Subgroup Analyses Not a Tool to Prove Efficacy of a Drug

Subgroup, or post-hoc, analyses of clinical trial results can be useful, especially if planned prospectively, but sponsors cannot use them to prove efficacy, except in the rarest of exceptions, an FDA official says. “These analyses cannot allow a drug that failed overall to be considered effective, but they can show consistency, possible inconsistency and are useful, but must be used very cautiously,” Robert Temple, deputy director for clinical science at CDER, said.
Clinical Trials Advisor