In a case-control study of oral contraceptive use and venous thromboembolism, investigators match controls to cases on the use of any hormonal medication in the preceding year. The resulting odds ratio is close to 1 despite a known strong association in prior literature. This loss of a detectable association is best explained by:

• A Recall bias introduced by the matching variable
• B Overmatching on a variable closely linked to the exposure ✓
• C Berkson's bias from hospital-based control selection
• D Differential misclassification of the outcome

Explanation

Overmatching occurs when the matching variable is itself strongly associated with, or lies on the pathway to, the exposure of interest. By forcing cases and controls to be similar on such a variable, real differences in exposure between them are erased and the odds ratio is biased toward the null. Berkson's bias concerns admission rate differences in hospital studies, and neither recall nor outcome misclassification explains systematic null findings.

Reference: Gordis Epidemiology, 6th ed.

High-yield for: NEET PGINI-CETNExTFMGEUSMLEPLABMRCP

Written and medically reviewed by the StethoPrep medical team.