|
Activity Number:
|
422
- SPEED: Section on Bayesian Statistical Science
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 1, 2017 : 2:00 PM to 2:45 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
|
Abstract #325177
|
|
|
Title:
|
Adaptive Bayesian Modeling and Prediction of Patient Accrual with Varying Activation Time in Multicenter Clinical Trials
|
|
Author(s):
|
Junhao Liu* and Yu Jiang and Jo Wick and Byron Gajewski
|
|
Companies:
|
University of Kansas Medical Center and School of Public Health, University of Memphis and University of Kansas Medical Center and Department of Biostatistics, University of Kansas Medical Center
|
|
Keywords:
|
Multicenter clinical trials ;
Activation time ;
Bayesian Poisson-gamma model ;
Sites Decision
|
|
Abstract:
|
Understanding patterns of patient accrual is critical in multicenter clinical trials. We propose a Bayesian hierarchical model with varying activation time for each center to estimate the rate of patient accrual in the process of a multicenter clinical study. Activation time is defined as the amount of time elapsed between a center's activation (officially on study) and the enrollment of its first study patient. The difference in activation times between the centers is assumed to follow an exponential distribution. We evaluate model performance and precision of accrual prediction via simulation using non-informative and informative uniform and inverse-gamma priors respectively. The results show improved accuracy of prediction with informative priors when compared to non-informative priors. We also apply our proposed model to the PAIN-CONTRoLS study, a multicenter clinical trial with 40 centers. Overall, the Bayesian accrual model demonstrates power for prediction of patient recruitment while accounting for both center- and individual patient-level variation.
|
Authors who are presenting talks have a * after their name.
