COVID-19 infections in German long-term care facilities: a descriptive three-level analysis using claims and...

Although many studies have investigated COVID-19 outbreaks in long-term care facilities (LTCFs), evidence that combines multiple clustered levels is scarce. We aimed to describe individual, LTCF, and regional-level factors associated with COVID-19 infections.

We conducted a nationwide study using insurance claims data from Germany between 1st October 2020 and 31st March 2021. The sample comprised 284,186 residents over 60 years in 9,869 LTCFs across all of Germany's 400 districts. We used multilevel logistic regression to model associations between individual, LTCF, and district-level factors, and the probability of a COVID-19 infection.

A total of 44,042 (15.5%) COVID-19 infections were recorded during the study period. On the individual level, male sex (OR 1.15; 95% CI 1.12-1.18), dementia (OR 1.09; CI 1.06-1.11), medium-severe care dependency level 3 and 4 (OR 1.17; CI 1.12-1.22 / OR 1.21; CI 1.16-1.26) were associated with greater risk of infection. At the LTCF level, infection risks increased with the mean age of residents (OR 1.09; CI 1.03-1.15) and higher resident numbers (OR 1.20; CI 1.14-1.27). On the district level, a higher proportion of public LTCFs was associated with lower infection risks (OR 0.90; CI 0.84-0.97), while a higher mean number of residents (OR 1.16; CI 1.05-1.28), and the district-level SARS-CoV-2 incidence rate among the general population (OR 1.54; CI 1.41-1.67) was associated with higher risks. A cross-level interaction between facility size and COVID-19 prevalence was not significant (p > 0.5).

We found evidence of individual, facility, and regional levels factors associated with COVID-19 infections among older adults in LTCFs. Future measures to combat infections, outbreaks, and pandemics should take an orchestrated multilevel approach.