Background: Active monitoring following COVID-19 vaccinations under the U.S. Food and Drug Administration Biologics Effectiveness and Safety (BEST) Initiative assesses potential associations between vaccine exposure and adverse events (AEs) in near-real time. As vaccine authorizations were expanded, this framework was adapted to monitor COVID-19 vaccine safety in the pediatric population.
Objectives: To conduct near-real time monitoring of AEs following BNT162b2 COVID-19 primary series and monovalent booster vaccinations in the U.S. population aged 5-17 years.
Methods: Active monitoring of 20 pre-specified AEs is ongoing among vaccinated individuals from three commercial claims databases with linkages to Immunization Information Systems registries: CVS Health, Optum, and HealthCore. Starting from the earliest date of emergency use authorization in the relevant age groups, we performed monthly sequential testing using the Poisson Maximized Sequential Probability Ratio Test to generate incidence rate ratios (IRR) of observed outcome rates compared to database-specific historical (expected) rates pre-COVID-19 vaccine availability. Exposure was defined as the receipt of a BNT162b2 COVID-19 vaccine, and doses were assigned chronologically. We stratified dose-specific analyses for primary series and monovalent booster doses by age categories: 5-11, 12-15, and 16-17. We are conducting medical record review for myocarditis/pericarditis (myo/peri) to assess case status identified by claims-based algorithms.
Results: Approximately 5.9 million BNT162b2 vaccine doses were administered to over three million commercial health plan enrollees aged 5-17 years. Based on data through 5/2022 (HealthCore, CVS Health) and 6/2022 (Optum), 12 of the 13 AEs with sufficient cases for sequential testing did not meet the threshold for a statistical signal in any of the three databases. Myo/peri (153 claims-based cases identified) signaled in all databases for ages 12-17 years following BNT162b2 primary series vaccinations. Following third/monovalent booster doses, myo/peri signals were detected in ages 16-17 years in CVS Health and ages 12-15 years in HealthCore databases. We observed no signals in ages 5-11 years in any of the databases. Of the medical records (n=37) reviewed for the myo/peri cases in ages 12-17 years, 27 (73.0%) were confirmed as true cases, of which 23 were male (92.6%).
Conclusions: Our near-real time monitoring identified signals following BNT162b2 vaccination for myo/peri in children aged 12-17 years. Results do not imply a causal effect, and IRRs do not necessarily represent the true magnitude of the vaccine association with AE risk. Signals may be further evaluated in more robust epidemiologic studies.
