A meta-analysis of 8 RCTs evaluating aspirin for prevention of myocardial infarction shows significant heterogeneity with I² = 75%. Which of the following is the MOST appropriate statistical model and interpretation?

• A Fixed-effects model; the heterogeneity is due to random sampling variation
• B Random-effects model; there is substantial heterogeneity suggesting true between-study variability ✓
• C Fixed-effects model; I² above 50% indicates publication bias only
• D Random-effects model; I² of 75% means 75% of studies have conflicting results

Explanation

I² quantifies the proportion of total variation in effect estimates attributable to true heterogeneity rather than chance; values of 75% indicate substantial heterogeneity. When I² exceeds 50–60%, a random-effects model is preferred because it accounts for between-study variance (τ²) and provides wider, more honest confidence intervals. I² is not equivalent to the proportion of conflicting studies and is not a direct measure of publication bias — funnel plot asymmetry assesses publication bias.

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.