Ecological Analysis of the Decline in Incidence Rates of COVID-19 Among Nursing Home Residents Associated with Vaccination, United States, December 2020-January 2021.

The study was ecological and facility-level, so it could not separate the effect of resident vaccination from infection prevention, testing, staff infections, community spread, or other concurrent measures. Resident case rates in nursing homes were already declining before vaccination clinics began. This leaves residual confounding. Participation in the Pharmacy Partnership Program was voluntary. This may have introduced selection bias between facilities that vaccinated early and those that did not. The comparison group was not guaranteed to remain unvaccinated after the first week, which may have reduced the observed difference. The main analysis says 6051 nursing homes, while Table 1 lists 2843 vaccinated and 3216 comparison facilities, so the exact analysed denominator is unclear. Staff vaccination was not included in the model because it was collinear with resident vaccination, and staff coverage was low at 38%. The analysis used first-dose clinic status and short follow-up, not individual patient outcomes or the full two-dose regimen.

The supplied text does not report author conflicts of interest or external funding. It states the activity was reviewed by CDC and used federal NHSN and Pharmacy Partnership Program data.

Do not conclude that vaccination alone caused the decline in nursing home resident cases. The decline began before clinics started, the comparison group likely received vaccine later, and the analysis was facility-level and could not separate vaccination from infection prevention, testing, staff cases, or community spread.