Abstract
Background
The largest German study on place of death was performed for deaths in selected regions of Westphalia in the years 2001 and 2011. In the period thereafter, provision of palliative care was regionally expanded. This upgrade included the establishment of palliative medicine consultation services (PMCS), combining general and specialized palliative care on an outpatient basis. A follow-up place of death survey took place in 2017. The aim was to depict the trends in place of death between 2001 and 2017. A second goal was to determine how often outpatient PMCS were used by persons who died in 2017.
Methods
Descriptive analysis of place of death as specified in all death certificates (2001, 2011, 2017) issued in the cities of Bochum and Münster and the districts of Borken and Coesfeld. Comparison of pseudonymized data on deceased patients (2017) treated by the PMCS of Münster and Coesfeld with the place of death database to ascertain the rate of PMCS care at the end of life.
Results
A total of 38 954 death certificates were analyzed, and 5887 deaths were compared with PCMS data. The distribution of place of death was as follows: (2001, 2011, 2017; age standardized; %): own residence 27.8; 23.3; 21.3; hospital: 55.8; 51.8; 51.8; palliative care unit: 0.0; 1.0; 6.2; hospice: 1.9; 4.5; 4.8; nursing home: 13.1; 18.6; 20.4; other: 1.2; 1.2; 1.5. The rate of PMCS use was 28.8% (1694/5887).
Conclusion
Over the period 2001–2017, the proportion of people who died at home or in the hospital went down, while the number who died in a palliative care unit, hospice, or nursing home increased. In the city of Münster and the district of Coesfeld, one fourth of the people who died in 2017 received PMCS care at the end of life.
Every year, slightly more than one percent of the German population of more than 83 million residents die (1). Representative surveys on “Dying in Germany“ have shown that the vast majority of Germans wish to die at home (2012/2017): at home: 66/57%, specialized facility for end-of-life care: 18/27%; hospital: 3/4%; nursing home: 1/1%, no information: 12/10% (2). This result is in line with international studies on the preferred place of death (3– 5).
In actual fact, only a minority of people die outside of institutions in Western industrialized countries (table 1) (6– 16). In Germany, the largest place of death studies were conducted in the years 1995 (17) and 2001/2011 (18). They found that 39.8% and 27.8%/23.3%, respectively, died at home, while 56.9% and 70.8%/75.9% (1995 and 2001/2011, respectively) died in institutions.
Table 1. Places of death—international.
| Country | Year |
Population (N) [in million] |
Deaths (N) | At home (%) | Hospital (%) | Nursing home (%) | Hospice (%) | Other (%) |
| Austria (6) | 2016 | 8.7 | 80 669 | 26.4 | 49.1 | 18.6 | n.s. | 4.1 |
| Switzerland (7) | 2010 | 7.9 | 61 786 | n.s. | 39.8 | 40.8 | n.s. | k.A. |
| Belgium (8) | 2007 | 10.7 | 65 435*1 of 101 384 | 22.5 | 51.7 | 22.6 | n.s. | 3.1 |
| France (9) | 2018 | 65.0 | 609 648 | 23.6 | 53.3 | 12.7 | n.s. | 1.1 |
| Portugal (10) | 2010 | 10.6 | 105 471 | 29.6 | 61.7 | 8.6 | ||
| Czech Republic (11) | 2011 | 10.5 | 102 385 | 20.4 | 58.4 | 12.2 | n.s. | n.s. |
| Slovakia (11) | 2011 | 5.4 | 51 903 | 30.0 | 54.8 | 3.0 | n.s. | n.s. |
| Ireland (12) | 2018 | 4.8 | 31 140 | 22.5 | 51.4 | 15.7 | 7.6 | 2.7 |
| England (13) | 2019 | 56.3 | 505 859 | 24.4 | 44.9 | 22.5 | 5.8 | 2.5 |
| Sweden (14) | 2012 | 9.5 | 83 712*2 of 91 938 | 17.8 | 43.1 | 38.1 | n.s. | n.s. |
| Canada (15) | 2019 | 37.4 | 284 706 | n.s. | 58.9 | k.A. | n.s. | n.s. |
| USA (16) | 2017 | 325.1 | 2.3 *3 of 2.8 | 30.7 | 29.8 | 20.8 | 8.3 | n.s. |
*1 All deaths aged = 1 year in the region of Flanders and the city of Brussels; *2 deaths in Sweden with registered place of death; *3 natural deaths; n.s.; not stated
In the modern hospice and palliative care movement, the place of death is regarded as a quality indicator for needs-oriented care of the dying. Usually, a distinction is made between general and specialized palliative care on an outpatient basis. It is estimated that about 10% to 20% of all dying persons require specialized palliative care (19, 20).
In the Westphalia-Lippe region of Germany, the Association of Palliative Care Physicians in Westphalia-Lippe, various health insurances and the Association of Statutory Health Insurance Physicians in Westphalia-Lippe signed a palliative care contract which combines general and specialized palliative care on an outpatient basis, a unique approach in Germany (21). The aim of the contract was to provide the broadest possible spectrum of palliative care on an outpatient basis. Specifically, the coordinating general practitioner agrees to collaborate with a so-called palliative medicine consultation service (PMCS). The PMCS meets both staffing and professional quality standards and additionally is accessible 24/7 to ensure the availability of palliative medicine consultation services whenever they are needed. However, general practitioners participating in the care contract are excluded from billing specific billable items of the Uniform Value Scale (EBM) fee catalogue which are related to palliative home care.
The data presented here are results from a follow-up survey of the study on place of death conducted by Dasch et al. (18) for the year 2017.
The aim of this study was to depict the trends in place of death between 2001 and 2017. A second aim was to determine, based on pseudonymized PMCS data, how often outpatient PMCS were utilized at the end of life by persons who died in 2017.
Methods
Study design
The study was designed as a population-based cross-sectional survey and is a trend study, depicting changes in the distribution of place of death at various points in time.
Study region
The study region comprised selected regions of Westphalia (the cities of Bochum and Münster and the districts of Coesfeld and Borken). The structural data of the study region are summarized in Table 2. The study regions were selected to allow for urban-rural comparisons. In addition, for reasons of data protection, data collection had to be undertaken in the offices of the local health authorities. This task was performed by the principal investigator and study staff. In order to keep travelling distances to a minimum, regions in close proximity to where the team members lived were selected.
Table 2. Structural data of the study region.
| 2001 | 2011 | 2017 | |
| Population (in 1 000) | 1282 | 1298 | 1343 |
| Hospitals | 30 | 28 | 26 |
| Hospital beds | 9985 | 9805 | 9865 |
| Palliative care units | 0 | 7 | 13 |
| Palliative care beds | 0 | 40 | 79 |
| Palliative care beds per 1 million inhabitants | 0 | 30.8 | 58.8 |
| Hospice | 3 | 6 | 6 |
| Hospice beds | 27 | 50 | 52 |
| Hospice beds per 1 million inhabitants | 21.1 | 38.5 | 38.7 |
| Nursing homes | 122 | 166 | 209 |
| Nursing home beds | 9592 | 11 276 | 12 991 |
| Palliative medicine consultation service (PMCS)* | 0 | 5 | 6 |
* PMCS year established: Bochum (2007), Münster (2006), Borken I + II (2010) + III (2015), Coesfeld (2010)
Definition of place of death
The place of death was divided into the following categories: at home, hospital, palliative care unit, hospice, nursing home, other. For detailed information about the definition of place of death, please refer to the initial study (18).
Death certificates
The analysis is based on the full dataset of archived death certificates for the years 2001, 2011 and 2017. Data on the following variables were collected: place of death, date of death, time of death, place of residence, age, gender, mode of death, selected information about the cause of death (such as tumor disease, dementia, among others). All information about place of death was checked for plausibility.
Pseudonymized data of selected PMCS
All PMCS were asked to provide pseudonymized data (gender, age of death, date of death) of deceased patients who died in 2017. The PMCS of Münster and Coesfeld promised to support the study, in keeping with data protection requirements. The other PMCS declined to participate, citing time constraints or changes in data protection requirements as reasons. It was possible to compare over 90% of the data with the place of death database. Most cases where matching failed were due to errors in the documentation of the date of death.
Statistical analyses
The distribution of place of death was depicted both in absolute numbers and percentages and tested for differences between the survey years 2001, 2011 and 2017. The place of death was stratified for gender and age. Deaths were tested for differences over time, focusing on the variables age, gender and place of residence. The following statistical tests were used: continuous data, parametric > unpaired t-test, non-parametric > Mann–Whitney U test; categorical data > chi-square test. In order to minimize the multiple comparison-related increase in the chance of making a type I error, the significance level of p = 0.05 was subjected to a Bonferroni correction. We carried out a direct age standardization to adjust for confounding by age. We calculated the expected deaths in the 2001, 2011 and 2017 populations based on the assumption that these populations had the same age distribution as the standard population. The standard population was defined by us as the study population of all three middle-aged cohorts (2001, 2011 and 2017).
The frequency of PMCS care for persons at the end of life (2017) was determined based on pooled data from the regions Münster and Coesfeld. This was calculated as the ratio of deaths with PMCS care to total deaths.
The statistical software package SPSS, version 25, was used to analyze the data.
Ethics committee approval
The study was submitted to the Ethics Committee of the Ruhr University Bochum (Germany) and approved after review. (Registration no. 17–6330).
Results
Deaths
A total of 38 954 death certificates were analyzed. The absolute number of deaths increased continuously over the years (11 963, 12 914, 14 077).
On average, women died at the age of 79.6 years and men at the age of 72.5 years. The mean age at death increased from 2001 to 2017 in women by 2.6 years, in men by 4.7 years. The proportion of persons aged over 80 years increased from 42.5% (2001) to 54.3% (2017) (table 3).
Table 3. Characterization of deaths by age, gender and place of residence 2001. 2011. 2017.
| 2001 | 2011 | 2017 | Total | ||||
| N=11 963 | 2001 vs. 2011 | N=12 914 | 2001 vs. 2017 | N=14 077 | 2011 vs. 2017 | N=38 954 | |
|
Women [95% CI] (N) |
53.1% [52.2; 54.0] (6355) |
Χ2 (1) = 9.928 p = 0.007 φ = 0.020 |
51.3 % * [50.4; 52.2] (6626) |
Χ2 (1) = 16.277 p < 0.001 φ = 0.025 |
51.3% [50.5; 52.1] (7227) |
Χ2 (1) = 1.968 p = 0.374 φ = 0.009 |
51.8% [51.3; 52.3] (20 188) |
|
Men [95% CI] (N) |
46.9% [46.0; 47.8] (5614) |
Χ2 (1) = 9.928 p = 0.007 φ = 0.020 |
48.7 % * [47.8; 49.6] (6281) |
Χ2 (1) = 16.277 p < 0.001 φ = 0.025 |
48.7% [47.9; 49.5] (6847) |
Χ2 (1) = 1.968 p = 0.374 φ = 0.009 |
48.2 % [47.7; 48.7] (18 742) |
|
Age, total M +/– SD [95% CI] (N) |
74.1 +/–16.3* (U)
[73.8; 74.4] (11 942) |
p < 0.001 r = 0.074 |
76.5 +/–14.7*(U)
[76.2; 76.8] (12 833) |
p < 0.001 r = 0.122 |
77.6 +/– 15.0*(U)
[77.3; 77.8] (13 997) |
p < 0.001 r = 0.051 |
76.2 +/−15.4 [76.0; 76.3] (38 772) |
|
Age, women M +/– SD [95% CI] (N) |
78.0 +/–15.7* (U)
[77.6; 78.4] (6326) |
p < 0.001 r = 0.052 |
79.9 +/–13.9*(U)
[79.5; 80.2] (6590) |
p < 0.001 r = 0.092 |
80.6 +/–14.0*(U)
[80.3; 80.9] (7192) |
p < 0.001 r = 0.041 |
79.6 +/−14.6 [79.4; 79.8] (20 108) |
|
Age, men M +/– SD [95% CI] (N) |
69.7 +/–15.9* (U)
[69.3; 70.1] (5603) |
p < 0.001 r = 0.123 |
73.0 +/–14.8*(U)
[72.6; 73.3] (6236) |
p < 0.001 r = 0.182 |
74.4 +/–15.3*(U)
[74.0; 74.7] (6804) |
p < 0.001 r = 0.069 |
72.5 +/−15.4 [72.3; 72.7] (18 643) |
|
Age < 18 years [95% CI] (N) |
1.2 % * [1.0; 1.4] (147) |
Χ2 (1) = 17.814 p < 0.001 φ = 0.027 |
0.7 % [0.6; 0.8] (91) |
Χ2 (1) = 9.996 p = 0.002 φ = 0.020 |
0.9% [0.7; 1.1] (127) |
Χ2 (1) = 1.427 p = 0.235 φ = 0.007 |
0.9% [0.8; 1.0] (365) |
|
Age 18–39 years [95% CI] (N) |
2.4 % * [2.1; 2.7] (287) |
Χ2 (1) = 37.161 p < 0.001 φ = 0.039 |
1.3 % * [1.1; 1.5] (174) |
Χ2 (1) = 29.992 p < 0.001 φ = 0.034 |
1.5% [1.3; 1.7] (207) |
Χ2 (1) = 0.644 p = 0.438 φ = 0.005 |
1.7 % [1.6; 1.8] (668) |
|
Age 40–64 years [95% CI] (N) |
18.4 % * [17.7; 19.1] (2206) |
Χ2 (1) = 43.906 p < 0.001 φ = 0.042 |
15.2 % * [14.6; 15.8] (1966) |
Χ2 (1) = 98.432 p < 0.001 φ = 0.062 |
13.9 % * [13.3; 14.5] (1951) |
Χ2 (1) = 10.243 p = 0.001 φ = 0.020 |
15.7% [15.3; 16.1] (6123) |
|
Age 65–79 years [95% CI] (N) |
35.3 % * [34.4; 36.2] (4220) |
Χ2 (1) = 37.739 p < 0.001 φ = 0.039 |
31.5 % * [30.7; 32.3] (4062) |
Χ2 (1) = 112.608 p < 0.001 φ = 0.066 |
29.0 % * [28.3; 29.7] (4080) |
Χ2 (1) = 19.511 p < 0.001 φ = 0.027 |
31.7% [31.2; 32.2] (12 362) |
|
Age ≥ 80 years [95% CI] (N) |
42.5 % * [41.6; 43.4] (5083) |
Χ2 (1) = 175.506 p < 0.001 φ = 0.084 |
50.7 % * [49.8; 51.6] (6541) |
Χ2 (1) = 373.376 p < 0.001 φ = 0.120 |
54.3 % * [53.3; 55.1] (7642) |
Χ2 (1) = 35.350 p < 0.001 φ = 0.036 |
49.5% [49.0; 50.0] (19 266) |
|
Urban region [95% CI] (N) |
63.2% [62.3; 64.1] (7562) |
Χ2 (1) = 4.764 p = 0.030 φ = 0.014 |
61.9 % * [61.1; 62.7] (7990) |
Χ2 (1) = 23.121 p < 0.001 φ = 0.030 |
60.4% [59.6; 61.2] (8499) |
Χ2 (1) = 6.953 p = 0.008 φ = 0.016 |
61.7% [61.2; 62.2] (24 041) |
|
Rural region [95% CI] (N) |
36.8% [35.9; 37.7] (4401) |
Χ2 (1) = 4.764 p = 0.030 φ = 0.014 |
38.1 % * [37.3; 38.9] (4924) |
Χ2 (1) = 23.121 p < 0.001 φ = 0.030 |
39.7% [38.9; 40.5] (5588) |
Χ2 (1) = 6.953 p = 0.008 φ = 0.016 |
38.3% [37.8; 38.8] (14 913) |
*Significant test result (Bonferroni corrected); T, unpaired t-test; U; Mann-Whitney U test;
effect size: small: from 0.1, medium: from 0.3, large: from 0,5 φ Cramer’s phi (chi-square test); M, mean; r, correlation coefficient unpaired t-test, Mann-Whitney U test
95% CI; 95% confidence interval; SD, standard deviation
Place of death
The hospital was the most common place of death, accounting for more than 50% of all deaths. In the place of death ranking, the home environment rank second with percentages between 27.8% (2001) and 21.3% (2017), while the nursing home ranked third with percentages between 13.1% (2001) and 20.4% (2017) (age-standardized rates). The trend over time showed a decrease in the frequency of deaths at home and in hospital and an increase in deaths in palliative care units, hospices and nursing homes (table 4). The analysis of the frequency of the place of deaths based on raw data differed only slightly from the age-adjusted analysis (eTabelle). Women died more frequently in nursing homes, while men died more frequently at home or in hospital (Figure 1a, 1b).
Table 4. Age-standardized data of the places of death in the years 2001, 2011, 2017.
| 2001 | 2011 | 2017 | ||||
| N=11 963 | 2001 vs 2011 | N=12 914 | 2001 vs 2017 | N=14 077 | 2011 vs 2017 | |
|
At home [95% CI] (N) |
27.8% * [27.0; 28.6] (3326) |
Χ2 (1) = 66.3 p<0.001 φ = 0.052 |
23.3% * [22.5; 24.1] (3009) |
Χ2 (1) = 148.6 p<0.001 φ = 0.076 |
21.3% * [20.6; 22.0] (2998) |
Χ2 (1) = 15.6 p<0.001 φ = 0.024 |
|
Hospital [95% CI] (N) |
55.8% * [54.9; 56.7] (6675) |
Χ2 (1) = 40.0 p<0.001 φ = 0.040 |
51.8% * [50.9; 52.7] (6689) |
Χ2 (1) = 41.6 p<0.001 φ = 0.040 |
51.8% [50.9; 52.7] (7292) |
Χ2 (1) = 0.0 p= 0.998 φ = 0.001 |
|
Palliative care unit [95% CI] (N) |
0.0% [0.0; 0.0] (0) |
– | 1.0% [0.8; 1.2] (129) |
– | 6.2% * [5.8; 6.6] (873) |
Χ2 (1) = 509.5 p<0.001 φ = 0.137 |
|
Hospice [95% CI] (N) |
1.9% * [1.7; 2.1] (227) |
Χ2 (1) = 133.5 p<0.001 φ = 0.073 |
4.5% * [4.1; 4.9] (581) |
Χ2 (1) = 162.5 p<0.001 φ = 0.079 |
4.8% [4.4; 5.2] (676) |
Χ2 (1) = 1.4 p = 0.243 φ = 0.007 |
|
Nursing home [95% CI] (N) |
13.1%* [12.5; 13.7] (1567) |
Χ2 (1) = 140.1 p<0.001 φ = 0.075 |
18.6% * [17.9; 19.3] (2402) |
Χ2 (1) = 243.7 p<0.001 φ = 0.097 |
20.4% * [19.7; 21.1] (2872) |
Χ2 (1) = 13.9 p<0.001 φ = 0.023 |
|
Other [95% CI] (N) |
1.2% [1.0; 1.4] (144) |
Χ2 (1) = 0.0 p = 0.998 φ = 0.001 |
1.2% [1.0; 1.4] (155) |
Χ2 (1) = 4.3 p = 0.037 φ = 0.013 |
1.5% [1.3; 1.7] (211) |
Χ2 (1) = 4.5 p = 0.033 φ = 0.013 |
|
Not stated [95% CI] (N) |
0.2% * [0.1; 0.3] (24) |
Χ2 (1) = 15.8 p<0.001 φ = 0.025 |
0.5% [0.4; 0.6] (65) |
Χ2 (1) = 2.6 p = 0.110 φ = 0.010 |
0.3% [0.2; 0.4] (42) |
Χ2 (1) = 6.8 p = 0.009 φ = 0.016 |
*Significant test result (Bonferroni corrected); effect size: small: from 0.1, medium: from 0.3, large: from 0.5. Cramer’s phi (φ) (chi-square test)
The hospital frequency data include palliative care units.
vs, versus; 95% CI, 95% confidence interval
eTable. Raw data of the places of death—in the years 2001, 2011, 2017.
| 2001 | 2011 | 2017 | ||||
| N=11 963 | 2001 vs 2011 | N=12 914 | 2001 vs 2017 | N=14 077 | 2011 vs 2017 | |
| At home | 27.7% * [26.9; 28.5] (3316) |
Χ2 (1) = 63.3 p < 0.001 φ = 0.051 |
23.3% * [22.5; 24.1] (3003) |
Χ2 (1) = 153.4 p < 0.001 φ = 0.077 |
21.1% * [20.4; 21.8] (2974) |
Χ2 (1) = 17.7 p < 0.001 φ = 0.026 |
| Hospital | 56.7% * [55.8; 57.6] (6783) |
Χ2 (1) = 60.1 p < 0.001 φ = 0.049 |
51.8% * [50.9; 52.7] (6689) |
Χ2 (1) = 76.2 p < 0.001 φ = 0.054 |
51.3% [50.4; 52.2] (7220) |
Χ2 (1) = 0.7 p = 0.407 φ = 0.005 |
| Palliative care unit | 0.0% [0.0; 0.0] (0) |
– | 1.0% [0.8; 1.2] (126) |
– | 6.0% * [5.6; 6.4] (842) |
Χ2 (1) = 488.1 p < 0.001 φ = 0.134 |
| Hospice | 1.9% * [1.7; 2.1] (233) |
Χ2 (1) = 125.2 p < 0.001 φ = 0.071 |
4.5% * [4.1; 4.9] (577) |
Χ2 (1) = 138.3 p < 0.001 φ = 0.073 |
4.6% [4.2; 5.0] (646) |
Χ2 (1) = 0.2 p = 0.639 φ = 0.003 |
| Nursing home | 12.1% * [11.5; 12.7] (1445) |
Χ2 (1) = 213.9 p < 0.001 φ = 0.093 |
18.8% * [18.1; 19.5] (2429) |
Χ2 (1) = 395.2 p < 0.001 φ = 0.123 |
21.4%* [20.7; 22.1] (3011) |
Χ2 (1) = 27.9 p < 0.001 φ = 0.032 |
| Other | 1.4% [1.2; 1.6] (165) |
Χ2 (1) = 2.0 p = 0.158 φ = 0.009 |
1.2% [1.0; 1.4] (152) |
Χ2 (1) =0.04 p = 0.873 φ = 0.001 |
1.3% [1.1; 1.5] (190) |
Χ2 (1) = 1.6 p = 0.211 φ = 0.008 |
| Not stated | 0.2* [0.1; 0.3] (21) |
Χ2 (1) = 18.7 p < 0.001 φ = 0.027 |
0.5% [0.4; 0.6] (64) |
Χ2 (1) = 1.9 p = 0.185 φ = 0.009 |
0.3% [0.2; 0.4] (36) |
Χ2 (1) = 10.5 p = 0.001 φ = 0.020 |
*Significant test result (Bonferroni corrected); effect size: small: from 0.1, medium: from 0.3, large: from 0.5. Cramer’s phi (f) (chi-square test)
The hospital frequency data include palliative care units. vs, versus
Figure 1a.
Distribution of place of death over time—stratified according to gender and age
Utilization of PMCS care among the deceased
Of the 5887 deaths in the city of Münster and the district of Coesfeld in 2017, 1694 (28.8% [95% CI: (27.6–30.0%)]) received PMCS care at the end of life. Figure 2 depicts the deaths who received PMCS care, stratified by age and place of death. A corresponding breakdown by gender is shown in the eFigure.
Figure 2.
Deaths 2017 (city of Münster and district of Coesfeld) with PMCS care—stratified according to age and place of death
eFigure.
Deaths 2017 (city of Münster and district of Coesfeld) with PMCS care—stratified according to gender, age and place of death
Discussion
The hospital is the most common place of death, accounting for more than half of all deaths. In 2017, only 21.3% of deaths occurred at home, while 51.8% occurred in hospital and 6.2% in a palliative care unit, 4.8% in a hospice, 20.4% in a nursing home, and 1.5% in other places. In the period from 2001 to 2017, less people died at home or in hospital, while more people died in a palliative care unit, hospice, or nursing home. One fourth of the people who died (2017) in the city of Münster and the district of Coesfeld received outpatient PMCS care at the end of life.
Thus, our study data are in line with place of death data from other industrialized countries which also identify the hospital as the most common place of death (6, 8– 15). In addition, the data reveal a continuous increase in age of death for both sexes, which is very likely to be understood in the context of demographic change. The proportion of persons who died at age ≥ 80 years increased from 42.5% (2001) to 54.3% (2017). The likelihood of comorbidities, frailty and nursing care needs increases with increasing age; at the same time, the risk of dementia increases (22). This, of course, has an effect on the place of deaths. The percentage of those who died in a nursing home was 13.1% in 2001 and then increased markedly to 20.4% in 2017. Women died more frequently in a nursing home, while men died more frequently at home or in hospital. This observation is explained by gender differences in life expectancy (18).
While the majority of the surveyed persons (2– 5) wished to die at home, only one in four to five persons actually died in their own residence. Here, a steady downward trend was noted. Social changes, directed away from the care-providing (extended) family towards an increasing number of people living alone, may have been a key driver of this development (23). While most persons in need of care are still being looked after by family members in a private setting, the general social conditions have changed. Declining birth rates, increasing female employment and larger distances between the residence of adult children and that of their parents are just some of the factors which make it more difficult for family members to provide care in a home setting (24, 25).
Hospices and especially palliative care units were significantly more frequently identified as places of death. While in 2001 palliative care units were non-existent and hospices existed only in cities, the availability of these care facilities had increased by 2017 to 58.8 palliative care beds and 38.7 hospice beds per 1 million inhabitants. Thus, the demand estimate of the European Association for Palliative Care (EAPC) of 40–50 palliative care beds and 40–50 hospice beds per 1 million inhabitants was exceed and not met, respectively (26). The number of palliative care units was found significantly increased, especially in the two university cities. The high density of hospitals with numerous oncology departments may have greatly contributed to this development. After 2011, by contrast, the regional expansion of hospice beds came to a halt.
Despite the fact that the primary mission of hospitals is to cure patients, hospitals ranked first by a wide margin in the distribution of place of death. Information about the morbidity profile of patients who died in hospital is not available for Germany. However, the hospital patient population is likely to be very heterogeneous, including, among others, trauma patients, patients with other acute emergencies, patients with complicated courses of disease, as well as patients with incurable disease wishing to receive further treatment. In many cases, the disease no longer allows patients to decide on the place of death themselves. Overall, a slight downward trend in the number of in-hospital deaths was noted which is certainly linked to an improved service offering by hospices and palliative care units, but may also be related to cost pressure in the inpatient healthcare sector.
One fourth of the people who died in the city of Münster and the district of Coesfeld in 2017 received outpatient PMCS care at the end of life. This finding validates the intention of the general and specialized palliative care contract in Westphalia to establish a broad outpatient palliative care offering (21). PMCS care was received by 45.3% of all persons who died at home and 51.8% of all persons who died in a nursing home. Of the hospice patients, 90.5% died receiving PMCS care. This is likely to be due to mature network structures. The high rates of PMCS care provided show how labor-intensive the provision of outpatient palliative care has become in Westphalia (27). These rates also point to the limitations of this model of care, which requires extensive human resources. The qualified palliative medicine physicians of the PMCS are mostly community-based medical care providers, engaged in the PMCS on a part-time basis. It remains to be seen to what extent the development of a nationwide specialized palliative care framework contract will lead to structural changes in the care model in Westphalia.
Strengths and weaknesses
With 38 954 analyzed death certificates, this study is now the largest German place of death study. Since the analysis was based on selected regions in Westphalia, its results are not representative for Germany. The validity of the place of death information was confirmed. Almost all death certificates of hospitals had an institute stamp on them; all hospice death certificates and most of the nursing home death certificates were identifiable by their respective postal address. Hospital deaths in palliative care units could not always be differentiated. In unclear cases, the hospitals were contacted and a pseudonymized plausibility check performed. In isolated cases, it was found that the lack of an institute address (mainly of a nursing home) resulted in the misclassification of the death as a death at home. Because of this, the number of deaths at home might be slightly overestimated and the number of deaths in institutions—especially nursing homes—slightly underestimated. The pseudonymized data do not provide information on the duration and intensity of the PMCS care provided. Only two palliative medicine consultation services (Münster, Coesfeld) participated in this study.
Figure 1b.
Acknowledgments
Translated from the original German by Ralf Thoene, MD.
Footnotes
Conflict of interest
The authors declare that no conflict of interest exists.
References
- 1.Statistisches Bundesamt (Destatis) Bevölkerung - Sterbefälle und Lebenserwartung. www.destatis.de/DE/Themen/Gesellschaft-Umwelt/Bevoelkerung/Sterbefaelle-Lebenserwartung/_inhalt.html (last accessed on 20 December 2020) [Google Scholar]
- 2.Deutscher Hospiz- und PalliativVerband e.V. (DHPV) Bevölkerungsbefragung „Sterben in Deutschland - Wissen und Einstellungen zum Sterben“ 2017. www.dhpv.de/service_forschung_detail/items/bevoelkerungsbefragung-sterben-in-deutschland-wissen-und-einstellungen-zum-sterben-2017.html (last accessed on 20 December 2020) [Google Scholar]
- 3.Billingham MJ, Billingham SJ. Congruence between preferred and actual place of death according to the presence of malignant or non-malignant disease: a systematic review and meta-analysis. BMJ Support Palliat Care. 2013;3:144–154. doi: 10.1136/bmjspcare-2012-000292. [DOI] [PubMed] [Google Scholar]
- 4.Gomes B, Higginson IJ, Calanzani N. 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;23:2006–2015. doi: 10.1093/annonc/mdr602. [DOI] [PubMed] [Google Scholar]
- 5.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;12 doi: 10.1186/1472-684X-12-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Lang A. Sterben in Österreich: Rahmenbedingungen und Stand der Forschung Sterbewelten in Österreich. Die Perspektive der Betroffenen auf „gutes Sterben“. In: Heimerl K, Egger B, Schuchter P, Wegleitner K, editors. IFF-Alpen Adria Universität Klagenfurt. Wien, Klagenfurt: 2018. [Google Scholar]
- 7.Schweizerische Eidgenossenschaft. Bundesamt für Gesundheit OFSP (Schweiz). Palliative Care. Statistische Daten. Entwicklung der Anzahl der Todesfälle nach Ort, 2006-2011. www.bag.admin.ch/bag/fr/home/das-bag/publikationen/forschungsberichte/forschungsberichte-palliative- care/datensituation-zu-palliative-care.html (last accessed on 20 December 2020) [Google Scholar]
- 8.Houttekier D, Cohen J, Surkyn J, Deliens L. Study of recent und future trends in place of death in Belgium using death certificate data: a shift from hospital to care homes. BMC Public Health. 2011;11 doi: 10.1186/1471-2458-11-228. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Institut national de la statistique et des etudes economiques (Frankreich) Les dces en 2018. www.insee.fr/fr/statistiques/4204058?sommaire=4204068 (last accessed on 20 December 2020) [Google Scholar]
- 10.Gomes B, Sarmento VP, Ferreira PL, Higginson IJ. Epidemiological study of death in Portugal in 2010 and comparison with the preferences of the portuguese popuation. Acta Med Port. 2013;26:327–334. [PubMed] [Google Scholar]
- 11.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 doi: 10.1186/1472-684X-13-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Central Statistics Office (Irland) Home/Births, Deaths & Marriages/Deaths Occurrence. www.data.cso.ie (last accessed on 20 December 2020) [Google Scholar]
- 13.Public Health England (England) Palliative and End of Life Care Profiles. www.fingertips.phe.org.uk/profile/end-of-life/data#page/0/gid/1938132883/part/15/par/E12000004/ati/6/cid/4/tbm/1/page-options/ovw-do-0 (last accessed on 20 December 2020) [Google Scholar]
- 14.Håkanson C, Öhlén J, Morin L, Cohen J. A population-level study of place of death and associated factors in Sweden. Scand J Public Health. 2015;43:744–751. doi: 10.1177/1403494815595774. [DOI] [PubMed] [Google Scholar]
- 15.Statistics Canada (Kanada), Deaths by place of death (hospital or no-hospital) www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1310071501 (last accessed on 20 December 2020) [Google Scholar]
- 16.Cross SH, Warraich HJ. Changes in the place of death in the United States. N Engl J Med. 2019;381:2369–2370. doi: 10.1056/NEJMc1911892. [DOI] [PubMed] [Google Scholar]
- 17.Ochsmann R, Slangen K, Feith G, Klein T, Seibert A. Sterbeorte in Rheinland-Pfalz Zur Demographie des Todes. Interdisziplinärer Arbeitskreis Thanatologie. Beiträge zur Thanatologie. Mainz: Johannes Gutenberg-Universität Mainz. 1997 Heft 8:1–39. [Google Scholar]
- 18.Dasch B, Blum K, Gude P, Bausewein C. Place of death: trends over the course of a decade—a population-based study of death certificates from the years 2001 and 2011. Dtsch Arztebl Int. 2015;112:496–504. doi: 10.3238/arztebl.2015.0496. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Murtagh FE, Bausewein C, Verne J, Groeneveld EI, Kaloki YE, Higginson IJ. How many people need palliative care? A study developing and comparing methods for population-based estimates. Palliat Med. 2014;28:49–58. doi: 10.1177/0269216313489367. [DOI] [PubMed] [Google Scholar]
- 20.Addington-Hall J, Fakhoury W, McCarthy M. Specialist palliative care in nonmalignant disease. Palliat Med. 1998;12:417–427. doi: 10.1191/026921698676924076. [DOI] [PubMed] [Google Scholar]
- 21.Kassenärztlich Vereinigung Westfalen-Lippe (KVWL) Verträge - Palliativmedizinische Versorgung - Kooperation mit palliativmedizinischen Konsiliardiensten (PKD) www.kvwl.de/arzt/qsqm/genehmigung/antrag/palliativmedizin/index.htm (last accessed on 20 December 2020) [Google Scholar]
- 22.Robert Koch-Institut (Hsg.) Welche Auswirkungen hat der demografische Wandel auf Gesundheit und Gesundheitsversorgung? In: Gesundheit in Deutschland. Gesundheitsberichterstattung des Bundes. Gemeinsam getragen von RKI und Destatis Berlin. Robert Koch Institut. 2015 [Google Scholar]
- 23.Umweltbundesamt (UBA) Bevölkerungsentwicklung und Struktur privater Haushalte. www.umweltbundesamt.de/daten/private-haushalte-konsum/strukturdaten-privater-haushalte/bevoelkerungsentwick lung-struktur-privater#83-millionen-menschen (last accessed on 20 December 2020) [Google Scholar]
- 24.Thönnes M. Sterbeorte in Deutschland: Eine soziologische Studie. Frankfurt: Peter Lang Verlag. 2013 [Google Scholar]
- 25.Bundesministerium für Gesundheit (BMG) Sechster Bericht der Bundesregierung über die Entwicklung der Pflegeversicherung und den Stand der pflegerischen Versorgung in der Bundesrepublik Deutschland. www.bundesgesundheitsministerium.de/fileadmin/Dateien/5_Publikationen/Pflege/Berichte/6.Pflegebericht.pdf (last accessed on 20 December 2020) [Google Scholar]
- 26.Radbruch L, Payne S. Standards und Richtlinien für Hospiz- und Palliativversorgung in Europa: Teil 2 Weißbuch zu Empfehlungen der Europäischen Gesellschaft für Palliative Care (EAPC) Z Palliativmed. 2011;12:260–270. [Google Scholar]
- 27.Dasch B, Blum K, Bausewein C. [Estimation of the palliative care needs and the extent of coverage by specialized outpatient palliative care teams in selected regions of Westphalia-Lippe] Gesundheitswesen. 2017;79:1036–1042. doi: 10.1055/s-0041-110529. [DOI] [PubMed] [Google Scholar]