Transcript Introduction of Epidemiology

Case-control study
Chihaya Koriyama
August 17 (Lecture 1)
Study design in epidemiology
Observational
study
individual
Case-control
study
intervention
population
Cohort
study
Ecological
study
Why case-control study?
• In a cohort study, you need a large number
of the subjects to obtain a sufficient number
of case, especially if you are interested in a
rare disease.
– Gastric cancer incidence in Japanese male:
128.5 / 100,000 person year
• A case-control study is more efficient in
terms of study operation, time, and cost.
Case-control study - subjects
• Start with identifying the cases of your
research interest.
– If you can identify the cases systematically,
such as a cancer registration, that would be
better.
– Incident cases (newly diagnosed cases) are
better than prevalent cases (=survivors).
• Recruitment of appropriate controls
– From residents, patients with other
disease(s), cohort members who do not
develop the disease yet.
Various types of case-control studies
１）a population-based case-control study
Both cases and controls are recruited from the
population.
２）a case-control study nested in a cohort
Both case and controls are members of the cohort.
３）a hospital-based case-control study
Both case and controls are patients who are
hospitalized or outpatients.
Who will be controls?
• Control
≠
non-case
– Controls are also at risk of the disease
in his(her) future.
– In a case-control study of gastric
cancer, a person who has received the
gastrectomy cannot be a control.
– In a case-control study of car accident,
a person who does not drive a car
cannot be a control.
Case-control study - information
• Collection of the information (past information)
by interview, biomarkers, or medical records
– Exposure (your main interest)
– Potential confounding factors
• Bias & Confounding
– Selection bias
– Information bias (recall bias)
– confounding
Selection bias
 Sampling is required in a case-control
study (since we cannot examine all!)
 We need to chose appropriate subjects.
Selection bias is “Selection of cases and
controls in a way that is related to exposure
leads to distortions of exposure prevalence”.
Error & Bias
• Error: random error
• Bias：systematic error
– differential misclassification
This is a problem!
– non-differential misclassification
An example of non-differential
misclassification in an exposure variable
 We want to compare mean of blood
pressure levels between cases and
controls.
 The blood pressure checker has a
problem and always gives 5mmHghigher than true values.
 All subjects were examined by the
same blood pressure checker.
→ no problem for internal
comparison
An example of non-differential
misclassification in the
ascertainment of exposure
True (nobody
knows)
Results of
test＊
Exp +
Exp -
Observed risk
estimate always
comes close to
“1(null)”
Case
Control Odds ratio
1 50 10 9 5010
(50*90) /
(50*10) =9
10
90
Exp +
41
49
Exp -
19
91
(41*91) /
(49*19)=4.01
＊Sensitivity 80% (80% of the exposed subjects are correctly
diagnosed)
Specificity 90% (90% of the un-exposed subjects are
correctly diagnosed)
Differential misclassification
• Selection bias
• Detection bias
• Information bias
– Recall bias
– Family information bias
Confounding
 Confounders are risk factors for
the outcome.
 Confounders are related to
exposure of your interest.
 Confounders are NOT in the
process of causal relationship
between the exposure and the
outcome of your interest.
Example of confounder
- living in a HBRA is a confounder HBRA: high background radiation area
High infant death
Causation ?
Exposure to
radiation in uterus
Low
socio-economical
A surrogate
marker of low
status
in HBRA
socio-economic
status
Living in a
HBRA
Example of confounder
- smoking is a confounder Myocardial
infarction
Causation ?
(We observe an association)
Radiation
Smoking is a risk factor of MI
smoking
related by chance
Example of “not” confounder
- pineal hormone is not a confounder EMF: electro-magnetic field
Breast cancer
Causation ?
EMF
Decrease of pineal hormone
may be the risk of breast ca.
Down regulation
of pineal hormone
EMF exposure induces down
regulation of pineal hormone
If EMF exposure cause breast cancer only through down regulation of
pineal hormone, this is not a confounder.
Why do we have to consider
confounding?
We want to know the “real”
causal association but a
distorted relationship
remains if you do not adjust
for the effects of
confounding factors.
How can we solve the problem of
confounding?
“Prevention” at study design
Limitation
Randomization in an
intervention study
Matching in a cohort study But
not in a case-control study
How can we solve the problem of
confounding?
“Treatment “ at statistical analysis
Stratification by a confounder
Multivariate analysis
Case ascertainment
• Who is your case?
– Patient?
– Deceased person?
• What is the definition of the case?
– Cancer (clinically? Pathologically?)
– Virus carriers (Asymptomatic patients)
→ You need to screen the antibody
Incident or Prevalent cases with
chronic disease(s)
Incident case
• You recruit cases
prospectively.
Prevalent case
• You recruit cases crosssectionaly.
• Newly diagnosed cases
• Mixed cases with
diagnosed recently and
long time ago.
• You miss patients who
died before study.
• All cases are alive.
– Only survivors
Cases with better prognosis!
Matching in a case-control study
• Matched by confounding
factor(s)
– Sex, age ・・・・
• Cannot control confounding
– Conditional logistic analysis is
required.
• To increase the efficiency of
statistical analysis
Over matching
• Matched by factor(s) strongly
related to the exposure which is
your main interest
– CANNOT see the difference in
the exposure status between
cases and controls
a case-control study
Cases
Controls
(brain tumor)
N=100
N=100
The incubation period
of tumor is a few years
at least.
Mobile phone users (NOT recently started)
↓
↓
50
10
Yes
Cases
50
Controls
10
No
50
90
Risk measure in a case-control study
Odds = prevalence / (1－ prevalence)
Odds ratio = odds in cases / odds in controls
Disease
+（case）
－（control）
+
a
c
Exposure －
b
d
Exposure odds in cases ＝a / b
Exposure odds in controls＝c / d
Odds ratio＝(a / b) / (c / d) ＝ a * d / b * c
Comparison of the study design
Case-control
Rare diseases
suitable
Number of disease
1
Sample size
relatively small
Control selection difficult
Study period
relatively short
Recall bias
yes
Risk difference
no available
Cohort
not suitable
1<
need to be large
easier
long
no
available