Background: As of December 2022, 1.59 billion doses of mRNA-1273 have been distributed to 91 countries. Spontaneous reports of adverse events (AE) from consumers, care givers or healthcare workers from countries with mRNA-1273 approval are captured in the Moderna global safety database (GSD). As of December 2022, >2.5 million AE were reported in GSD, enabling open-ended hypothesis generation. The incidence of AE cannot be calculated from the database due to lack of precise exposure data; instead, large cohort studies are required. During the COVID-19 vaccine campaign, many countries posted vaccine uptake counts allowing improved estimation of AE reporting rates. Whether these improvements enhanced the quality of observed to expected (OE) analyses remains an open question.
Objectives: To assess precision and accuracy of OE analyses of GSD compared to results from a cohort study in the United States.
Methods: As part of routine pharmacovigilance, Moderna performed OE analyses for AE of special interest (AESI) identified by regulatory agencies and vaccine experts. Reporting rates were estimated by dividing cases in the GSD by person-time at risk (21 days after each dose based on estimated global exposure to mRNA-1273). The observed reporting rates were then compared to background incidence obtained from published literature and assumed applicable to the vaccinated population. The rate ratio and 95% confidence interval (CI) were then calculated stratified by age by sex. Sensitivity analyses were performed assuming 50% and 25% capture of reported cases. Evaluation of AESI from GSD was compared to a cohort study including >20 million mRNA-1273 exposed individuals in the US HealthVerity (HV) database.
Results: As of December 2022, ≥50 AESI were routinely analyzed in GSD and 38 AESI were analyzed in HV. Of the 31 AESI in both systems, interpretation of the results of all but CVST and Chilblain’s like lesions in adults (further analyzed in the observational cohort study) was consistent. Both sources identified an increased risk of myocarditis and pericarditis, which are labelled events. The results for AE not associated with mRNA-1273 in GSD were also consistent with those from the cohort study (e.g., aseptic arthritis, Bell’s palsy, and Guillain-Barre Syndrome). However, for some non-associated AE such as renal failure and aseptic meningitis, GSD analyses lacked precision and accuracy, limiting interpretability in the absence of the HV cohort study.
Conclusions: Although there are factors that may make analyses of GSD challenging including estimated denominators and potential under or stimulated reporting, the GSD is a rich data source that can be used to provide timely insights into safety of vaccines
