
 Tampon use and toxic shocks
syndrome
 L-tryptophan and eosinophilia-
myalgia syndrome
 AIDS and sexual practices
1990’s
o Vaccine effectiveness
o Diet and cancer
Nested case-control studies
are case-control studies done in
the population of an ongoing
cohort study. The case-control
study is thus said to be “nested”
inside the cohort study.
NESTED CASE CONTROL
STUDIES
TIME 1
YEARS
TIME 2
Study Population
Develop
Disease
Do Not
Develop
Disease
CASES CONTROLS
CASE-CONTROL STUDY
Obtain
interviews,
bloods,
urines, etc.
Nested Case-Control Studies
How is it possible to measure the
exposure retroactively in a
cohort study? Usually when the
following situations occur:
A specimen (e.g. of blood) had
been stored but not analyzed.
An interview is conducted to
ask about at an exposure not
assessed at baseline
Records of exposure (medical
records, occupational records)
are retrieved that were not
thought about when the cohort
was first assembled
NESTED CASE-CONTROL
Analysis is just like any
case-control study with
computation of an odds-
ratio, adjustment for
confounding, etc.
NESTED CASE-CONTROL
Example
In ARIC (Atherosclerosis Risk in
Communities) study, a cohort of 16
thousand men, all men provided serum
samples at the outset which were saved.
The cohort is observed for CHD.
After 5 years we have 246 cases of CHD.
We randomly choose 500 participants to
be controls.
We only measure Chlamydia antibody in
the stored sera from these 246 + 500
subjects.
We compare the cases (CHD) to the
controls (no CHD) with regard to the
presence of exposure (Chlamydia) which
preceded the outcome.
Reference
K.Park-T.B of P.S.M..
Mahajan – T.B. of P.S.M
W.H.O – Basic Epidemiology
Oxford – T.B. of Public Health
Mc.Graw Hill
www.pubmed.gov
www.cdc.gov
www.who.int
C03 P06 CASE CONTROL STUDY.ppt

C03 P06 CASE CONTROL STUDY.ppt

  • 1.
  • 2.
    Definition CASE CONTROL STUDYIS AN ANALYTICAL AND COMPARATIVE METHOD OF OBSERVATIONAL NATURE TO TEST THE CAUSAL HYPOTHESIS. To establish causal association & its strength Purpose
  • 3.
  • 6.
    Types of CaseControl Study Unmatched Case Control Study Matched Case Control Study Nested Case Control Study
  • 7.
    Basic Steps Selection ofCases and Controls Matching Measurement of exposure Analysis and Interpretation
  • 8.
    Selection of Cases TheSpecifications Diagnostic Criteria 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 cases].
  • 9.
    Sources of Cases Hospitals Generalpopulation The cases should be representative of all cases in the community.
  • 10.
    1. The prevalenceof exposure among controls should reflect the prevalence of exposure in the source population. 2. Controls should come from the same source population as cases (e.g. would have been cases if diagnosed with the disease). Selection of Controls
  • 11.
    3. The timeduring which a subject is eligible to be a control should be the time in which the individual is also eligible to be a case. If #1, #2, or #3 are not met = Selection Bias Selection of Controls
  • 12.
    Sources of Controls HospitalControls Relatives Neighborhood Controls General population Random digit dialing AXIOM: The benefit of increased sample size is not as relevant past the 1:4 ratio (e.g increase in statistical power).
  • 13.
    Matching Matching is acomparative technique of neutralising all other variables present in cases and controls, except the variable [Disease] under study, to eliminate the systematic errors [biases] while conducting the study. Group matching Matching in pairs
  • 14.
    Matching Advantages Disadvantages May increasethe precision of case-control comparisons and thus allow a smaller study May be time-consuming and expensive to perform The sampling process is easy to understand and explain Some potential cases and controls may be excluded because matches cannot be made If analyzed correctly, provides reassurance that matched variables cannot explain case-control differences in the risk factor of interest The matched variables cannot be evaluated as risk factors in the study population
  • 15.
    Confounding is a distortionof results that occurs when the apparent effects of the exposure of interest are attributable entirely or in part to the effects of an extraneous variable. Example.age Confounding factors
  • 16.
  • 17.
    Measurement of exposure By Interviews Questionnaires Pastrecords Hospital records Employment records etc.
  • 18.
  • 19.
    Unmatched Case Control Datais expressed in a four- fold table, and an odds ratio is calculated Cases Controls Exposed a b Unexposed c d Odds Ratio = ad/bc
  • 20.
    Odds Ratio Odds ofexposure in cases: A / C Odds of exposure in controls: B / D Odds ratio: [A / C] / [B/ D)] Odds ratio: AD / BC
  • 21.
    Case Control E E EE E N E N E N E N N E N N N N Case Control Exposed Not exposed 6 3 4 7 Odds ratio = 6/4 =3.5 3/7 Odds ratio in unmatched case-control
  • 22.
    Odds Ratio inMatched Case Control Concordant pairs Both case and control were exposed Neither case nor control were exposed Discordant pairs Case exposed but control not exposed Case not exposed but control exposed
  • 23.
    Cont.. Odds Ratio =Ratio of Discordant pair = b/c Exposed Not exposed Exposed a b Not exposed c d Control Case
  • 24.
    Matched Case-Control Studyof Association Between Use of Oral Conjugated Estrogens and Cervical Cancer Estrogen Use Present Absent Total Present 12 43 55 Absent 7 121 128 Total 19 164 183 • OR=43/7=6.14
  • 25.
    Interpretation OR > 1indicates a positive association between the factor and the disease. OR < 1 indicates the factor is protective OR = 1 indicates no association
  • 26.
    Bias Confounding bias Memory bias Selectionbias Berkesonian bias Interviewer's bias
  • 27.
    Advantages Quick and Inexpensive Optimalfor rare diseases Useful for diseases of long latency from exposure to disease development Can evaluate multiple risk factors
  • 28.
    Disadvantages Cannot calculate incidence rate Notsuited for rare exposures May not elucidate temporal sequence of exposure and disease Susceptible to bias (selection & recall)
  • 29.
    Examples of CaseControl Study Doll’s 1951 study of smoking and lung cancer. Doll
  • 30.
    Male lung cancer patients Malepatients other diseases Smokers 647 (99.7%) 622 (95.8%) Nonsmokers 2 27 Total 649 649 OR = 647 x 27 = 14.04 622 x 2 Doll and Hill’s Data
  • 31.
    Thalidomide tragedy A retrospectivestudy of 46 mothers who delivered deformed babies were found to have Thalidomide during their early pregnancy.This was compared with a control of 300 mothers who delivered normal babies.
  • 32.
    Cont…. 1950’s Cigarette smoking andlung cancer 1970’s Diethyl stilbestrol and vaginal adenocarcinoma Post-menopausal estrogens and endometrial cancer
  • 33.
    Cont…. 1980’s  Aspirin andReyes syndrome  Tampon use and toxic shocks syndrome  L-tryptophan and eosinophilia- myalgia syndrome  AIDS and sexual practices 1990’s o Vaccine effectiveness o Diet and cancer
  • 35.
    Nested case-control studies arecase-control studies done in the population of an ongoing cohort study. The case-control study is thus said to be “nested” inside the cohort study. NESTED CASE CONTROL STUDIES
  • 36.
    TIME 1 YEARS TIME 2 StudyPopulation Develop Disease Do Not Develop Disease CASES CONTROLS CASE-CONTROL STUDY Obtain interviews, bloods, urines, etc. Nested Case-Control Studies
  • 37.
    How is itpossible to measure the exposure retroactively in a cohort study? Usually when the following situations occur: A specimen (e.g. of blood) had been stored but not analyzed. An interview is conducted to ask about at an exposure not assessed at baseline Records of exposure (medical records, occupational records) are retrieved that were not thought about when the cohort was first assembled NESTED CASE-CONTROL
  • 38.
    Analysis is justlike any case-control study with computation of an odds- ratio, adjustment for confounding, etc. NESTED CASE-CONTROL
  • 39.
    Example In ARIC (AtherosclerosisRisk in Communities) study, a cohort of 16 thousand men, all men provided serum samples at the outset which were saved. The cohort is observed for CHD. After 5 years we have 246 cases of CHD. We randomly choose 500 participants to be controls. We only measure Chlamydia antibody in the stored sera from these 246 + 500 subjects. We compare the cases (CHD) to the controls (no CHD) with regard to the presence of exposure (Chlamydia) which preceded the outcome.
  • 40.
    Reference K.Park-T.B of P.S.M.. Mahajan– T.B. of P.S.M W.H.O – Basic Epidemiology Oxford – T.B. of Public Health Mc.Graw Hill www.pubmed.gov www.cdc.gov www.who.int