Place of Death with Dementia--Dutch People Rarely Die in the Hospital.

Abstract

In the Netherlands, the majority of people with dementia die in a long-term care facility (LTCF).1 People with dementia2, 3 and care professionals4 prefer that people with dementia die at home. The objective of this study was to examine the personal characteristics associated with location of death for people with dementia. Data were from the 2006 Statistics Netherlands cause of death registry.5 Individuals who were aged 65 and older older at death; who died at home, in a hospital, or in a psychiatric unit, nursing home, or elderly home (LTCF); and for whom one or more of the causes of death was Alzheimer's disease (International Classification of Diseases, Tenth Revision (ICD-10) code G30), vascular dementia (F01), or unspecified dementia (F03)) were included. The data were analyzed using a multinomial logistic regression model with location of death categorized into three categories as the dependent variable. First, univariate multinomial logistic regression was performed to examine associations between location of death and each of the independent variables. Variables that had a relative risk ratio (RRR) with P < .10 were entered into the multivariate model. Independent variables were then removed from the model one at a time through a manual backward elimination procedure, starting with the variable with the highest P-value, until all variables had P < .05. Participant characteristics are described in Table 1. In summary, 707 (4%) persons died at home, 756 (4%) in a hospital, and 16,351 (92%) in a LTCF (14,692 (90%) in a nursing home, 1,659 (10%) in an elderly care home). After correcting for marital status, type of household, and income, it was determined that people aged 75–84 were more likely than those aged 65–74 to die in a hospital or LTCF than at home. People who were married or cohabitating were less likely to die in a hospital or a LTCF than those who were not married, cohabitating, or widowed. The relative risk ratio of dying in the hospital or LTCF compared with at home was higher for subjects who lived in an institution than for those living at home. People with an income of 14,410 euro or more were more likely to die at home than those with an income of less than 14,410 euro. Age, marital status, living status, and income were all associated with place of death. In the Netherlands, the proportion of persons dying in a LTCF was substantially higher than in previous studies in the United Kingdom6 and the United States.7 The differences in how medical care is organized in nursing homes may explain this. Dutch nursing homes have in-home physicians and skilled nurses who focus on supportive care. In contrast, elderly care homes are more like nursing homes in an international context and have nurse helpers and external physicians, with a focus on assisted living. Thus, Dutch nursing homes are potentially better equipped to offer medical and palliative care until death without hospital admission than nursing homes in other countries because of the in-home physicians. Strengths of this study include its large sample size and the ability to examine hospitalization data and identify transitions in living situation. Nursing home residents generally require more-complex care than elderly care home residents. Therefore, a possible limitation is that nursing and elderly care homes were not separately distinguished, but this was done because there was not a difference in characteristics between these groups (data not shown). In the Netherlands, most people with dementia as a cause of death died in a LTCF, suggesting that end-of-life care for persons with dementia is mostly handled within LTCFs. Because of new Dutch legislation, fewer people are admitted into LTCF, so more people are expected to stay at home longer. Future studies should investigate whether this leads to an increase in people with dementia dying in the hospital. We would like to thank Statistics Netherlands for their help. Author Contributions: Vroomen, Bosmans: design, analysis and interpretation of data, drafting and critical revision of article, final approval for publication. Holman: design, analysis and interpretation of data, critical revision of article, final approval for publication. van Rijn: data management, critical revision of article, final approval for publication. Buurman: critical revision of article, final approval for publication. van Hout: analysis, critical revision of article, final approval for publication. De Rooij: conception and design, interpretation of data, critical revision of article, final approval for publication. Sponsor's Role: None.