Determinants of excess mortality rates in nursing homes (Code)

This research is based on our own calculations using non-public microdata from Statistics Netherlands. Under certain conditions, these microdata are accessible for statistical and scientific research. For further information, contact microdata@cbs.nl.

Our study population consists of all individuals who lived in a nursing home in the Netherlands on January 1st in 2020 or 2021. To predict the individual level probability of dying in 2020 and 2021 had there not been a COVID-19 pandemic, we construct a second sample consisting of individuals who lived in a nursing home on January 1st in the years 2016-2019. We select nursing home residents who are i) at least 65 years old; ii) with complete information on individual level characteristics, which excludes individuals whose nursing home admission was before the year 2015 as data on hospitalizations is less reliable prior to 2015; and iii) who are eligible for a permanent stay, which excludes those who receive palliative and rehabilitation care.

For all these nursing home residents we use data at the individual level consisting of information on: long-term care use and care profile (Dutch: Zorgzwaarte pakket) from CAK; the reason for eligibility for long-term care from CIZ; date and cause of death from death registries; income and wealth data from tax registries; date of birth and death, sex, and place of residence from municipal records; medicine use based on health insurance data; date and diagnose of inpatient hospital admissions from Dutch Hospital Data and healthcare expenditures from Vektis.

We use individual-level information on nursing home provider codes and addresses to link individuals to the nursing home facility they were admitted to. The data at the nursing home resident level is linked to publicly available data on nursing home organizations from the annual reports and on facilities from Dutch Healthcare Institute.