Although numerous COVID-19 vaccines are effective against COVID-19 infection and variants of concern (VOC) in the real world, it is imperative to obtain evidence of the corresponding vaccine effectiveness (VE). This study estimates the real-world effectiveness of the BNT162b2 and mRNA-1273 vaccines against COVID-19 infection and determines the influence of different virus variants on VE by using test-negative design (TND) studies. We systematically searched for published articles on the efficacy of BNT162b2 and mRNA-1273 against COVID-19 infection. Two researchers independently selected and extracted data from eligible studies. We calculated the VE associated with different vaccine types, SARS-CoV-2 variants, and vaccination statuses, using an inverse variance random-effects model. We selected 19 eligible studies in the meta-analysis from 1651 records. For the partially vaccinated group, the VE of BNT162b2 and mRNA-1273 was 61% and 78% against COVID-19 infection, respectively. For the completely vaccinated group, the VE of BNT162b2 and mRNA-1273 was 90% and 92% against COVID-19 infection, respectively. During subgroup analyses, the overall VE of BNT162b2 and mRNA-1273 against the Delta variant was 53% and 71%, respectively, for the partially vaccinated group; the respective VE values were 85% and 91% for the fully vaccinated group. Irrespective of the BNT162b2 or mRNA-1273 vaccines, the Delta variant significantly weakened vaccine protection for the partially vaccinated group, while full vaccination was highly effective against COVID-19 infection and various VOC. The mRNA-1273 vaccine is more effective against COVID-19 infection and VOC than the BNT162b2 vaccine, especially for the partially vaccinated group. Overall, the results provide recommendations for national and regional vaccine policies.
Keywords: BNT162b2 vaccine; COVID-19 infection; mRNA-1273 vaccine; meta-analysis; test-negative design; vaccine effectiveness.