Reducing Confounding During Study Design

Lead Author(s): Jeff Martin, MD

Methods for Reducing Confounding During Study Design

The causal diagram below presents ways to prevent or manage confounding during study design.

0402_3dags.JPG

There are three main ways to reduce confounding during study design:

1. Randomization

Definition of Randomization: This will block all association between the confounder and the exposure:

0402_dag_ran.JPG

2. Restriction

Definition of Restriction: Restrict enrollment to only those subjects who have a specific value/range of the confounding variable

This will eliminate: 0402_4restrict.JPG

3. Matching

Definition of Matching: Cohort Study: unexposed subjects are matched to exposed subjects according to their values for the potential confounder. 0407_3amatch.JPG

Case Control Study: non-diseased controls are matched to diseased cases on the confounding variable. 3bmatch.JPG