Effectiveness of BNT162b2 and ChAdOx-1 vaccines in residents of long-term care facilities in England using a time-varying proportional hazards model

Introduction: residents of long-term care facilities (LTCFs) are at high risk of adverse outcomes from SARS-CoV-2. We aimed to estimate the vaccine effectiveness (VE) of one and two doses of BNT162b2 and ChAdOx-1 against SARS CoV-2 infection and COVID-19-related death in residents of LTCFs. Methods: this observational study used testing, vaccination and mortality data for LTCF residents aged ≥ 65 years who were regularly tested regardless of symptoms from 8 December 2020 to 30 September 2021 in England. Adjusted VE, calculated as one minus adjusted hazard ratio, was estimated using time-varying Cox proportional hazards models for infection and death within 28 days of positive test result. Vaccine status was defined by receipt of one or two doses of vaccine and assessed over a range of intervals. Results: of 197,885 LTCF residents, 47,087 (23.8%) had a positive test and 11,329 (5.8%) died within 28 days of a positive test during the study period. Relative to unvaccinated individuals, VE for infection was highest for ChAdOx-1 at 61% (40-74%) 7at 1-4 weeks and for BNT162b2 at 69% (52-80%) at 11-15 weeks following the second dose. Against death, VE was highest for ChAdOx-1 at 83% (58-94%) at 1-4 weeks and for BNT162b2 at 91% (75-97%) at 11-15 weeks following second dose. Conclusions: compared with unvaccinated residents, vaccination with one dose of BNT162b2 or ChAdOx-1 provided moderate protection against infection and death in residents of LTCFs. Protection against death improved after two doses. However, some waning of protection over time was noted.