Measurement Error in the Timing of Events: Effect on Survival Analyses in Randomized Clinical Trials

• Activity Number: 71
• Type: Topic Contributed
• Date/Time: Sunday, August 4, 2013 : 4:00 PM to 5:50 PM
• Sponsor: ENAR
• Author(s): Lori Dodd*+
• Companies: National Institutes of Health
• Keywords: Progression-free survival ; Measurement error ; time-to-event endpoint

Abstract

Variability of progressive disease measurement in oncology trials is an established phenomenon. True times of progression cannot be known with the same precision as other event times, such as time of death. The repercussions of this on the interpretation of phase III clinical trials are complex, and depend on several factors, including the magnitude of the variability, and whether there is a systematic reader evaluation bias favoring one treatment arm. To investigate the effects of non-differential measurement error in time-to-event outcomes on the conclusions of a proportional hazards analysis of a randomized clinical trial, we performed simulation studies. Differential measurement error is briefly considered. In some simulations, the measurement error structure was approximated based on data from a phase III clinical trial; for other simulations, larger or smaller measurement error was considered. The bias in estimating the hazard ratio was very small using parameters similar to the clinical trial. Even with a larger non-differential measurement error, the bias remained small. There was no or little effect on the variability of the estimated treatment effect.