All Times EDT
Keywords: safety, Bayesian method, risk ratio, absolute risk, log-regression
Pre-market safety as presented in regulatory submission is usually integrated across studies to gain more precision in the assessments of serious adverse events that suggest important problems with a drug--specifically and assessment of common (usually non-serious) adverse events that are, or may be, causally related to the use of the drug. While crude metrics, such as incidence proportions, are typical for presenting rates of these AEs, models are increasingly used to estimate them to adjust for attendant risks and exposures appropriately. We offer a Bayesian approach for estimating risk ratios in contradistinction to estimating absolute risk difference. The rationale for the use of Bayesian methodology is to characterize tolerable uncertainty with accumulating data in the context of benefit and unmet need.