In the forest plot accompanying a meta-analysis of antidiabetic drugs, the smaller studies cluster on one side of the pooled estimate while larger studies cluster symmetrically near it, producing an asymmetric inverted funnel shape. This finding most strongly suggests:
- A High statistical heterogeneity requiring a fixed effects model
- B An error in weighting of individual studies
- C Publication bias with selective non-publication of small negative studies ✓
- D Excessive precision of the pooled estimate
Correct answer: C. Publication bias with selective non-publication of small negative studies
Explanation
A symmetric inverted funnel plot suggests absence of publication bias because precision increases with study size. Asymmetry, where small studies scatter away from the pooled estimate on one side, classically reflects that small studies with negative or unfavourable results remained unpublished. Heterogeneity is assessed by Cochran Q and I squared statistics, not funnel shape, and the plot does not indicate any weighting miscalculation.
Reference: Gordis Epidemiology, 6th ed.
High-yield for: NEET PGINI-CETNExTFMGEUSMLEPLABMRCP
Written and medically reviewed by the StethoPrep medical team.