Weighted false discovery rate controlling procedures for clinical trials.

Area of Science:

• Biostatistics
• Clinical Trials
• Medical Research

Background:

• Replicability in early-phase clinical research is often compromised by selective inference.
• Existing methods struggle to adequately control for multiplicity in complex trial designs.

Purpose of the Study:

• To develop novel statistical procedures for controlling false discoveries in clinical research.
• To enhance the power of hypothesis testing for both primary and secondary endpoints.

Main Methods:

• Introduction of a hierarchical weighted false discovery rate (FDR) controlling procedure.
• Development of a single-level weighted FDR procedure tailored for clinical research.
• Assigning weights to reflect endpoint importance in error control.

Main Results:

• The proposed hierarchical and single-level weighted FDR procedures offer higher power than current gatekeeper methods.
• The hierarchical method tests primary endpoints and a weighted intersection hypothesis for secondary endpoints.
• Individual secondary endpoints are tested if the intersection hypothesis is rejected.

Conclusions:

• The new weighted FDR procedures improve power for discovering effects on secondary endpoints.
• These methods maintain statistical rigor while addressing multiplicity in clinical trials.
• The proposed procedures offer an advantage over existing familywise error-rate controlling methods for drug approval.