CMStatistics 2021: Start Registration
View Submission - CMStatistics
B1518
Title: A Bayesian hierarchical model for MedDRA coded adverse events in RCTs Authors:  Alma Revers - Amsterdam University (Netherlands) [presenting]
Michel Hof - Amsterdam university medical center (Netherlands)
Koos Zwinderman - Amsterdam university medical center (Netherlands)
Abstract: Patients participating in randomized controlled trials (RCTs) often report a wide range of different adverse events (AE) during the trial. MedDRA is a hierarchical standardization terminology to structure the AEs reported in an RCT. The lowest level in the hierarchy is a single medical event, and every higher level is the aggregation of the lower levels. Currently, the AE data of an RCT are often reported as crude rates or exposure-adjusted incidence rates. These rates have limited statistical power to detect rare AEs, leading to a high rate of false negatives. Therefore, we propose a hierarchical Bayesian model for identifying MedDRA coded AEs relative risks (RRs) in an RCT. We started by specifying a hierarchical Binomial and Poisson model, as done by others and extended the models to group the AEs to the complete hierarchy of the MedDRA. We developed our multi-stage hierarchical model, including the complete multiaxial MedDRA structure and developed a Bayesian algorithm to estimate posterior probabilities. We illustrate our model with AE-data from a large RCT ($n=2658$), and we compare results with other methods for analyzing AEs.