Abstract
BACKGROUND: Current coronavirus disease 2019 (COVID-19) vaccine recommendations in the United States (US) provide guidance for adults to receive at least annual variant-targeted vaccination. We sought to estimate the strength and durability of protection from annual variant-targeted vaccination against severe COVID-19 illness in individuals with vaccine-derived and hybrid immunity. METHODS: We emulated a target trial using an electronic health record-based, propensity score-matched (1:1) cohort of US Veterans. Booster-vaccinated adults were eligible for a variant-targeted messenger RNA (mRNA) booster starting 1 September 2022. Matched sets of those who did and did not receive the variant-targeted booster dose were identified on a weekly basis, and the cohort was followed until 31 August 2023. Outcomes were hospitalization due to COVID-19 pneumonia and in-hospital severe illness. We fit Cox models, overall and stratified by last documented severe acute respiratory syndrome coronavirus 2 infection (pre-Omicron, Omicron), to estimate relative vaccine effectiveness (rVE). RESULTS: The propensity score-matched cohort consisted of 1 576 626 COVID-19 booster-vaccinated adults. Estimates of rVE from variant-targeted mRNA booster against hospitalization due to COVID-19 pneumonia were significant and similar in the cohort with vaccine-derived immunity (rVE, 29% [95% confidence interval {CI}, 25%-34%]) and cohort with hybrid immunity (rVE, 38% [95% CI, 27%-47%]). These protective gains were significant from 0 to 6 months but not 6 to 12 months after vaccination and during pre-XBB and XBB variant eras. Findings were similar for in-hospital severe illness. CONCLUSIONS: In cohorts with vaccine-derived and hybrid immunity, modest but significant gains in protection against hospitalization and severe COVID-19 illness were conferred by the annual variant-targeted booster dose but not sustained beyond 6 months.