Flow chart describing the selection of participants for the analysis.

The number of individuals in this figure is not the same as some of the numbers in Table 1 because of missing data in key variables. Note that in the bottom part of the chart, related to secondary analyses, the boxes represent overlapping sets of study participants; in other words, some individuals included in the secondary analyses that correspond to the left box were also included in analyses corresponding to the box on the right.

Characteristics of the study population and reinfection rate by post-infection vaccination status.

Here, reinfection rate refers to the percentage of the relevant study subpopulation with evidence of reinfection between December 1, 2022 and January 3, 2023. Note that for the variables on region, occupation, and clinical severity, data are missing for large fractions of the study population. Note also that information was only available on sex at birth, but not on gender.

Vaccination coverage and cumulative incidence of SARS-CoV-2 reinfection in the study population.

Panel A presents the percentages of the study population vaccinated over time. The cumulative incidence of SARS-CoV-2 reinfections is presented by the number of vaccination doses before (panels) and after (lines) first infection (panels B-C). Shaded regions: 95% CIs. 1V-I, 2V-I represented 1, and 2 vaccine doses before infection, respectively; 1V-I-V, 2V-I-V corresponds to 1 and 2 doses before infection, then post-infection vaccination, respectively. As mentioned in the Results section, 142 and 144 study participants who received one and two pre-infection vaccine doses received two post-infection vaccine doses. We do not show the corresponding plot for those individuals who received 3 pre-infection doses as their post-infection dose was after the start of the follow-up, in December.

Effect of post-infection vaccination on SARS-CoV-2 reinfection stratified by pre-infection vaccination.

Error bars (95% CIs) and circles represent aHR for SARS-CoV-2 reinfection estimated using Cox proportional hazards models. V-I-V, 1V-I-V, 2V-I-V, 3V-I-V corresponds to any pre-infection vaccination, 1, 2 and 3 vaccine doses before infection, then vaccination, respectively; they were compared to V-I, 1V-I, 2V-I, 3V-I, respectively. Partial V-I-V, Full V-I-V and Booster V-I-V represent partial vaccination, full vaccination and booster vaccination before infection, followed by post-infection vaccination, respectively. The number of doses received by individuals with partial versus full (and full with booster) vaccination depends on the type of SARS-CoV-2 vaccine received; in Table S3 we present a cross-classification of participants in the analytic population by these vaccination-related categorical variables.

Vaccine-related protection against SARS-CoV-2 reinfection stratified by demographic characteristic.

The vertical dotted line at 1.0 indicates no effect on protection.