Evaluation of spontaneous reports of adverse events following the administration of CBER-regulated medical products remains a key component of CBER’s safety surveillance strategy. Continuing to improve the efficiency of analysis of this data and the validity of inferences drawn from it are important goals of CBER’s work in this area. CBER conducts disproportionality analysis (i.e., commonly referred to as data mining) of spontaneous reports for CBER regulated products in the Vaccine Adverse Event Reporting System (VAERS) and FDA Adverse Event Reporting System (FAERS) using Empirical Bayesian methods.
Data mining in the context of product safety surveillance refers to the use of statistical or mathematical tools to discover patterns of associations or unexpected occurrences in large databases, such as FAERS and VAERS. It involves the systematic examination of reported adverse events (AEs) to identify an excess of AEs reported for a product relative to other products in the database (disproportionality). Results from data mining are considered hypothesis generating and do not, by themselves, demonstrate causal associations. By applying data mining techniques, FDA may be able to identify unusual or unexpected product-event combinations that warrant further investigation.
Potential limitations of data mining include spurious alerts from statistical interaction (e.g., due to concomitant exposures), dictionary constraints, confounding (e.g., by indication), and limitations inherent to passive reporting systems (e.g., stimulated reporting impacts, variable spontaneous reporting of different exposures and outcomes). Unexpectedly high reporting associations may generate a hypothesis that there may be an association between the AE and the product. However, the absence of disproportionality does not confirm the absence of a safety signal nor negate a signal detected by other methods.