Transcript Fig 15-1

Ch 15 Bias, Confounding, and
Interaction
Bias
“Any systematic error in the design,
conduct or analysis of a study that
results in a mistaken estimate of an
exposure's effect on the risk of disease."
1
Selection bias



In reality, exposure and disease are not
associated
However, the way in which individuals
were selected is such that an apparent
association is observed
The apparent association is the result of
selection bias.
2
Study by Ronmark et al. (1999)




A prevalence study of asthma, chronic
bronchitis, and respiratory symptoms
Studied the characteristics of nonresponders
and the reasons for nonresponse
9,132 people living in Sweden were invited to
participate
Data were obtained by a mailed questionnaire,
and the response rate was 85%
3
Study by Ronmark et al. (1999)



A sample of nonresponders was contacted by
telephone
Found a significantly higher proportion of
current smokers and manual laborers among
the nonresponders than among the responders
The prevalence rates of wheezing, chronic
cough, sputum production, attacks of
breathlessness, and asthma and use of
asthma medications were significantly higher
among the nonresponders
4
Selection bias


Nonresponse may introduce a serious
bias that may be difficult to assess
It is important to keep nonresponse to a
minimum.
5
Information bias


Can occur when the means for obtaining
information about the subjects are
inadequate
As a result some of the information
gathered regarding exposures and/or
disease outcome is incorrect
6
Information bias
Misclassification bias
 Example: In a case-control study, some people
who have the disease (cases) may be
misclassified as controls, and some without the
disease (controls) may be misclassified as
cases
 This may result, for example, from limited
sensitivity and specificity of the diagnostic tests
involved or from inadequacy of information
derived from medical or other records
7
Information bias
Misclassification bias
 If exposure data are based on
interviews, for example, subjects may
either not be aware of their exposure or
may erroneously think that it did not
occur
 If ascertainment of exposure is based on
old records, data may be incomplete
8
Confounding

In a study of whether factor A is a cause
of disease B, we say that a third factor,
factor X is a confounder if the following
are true:
• Factor X is a known risk factor for disease B
• Factor X is associated with factor A, but is not
a result of factor A
9
Confounding
Example (in Chap 10)
 The relationship between coffee and
cancer of the pancreas
 Found an apparent dose-response
relationship between coffee and cancer
of the pancreas, particularly in women
 It was rare to find a smoker who does
not drink coffee
10
Confounding
Example (in Chap 10)
 Smoking was a confounder, because although
we were interested in a possible relationship
between coffee consumption (factor A) and
pancreatic cancer (disease B), the following
are true of smoking (factor X):
•
•
Smoking is a known risk factor for pancreatic cancer
Smoking is associated with coffee drinking, but is not
a result of coffee drinking
11
Confounding

So if an association is observed
between coffee drinking and cancer of
the pancreas, it may be
(1) that coffee actually causes cancer of the
pancreas, or
(2) that the observed association of coffee
drinking and cancer of the pancreas may be
a result of confounding by cigarette smoking
12
Confounding

We observe the association of coffee
drinking and pancreatic cancer
because cigarette smoking is a risk
factor for pancreatic cancer and
cigarette smoking is associated with
coffee drinking (Fig 15-1)
13
Confounding

When we observe an association
• we ask whether it is causal (Fig. 15-1A) or
• whether it is a result of confounding by a third
factor that is both a risk factor for the disease
and is associated with the exposure in
question (see Fig. 15-1B).
14