In a forest plot of a meta-analysis examining a new antidiabetic drug, small trials cluster symmetrically at the base showing large effects while larger trials sit near the top close to the null value, producing a visibly skewed plot. The most likely explanation for this asymmetry is:
- A Excessive statistical heterogeneity requiring a random-effects model
- B Publication bias, with small negative trials remaining unpublished ✓
- C Ecological fallacy arising from pooling individual patient data
- D Regression dilution bias from measurement error in HbA1c
Correct answer: B. Publication bias, with small negative trials remaining unpublished
Explanation
The funnel plot plots study precision against effect size; in its absence of bias, points scatter symmetrically around the pooled estimate. Small studies showing spuriously large effects with larger ones clustering near the truth produces the classic asymmetric funnel, suggesting small negative or null trials went unpublished. Heterogeneity is assessed by Cochran Q and I squared statistics, not by funnel shape, and ecological fallacy does not apply to trial-level pooling.
Reference: Park's Textbook of Preventive and Social Medicine, 27th ed.
High-yield for: NEET PGINI-CETNExTFMGEUSMLEPLABMRCP
Written and medically reviewed by the StethoPrep medical team.