Evidence Synthesis Gone Awry: The Perils of Aggregating Ineffective or Unsafe Doses in Alopecia Areata Reviews

The document discusses the risks of network and conventional meta-analyses in alopecia areata reviews when they include ineffective, unsafe, or unapproved dosing regimens. It highlights how pooling such regimens with licensed therapies can distort efficacy and safety estimates, leading to misleading rankings and inappropriate causal inferences. The authors use examples like deuruxolitinib and ritlecitinib to illustrate these issues. They recommend methodological safeguards, such as excluding unapproved doses, transparently reporting approval status, avoiding causal attributions from small strata, and conducting sensitivity analyses. These practices aim to maintain the benefits of meta-analysis while preventing misleading conclusions.