Introduction
Evidence is limited on the effectiveness of a fourth vaccine dose against coronavirus disease 2019 (COVID-19) in populations with prior severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infections. We estimated the risk of COVID-19 deaths and SARS-CoV-2 infections according to vaccination status in previously infected individuals in Austria.
Methods
This is a nationwide retrospective observational study. We calculated age and gender adjusted Cox proportional hazard ratios (HRs) of COVID-19 deaths (primary outcome) and SARS-CoV-2 infections (secondary outcome) from 1 November to 31 December 2022, primarily comparing individuals with four versus three vaccine doses. Relative vaccine effectiveness (rVE) was calculated as (1-HR) X 100.
Results
Among 3,986,312 previously infected individuals, 281,291 (7,1%) had four and 1,545,242 (38.8%) had three vaccinations at baseline. We recorded 69 COVID-19 deaths and 89,056 SARS-CoV-2 infections. rVE for four versus three vaccine doses was −24% (95% CI: −120 to 30) against COVID-19 deaths, and 17% (95% CI: 14–19) against SARS-CoV-2 infections. This latter effect rapidly diminished over time and infection risk with four vaccinations was higher compared to less vaccinated individuals during extended follow-up until June 2023. Adjusted HR (95% CI) for all-cause mortality for four versus three vaccinations was 0.79 (0.74–0.85).
Discussion
In previously infected individuals, a fourth vaccination was not associated with COVID-19 death risk, but with transiently reduced risk of SARS-CoV-2 infections and reversal of this effect in longer follow-up. All-cause mortality data suggest healthy vaccinee bias.
© 2023 The Authors. European Journal of Clinical Investigation published by John Wiley & Sons Ltd on behalf of Stichting European Society for Clinical Investigation Journal Foundation.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.