The Rhineland-Palatinate Statistical Office provided population counts for the reference date December 31st, 2019 as well as weekly overall mortality counts from January to June 2021, both stratified by sex and age group (15–34, 35–59, 60–79, 80 + years).
The Robert Koch Institute published weekly counts of COVID-19 deaths and SARS-CoV-2 infections as reported according to the Infection Protection Act [7
], both stratified by sex and age group (15–34, 35–59, 60–79, 80 + years). Deaths were assigned to the location of the reporting municipal health authority.
During the initial vaccine rollout in 2021, the Division of Vaccine Documentation of the Cancer Registry Rhineland-Palatinate was put in charge of documenting all individual vaccinations with a COVID-19 vaccine provided by the state. This included vaccinations in dedicated centers as well as those carried out by mobile teams for nursing home residents, and by some company doctors. For each vaccinated person, the data set included information on sex, age, location of residence, vaccination dates, and the status of partial versus full vaccination. Individuals were considered fully vaccinated at the date of their second shot, at the date of their first shot after recovery from a previous SARS-CoV-2 infection, or at the date of their first shot with a vaccine requiring only one dose.
Vaccination data was preprocessed to exclude out-of-state residents and implausible observations (Online Resource). To obtain the same aggregation level as for the remaining data, individual data on vaccinations was aggregated by calendar week, sex, and age group (15–34, 35–59, 60–79, 80 + years). The data set used for all analyses resulted from merging the aggregated data sets on the resident population, all-cause mortality, SARS-CoV-2 infections, COVID-19 fatalities, and vaccinations, respectively. The units of observation were formed by all 160 combinations of calendar week, sex, and age group (Online Resource).
The aggregated data set contained the vaccination coverage as the percentage of the population fully vaccinated per calendar week, sex, and age group. The same information was included for vaccination dates that were time-lagged with a period of 14 days to account for the delay until reaching full protection.
In each calendar week, the sex-specific proportion of COVID-19 fatalities and of reported SARS-CoV-2 infections formed by each age group was calculated.
The association between vaccination coverage and COVID-19 mortality was explored by plotting the time course of these two measures together: (1) vaccination coverage, defined as the proportion of the population either at least partially or fully vaccinated and (2) the proportion of COVID-19 fatalities formed by each age group. A multivariable binomial logistic regression model was used to examine the association between the time-lagged vaccination coverage and the sex-specific proportion of weekly COVID-19 fatalities formed by each age group while adjusting for sex and age group (Online Resource). A second model for the sex-specific proportion of reported weekly SARS-CoV-2 infections formed by each age group used the same covariates. As a sensitivity analysis to test whether associations were specific to COVID-19 endpoints, a third analogous model for all-cause mortality was fitted (Online Resource). We report adjusted odds ratios together with 95% confidence intervals and p
values from Wald-tests based on heteroscedasticity and autocorrelation-consistent standard errors [8
] to account for autocorrelation of the time series.
We evaluated data from calendar week 1 to 20 in 2021 because fewer than 10 COVID-19 deaths per week were reported in Rhineland-Palatinate thereafter. Furthermore, rollout of vaccinations in private practices that were not documented in the available database started to increase in calendar week 15, leading to potential underestimation of vaccination coverage. Data analysis was carried out in the statistical environment R [9