ABSTRACT
There are several advantages of the use of medical claims data for postmarketing surveillance. First, they represent person-level data, similar to RCTs and spontaneous reports, but unlike spontaneous reports, we know the population at risk. Second, several medical claims databases such as the Veterans Administration (VA) or PharMetrics databases contain longitudinal information on adverse events, concomitant medications, and comorbid diagnoses both before and after the drug exposure. Third, the populations that can be sampled are often large enough to study even the rarest of events such as suicide attempts and completion, whereas meta-analyses of RCTs are often restricted to more distal measures such as suicidal ideation, which may have little to do with suicide completion, the adverse event of interest. The primary limitation of medical claims data is that they are observational, and any association identified may or may not represent a causal link between the drug and the adverse event. The primary objective is to design and analyze an observational study such that many of the benefits of a randomized study are preserved. In the following, we describe some of the experimental strategies relevant to analysis of observational pharmacoepidemiologic data and the associated statistical methods that can be used in conjunction with such designs.
