A systematic review pooling five randomized controlled trials on a new antihypertensive shows an I² statistic of 75%. The most appropriate next step in the meta-analysis is:

• A Use a fixed-effects model because the overall sample size is large
• B Exclude the two studies contributing most to heterogeneity
• C Report the pooled estimate without concern as I² < 90%
• D Use a random-effects model and explore sources of heterogeneity ✓

Explanation

An I² of 75% indicates substantial heterogeneity (I² 25-50% = low, 50-75% = moderate, >75% = considerable by Cochrane classification; some sources classify >50% as substantial). With high heterogeneity, a random-effects model is more appropriate than fixed-effects, as it accounts for between-study variance. Exploring sources through subgroup analysis or meta-regression is mandated before drawing conclusions. Arbitrarily excluding studies introduces bias, and ignoring heterogeneity would produce misleading pooled estimates.

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.