Sequential FDR and pFDR control under arbitrary dependence, with application to pharmacovigilance database monitoring

TL;DR

This paper introduces new sequential testing procedures that control FDR and pFDR under any dependence structure, optimizing error bounds and demonstrating efficiency gains in pharmacovigilance data analysis.

Contribution

It develops novel sequential multiple testing methods that handle arbitrary dependence, controlling error rates and reducing sample sizes compared to traditional fixed-sample approaches.

Findings

Achieved 45-65% reduction in average sample size.

Controlled FDR and pFDR under arbitrary dependence.

Successfully applied methods to pharmacovigilance data.

Abstract

We propose sequential multiple testing procedures which control the false discover rate (FDR) or the positive false discovery rate (pFDR) under arbitrary dependence between the data streams. This is accomplished by "optimizing" an upper bound on these error metrics for a class of step down sequential testing procedures. Both open-ended and truncated versions of these sequential procedures are given, both being able to control both the type~I multiple testing metric (FDR or pFDR) at specified levels, and the former being able to control both the type I and type II (e.g., FDR and the false nondiscovery rate, FNR). In simulation studies, these procedures provide 45-65% savings in average sample size over their fixed-sample competitors. We illustrate our procedures on drug data from the United Kingdom's Yellow Card Pharmacovigilance Database.