Penalized Logistic Regression Likelihood Ratio Test Analysis to Detect Signals of Adverse Events From Interactions in Postmarket Safety Surveillance
While administration of a vaccine may be associated with a particular adverse event (AE), a vaccine interaction adverse event (VIAE) occurs when the relationship between administration of one vaccine and an AE is influenced by the administration of an additional vaccine or vaccines in a nonadditive manner. In clinical trials, it is often challenging to detect AEs from vaccine or drug interactions due to limited sample size and limited comparison of treatment groups which can result in AEs not being detected until the postmarket stage. The Vaccine Adverse Event Reporting System (VAERS) is a national vaccine safety surveillance program co-sponsored by the Centers for Disease Control and Prevention (CDC) and the Food and Drug Administration (FDA). The VAERS database contains reports of adverse events associated with immunization, and disproportionality analyses can be used to explore vaccine interaction adverse events (VIAEs). In this article, we develop a penalized logistic regression-based likelihood ratio test for detecting data mining signals due to interactions, and we contrast and compare our method with other methods for exploring interactions in passive surveillance systems such as VAERS. We apply our procedure to well-known safety profiles to examine its performance in detecting potential VIAEs, and we further evaluate our method with a simulation study.