Link to HAL – pasteur-04644962
Link to DOI – 10.2807/1560-7917.es.2024.29.13.2300403
Eurosurveillance, 2024, 29 (13), pp.2300403. ⟨10.2807/1560-7917.es.2024.29.13.2300403⟩
Background Scarce European data in early 2021 suggested lower vaccine effectiveness (VE) against SARS-CoV-2 Omicron lineages than previous variants. Aim We aimed to estimate primary series (PS) and first booster VE against symptomatic BA.1/BA.2 infection and investigate potential biases. Methods This European test-negative multicentre study tested primary care patients with acute respiratory symptoms for SARS-CoV-2 in the BA.1/BA.2-dominant period. We estimated PS and booster VE among adults and adolescents (PS only) for all products combined and for Comirnaty alone, by time since vaccination, age and chronic condition. We investigated potential bias due to correlation between COVID-19 and influenza vaccination and explored effect modification and confounding by prior SARS-CoV-2 infection. Results Among adults, PS VE was 37% (95% CI: 24–47%) overall and 60% (95% CI: 44–72%), 43% (95% CI: 26–55%) and 29% (95% CI: 13–43%) < 90, 90–179 and ≥ 180 days post vaccination, respectively. Booster VE was 42% (95% CI: 32–51%) overall and 56% (95% CI: 47–64%), 22% (95% CI: 2–38%) and 3% (95% CI: −78% to 48%), respectively. Primary series VE was similar among adolescents. Restricting analyses to Comirnaty had little impact. Vaccine effectiveness was higher among older adults. There was no signal of bias due to correlation between COVID-19 and influenza vaccination. Confounding by previous infection was low, but sample size precluded definite assessment of effect modification. Conclusion Primary series and booster VE against symptomatic infection with BA.1/BA.2 ranged from 37% to 42%, with similar waning post vaccination. Comprehensive data on previous SARS-CoV-2 infection would help disentangle vaccine- and infection-induced immunity.