Retrospective cohort study of mRNA COVID-19 vaccine-associated enhanced disease in the Vaccine Safety Datalink, 2021-23.

Vaccine-associated enhanced disease (VAED) was recognized as a theoretical risk when COVID-19 vaccines were authorized. Epidemiologic data can identify a signal for VAED if vaccine effectiveness (VE) is negative for preventing serious illness. In 2021, we initiated a cohort study to examine VAED indirectly by estimating VE against COVID-19 hospitalization and severe illness in the Vaccine Safety Datalink. The study cohort included VSD enrollees who were age-eligible to receive mRNA COVID-19 vaccine from February 2021-March 2023. Outcomes were initial COVID-19-associated hospital admission and severe in-hospital complications (acute respiratory distress or failure). VE was estimated by dose, variant period, and for early (14-119 d) and late (≥120 d) post-vaccination periods. VE was defined as 100*(1- adjusted hazard ratio). Relative VE (rVE) was estimated for bivalent vaccine versus ≥2 monovalent vaccine doses. The hospitalization analysis included all 16,027 enrollees with COVID-19 hospitalization and a 5% random sample of non-hospitalized enrollees who contributed person-time (n = 138,238). Two-dose original monovalent VE against hospitalization exceeded 90% in most age groups during the pre-Delta and Delta periods. VE was lower during the Omicron period. During the Omicron BA.2/BA.4/BA.5/XBB period, a fourth monovalent dose or bivalent dose generated VE of 59%-78% across age groups with modest waning. rVE was 48%-60% across age groups for bivalent vaccine vs. ≥2 original monovalent doses. Overall, the results were similar for VE against severe in-hospital complication. In conclusion, there was no signal for VAED during any variant period or for any specific dose of monovalent or bivalent vaccine.