skip to main content

Trial Design — SCE Medical Oncology MCQ

Instant feedback + full explanation. One question, done properly.

HardClinical Trials & StatisticsTrial DesignSCE Medical Oncology

A researcher is designing a phase III trial comparing a new immunotherapy combination against standard chemotherapy for first-line treatment of advanced NSCLC. The trial has co-primary endpoints of PFS and OS. What statistical consideration is required when testing two primary endpoints?

Educational content. Not a substitute for clinical judgement or local policy.

Reveal the answer and explanation

Correct answer: A — Pre-specify a multiplicity strategy controlling the family-wise type-I error

Two confirmatory primary endpoints require a prospective gatekeeping, alpha-splitting or equivalent procedure that controls family-wise type-I error. Pre-specify a Bayesian prior controlling the per-protocol treatment effect: a prior alone does not control frequentist family-wise error across co-primary tests. Pre-specify a subgroup hierarchy controlling the false-discovery proportion: subgroup multiplicity is a different problem from two confirmatory primary endpoints. Pre-specify an interim boundary controlling conditional statistical power: interim stopping boundaries do not themselves allocate alpha across endpoints. Pre-specify a non-inferiority margin controlling assay sensitivity: a non-inferiority margin addresses a different trial objective.

Reference: Multiplicity considerations for confirmatory clinical trials (Published December 2016): https://pubmed.ncbi.nlm.nih.gov/28025257/; GMC good practice in research and consent (Effective April 2024): https://www.gmc-uk.org/professional-standards/the-professional-standards/good-practice-in-research-and-consent/good-practice-in-research