ABSTRACT
This chapter discusses the use of Bayesian methods to incorporate real-world evidence to assess the safety of drugs. McCandless et al. considered Bayesian sensitivity analysis for the effect of an observed confounder when the exposure, response, and confounder are all binary variables. Meta-analysis is a suite of statistical methods for systematic literature review with the goal to aggregate and contrast findings from several related studies. If the basis of brodalumab causing suicide can be ruled out both scientifically and statistically, the observed suicidal risk in the brodalumab might be simply a false alarm. The chapter considers two case studies. The first one is using the Bayesian method to examine the impact of an unobserved confounder on the safety of spermicide. The second case is to assess whether a monoclonal antibody drug causes excess suicidal attempts for psoriasis patients.
