Interim futility analysis with intermediate endpoints.

Publication Type:

Journal Article

Source:

Clinical trials (London, England), Volume 5, Issue 1, p.14-22 (2008)

Keywords:

2008, Clinical Trials, Phase III as Topic, Computer Simulation, Decision Making, Endpoint Determination, Humans, Medical Futility, Neoplasms, Public Health Sciences Division, Research Design, Survival Analysis

Abstract:

Interim analysis of Phase III trials typically includes testing for both efficacy and futility. Futility testing is commonly performed on the primary outcome at very low levels (e.g., one-sided alpha=0.0025) at one or two times before final analysis. When overall survival is the primary outcome and events accrue slowly, and if a suitable intermediate endpoint is available, then using this endpoint for interim futility testing may yield a higher probability of stopping early for futility in the absence of any treatment effect.