This cohort study has been described previously (9,10). Briefly, in mid-2021, we enrolled HCWs from seven hospitals in Baku and collected information on sociodemographics, occupational exposure, and COVID-19 vaccination status. Participants completed weekly symptom questionnaires, those reporting any respiratory symptom were tested for SARS-CoV-2 by RT-PCR. We administered questionnaires to PCR-positive participants 30 days post-symptom onset to obtain information about their illness course. Data on vaccination and SARS-CoV-2 PCR testing results were validated using national databases.

We restricted the analysis to participants eligible for booster doses according to the Azerbaijan government: those who had received PVS at least five months (150 days) prior. During January 1 – August 31, 2022, we estimated the rVE of a first booster compared to a PVS series against two outcomes: i) symptomatic PCR-confirmed SARS-CoV-2 infection (COVID-19) and ii) medically attended COVID-19, defined as an outpatient visit because of COVID-19. Participants were considered fully vaccinated 14 days after receiving their PVS and 7 days after receiving their booster dose. We assessed the effect of time since vaccination on rVE against symptomatic infection over the time intervals: 7–89, 90–150 and ≥150 days following booster dose.

We evaluated rVE during the overall study period and separately for the period in which SARS-CoV-2 Omicron BA.1 and BA.2 variants were predominantly circulating (January 1–March 31, 2022), which we defined using whole genome sequencing data from study samples and publicly available data from GISAID (Figure S2) (9,11).

Relative VE was estimated as (1–adjusted hazard ratio)*100. Hazard ratios comparing vaccinated and unvaccinated participants were estimated using Cox proportional hazards models with vaccination as a time-varying exposure and a calendar timescale. Hospital location (urban or sub-urban) and previous SARS-CoV-2 infection were included in the model as fixed effects. Additional details about study methods are included in the supplemental material. The study was approved by the Azerbaijan and WHO Research Ethics Review Committees. All study participants provided written informed consent.

At the study start, there were 1367 participants (Figure S1). The median age was 48 (IQR: 39–57), and 1276 (93%) were female. Overall, 506 (37%) were nurses, 349 (26%) were physicians, and 512 (37%) were other professions. Overall, 375 (27%) had received PVS, 939 (69%) had received a booster, 30 (2%) received one vaccine dose, and 23 (2%) remained unvaccinated. Only participants who received CoronaVac vaccine were included in the analysis (Table 1). The median time from receiving the last vaccine dose until the start of follow-up was 66 days (IQR: 56–96) for booster dose and 150 days (IQR: 150–187) for PVS. At the study end, 1186 (87%) participants had received a CoronaVac booster dose (Table S1).

During the study period, 74 participants (5%) had PCR-confirmed COVID-19, of whom 53 (72%) had been vaccinated with a booster dose and 17 (23%) with PVS. Most cases [66 (89%)] occurred between January and March 2022, mirroring the trends in COVID-19 incidence in Azerbaijan (Figure 1). Among the participants with PCR-confirmed SARS-CoV-2, 24 (32%) sought outpatient medical care. There were no emergency room visits, hospitalisations, or deaths among participants.

For the whole analysis period (January–August 2022), the rVE was 59% (95% CI: 24–78) against symptomatic infection and 78% (95% CI: 41–92) against medically attended infection. During the Omicron BA.1/BA.2-predominant period (January–March 2022), rVE of a CoronaVac booster was 60% (95% CI: 25–79) against symptomatic infection and 79% (95% CI: 44–92) against medically attended COVID-19 (Figure 2, Table 2). Relative VE against symptomatic infection was 50% (95% CI: -3–75) 7-89 days post-booster, 67% (95% CI: 35–83) 90-149 days post-booster, and 48% (95% CI: -190–91) ≥150 days post-booster during the whole study period. (Figure 2, Table 2).