Quantifying the direct and indirect impact of COVID-19 vaccination: evidence from Victoria, Australia

Abstract

Vaccines not only directly protect vaccinated individuals but also contribute to protecting the entire population via indirect herd-immunity benefits. However, researchers have long struggled to quantify these indirect effects at the population level, hindering the assessment of vaccination programme effectiveness. We developed a new method to estimate these effects, thereby markedly improving measures of the number of infections, hospitalizations and deaths averted by vaccination. Our population-based analysis of 6 440 000 residents of Victoria, Australia, reveals strong indirect effects during the Delta outbreak (September–November 2021). By modelling a non-vaccination counterfactual, we conservatively estimate 315 000 infections were averted (95% Bayesian credible interval (BCI): 231k–406k), as well as 33 400 hospitalizations (95% BCI: 22k–46k), and 4900 deaths (95% BCI: 2.8k–7.2k). These are 4.0, 7.5 and 8.0 times higher, respectively, than observed. Half of the averted infections and around one-quarter of hospitalizations and deaths were attributable to indirect protection. Homogeneous vaccination across local government areas could have reduced outcomes by approximately 25%.