Bias and Confounding

scientific presentation about the bias and confounding

1. Bias and Confounding
Dr. Faiza Abou El-Soud
Prof. of Community Health Nursing
Menoufiya University -Egypt

2. Learning Objectives
§Define bias, confounder and related terms
§ Discuss types of bias in the epidemiological
study
§ Elaborate methods of handling confounder

3. Definition of Bias
•“Any systemic error (design, data
collection, analysis or reporting of a study)
in epidemiological study that results in
incorrect the estimation of the association
between exposure and outcome”
•“Deviation of results or inferences from
the truth”

4. Properties of Measurement
•alidity (Accuracy): Validity is how much a
test measures what it is supposed to
measure.
•eliability (precision): Reliability is the
consistency of test results

5. Types of Bias
I.SELECTION BIAS
1.
2.
3.
II. INFORMATION BIAS
A.
B.
C.
D.

6. I. Selection Bias
• Occurs when the two groups being compared
differ systematically
• That is, there are differences in the characteristics
between those who are selected for a study and
those who are not selected
• Most common type of bias in health research
• Seen in observational and analytical studies

7. I. Sources of Selection Bias
1. Volunteer Bias
•Occurs when the participants select themselves for a study,
either because they are unwell or because they are
particularly worried about an exposure.
• , that people who respond to an invitation to
participate in a study on the effects of smoking in their
smoking habits non responders, the latter are usually
heavier smoker.

8. 2.Non-response Bias
•Occurs because individuals who do not respond to a call or
mailed questionnaire to participate in research studies are
generally from those who do respond.
• , smokers are less likely to return questionnaire
than are non-smokers

9. 3. Exclusion Bias
• Occurs when in certain circumstances epidemiologic studies
exclude participants to prevent confounding.
• , when exclusion criteria is different for cases and
control, or exposed and non-exposed

10. II. Information Bias
is
inappropriate and yields systemic errors in the
measurement of or
Affects
the nature of true association

11. Information Bias
•Occurs as a result of misclassification of
or status.
•For example,
•The figure below shows a two-by-two
contingency table in which apparent
that subjects are in the cell
of the contingency table, but there are
who have been misclassified and
are in an cell.

12. 1. Interviewer Bias (Abstract bias)
An interviewer’s knowledge may influence and the
, which may influence responses.
If an interviewer has a about the hypothesis being tested,
he or she might consciously or unconsciously interview case subjects differently
than control subjects.
If a reviewer believes that the research hypothesis was , the medical
record of a case subject might be looking at more thoroughly to find evidence of
exposure.
interviewers who believe that there is an association might
question case subjects more strictly in order to encourage cases to a past
exposure, while not prompting controls in the same way.

13. 2. Recall Bias
• , the subjects who are with a particular outcome or exposure
may more clearly.
, in case-control studies, this bias occurs when certain
information recalled by the (cases) compared with (controls).
Therefore, this missed of information, such as a potentially relevant
exposure, may be recalled by the (case) but forgotten by the (control).
•For example, congenital disorder remembers
every events clearly during pregnancy

14. 3.Reporting Bias
•Occurs when the participants can collaborate with researchers
and give answers in the directions they perceive are of
interest.
For example,the participant (either among the cases or among
the controls) may be to report an
exposure/event he is of because of
Therefore, this report bias
may affect the result.

15. 4.Surveillance Bias
• It is called or
•Occurs when the study group with known exposure or outcome may be
followed closely or longer the comparison group.
•It occurs when subjects in one exposure group are more likely to have
the study ourcome detected because they receive increased surveillance,
screening or testing as a result of having some other medical condition
for which they are being followed.
• obese patients are more likely to undergo medical
examinations, blood tests, and imaging studies than non-obese people.
If obese subjects were being compared to non-obese subjects

The early cancers reseraches would be more likely to be found in the
obese group, an of the .

16. 5. Withdrawal Bias ( )
• is other sources of bias that may found
some participants those are lost to follow up or who
withdraw from study may be different from those who are
followed in the study.

17. •This Hawthorne effect is other sources of bias that found
among some people act differently if they know they are
being watched.

•For example, One study was performed at a factory to
if change in lighting would affect
productivity.
•Therefore, productivity did increase but only because of
increased attention due to the study, as soon as study had
ended productivity decreased again.

18. Confounding
When another exposure exists in the study population (besides
the one being studied) and is associated both with disease and
the exposure being studied.
Confounder must be…….
1. Risk factor for the disease independently
2. Associated with exposure under study
3. It is not casual pathway between exposure and disease
For example,
age , sex ,living condition

19. Example:

20. Example:

21. Control of Confounding
At design stage:
•-Restriction
•-Matching
•-Randomisation
At analysis stage
• -Stratification
•-Multivariate analysis

22. Restriction
•Subject chosen for study are restricted to only those
possessing a narrow of characteristics , to equalize
important extraneous factors.
For example,
•Restrict study to women having

23. Matching
•For each patient in study group there is one or more patients
in comparison group with same characteristics, for the
factor of interest.
For example,
•Matching done for age ,sex ,race etc.

24. Randomization
•Subjects of study are randomly selected to even out unknown
confounders.

25. •The process of separating a sample into several sub-samples
according to specified criteria such as age group ,
socioeconomic status etc.
Stratification

26. The statistical analysis of data collected on more than one
variable.
For example,
People age ,weight, body fats . Skull length, width and cranial
capacity

27. Thank You