LAUSR

Abstract Background Vaccine effectiveness against the severe acute respiratory syndrome coronavirus 2 Omicron BA.2 sublineage in Japan is unknown. We assessed the effectiveness of a third dose of COVID-19 mRNA vaccine compared with that of 2 doses. Methods We performed a population-based cohort study using a municipality database located in the Chubu region of Japan during the Omicron BA.1- and BA.2-predominant periods (January 1–March 31, 2022 and April 1–27, 2022, respectively). We included residents aged ≥16 years who received a second vaccine dose at ≥14 days before the start of each period, regardless of the third dose. We compared the data at 14 days after the second and third dose and at 2-week intervals from 14 days to 10 weeks after the third dose using a Cox regression model. Vaccine effectiveness was defined as (1 – hazard ratio) × 100 (%). Results In total, 295 705 and 288 184 individuals were included in the BA.1- and BA.2-predominant periods, respectively. The effectiveness of a third dose against infection was 62.4% and 48.1% in the BA.1- and BA.2-predominant periods, respectively. Vaccine effectiveness at 2–3 weeks and ≥10 weeks after the third dose decreased from 63.6% (95% confidence interval [CI], 56.4–69.5%) to 52.9% (95% CI, 41.1–62.3%) and from 54.5% (95% CI, 3.0–78.7%) to 40.1% (95% CI, 15.1–57.7%) in the BA.1- and BA.2-predominant periods, respectively. Conclusions A third dose was moderately effective against BA.1 and BA.2 sublineages, but its effectiveness decreased by approximately 10% age points from 2–3 weeks to ≥10 weeks after the third vaccination.