Modelling direct and herd protection effects of vaccination against the SARS-CoV-2 Delta variant in Australia

Objectives: To analyse the outcomes of COVID-19 vaccination by vaccine type, age group eligibility, vaccination strategy, and population coverage. Design: Epidemiologic modelling to assess the final size of a COVID-19 epidemic in Australia, with vaccination program (Pfizer, AstraZeneca, mixed), vaccination strategy (vulnerable first, transmitters first, untargeted), age group eligibility threshold (5 or 15 years), population coverage, and pre-vaccination effective reproduction number (R-eff((v) over bar) for the SARS-CoV-2 Delta variant as factors. Main outcome measures: Numbers of SARS-CoV-2 infections; cumulative hospitalisations, deaths, and years of life lost. Results: Assuming R-eff((v) over bar = 5, the current mixed vaccination program (vaccinating people aged 60 or more with the AstraZeneca vaccine and people under 60 with the Pfizer vaccine) will not achieve herd protection unless population vaccination coverage reaches 85% by lowering the vaccination eligibility age to 5 years. At R-eff((v) over bar = 3, the mixed program could achieve herd protection at 60-70% population coverage and without vaccinating 5-15-yearold children. At R-eff((v) over bar = 7, herd protection is unlikely to be achieved with currently available vaccines, but they would still reduce the number of COVID-19-related deaths by 85%. Conclusion: Vaccinating vulnerable people first is the optimal policy when population vaccination coverage is low, but vaccinating more socially active people becomes more important as the R-eff((v) over bar declines and vaccination coverage increases. Assuming the most plausible R-eff((v) over bar of 5, vaccinating more than 85% of the population, including children, would be needed to achieve herd protection. Even without herd protection, vaccines are highly effective in reducing the number of deaths.