While assessing a meta-analysis, a reviewer plots the standardized effect size of each trial against its standard error. Small trials cluster symmetrically around the pooled estimate while several larger trials are conspicuously absent from the lower part of the plot. What does this pattern suggest?

• A Overmatching
• B Effect modification
• C Ecological fallacy
• D Publication bias ✓

Explanation

D funnel plot should show smaller studies scattered widely and symmetrically on both sides of the pooled effect, with larger studies clustering tightly near it. Asymmetry, especially absence of small studies showing no effect, suggests that negative or null trials were not published and that the pooled estimate may be inflated. Effect modification and overmatching are analytic issues unrelated to the distribution of trial estimates.

Reference: Cochrane Handbook for Systematic Reviews of Interventions, Version 6 ed.

High-yield for: NEET PGINI-CETNExTFMGEUSMLEPLABMRCP

Written and medically reviewed by the StethoPrep medical team.