Comparative Effectiveness and Safety of BNT162b2 and CoronaVac in Hong Kong: A Target Trial Emulation

Abstract

OBJECTIVES: To evaluate the difference between BNT162b2 and CoronaVac in vaccine effectiveness and safety. METHODS: This target trial emulation study included individuals aged ≥ 12 during 2022. Propensity score matching was applied to ensure group balance. The Cox proportional hazard model was used to compare the effectiveness outcomes including COVID-19 infection, severity, 28-day hospitalization and 28-day mortality after infection. Poisson regression was used for safety outcomes including 32 adverse events of special interests between groups. RESULTS: 639,818 and 1,804,388 individuals were identified for the 2-dose and 3-dose comparison, respectively. In 2-dose and 3-dose comparison, the hazard ratios (HRs) (95% confidence intervals [CI]) were 0.844 [0.833-0.856] and 0.749 [0.743-0.755] for COVID-19 infection, 0.692 [0.656-0.731] and 0.582 [0.559-0.605] for hospitalization, 0.566 [0.417-0.769] and 0.590 [0.458-0.76] for severe COVID-19, and 0.563 [0.456-0.697] and 0.457 [0.372-0.561] for mortality for BNT162b2 recipients versus CoronaVac recipients, respectively. Regarding safety, 2-dose BNT162b2 recipients had a significantly higher incidence of myocarditis (Incidence rate ratio[IRR][95% CI]: 8.999 [1.14-71.017]) versus CoronaVac recipients, but the difference was insignificant in 3-dose comparison (IRR [95% CI]: 2.000 [0.500-7.996]). CONCLUSIONS: BNT162b2 has higher effectiveness among individuals aged ≥ 12 against COVID-19-related outcomes for SARS-CoV-2 omicron compared to CoronaVac, with almost 50% lower mortality risk. (200 words).