A meta-analysis on the effectiveness of a new weight-loss medication finds that small trials with large benefits are common, while larger trials show smaller effects. The authors did not search clinical trial registries and did not assess for missing negative studies. What type of bias could affect the validity of this meta-analysis?
- Publication bias leading to an exaggerated pooled treatment effect (correct answer)
- Attrition bias because meta-analyses lose participants during follow-up
- Recall bias because weight is self-reported in all trials
- Selection bias because randomization was not used in meta-analyses
Explanation: This question tests publication bias in meta-analyses without registry searches. The specific bias is publication bias, as missing negative studies exaggerate the weight-loss effect, especially with small positive trials. Choice A is the best answer because it accurately identifies how this affects the pooled estimate. Choice B is incorrect as attrition is a primary study issue; choice C is wrong because weight measurement varies but recall is not specified; choice D is suboptimal as selection bias applies to participants, not studies. Assess funnel plots and registries to detect publication bias. In pharmacy practice, this promotes critical appraisal of weight-loss drug evidence for realistic patient expectations.