Indirect protection and long-term effectiveness of inactivated COVID-19 vaccine: a stepped-wedge randomised trial in Serrana, Brazil

Currently approved vaccines have shown wide ranges of estimated efficacy. Despite the well-known benefits of inactivated virus vaccines, little information on their effectiveness has been robustly provided to monitor their specific impact. A stepped-wedge trial is an approach for assessing vaccine effectiveness and indirect protection in the real world. By the end of the study, all participants can receive the intervention, eliminating the ethical dilemma of placebo, especially during a pandemic. The aim of this study was to assess the effectiveness of an inactivated COVID-19 vaccine in a massive urban population during the uncontrolled COVID-19 epidemic. We evaluated the long-term vaccine effectiveness (VE) of CoronaVac in Serrana, Brazil, using a stepped-wedge randomised trial. An indirect effect was also inferred from the temporal association between increasing vaccine coverage and the concurrent decline in SARS-CoV-2 incidence among participants who were not yet fully vaccinated. The city was divided into 25 subareas, clustered into four groups, and randomised to receive CoronaVac in a two-dose scheme. After 6 months, a booster dose was offered by the Brazilian Immunization Program. Participants were followed for one year, divided into four periods: February-May 2021; May-August 2021; August-November 2021; November 2021-February 2022. Vaccination occurred between February 14 and April 11, 2021. Up to 27,390 participants received the CoronaVac first dose, corresponding to 82.8% of the adult urban population. In the 1^st period, overall VE was 52.8% (95% CI, 44.7 to 59.7) for preventing symptomatic COVID-19 and direct VE was 81.6% (95% CI, 76.4 to 85.6). When approximately 50% of the adult population was fully vaccinated, a reduction in symptomatic COVID-19 was also observed among participants who were not yet fully vaccinated, suggesting an indirect protective effect. Two-dose VE for COVID-19-related hospitalisation and death was 89.2% (95% CI, 68.1 to 96.3), 86.8% (95% CI, 72.2 to 93.7), 85.2% (95% CI, 66.1 to 93.6), 80.4% (95% CI, 47.6 to 92.7) in each period, respectively. The booster dose increased VE to 94.9% (95% CI, 59.1 to 99.4) and 84.1% (95% CI, 64.1 to 92.9) in the 3^rd and 4^th periods, respectively. CoronaVac induced long-term protection for severe cases. Unvaccinated individuals benefited from high vaccine coverage levels.Trial registration: https://ClinicalTrials.gov (NCT04747821).