BNT162b2 vaccination reduced infections and transmission in a COVID‐19 outbreak in a nursing home in Germany, 2021

Abstract Background A SARS‐CoV‐2 outbreak was detected in a nursing home in February 2021 after residents and staff had received two doses of BNT162b2 vaccine in January 2021. Methods Nursing home staff, long‐term residents and day‐care receivers were included in a retrospective cohort study. We calculated attack rates (AR), secondary AR (SAR) and their 95% binomial confidence interval (CI), and we compared them using Fisher's exact test or chi‐squared test, depending on the sample size. We used Poisson regression with robust error estimates to calculate vaccine effectiveness against SARS‐COV‐2 infections. We selected variables based on directed acyclic graphs. As a proxy for viral load at diagnosis, we compared the mean Ct values at diagnosis using t tests or Mann–Whitney U tests. Results The adjusted vaccine effectiveness against infection was 56% (95% CI: 15–77%, p = 0.04). Ct values at diagnosis were higher when intervals after receiving the second vaccination were longer (>21 vs. ≤21 days: 4.48 cycles, p = 0.08). The SAR was 67% lower in households of vaccinated (2/9 [22.2%]) than of unvaccinated infected staff (12/18 [66.7%]; p = 0.046). Vaccination rates were lowest among staff with close physical contact to care‐receivers (46%). The highest AR in vaccinated staff had those working on wards (14%). Conclusions Vaccination reduced the risk for SARS‐CoV‐2 infection, viral load and transmission; however, non‐pharmaceutical interventions remain essential to reduce transmission of SARS‐CoV‐2 infections, even for vaccinated individuals. Vaccination coverage of staff ought to increase reduction of infections among themselves, their household members and residents.


| INTRODUCTION
While COVID-19 case fatality was <0.1% in under the age of 50 years, 1 it was 13% in outbreaks of SARS-CoV-2 in nursing homes in Germany between January 2020 and February 2021. 2 Hence, nursing homes were prioritized for COVID-19 vaccination, which began in Germany in December 2020. 3 However, the vaccine effectiveness (VE) in such institutionalized setting, particularly in aged and comorbid care receivers, was unknown.
We report on a SARS-CoV-2 Alpha (B.1.1.7) outbreak among day-care receivers and long-term residents and staff of a nursing home in Osnabrück, Germany, January to March 2021. Due to vaccine prioritization policies, only residents of long-term care wards and all staff of nursing homes were eligible for voluntary  vaccination. 4 Hence, all day-care receivers, some long-term residents and some staff were unvaccinated during the outbreak.
This study describes the epidemiology of the outbreak and the undertaken control measures. We assess vaccination coverage (VC) and attack rates (AR) by occupational groups, VE against SARS-CoV-2 Alpha infections and vaccine effects on viral load and secondary transmission.

| Study population and definitions
We conducted a retrospective cohort study and included all care-receivers and staff working in the nursing home between 03 January (symptom onset of first case) and 18 March 2021 (2 weeks after diagnosing the last case). We defined cases of the outbreak as care-receivers or staff who had a positive SARS-CoV-2-PCR result between 03 January 2021 and 18 March 2021. We defined the date of diagnosis as the earlier date of either symptom-onset or sampling of the first positive test (PCR or rapid antigen detection tests [RADT]). We grouped staff according to their degree of contact with care-receivers into staff with close physical contact (nursing specialist, nursing assistants, nursing trainees and interns), staff working on wards (chaplains, administrative care and cleaner) and staff working outside of wards (administration, kitchen and housekeeping).
To calculate secondary attack rates (SAR) in households of infected staff, we collected the number of household members and their PCR results from the local health authority database. Secondary cases were defined as SARS-CoV-2 PCR-positive household members diagnosed 1-14 days after the diagnosis of the corresponding staff index case. All household members of infected staff were PCR-tested on days 5 and 10 of their quarantine. We excluded household members with a PCR-confirmed SARS-CoV-2 infection within 6 months prior or who quarantined separately from infected staff, of which none were diagnosed with a SARS-CoV-2 infection prior to this outbreak. One household outside of the administrative district was excluded because data was unavailable.

| Descriptive and analytical epidemiology
We estimated AR of SARS-CoV-2 infections with their 95% confidence interval (CI) and compared AR between the vaccinated and unvaccinated cohort using chi-squared or Fisher's exact tests, respectively.
We estimated VE as VE = 1 À RR, for which the relative risk (RR) for SARS-CoV-2 infection in vaccinated versus unvaccinated individuals was estimated via Poisson regression. 5 We calculated 95% CI with robust standard errors. Our variable selection included vaccination status as the exposure and SARS-CoV-2 infection as the outcome. Based on directed acyclic graph modelling (Supporting Information S1), we further included the status (long-term residents, day-receivers or staff) and sex. We could not include the risk perception for an infection before vaccination because the respective data were unavailable.

| Determination of viral load
We assessed viral load at diagnosis using the Ct value for the ORF1AB gene at the first positive PCR as a proxy. Depending on the data distribution, we used t tests or Mann-Whitney U tests to compare if the mean viral load differed between vaccinated versus unvaccinated, >14 days versus ≤14 days and >21 days versus ≤21 days between receiving the second vaccination and sampling.

| Contact network analysis
Using the date of diagnosis with contact information collected by the local public health authority (LPHA), we inferred a possible contact network and presumable transmission chains which were visualized with the R package epicontacts. 6 3 | RESULTS

| Study population
The nursing home comprised one ward for day-care and seven for long-term care with a total of 128 staff members, 100 long-term care residents and 24 day-care receivers.

| Potential source of the outbreak
Typing detected two SARS-CoV-2 strains (one wild-type and 27/28 Alpha variant), suggesting at least two introductory events. The vaccination team tested negative with daily RADT and with weekly PCR. Therefore, we excluded the vaccination team as a source.
Potentially, the outbreaks in the day-care and the permanent care  We assessed the use of RADT in a separate publication. 19

CONFLICT OF INTEREST
No conflict of interest declared.

ETHICS STATEMENT
This outbreak investigation was conducted as part of an administrative assistance procedure under the German Infection Protection Act §4 and thus did not require ethical approval.

AUTHOR CONTRIBUTIONS
Emily

PEER REVIEW
The peer review history for this article is available at https://publons. com/publon/10.1111/irv.13051.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.