Background: Mass vaccination has the potential to bring the COVID-19 pandemic to a halt by not only protecting individuals who have been vaccinated but also by providing cross-protection to unvaccinated individuals, such as children. However, this indirect protection effect from a vaccinated population onto an unvaccinated group is extremely difficult to observe in real-world situations.
Methods: We studied cross-protection to unvaccinated individuals following an unprecedented rapid mass vaccination campaign in Europe. After a large outbreak of B.1.351 (Beta) in the district of Schwaz in Austria, the government offered every adult (16+) citizen of the district a vaccination with BNT162b2 between the 11th and 16th of March 2021. After this week, around 70% of the adult population of Schwaz had received their first dose, which made Schwaz the first widely inoculated region in Europe. The cohort of children under the age of 16 remained entirely unvaccinated (EMA only approved the vaccine for 12-15 year-olds on the 28th of May). This local mass vaccination campaign created a situation in which the vaccination coverage of the adult population sharply differed at the district border of Schwaz, while the coverage of those below the age of 16 remained the exact same. We compared SARS-CoV-2 cases among the adult population as well as children in Schwaz with case numbers of the same age cohorts from control regions. First, we compared Schwaz with a control group of other Austrian districts highly similar to Schwaz in many socio-demographic characteristics as well as in infection spread prior to the mass vaccination campaign. Second, we compared local populations residing along the border of Schwaz which live in the very same geographic area but with different vaccination coverage because they were not included in the vaccination campaign.
Interpretation: Prior the mass vaccination campaign, we observed very similar infection spread across all age cohorts in Schwaz and the control regions. Around 3-4 weeks after the campaign, infections started to diverge between Schwaz and the control regions. While the difference was largest among the population aged 16–50 years (which was offered vaccination in the campaign), we also observed a statistically significant reduction in cases among the group of unvaccinated children. Our findings are robust to changes in the control group, as well as controls of a rich set of time and region specific effects.
Policy implications:
Our results constitute one of the first evidence of an indirect cross-protection effect from a group of vaccinated individuals to an unvaccinated group (in our case children). Given that in many countries the proposition to keep schools open during the academic year 2021/22 is a top priority, this evidence of community-protection is highly policy relevant.