Comparative effectiveness of ChAdOx1 versus BNT162b2 COVID-19 vaccines in Health and Social Care workers in England: a cohort study using OpenSAFELY

Hulme, W. J.; Williamson, E. J.; Green, A. C.; Bhaskaran, K.; McDonald, H. I.; Rentsch, C. T.; Schultze, A.; Tazare, J.; Curtis, H. J.; Walker, A. J.; Tomlinson, L.; Palmer, T. M.; Horne, E.; MacKenna, B.; Morton, C. E.; Mehrkar, A.; Fisher, L.; Bacon, S. C.; Evans, D.; Inglesby, P.; Hickman, G.; Davy, S.; Ward, T.; Croker, R.; Eggo, R. M.; Wong, A. Y.; Mathur, R.; Wing, K.; Forbes, H.; Grint, D. J.; Douglas, I. J.; Evans, S. J.; Smeeth, L.; Bates, C.; Cockburn, J.; Parry, J.; Hester, F.; Harper, S.; Sterne, J. A.; Hernan, M. A.; Goldacre, B.

Background: The UK COVID-19 vaccination programme delivered both the BNT162b2 mRNA (Pfizer-BioNTech) and the ChAdOx1 (Oxford-AstraZeneca) vaccines during overlapping periods, providing a rare opportunity to emulate a trial that directly compares both vaccines using routinely-collected NHS data. Frontline Health and Social Care workers comprise a useful population to assess comparative effectiveness due to early vaccine eligibility and relatively high post-vaccination transmission risk due to occupational exposure. Methods: With the approval of NHS England we used the OpenSAFELY-TPP database, covering 40% of GP practices in England and linked to national coronavirus surveillance, hospital episodes, and death registry data, to compare the effectiveness of ChAdOx1 versus BNT162b2 in 1/3 million health and social care workers vaccinated between 4 January and 28 February 2021. Recipients were followed-up for 20 weeks. Second-dose effects were estimated under an intention-to-treat strategy. Primary outcomes were recorded SARS-CoV-2 infection, COVID-19-related accident and emergency attendance, and COVID-19-related hospital admission. Results: The cumulative incidence of each outcome was similar for both vaccines during the first 20 weeks post-vaccination. The cumulative incidence of recorded SARS-CoV-2 infection 6 weeks after vaccination with BNT162b2 was 19.2 per 1000 people (95%CI 18.6 to 19.7) and with ChAdOx1 was 18.9 (95%CI 17.6 to 20.3), representing a difference of -0.24 per 1000 people (95%CI -1.71 to 1.22). The difference in the cumulative incidence of COVID-19 accident and emergency attendance at 6 weeks was 0.01 per 1000 people (95%CI -0.27 to 0.28). For COVID-19 hospital admission, this difference was 0.03 per 1000 people (95%CI -0.22 to 0.27). Conclusion: In this cohort of healthcare workers where we would not anticipate vaccine type to be related to health status, we found no substantial differences in the incidence of SARS-CoV-2 infection or COVID-19 disease up to 20 weeks after vaccination. Incidence dropped sharply after 3-4 weeks and there were very few COVID-19 hospital attendance and admission events after this period. This is in line with expected onset of vaccine-induced immunity, and suggests strong protection against COVID-19 disease for both vaccines.