Research Article: Individual and country-level determinants of nursing home admission in the last year of life in Europe

Date Published: March 14, 2019

Publisher: Public Library of Science

Author(s): Erwin Stolz, Hannes Mayerl, Éva Rásky, Wolfgang Freidl, Sukumar Vellakkal.


Previous research has focussed on individual-level determinants of nursing home admission (NHA), although substantial variation in the prevalence of NHA between European countries suggests a substantial impact of country of residence. The aim of this analysis was to assess individual-level determinants and the role of country of residence and specifically a country`s public institutional long-term care infrastructure on proxy-reported NHA in the last year of life.

We analysed data from 7,018 deceased respondents (65+) of the Survey of Health, Ageing and Retirement in Europe (2004–2015, 16 countries) using Bayesian hierarchical logistic regression analysis in order to model proxy-reported NHA.

In total, 14% of the general older population utilised nursing home care in the last year of life but there was substantial variation across countries (range = 2–30%). On the individual-level, need factors such as functional and cognitive impairment were the strongest predictors of NHA. In total, 18% of the variance of NHA was located at the country-level; public expenditure on institutional care strongly affected the chance of NHA in the last year of life.

On the individual-level, the strong impact of need factors indicated equitable access to NHA, whereas differences in public spending for institutional care indicated inequitable access across European countries.

Partial Text

Against the backdrop of ageing societies in Europe, research on determinants of nursing home admission (NHA) in old age is important in order to assess what drives costly institutionalisation at the end of life and to which degree the comprehensive personal care services provided therein are allocated according to need (= equity)[1]. Two meta-analyses[2,3] found evidence for the following risk factors of NHA (classified according to[1]): older age (predisposing); living alone, low social support, being a tenant rather than a house owner (enabling); poor self-rated health, functional impairment, cognitive impairment, dementia, a high number of prescriptions, prior NHA (need factors). In contrast to these well-documented individual-level risk factors, we are not aware of any cross-national studies assessing the role of country-level characteristics for NHA, as existing studies have focussed exclusively on individual-level determinants (e.g.[4–10]). This is all the more surprising, since national differences in public expenditure and private out-of-pocket costs with regard to long-term institutional care in Europe have been repeatedly reported to be substantial[11–14]. For example, public expenditure on institutional care (2010) in Greece or Estonia was estimated to be below 0.2% of their gross domestic product (GDP), whereas this amounted to around 2% in Sweden or the Netherlands[15]. These differences in public expenditure result in a highly varied nursing and care home infrastructure[16], which ought to affect the chance of NHA in the last year of life. A recent study focussing on inpatient care facilities which includes hospital and hospice next to nursing home admissions, for example, reported considerable between-country variation in such admissions and suggested that these are at least partly driven by system-level or cultural factors[17]. Finally, it is unknown, whether or not the impact of the well-established individual-level determinants varies between countries, and if, whether this variation is linked to a country’s public institutional care infrastructure. For example, older adults in more generous public long-term care systems might access nursing home care sooner, that is, with fewer functional limitations compared to older adults in countries with a more restricted public institutional long-term care infrastructure.

51.8% of the total sample were men, and the average age at death was 79.9 (SD = 7.8, range = 65–104) for men and 82.7 (SD = 8.4, range = 65–104) for women. Sample size by country varied between 94 in Slovenia (waves 4–6 only) and 961 in Spain. Proxy-respondents were mostly partner (39.8%) or children (32.8%) of the deceased.

In this study, we analysed cross-national data from deceased survey respondents and from post-mortem proxy-interviews from 16 European countries using Bayesian hierarchical logistic regression models. We assessed both the impact of individual-level predisposing, enabling and need factors as well as of country of residence, respectively public long-term care expenditure on institutional care on proxy-reported NHA in the last year of life. We are not aware of any other study attempting to assess the impact of country-level characteristics on NHA using nationally-representative data from multiple countries so far. Assessing determinants of NHA in the last year of life is important, as institutional nursing care is the most expensive type of long-term care for the public, the clients and their relatives[11,12,14,15]. Access to nursing homes should be equitable[1], that is, primarily according to the need for care.




Leave a Reply

Your email address will not be published.