Place of Death
Trends Over the Course of a Decade: A Population-Based Study of Death Certificates From the Years 2001 and 2011
; ; ;
Background: In Germany, data on place of death is recorded from death certificates, but not further analyzed. Consequently, hardly any information is available at the population level regarding the distribution of place of death (e.g. home, hospital, palliative care unit, nursing home, hospice).
Methods: We carried out a descriptive statistical analysis of the registered places of death in evaluated death certificates from selected areas of Westphalia–Lippe for the years 2001 and 2011. Factors affecting the place of death were determined with binary multivariate regression.
Results: We analyzed 24 009 death certificates (11 585 for 2001 and 12 424 for 2011). The distribution of places of death for the overall population was as follows (2001 vs. 2011): at home, 27.5% vs. 23.0% (p<0.001); in the hospital, 57.6% vs. 51.2% (p<0.001); on a palliative care unit, 0.0% vs. 1.0%, in a care or nursing home, 12.2% vs. 19.0% (p<0.001); in a hospice, 2.0% vs. 4.6% (p<0.001); elsewhere, 0.6% vs. 0.6% (p = 0.985); not indicated, 0.1% vs. 0.6% (p<0.001). Independent factors affecting the place of death were age, sex, place of residence, and the presence of cancer or of dementia.
Conclusion: Most people in Germany die in institutions; the most common place of death is still the hospital, where more than half of all deaths take place. Only one death in four occurs at home. There is a marked secular trend away from dying at home or in the hospital, in favor of dying in a care or nursing home; death in palliative care units and hospices is also becoming more common.
According to information from the Federal Statistical Office (Statistisches Bundesamt), 852 328 persons died in Germany in the year 2011, out of a population of 80.3 million (1), indicating that the number of the deceased was 1.1% of the overall population.
Research on the subject of place of death shows that most people prefer to die at home (2–4). Dying in an institution such as a hospital or a care or nursing home is viewed negatively. Hospices and palliative care units are institutions that are consciously devoted to the issue of dignity in dying. Since the end of the 1990s, there has been a marked increase in the number of these care facilities in Germany.
In Germany, the place of death is recorded but not statistically analyzed. Official statistics on place of death relate only to hospital deaths (5).
The present study analyzes the places of death for natural deaths occurring during the years 2001 and 2011 in selected rural and urban areas of Westphalia on the basis of death certificate analysis. The aim is to draw conclusions about changes in the distribution of place of death over time. The study also determines sociodemographic factors that influence place of death.
The study is a cross-sectional study based on the analysis of death certificates from the years 2001 and 2011 in selected areas of Westphalia–Lippe, Germany.
The area studied includes the cities of Bochum and Münster and the rural areas of Borken and Coesfeld in the Westphalia–Lippe region (in the state of North Rhine–Westphalia). Both urban and rural areas were purposely included in the study in order to allow comparative analyses. Because of the requirements of data protection, data collection took place on site in the local health authority offices.
On 31 December 2000, 1 231 222 persons lived in the study region. Over the course of the 10-year observation period, the population rose by 1.03%. The populations of the city of Münster and the two rural areas of Borken and Coesfeld increased, while that of Bochum decreased (6). The number of hospital beds decreased from 8.1 to 7.8 per 1000 head of population, while the number of nursing home beds increased from 7.6 to 8.9 per 1000 population (7). In 2001, there were three hospices; by 2011 they had doubled in number to six. Three palliative care units were opened between 2001 and 2011.
Definition of place of death
In this study, the place of death was divided into the following categories: “home,” “hospital,” “palliative care unit,” “care or nursing home,” “hospice,” and “other.” The category “home” included both the private home of the deceased and other private homes that were not the deceased’s own home. Hospitals, psychiatric clinics, and sanatoriums were all subsumed under the category “hospital.” Palliative care units were given a category of their own. The category “care or nursing home” included all facilities provided by care homes, nursing homes, assisted living residences, and short-term care. “Other” included (among others) public places, doctors’ offices, and leisure facilities.
At the time when the study was carried out, all death certificates from 2001 and 2011 were available for the study. None of the documents had been destroyed because their archiving period had expired. Data collected were: place of death, mode of death, sociodemographic details, and medical notes mentioning malignant disease or dementia as underlying or contributory cause of death. The analysis included all deaths for which a natural cause of death was recorded in the death certificate. Documents for which no information was provided about age or sex were excluded from the analysis. This was the case for 126 death certificates. There were grounds for suspecting an unnatural cause of death in 742 cases.
The study population was investigated for differences over time (2001 vs. 2011) in terms of sociodemographic details and the distribution of places of death. In addition, both the crude and the age-adjusted mortality rate were calculated. The frequency distributions of places of death were represented both in percentages and as absolute numbers. Stratification by age and sex was carried out.
Continuous variables were tested for normal distribution using the Kolmogorov–Smirnov test and, if a normal distribution was confirmed, analyzed using the unpaired t-test. If the data was not normally distributed, the nonparametric Wilcoxon–Mann–Whitney test was used. Categorical variables were investigated using the χ2 test. To minimize the global increase in the probability of a type 1 error (alpha) due to multiple testing of the same sample, the significance level was set at p<0.01 (two-sided).
Regarding mortality rates, direct adjustment for age was carried out. The European Standard Population of 1976 (ESP1976) was used as the standard population.
Effect sizes on the place of death were calculated by means of binary logistic regression, with place of death as the dependent variable. The following were included as independent variables in the statistical model:
- Place of residence
- Malignant disease
- Dementia (when recorded in the death certificate as underlying or contributory cause of death).
Odds ratios (OR) with their concomitant 95% confidence intervals (CI) were generated as the measurement units of the multivariate analyses. SPSS version 21 was used for the analysis.
Data protection and ethical approval
Access to archived death certificates, retrieval of the information contained in them, and the analysis of that information were permitted by the relevant health authorities on condition that data protection regulations were followed. The study was approved by the ethics committee of the Ruhr-Universität Bochum (decision no. 4338–12).
A total of 24 009 death certificates were analyzed, 11 585 from the year 2001 and 12 424 from 2011. The average age at death of the study population was 75.1 years in 2001 and within 10 years rose to 77.2 years. In 2001, men died at a mean age of 70.9 years and women at 78.7. Ten years later, the mean age at death for men was 73.8 years and that for women 80.4 years. The proportion of men and women was almost equal, with slightly more women than men dying. Malignant disease was recorded in one in three death certificates. The percentage of those in which dementia was recorded rose from 6.3% in 2001 to 13.2% in 2011 (Table 1).
The crude mortality rate of death from natural causes rose over the 10-year observation period from 940.9 to 998.7 deaths per 100 000 head of population. The age-adjusted death rate decreased from 723.2 to 648.3 deaths per 100 000 population.
Between 2001 and 2011, the rate of deaths at home fell by 4.5%. In 2011, fewer than 1 in 4 people (23%) died at home. Hospitals remained the most common place of death, with over 50%, but here too a drop of 6.4% was observed. In contrast to this, in 2011 more people died on a palliative care unit or in a hospice. In 2011, the proportion of those who died in a hospice was less than 5%. Care and nursing homes showed the largest rise as place of death (12.2% vs. 19.0%; p<0.001). Thus, in 2011 almost one person in five died in a care or nursing home (Table 2).
In absolute numbers, more men died at home, in hospital, or in “other places.” In contrast to this, there were more female deaths in care or nursing homes and in hospices. Stratification by age showed that male deaths predominated until the age of 80. Deaths at 80 years or older were predominantly among women (Table 3).
Men were 1.2 times more likely than women to die at home. In contrast, the likelihood of dying in a care or nursing home was 50% lower for men than for women. For hospital patients, those below the age of 80 were twice as likely to die in hospital compared to those aged 80 or over. People who lived in an urban area were more likely to die in hospital than those living in a rural area, and less likely to die at home or in a care or nursing home.
Malignant disease was strongly associated with the probability of dying in a hospice. Likewise, there was a strong correlation between dementia and an increased probability of dying in a care or nursing home (Table 4).
This study showed that hospital is by far the most common place of death, followed by the deceased’s own home, a care or nursing home, and a hospice or palliative care unit. In the comparison over time (2001 vs. 2011), in 2011 fewer people died at home or in hospital, more people in a care or nursing home, a hospice, or a palliative care unit. The greatest increase was in the number of those who died in care or nursing homes.
Research into place of death in Germany
Some scattered information already exists from population-based studies of place of death in Germany (8–19). Table 5 provides an overview of these studies. The most comprehensive study so far is that by Ochsmann et al. (10). This study analyzed 12 217 death certificates from the year 1995 in selected health authority areas in Rheinland–Palatinate. No information was available for hospices and palliative care units. In the study area, during the observation period, 39.8% died at home, 44.1% in hospital, 12.8% in a care home, and 1.7% in another place. In 1.7% of death certificates, no details were given of the place of death. At 39.8%, the percentage of deaths that took place at home is notably high.
International studies confirm the predominance of deaths in institutions, but differ as to the information they give (20–36). Table 6 briefly summarizes important international studies on place of death (Table 6). Cohen et al. (28) investigated hospital deaths as a proportion of all deaths during 2003 in various European countries. Percentages varied from 33.9% (Netherlands) to 62.8% (Wales). The care or nursing home is becoming increasingly important as a place of death, as is shown by an example from Belgium (27) showing that, from 1998 to 2007, the percentage of deaths that took place in a care or nursing home rose from 18.3% to 22.6%. The percentages of deaths occurring at home also varied between studies, but most lay between 20% and 30% (20–22, 24–27, 30–32). Over the past 40 years, various tendencies in the distribution of place of death have been observed. For example, the number of deaths at home increased in Canada (31) and the USA (33), whereas in most European countries it decreased. At the same time, between 1994 and 2004, in Canada the frequency of deaths in hospital fell from 77.7% to 60.6%, and between 1980 and 1998 in the USA it fell from 54% to 41%. Investigations by Gomes et al. (25) show a steady drop in deaths at home in England and Wales since the 1970s, down to 18.3% in 2004. In England, the National End of Life Care Programme (2003–2006) was given special financial support, partly in order to counteract this trend, and indeed the percentage of deaths at home rose again to 20.8% in 2010.
Factors influencing place of death
In the present study, the likelihood of dying at home or hospital was greater for men; women were more likely to die in a care home or hospice. Because of their lower life expectancy, men living in a relationship have a higher probability of being survived by their partner, and this increases their chance of being cared for at home by relatives at the end of their life. The corollary of this is that, living to a greater age, women are more likely to be widowed or living alone in old age, and to need outside or institutional help as they become more debilitated. The findings of the 2011 microcensus by the Federal Statistical Office show a steady rise in the number of persons living alone in Germany (37), which had gone up by around 4 million since 1991. At the age of 80, more than half of women (56%) were living alone, against only 22% of men.
Compared to those living in the rural areas, people living in urban areas were more likely to die in hospital. On the other hand, their probability of dying at home or in a care or nursing home was lower. Differences in the provision of care may explain the difference. If the number of hospital beds is small, as in the rural areas selected for this study, the probability of dying at home will rise. If the availability of long-term care, hospice, and palliative care beds rises, they will increasingly be occupied. The percentage of those living alone can also have an effect here, since especially in urban areas many elderly people live on their own.
As would be expected, a clear association was found between the presence of malignant disease and an increased probability of dying in a hospice. Likewise, people with dementia more often died in care or nursing homes. Study data indicate that patients with non-solid tumors (leukemia, lymphoma) have an increased risk of dying in hospital (38).
Finally, public health cost considerations may be related to the observed distribution of place of death—specifically, with the fall in the number of hospital deaths. The growing pressure of cost containment in the hospital sector could be one factor that has led to a reduction in lengths of hospital stay since the introduction of the DRG (diagnosis-related group) funding system in German health care; but has at the same time led to earlier discharge of patients for whom no further inpatient treatment options remain.
Strengths and limitations of this study
A strength of this study is the large number of death certificates analyzed (n = 24 009)—the largest dataset for any study on this topic so far in Germany. Only 0.4% of the death certificates included lacked information about the place of death. The details given about the place of death were illegible or ambiguous only in a few rare cases; overwhelmingly, they were valid. No systematic misclassification in these few cases was identifiable. With regard to cause of death, the study looked only at whether malignant disease or dementia was present. It is not clear when the patients who died in hospital were admitted, or whether the admission was in fact pre-terminal for the purposes of life-preserving treatment. There is a limit to how far generalizations can be made on the basis of this study, because it is restricted to selected areas of Westphalia–Lippe and did not aim to represent the overall population of Germany.
Death occurs predominantly in institutions. It may be anticipated—not least owing to demographic changes, with increased life expectancies and a rise in the absolute number of deaths—that over the next few decades (40) dying in institutions will at least remain steady or, more likely, will increase.
At the same time, however, the hospice movement and palliative medicine are active in support of dying at home. These two areas of medicine are steadily becoming more accepted and valued by society in Germany. Against this background, it will be interesting to see whether this development even leads to a reversal of the trend: away from institutionalized dying and toward dying at home.
Data collection was ably supported by Laila Boutakmant, Hartmut Hofmeister, Marievonne Hofmeister, and Viola Willeke. Thanks are also due to the chief executives of the health authorities in Bochum, Borken, Coesfeld, and Münster.
Conflict of interest statement
Mr. Blum has had conference fees and travel and accommodation costs reimbursed by Mundipharma.
Dr. Dasch, Dr. Gude, and Prof. Bausewein declare that no conflict of interest exists.
Manuscript received on 20 February 2015, revised version accepted on
22 April 2015.
Translated from the original German by Kersti Wagstaff, MA.
Dr. Burkhard Dasch; MSE, MPH
Klinik und Poliklinik für Palliativmedizin
Klinikum der Universität München
Marchioninistr. 15, 81377 Munich, Germany
|1.||Statistisches Bundesamt. Gesundheit: Todesursachen in Deutschland. Fachserie 12 Reihe 4–2013. Statistisches Bundesamt, Wiesbaden 2013. www.destatis.de/DE/Publikationen/Thematisch/Gesundheit/AlteAusgaben/TodesursachenAlt.html (last accessed on 18 February 2015).|
|2.||Escobar Pinzon LC, Claus M, Zepf KI, Letzel S, Fischbeck S, Weber M: Preference for place of death in Germany. J Palliat Med 2011; 14: 1097–103 CrossRef MEDLINE|
|3.||Gomes B, Higginson IJ, Calanzani N, et al.: PRISMA: Preferences for place of death if faced with advanced cancer: a population survey in England, Flanders, Germany, Italy, the Netherlands, Portugal and Spain. Ann Oncol 2012; 23: 2006–15 CrossRef MEDLINE|
|4.||Gomes B, Calanzani N, Gysels M, Hall S, Higginson IJ: Heterogeneity and changes in preferences for dying at home: a systematic review. BMC Palliat Care 2013; 15: 7 CrossRef MEDLINE PubMed Central|
|5.||Federal Statistical Office, Germany. Statistisches Bundesamt. Gesundheit: Diagnosedaten der Patienten und Patientinnen in Krankenhäusern (einschl. Sterbe- und Stundenfälle) Fachserie 12 Reihe 6.2.1–2011. Statistisches Bundesamt, Wiesbaden 2013. www.destatis.de/DE/Publikationen/Thematisch/Gesundheit/Krankenhaeuser/DiagnosedatenKrankenhaus2120621117004.pdf?__blob=publicationFile (last accessed on 18 February 2015).|
|6.||State Office of Statistics, North Rhine-Westphalia. Information und Technik Nordrhein-Westfalen (IT.NRW). Publikationen – Bevölkerung in NRW; Bevölkerungsstand, Bevölkerungsbewegung. https://webshop.it.nrw.de/ssearch.php?kategorie=1200&prefix=A10 (last accessed on 18 February 2015).|
|7.||State Office of Statistics, North Rhine-Westphalia. Information und Technik Nordrhein-Westfalen (IT.NRW). Publikationen – Krankenhäuser und Vorsorge- oder Rehabilitationseinrichtungen in NRW. https://webshop.it.nrw.de/qsearch.php?keyword=A41 (last accessed on 18 February 2015).|
|8.||Weber M, Schütz J, Ochsmann R, Huber C: Sterbeorte in Mainz 1951–1995. Beiträge zur Thanatologie. Mainz: Johannes Gutenberg Universität Mainz 1997; Heft 12.|
|9.||Bickel H: Das letzte Lebensjahr. Eine Repräsentativstudie an Verstorbenen. Wohnsituation, Sterbeort und Nutzung von Versorgungsangeboten. Z Gerontol Geriat 1998; 31: 193–204 CrossRef MEDLINE|
|10.||Ochsmann R, Slangen K, Feith G, et al.: Sterbeorte in Rheinland-Pfalz. Zur Demographie des Todes. Interdisziplinärer Arbeitskreis Thanatologie. Beiträge zur Thanatologie. Mainz: Johannes Gutenberg Universität Mainz 1997; Heft 8.|
|11.||van Oorschot B, Klak L, Reichert L, Hildenbrand B, Wendt TG: Sterbeorte in Jena im Jahre 1999. www.hospiz-jena.de/pages/forschung/sterbeorte.html (last accessed on 18 February 2015).|
|12.||Papke J, Koch R: Places of death from cancer in a rural location. Onkologie 2007; 30: 105–8 CrossRef MEDLINE|
|13.||Kaluza J, Töpferwein G, eds.: Sterben begleiten. Zur Praxis der Begleitung Sterbender durch Ärzte und Pflegende. Eine empirische Studie. Berlin: Trafo Verlag 2006.|
|14.||Rodenstein M: Sterben in einer Großstadt – am Beispiel Frankfurt am Main. In: Bürgerinstitut e. V., Bautsch R, (eds.): Sterben in der Großstadt. Ein Beitrag zur Hospizgeschichte in Frankfurt am Main. Frankfurt am Main: Mabuse-Verlag 2008; 34–50.|
|15.||Nick MB: Todesumstände und Todesursachen der Bevölkerung der Gemeinden des Landkreises Harburg/Niedersachsen. Dissertationsschrift 2006. http://ediss.sub.uni-hamburg.de/volltexte/2007/3333/pdf/DissertationMNick.pdf (last accessed on 18 February 2015).|
|16.||Escobar Pinzón LC, Weber M, Claus M, et al.: Factors influencing place of death in Germany. J Pain Symptom Manage 2011; 41: 893–903 CrossRef MEDLINE|
|17.||Escobar Pinzon LC, Claus M, Perrar KM, Zepf KI, Letzel S, Weber M: Dying with dementia: symptom burden, quality of care, and place of death. Dtsch Arztbl Int 2013; 110: 195–202 VOLLTEXT|
|18.||Sauer S, Müller R, Rothgang H: Institutionalised dying in Germany. Trends in place of death distribution: home, hospitals and nursing homes. Z Gerontol Geriatr 2013; 24: 1–6.|
|19.||Jaspers B, Schindler T (2004): Stand der Palliativmedizin und Hospizarbeit in Deutschland und im Vergleich zu ausgewählten Staaten (Belgien, Frankreich, Großbritannien, Niederlande, Norwegen, Österreich, Polen, Schweden, Schweiz, Spanien). Gutachten im Auftrag der Enquete-Kommission des Bundestages „Ethik und Recht der modernen Medizin“. www.ethikrat.org/themen/medizin-und-pflege/palliativmedizin (last accessed on 18 February 2015).|
|20.||Schweizerische Eidgenossenschaft. Bundesamt für Gesundheit (BAG): Themen- Gesundheitspolitik-Strategie Palliative Care-Statistische Daten. www.bag.admin.ch/themen/gesundheitspolitik/13764/13777/13793/index.html?lang=de (last accessed on 18 February 2015).|
|21.||Bundesanstalt Statistik Österreich, Wien. STATcube-Statistische Datenbank von Statistik Austria. http://statcube.at/superwebguest/autoLoad.do?db=degestorbene_ext (last accessed on 18 February 2015).|
|22.||Gao W, Ho YK, Verne J, Gordon E, Higginson IJ: Geographical and temporal understanding in place of death in England (1984–2010): analysis of trends and associated factors to improve end-of-life care (GUIDE-Care) – primary research. Health Services and Delivery Research 2014; 2.|
|23.||Gao W, Ho YK, Verne J, et al.: Changing patterns in place of cancer death in England: a population-based study. PLoS Med 2013; 10: e1001410 CrossRef MEDLINE PubMed Central|
|24.||National End Of Life Care Intelligence Network: Variations in place of death in England. Inequalities or appropriate consequences of age, gender and cause of death? August 2010. file:///C:/Users/hp/Downloads/Variationsinplaceofdeath_FINAL_forlaunchAugust2010_1%20(5).pdf (last accessed on 18 February 2015).|
|25.||Gomes B, Calanzani N, et al.: Reversal of the British trends in place of death: time series analysis 2004–2010. Palliat Med 2011; 26: 102–7 CrossRef MEDLINE|
|26.||Gomes B, Higginson IJ: Where people die (1974–2030): past trends, future projections and implications for care. Palliat Med 2008; 22: 33–41 CrossRef MEDLINE|
|27.||Houttekier D, Cohen J, Surkyn J, et al.: Study of recent and future trends in place of death in Belgium using death certificate data: a shift from hospitals to care homes. BMC Public Health 2011; 11: 228 CrossRef MEDLINE PubMed Central|
|28.||Cohen J, Bilsen J, Addington-Hall J: Population-based study of dying in hospital in six European countries. Palliat Med 2008; 22: 702–10 CrossRef MEDLINE|
|29.||Cohen J, Houttekier D, Onwuteako-Philipsen B, et al.: Which patients with cancer die at home? A study of 6 European countries using death certificate data. J Clin Oncol Res 2010; 28: 2267–73 CrossRef MEDLINE|
|30.||Loucka M, Payne SA, Brearley G, EURO IMPACT: Place of death in the Czech Republic and Slovakia: a population based comparative study using death certificates data. BMC Palliat Care 2014; 13: 13 CrossRef MEDLINE PubMed Central|
|31.||Wilson DM, Truman CD, Thomas R, et al.: The rapidly changing location of death in Canada, 1994–2004. Soc Sci Med 2009; 68: 1752–8 CrossRef MEDLINE|
|32.||U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics. Health, United States, 2010: with special features on death and Dying. Hyattsville (MD); 2011 Feb. Report No. 2011–1232. www.cdc.gov/nchs/data/hus/hus10.pdf (last accessed on 18 February 2015).|
|33.||Teno JM, Gozalo PL, Bynum JP, et al.: Change in end-of-life care for medicare beneficiaries: site of death, place of care, and health care transitions in 2000, 2005, and 2009. JAMA 2013; 6: 309: 470–7 CrossRef MEDLINE PubMed Central|
|34.||Higginson IJ, Sarmento VP, Calanzani N, Benalia H, Gomes B: Dying at home-is it better: a narrative appraisal of the state of the science. Palliat Med 2013; 27: 918–24 CrossRef MEDLINE|
|35.||Fischer S, Bosshard G, Zellweger U, Faisst K: Place of death: where do people die in Switzerland nowadays?. Z Gerontol Geriatr 2004; 37: 467–74 CrossRef MEDLINE|
|36.||Broad JB, Gott M, Kim H, Boyd M, Chen H, Connolly MJ: Where do people die? An international comparison of the percentage of deaths occurring in hospital and residential aged care settings in 45 populations, using published and available statistics. Int J Public Health 2013; 58: 257–67 CrossRef MEDLINE|
|37.||Federal Statistical Office, Germany. Statistisches Bundesamt: Alleinleben in Deutschland. Ergebnisse des Mikrozensus 2011. Statistisches Bundesamt, Wiesbaden 2012. www.destatis.de/DE/PresseService/Presse/Pressekonferenzen/2012/Alleinlebende/begleitmaterial_PDF.pdf?__blob=publicationFile (last accessed on 18 February 2015).|
|38.||Gomes B, Higginson IJ: Factors influencing death at home in terminally ill patients with cancer: systematic review. BMJ 2006; 332: 515–21 CrossRef MEDLINE PubMed Central|
|39.||Thönnes M: Sterbeorte in Deutschland: Eine soziologische Studie. Frankfurt: Peter Lang Verlag 2013.|
|40.||Simon ST, Gomes B, Koeskeroglu P, Higginson IJ, Bausewein C: Population mortality and place of death in Germany (1950–2050)—implications for end-of-life care in the future. Public Health 2012; 126: 937–46 CrossRefMEDLINE|