Performance of Constant Accrual Model and Alternatives on Clinical Data and Simulation

• Activity Number: 414
• Type: Contributed
• Date/Time: Tuesday, August 5, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Yu Jiang*+ and Byron Gajewski and Steve Simon and Matthew S. Mayo
• Companies: University of Kansas Medical Center and University of Kansas Medical Center and P. Mean Consulting/University of Missouri-Kansas City and Kansas University Medical Center
• Keywords: Subject accrual ; clinical trials ; hedging prior ; Bayesian hierarchical model ; accelerated prior

Abstract

Slow recruitment in medical research leads to increased costs and resource utilization, including the goodwill contribution of patient volunteers. Careful monitoring of the accrual process is important. We propose two extensions, the accelerated prior and the hedging prior, to the Bayesian constant accrual model, which utilizes the researcher's previous experience and current accrual data to predict trial completion time. The performance of these models, including prediction precision, coverage probability, and correct decision-making ability, is evaluated using actual studies and simulation. The results showed that a constant accrual model with strongly informative priors works very well when accrual is on target or slightly off, but strongly biased when off target. Flat or weakly informative priors provide protection against on off target trial, but are less efficient when the accrual is on target. The accelerated prior performs similar to strong priors. The hedging prior performs much like weak priors when the accrual is off target, but closer to the strong priors when the accrual is on target or slightly off. Improvements in these models for future research are suggested.