Using a cohort of more than 2 million individuals, they showed that infection-associated immunity was 95% protective against subsequent reinfection during 20 months of follow-up compared with no immunity (adjusted hazard ratio aHR 0·05 95% CI 0·05–0·05 p<0·001). In particular, protection from previous infection was weaker in individuals aged 65 years and older, suggesting that vaccination might be particularly important in some high-risk populations; this is probably particularly true for protection against omicron, in which a third vaccine dose is key to eliciting cross-reactive neutralising antibodies.2,3 On balance, these data clearly show the benefits of two-dose vaccination for convalescent individuals, both in terms of the durability of immunity and protection from severe disease. Due to the immune evasion of the omicron variant driving an increasing number of breakthrough infections, combined with varying recommendations for third or fourth vaccine doses,6 many populations will exhibit a mixture of infection-elicited and vaccine-elicited immunity.