|
Abstract:
|
In clinical trials, AEs may occur repeatedly over the course of follow-up. Evaluation of first event alone often misses large amounts of potentially important data and may produce different results than evaluating of all event recurrences. Furthermore, the conventional summary and analysis methods based on frequency and count may be invalid and misleading for long-term trials. Mean function has been widely used to analyze recurrent event data. Two statistical issues arise in applying mean function for safety monitoring of recurrent adverse events. First, patient entry is group sequential in most randomized clinical trials. Sample size is small at early time of the trial and the incidence of adverse event is infrequent. Second, the background rate for exposure-adjusted analysis is time-dependent function. The sample estimator is step-wise function, provides poor visualization of the true function. Smooth function is needed to quantify population safety profile. In this research, we describe a Bayesian method for comparing mean function of recurrent adverse events between treatment groups.
|
