Example: Do Not Adjust - Lung Cancer and Matches

Lead Author(s): Jeff Martin, MD

Crude and Adjusted Data

Here is an example of what can happen when you adjust for something you don’t need to adjust for.
0417_4lung.JPG

Comparing Results

When we look at the effect of matches on the association between smoking and lung cancer.

In other words, matches had no effect on the association.

Reporting

In this case, we would report the crude estimate, only right?

Loss of Precision

In other words, the confidence interval of the crude estimate will be narrower than the CI of the adjusted measure - they both will have 21 as their point estimate but the crude association will be more precise. Hence, this illustrates why you don’t want to adjust for things that you don’t need to.