In a meta-analysis of trials comparing a new antidiabetic drug with placebo, the funnel plot shows most small trials scattered widely to the right of the pooled estimate while large trials cluster symmetrically near it. The most likely explanation for this pattern is:
- A Reverse causation
- B Publication bias favouring positive small trials ✓
- C Effect modification by trial size
- D Random error alone, since spread widens naturally in large trials
Correct answer: B. Publication bias favouring positive small trials
Explanation
B symmetrical inverted funnel is expected when spread reflects sampling variation alone, with imprecision increasing as trial size decreases. Asymmetry in which small trials report systematically larger effects suggests that small negative or null trials went unpublished, inflating the pooled estimate. Option D states the opposite of the statistical truth since precision improves, not worsens, with larger samples. Reverse causation is a temporal problem in observational studies and irrelevant to trial pooling.
Reference: Systematic Reviews in Health Care: Meta-analysis in Context (Egger), 2nd ed.
High-yield for: NEET PGINI-CETNExTFMGEUSMLEPLABMRCP
Written and medically reviewed by the StethoPrep medical team.