A systematic review meta-analysis on antihypertensive therapy shows significant heterogeneity (I² = 78%, p < 0.001). The MOST appropriate statistical model to pool the effect estimates is:
- A Fixed-effects model, as it is more conservative
- B Random-effects model, as it accounts for between-study variance ✓
- C Mantel-Haenszel method, as it requires no assumptions about heterogeneity
- D Meta-regression, which directly corrects heterogeneity before pooling
Correct answer: B. Random-effects model, as it accounts for between-study variance
Explanation
An I² of 78% indicates substantial heterogeneity among studies. The random-effects model (DerSimonian-Laird) incorporates between-study variance (tau²) in addition to within-study variance, providing wider confidence intervals that better reflect true uncertainty when studies differ in populations, interventions, or outcomes. The fixed-effects model assumes all studies estimate one true common effect — inappropriate when heterogeneity is high. Meta-regression explores sources of heterogeneity but does not replace the choice of pooling model.
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.