Abstract:
|
Patient recruitment is critical to the success of clinical trials. Slow patient accrual may lead to research delays, cost increases, or low statistical power. Our previously developed Bayesian method integrated researchers' clinical trials experience, which provides a reliable prediction model for the accrual rate for a single site participating in a clinical trial. In this current study, we proposed a Bayesian hierarchical model using adaptive priors to incorporate historical enrollment information for specific sites. Our model captures the specific characters of the each site, as well as correlations between different sits, therefore increases the reliability of a patient accrual prediction model. The performance of the models are evaluated using both simulations and actual multicenter clinical studies. The results showed that correlated model are more robust in the prediction of accrual process than independent model, with smaller MSE and higher percentages of correct coverage. Overall, prediction and modeling of patient accrual prior to initiating a trial or during a trial will help researchers, sponsors, and data monitoring committee to better evaluate and manage an ongoing st
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.