Abstract
BACKGROUND: We estimate effectiveness of 3 versus 2 vaccine doses against SARS-CoV-2 B.1.1.529 Omicron in a mostly infection-naiive but highly vaccinated Australian population. METHODS: Cohort study of adults aged 40+ years resident in Sydney followed from 1 January 2022 for SARS-CoV-2 infection and COVID-19 hospitalisation or death using linked immunisation, disease notification and hospitalisation registers. Adjusted hazard ratios (aHR) and corresponding relative vaccine effectiveness (rVE) were estimated comparing 3 to 2 vaccine dose recipients by time since dose receipt, vaccine brand, and prior infection. Absolute risk reductions and numbers needed to boost by age groups were calculated. RESULTS: 2,053,123 infection-naiive individuals (mean age 59 years) were followed for 327,272 person-years for infection and 224,269 person-years for severe outcomes (hospitalisation/death). There were 175,849 infections and 4113 hospitalisations/deaths. Compared to individuals receiving dose 2 within the last 3 months, rVE in dose 3 recipients was 7% (95% CI 5-9%) against infection and 65% (95%CI 61-69%) against hospitalisation/death. Almost all dose 3 recipients had an mRNA vaccine; there was little difference in dose 3 rVE by primary course vaccine brand (ChAdOx1 versus BNT162b2). Over the 6-week follow-up, we estimated one hospitalisation/death was avoided for every 192 adults aged ≥70 years boosted with dose 3 in the infection-naiive cohort. The aHR for hospitalisation/death from Omicron was 0.12 (95 %CI 0.07-0.23) for 2-dose recipients with a prior Delta infection compared with 2-dose recipients with no prior infection. CONCLUSIONS: Receipt of a third COVID-19 vaccine dose in adults aged 40 years and above significantly reduced hospitalisations and deaths from SARS-CoV-2 Omicron infections in a primarily infection-naiive population.