Article Tools
ORIGINAL REPORTS
Palliative and Supportive Care
Article Tools
End-of-Life Care and Circumstances of Death in Patients Dying As a Result of Cancer in Belgium and the Netherlands: A Retrospective Comparative Study
Abstract
To examine and compare end-of-life care in patients with cancer dying in Belgium and the Netherlands.
A mortality follow-back study was undertaken in 2008 via representative nationwide sentinel networks of general practitioners (GPs) in Belgium and the Netherlands. By using similar standardized procedures, GPs reported on aspects of end-of-life care and the circumstances of nonsudden death of patients with cancer in their practice.
Of the 422 reported patients with cancer, most resided at home during the last year of life (Belgium, 91%; the Netherlands, 95%). Death occurred at home in 34% (Belgium) and 61% (the Netherlands) and in the hospital in 29% (Belgium) and 19% (the Netherlands). In the last month of life, end-of-life issues were more often discussed in the Netherlands (88%) than in Belgium (68%). In both countries, physical problems were discussed most often (Belgium, 49%; the Netherlands, 78%) and spiritual issues least often (Belgium, 20%; the Netherlands, 32%). Certain end-of-life treatment preferences were known for 43% (Belgium) and 67% (the Netherlands) of patients. In the last week of life, treatment was most often focused on palliation (Belgium, 94%; the Netherlands, 91%). Physical distress was reported in 84% (Belgium) and 76% (the Netherlands) of patients and psychological distress in 59% and 36%. Most distressing was lack of energy (Belgium, 73%; the Netherlands, 71%) and lack of appetite (Belgium, 61%; the Netherlands, 53%). Two thirds of patients were bedridden (Belgium, 67%; the Netherlands, 69%).
Although place of death and communication about end-of-life issues differ substantially, a palliative treatment goal is adopted for the vast majority of patients in both countries. However, GPs reported that the majority of patients experienced symptom distress at the end of life, which suggests important challenges remain for improving end-of-life care.
Cancer therapy is successful for many patients with cancer. However, in advanced cancer, cure or prolonging life may no longer be possible, and palliative care becomes the only realistic treatment option. Palliative care is aimed at improving the quality of life of patients and their families by providing relief from physical, psychological, and spiritual problems.1 Although palliative care is well developed for patients suffering from cancer,2 there is consensus that further improvement is still possible and needed.3,4
Although large-scale studies demonstrate that many patients with cancer die at home,5,6 much information on end-of-life care and circumstances of dying for patients with cancer is often gathered from studies focusing on specific institutional care settings.7–13 To gain better insight into the care and the manner of dying of the total population of patients with cancer, nationwide studies irrespective of care setting or type of cancer are needed.
In both Belgium and the Netherlands, the existing nationwide representative networks of general practitioners (GPs) can be used to meet these goals. Since the provision of palliative care in both countries is often coordinated by the GP, almost all inhabitants have their own GP,14,15 and GPs are involved in end-of-life care in virtually all patients with cancer.16 This sentinel GP network has been used successfully in the past to study aspects of end-of-life care in both countries,17–23 but studies have never focused specifically on patients dying as a result of cancer.
Identical study designs were used in both countries, which provides an important opportunity to perform cross-national comparisons. In both Belgium and the Netherlands, there is a high level of accessibility to general health care services and a policy of integration of palliative care into the national health care system,24 but there are also important differences, making comparisons particularly relevant. Although Belgium has chosen to place multidisciplinary teams into different care settings, the Netherlands focuses more on the promotion of palliative care skills among regular caregivers supported by consultation teams.25 GPs in the Netherlands act as absolute gatekeepers to more specialized care by being the patient's first contact with the health care system and controlling access and referrals to specialists' services and clinicians. In addition, there are specialized nursing homes in the Netherlands but not in Belgium.26
The main objectives of this study are to describe the end-of-life care and circumstances of death of patients dying as a result of cancer in Belgium and the Netherlands and to make comparisons between the countries.
This study is an analysis of 1-year (2008) data collection within the Senti-Melc study, a study designed to monitor retrospectively the end of life of patients via existing national sentinel networks of general practitioners in Belgium and the Netherlands.27 In both countries, these networks of GPs are reliable surveillance systems with a long history of nationwide surveillance of a wide variety of health-related topics that use similar standardized procedures of data collection.27–33 In the past, several of the registrations (eg, stroke mortality, suicide incidences, place of death, care setting trajectories) have been compared with data from external resources and have resulted in good external validity of the sentinel network's registrations.28,34–36
In 2008, the Belgian network comprised 172 regularly participating practices compared with 45 in the Netherlands, each covering approximately 1% to 2% of the total registered patient population in each country and representative of GPs nationwide by sex, age, and geographic distribution.37,38
Participating GPs were asked to fill in retrospectively a standardized registration form for all deaths at age 1 year or more in their practice covering the dying process and the care received during the last months of life. In Belgium, GPs reported deaths on a weekly basis; in the Netherlands, because GP patient lists are registered centrally, the registration form was sent to them by the Netherlands Institute of Health Services Research within 1 week of receipt of an electronic death notice. When necessary, reminders were sent.
Patients dying as a result of cancer were identified by a question about the underlying cause of death. Those who had died totally unexpectedly and suddenly, as judged by the GP, were excluded from further analysis since our focus was on those patients with cancer for whom the provision of end-of-life care was a relevant consideration.20–22 All data were reported directly by the GP.
In both countries, the main settings for receiving end-of-life care and for dying are home, hospital, hospice or palliative care unit, and care home. In the Netherlands, there is a structural distinction between residential homes and nursing homes. Although both provide long-term home replacement care, residential homes provide basic assistance to older people who can no longer live independently under the care of the GP, although nursing homes provide more specialized geriatric care by specialist nursing home physicians.39 In Belgium, the GP is responsible for the care of both kinds of care home residents.21
Several comprehensive conceptual frameworks on quality of end-of-life care exist, some specifically for patients with cancer.3,40–51 On the basis of these models, we identified the following key domains of end-of-life care: medical care processes, communication processes, and circumstances of dying (Table 1 provides an overview of the topics surveyed). Most questions had been developed and pretested in previous research.20,21,23,27,52–56
|
| Medical care processes at the end of life |
| Number of general practitioner (GP)–patient encounters in the final 3 months of life (or encounters between GP and patient's relatives, regarding the patient) |
| Transitions, if any, between care settings in the final 3 months of life |
| In case of hospital death, the moment of admission |
| Patient's main treatment goal in the last week of life (cure, prolonging life, palliation) |
| Involvement of specialist palliative care services. In the Netherlands: a GP with palliative care training, team with a palliative care consultant (trained nurse or physician), hospital-based palliative units, nursing home–based palliative care units, and hospice day care facility. In Belgium: multidisciplinary palliative support home care teams, hospital-based mobile support teams, hospital-based palliative care units, palliative day care centers, and care home–based palliative care reference nurses |
| Palliative care services were considered to have been used if at least one service had been used during the last 3 months of life |
| Communication processes at the end of life |
| Content of the GP-patient conversations in the last month of life and before (about primary diagnosis, incurability of illness, life expectation, possible medical complications, physical symptoms, psychological problems [eg, sadness, worry, fear], social problems [eg, relationship problems, lack of social support, family not accepting the situation], spiritual problems [eg, difficulty in accepting situation, trouble with the meaning of life, angry at God], options for palliative care, burden of treatments) |
| Whether the patient had ever expressed wishes about any medical treatment that he or she would or would not want in the final phase of life and about the place of death |
| Circumstances of dying |
| Functional status during the last week of life using the Eastern Cooperative Oncology Group (ECOG) performance status scale: 0, fully active; 1, ambulatory, capable of work of a light nature; 2, capable of self-care but not work; 3, in bed ≥ 50% of the time, capable of only limited self-care; 4, completely bedridden, incapable of self-care |
| Whether the patient was in a coma or unconscious until death and, if so, for how long. Physical and psychological symptom distress during the last week of life was measured by the Memorial Symptom Assessment Scale-Global Distress Index (MSAS-GDI). GPs were asked whether the patient had the following symptoms during the last week of life (yes, no, unknown): lack of appetite, lack of energy, pain, feeling drowsy, constipation, dry mouth, difficulty breathing (physical symptoms); feeling sad, worrying, feeling irritable, feeling nervous (psychological symptoms). If yes, GPs were asked how much symptom distress the patient experienced (not at all, a little bit, somewhat, quite a bit, very much, unknown) for physical symptoms and how often a symptom occurred (rarely, occasionally, frequently, almost constantly, unknown) for psychological symptoms |
| Psychological symptoms were considered to have caused distress if patients appeared to feel this way “frequently” or “almost constantly” during the last week of life. Physical symptoms were considered to have caused distress if symptoms distressed the patient “quite a bit” or “very much.” Symptoms that were not present were considered to have caused no distress. The GDI was calculated as the mean of the item scores for all physical and psychological symptoms; patients for whom no answer was given on more than three symptom items were not included in this calculation55 |
| Whether the patient died at the place wished |
Several procedures were used to ensure data quality: collective development and pretesting of the registration form in both countries, automatic follow-up (in Belgium), and telephone contact with the GPs to prevent missing data and ensure data entry with range, consistency, and skip checks. Methodologic details are described elsewhere.27
Percentages were used to describe the proportion of patients with cancer for whom different aspects of end-of-life care were provided. Fisher's exact tests were used to determine which aspects of the dying process or the care received at the end-of-life were associated with residing in either Belgium or the Netherlands (P < .05). Associations were further explored in multivariate logistic regression analyses while adjusting for differences in place of death (because of the relationship with GP involvement and country), age, and sex, unless stated otherwise. Possible interaction effects were determined between these patient characteristics and the patient's country. In the tables, we display the odds ratios (ORs) in cases in which a relevant country effect was found.
To investigate the representativeness of the data for all nonsudden cancer deaths in Belgium and the Netherlands, we compared age, sex, and place of death of our sample with numbers identified in previous nationwide representative death certificate studies in both Belgium (Flanders; comparison data were not available for the French-speaking part of Belgium)57 and the Netherlands.58 In Belgium, no significant differences between studies were found for these characteristics (multinomial 95% CI). In the Dutch part of the sample, nursing home deaths were underrepresented, which was to be expected since nursing home physicians take over care from the GP. After leaving out these deaths, representativeness was reached. We used SPSS 17 (SPSS, Chicago, IL) and StatXact6 (Cytel Studio, Cambridge, MA) for statistical computations.
The Belgian and Dutch GPs reported 1,354 and 405 deaths, respectively. We excluded all patients who had died suddenly (35%; n = 616) leaving 1,143 patients. For 422 of these (37% in each country), cancer was reported as the underlying cause of death. The majority of patients in both countries resided at home during the last year of life (Belgium, 95%; the Netherlands, 91%; Table 2). Place of death was the only characteristic that was strongly associated with country (P < .001), with more hospital deaths in Belgium (29% v 19%) and more deaths at home in the Netherlands (59% v 34%).
|
| Characteristic | All Cancer Deaths (N = 422) | Belgium (n = 321) | The Netherlands (n = 101) | P* | |||
|---|---|---|---|---|---|---|---|
| No. | % | No. | % | No. | % | ||
| Age, years | |||||||
| ≤ 64 | 120 | 28.7 | 87 | 27.4 | 33 | 32.7 | .381 |
| 65-79 | 156 | 37.3 | 117 | 36.9 | 39 | 38.6 | |
| 80+ | 142 | 34.0 | 113 | 35.6 | 29 | 28.7 | |
| Mean | 71.7 | 72.2 | 70.4 | .247 | |||
| Sex | |||||||
| Male | 226 | 53.7 | 173 | 54.1 | 53 | 52.5 | .819 |
| Female | 195 | 46.3 | 147 | 45.9 | 48 | 47.5 | |
| Educational level | |||||||
| Elementary school/elementary or middle school (approximate age, ≤ 12 years) | 129 | 33.7 | 92 | 32.1 | 37 | 38.5 | .465 |
| Middle school or high school | 202 | 52.7 | 156 | 54.4 | 46 | 47.9 | |
| College/university | 52 | 13.6 | 39 | 13.6 | 13 | 13.5 | |
| Longest place of residence in last year | |||||||
| Home or with family | 382 | 92.0 | 287 | 91.1 | 95 | 95.0 | .289 |
| Care home | 33 | 8.0 | 28 | 8.9 | 5 | 5.0 | |
| Elsewhere (n = 6)† | |||||||
| Place of death | |||||||
| Home or with family | 169 | 40.0 | 109 | 34.0 | 60 | 59.4 | < .001 |
| Care home‡ | 48 | 11.4 | 38 | 11.8 | 10 | 10.0 | |
| Hospital | 113 | 26.8 | 94 | 29.3 | 19 | 18.8 | |
| Hospice/palliative care unit | 87 | 20.6 | 78 | 24.3 | 8 | 8.9 | |
| Elsewhere† | 5 | 1.2 | 2 | 0.6 | 3 | 3.0 | |
NOTE. Missing values for age at death, n = 4; sex, n = 1; level of education, n = 39; longest place of residence, n = 1.
*Fisher's exact test or t test for equality of means to test differences between patients with cancer dying in Belgium and the Netherlands.
†Not included in statistical analysis.
‡Care homes or residential homes for older people (n = 43); nursing homes (n = 5).
During the last 3 months of life, 38% of Belgian and 56% of Dutch patients had more than eight contacts with their GP (P = .001; Table 3); 66% of Belgian and 44% of Dutch patients were transferred to another care setting (P < .001), and 72% of Belgian and 34% of Dutch patients used palliative care services (P < .001).
|
| Characteristic | All Cancer Deaths (N = 422) | Belgium (n = 321) | The Netherlands (n = 101) | Fisher's Exact P | OR* | 95% CI | |||
|---|---|---|---|---|---|---|---|---|---|
| No. | % | No. | % | No. | % | ||||
| During the last 3 months of life: | |||||||||
| More than eight GP-patient contacts† | 179 | 42.4 | 122 | 38.0 | 57 | 56.4 | .001 | 0.62 | 0.37 to 1.02 |
| Transferred at least once between care settings | 250 | 61.1 | 209 | 66.3 | 41 | 43.6 | < .001 | 1.75 | 0.93 to 3.27 |
| Palliative care services initiated | 265 | 62.9 | 231 | 72.0 | 34 | 34.0 | < .001 | 6.73 | 3.86 to 11.75 |
| During last week of life: | |||||||||
| One or more GP-patient contacts | 316 | 74.9 | 228 | 71.0 | 88 | 87.1 | .001 | 0.48 | 0.23 to 0.996 |
| Transferred at least once between care settings | 82 | 20.0 | 70 | 22.2 | 12 | 12.8 | .056 | 1.52 | 0.71 to 3.28 |
| Terminal hospital admission | 113 | 26.8 | 94 | 29.3 | 19 | 18.8 | .040 | 1.89‡ | 1.08 to 3.30 |
| Admission during last week (v before last week) | 38 | 34.9 | 32 | 34.8 | 6 | 35.3 | 1.000 | 1.01 | 0.33 to 3.03 |
| Palliative treatment goal (v curative/prolonging life) | 385 | 93.0 | 294 | 93.6 | 91 | 91.0 | .372 | 1.92 | 0.77 to 4.77 |
NOTE. Missing values for transfer between care setting, n = 13; timing of terminal hospital admission, n = 9; treatment goal, n = 8; specialist palliative care services involved, n = 1.
Abbreviations: GP, general practitioner; OR, odds ratio.
*From multivariate logistic regression models. For these analyses, we compared end-of-life care between patients with cancer from the Netherlands (reference category) and those from Belgium while adjusting for differences in place of death, age, and sex.
†Dichotomized at its median value; the two groups differ in size because approximately 10% of patients had the median number of eight GP contacts.
‡Not controlled for place of death in multivariate analysis.
During the last week of life, contact with the GP was maintained by 71% of Belgian and 87% of Dutch patients (P = .001), and the treatment goal was palliation and not cure or prolonging life in 94% of Belgian and 91% of Dutch patients (P = .372).
Multivariate analyses showed that variations in the use of palliative care services (more in Belgium; OR, 6.73), the proportion of patient-GP encounters in the last week of life (more in the Netherlands; OR, 0.48), and hospital deaths (more in Belgium; OR, 1.89) remained statistically significant after controlling for differences in patient characteristics.
Before the last month of life, 74% (Belgium) and 71% (the Netherlands) of patients who died as a result of cancer had discussed one or more end-of-life care issues with their GP (P = .695; Table 4). Primary diagnosis, incurability of the illness, and physical symptoms were discussed most frequently in both countries, and spiritual issues were discussed least frequently.
|
| Characteristic | All Cancer Deaths (N = 422) | Belgium (n = 321) | The Netherlands (n = 101) | Fisher's Exact P | OR* | 95% CI | |||
|---|---|---|---|---|---|---|---|---|---|
| No. | % | No. | % | No. | % | ||||
| GP-patient conversations | |||||||||
| Prior to last month of life, about: | |||||||||
| Primary diagnosis | 239 | 58.4 | 192 | 60.8 | 47 | 50.5 | .094 | 1.71 | 1.04 to 2.82 |
| Incurability of illness | 213 | 51.7 | 157 | 49.7 | 56 | 58.3 | .162 | 0.73 | 0.45 to 1.19 |
| Life expectation | 170 | 41.5 | 127 | 40.3 | 43 | 45.3 | .407 | 0.92 | 0.56 to 1.50 |
| Possible medical complications | 178 | 43.5 | 136 | 43.3 | 42 | 44.2 | .906 | 1.01 | 0.62 to 1.65 |
| Physical symptoms | 240 | 58.3 | 187 | 59.2 | 53 | 55.2 | .555 | 1.23 | 0.76 to 2.00 |
| Psychological problems | 175 | 42.8 | 134 | 42.8 | 41 | 42.7 | 1.000 | 1.13 | 0.69 to 1.84 |
| Social problems | 128 | 31.4 | 101 | 32.4 | 27 | 28.4 | .529 | 1.21 | 0.71 to 2.05 |
| Spiritual problems | 77 | 19.2 | 61 | 19.8 | 16 | 17.0 | .654 | 1.34 | 0.70 to 2.53 |
| Options for palliative care | 147 | 35.8 | 106 | 33.8 | 41 | 42.3 | .146 | 0.67 | 0.41 to 1.09 |
| Burden of treatments | 179 | 43.4 | 143 | 45.3 | 36 | 37.5 | .197 | 1.53 | 0.93 to 2.52 |
| One or more of these topics was discussed | 302 | 72.9 | 233 | 73.5 | 69 | 71.1 | .695 | 1.22 | 0.72 to 2.08 |
| During the last month of life, about: | |||||||||
| Primary diagnosis | 155 | 37.9 | 98 | 31.0 | 57 | 61.3 | < .001 | 0.34 | 0.21 to 0.56 |
| Incurability of illness | 174 | 42.2 | 107 | 33.9 | 67 | 69.8 | < .001 | 0.28 | 0.17 to 0.46 |
| Life expectation | 169 | 41.2 | 100 | 31.7 | 69 | 72.6 | < .001 | 0.22 | 0.13 to 0.37 |
| Possible medical complications | 151 | 36.9 | 98 | 31.2 | 53 | 55.8 | < .001 | 0.43 | 0.26 to 0.70 |
| Physical symptoms | 231 | 56.1 | 156 | 49.4 | 75 | 78.1 | < .001 | 0.35 | 0.20 to 0.60 |
| Psychological problems | 207 | 50.6 | 140 | 44.7 | 67 | 69.8 | < .001 | 0.42 | 0.25 to 0.70 |
| Social problems | 139 | 34.2 | 90 | 28.8 | 49 | 51.6 | < .001 | 0.46 | 0.28 to 0.75 |
| Spiritual problems | 91 | 22.6 | 61 | 19.8 | 30 | 31.9 | .017 | 0.68 | 0.39 to 1.18 |
| Options for palliative care | 215 | 52.3 | 144 | 45.9 | 71 | 73.2 | < .001 | 0.40 | 0.24 to 0.67 |
| Burden of treatments | 153 | 37.1 | 98 | 31.0 | 55 | 57.3 | < .001 | 0.40 | 0.24 to 0.65 |
| One or more of these topics was discussed | 299 | 72.2 | 214 | 67.5 | 85 | 87.6 | < .001 | 0.37 | 0.19 to 0.72 |
| End-of-life preferences | |||||||||
| Patient ever expressed preference | |||||||||
| For place of death | 219 | 59.0 | 150 | 54.2 | 69 | 73.4 | .001 | 0.63 | 0.34 to 1.15 |
| About a medical end-of-life treatment | 184 | 48.8 | 119 | 42.5 | 65 | 67.0 | < .001 | 0.40 | 0.23 to 0.67 |
NOTE. Missing values on communication < 5%; missing values or indicated by the general practitioner (GP) as unknown for treatment preferences, n = 45; for preference for place of death, n = 51.
Abbreviation: OR, odds ratio.
*From multivariate logistic regression models. For these analyses, we compared end-of-life care between patients with cancer from the Netherlands (reference category) and those from Belgium while adjusting for differences in place of death, age, and sex.
During the last month of life, 68% (Belgium) and 88% (the Netherlands) of patients had discussed one or more end-of-life care issues with their GP (P < .001). In Belgium, less than a third had talked about spiritual issues, social problems, the burden of treatments, primary diagnosis, possible medical complications, or life expectation. Other issues were discussed more often but never by more than half the patients. In the Netherlands, only discussion of spiritual issues was rare, whereas all other topics were discussed by at least half the patients and most frequently concerned life expectation, options for palliative care, and physical problems. All differences between countries, except for spiritual issues, remained significant in the multivariate analyses after controlling for differences in age, sex, and place of death. After additionally adjusting for number of patient contacts during the last month of life (to evaluate whether contact frequency would explain differences in communication), significance was maintained.
About half (54%) of Belgian and 73% of Dutch patients expressed a preference for place of death (P = .001), and 43% of Belgian and 67% of Dutch patients had at some point expressed a wish concerning medical end-of-life treatment (P < .001). After adjustment for differences in age, sex, and place of death, country remained associated with the prevalence of a preference for end-of-life treatment (more in the Netherlands; OR, 0.40).
During the last week of life, about two thirds of patients with cancer in Belgium (67%) and the Netherlands (69%) were completely physically disabled and bedridden (P = .712; Table 5). Physical distress was reported in 84% of the patients in Belgium and in 76% in the Netherlands (P = .138) and psychological distress in 59% and 36%, respectively (P < .001). As judged by the GP, the most distressing symptoms were lack of energy (Belgium, 73%; the Netherlands, 71%) and lack of appetite (Belgium, 61%; the Netherlands, 53%). The difference in psychological distress, as judged by the GPs, remained after adjustment for differences in sex, age, and place of death (OR, 2.66). We additionally controlled for the occurrences of patient contact in the last week of life and prevalence of communication about psychological problems in the last month of life as possible confounders, but this did not alter the results.
|
| Variable | All Cancer Deaths (N = 422) | Belgium (n = 321) | The Netherlands (n = 101) | Fisher's Exact P | OR* | 95% CI | |||
|---|---|---|---|---|---|---|---|---|---|
| No. | % | No. | % | No. | % | ||||
| Clinical status in last week of life | |||||||||
| Physically completely disabled and bedridden† | 275 | 67.1 | 209 | 66.6 | 66 | 68.8 | .712 | 0.97 | 0.59 to 1.62 |
| One or more days unconsciousness until death | 214 | 56.0 | 171 | 58.0 | 43 | 49.4 | .177 | 1.26 | 0.76 to 2.08 |
| Physical symptom distress in last week of life | |||||||||
| GP could make estimation‡ | 362 | 85.8 | 282 | 87.9 | 80 | 79.2 | .034 | 3.58 | 1.76 to 7.13 |
| Distress from at least one symptom | 297 | 82.0 | 236 | 83.7 | 61 | 76.3 | .138 | 1.59 | 0.84 to 3.02 |
| Lack of appetite | 207 | 59.3 | 165 | 61.1 | 42 | 53.2 | .241 | 1.22 | 0.72 to 2.08 |
| Lack of energy | 254 | 72.8 | 197 | 73.2 | 57 | 71.3 | .775 | 1.09 | 0.61 to 1.96 |
| Pain | 80 | 23.2 | 61 | 23.0 | 19 | 23.8 | .881 | 0.98 | 0.52 to 1.87 |
| Feeling drowsy | 98 | 28.9 | 82 | 31.7 | 16 | 20.0 | .049 | 1.77 | 0.92 to 3.39 |
| Constipation | 34 | 11.0 | 29 | 12.7 | 5 | 6.3 | .147 | 2.37 | 0.85 to 6.64 |
| Dry mouth | 63 | 20.4 | 52 | 22.5 | 11 | 14.1 | .143 | 1.59 | 0.75 to 3.37 |
| Difficulty breathing | 86 | 25.3 | 74 | 28.5 | 12 | 15.0 | .018 | 2.14 | 1.04 to 4.39 |
| Psychological symptom distress in last week of life | |||||||||
| GP could make estimation‡ | 322 | 76.3 | 244 | 76.0 | 78 | 77.2 | .893 | 0.35 | 0.10 to 1.22 |
| Distress from at least one symptom | 172 | 53.4 | 144 | 59.0 | 28 | 35.9 | < .001 | 2.66 | 1.49 to 4.77 |
| Feeling sad | 102 | 34.1 | 88 | 38.9 | 14 | 19.2 | .002 | 2.57 | 1.29 to 5.13 |
| Worrying | 121 | 39.9 | 101 | 44.3 | 20 | 26.7 | .007 | 1.98 | 1.05 to 3.73 |
| Feeling irritable | 41 | 13.4 | 37 | 15.9 | 4 | 5.3 | .019 | 3.85 | 1.28 to 11.58 |
| Feeling nervous | 61 | 20.3 | 56 | 24.7 | 5 | 6.8 | .001 | 4.66 | 1.74 to 12.48 |
| Global Distress Index (mean) | 1.25 | 1.32 | 1.04 | .001 | 2.97 | 1.62 to 5.50 | |||
| Died at the place of wish§ | 167 | 76.3 | 106 | 70.7 | 61 | 88.4 | .004 | 0.31 | 0.14 to 0.71 |
NOTE. Additional missing values on symptom distress vary between n = 13 and n = 54; missing values for performance status, n = 12; for unconsciousness, n = 40.
Abbreviations: GP, general practitioner; OR, odds ratio.
*From multivariate logistic regression models. For these analyses, we compared end-of-life care between patients with cancer from Netherlands (reference category) and those from Belgium while adjusting for differences in place of death, age, and sex.
†Eastern Cooperative Oncology Group (ECOG) performance status scale: 0, fully active; 1, ambulatory, capable of work of a light nature; 2, capable of self-care but not work; 3, in bed ≥ 50% of the time, capable of only limited self-care; 4, completely bedridden, incapable of self-care. Figures in this row are % of patients for whom grade 4 (v 0 to 3) was assigned.
‡For at least one symptom.
§GP was informed by patient about the preferred place of death; variables controlled for in multivariate analysis: age, sex, longest place of residence.
Congruence between the preferred and actual place of death was reached in 71% of Belgian and in 88% of Dutch patients (P = .004). The chances of dying in the place of choice were three times lower in Belgium than in the Netherlands (OR, 0.31).
In both countries, more than 90% of patients were able to live at home for most of the time during their last year of life, had a palliative treatment goal in the last week of life, and discussed one or more end-of-life issues with their GP during the last months of life. During the last week of life, two thirds of patients were bedridden, and the majority experienced distress from at least one physical symptom. Several aspects of care and dying were found to be related to country.
Data were collected via existing representative national GP networks to provide a nationwide general practice cancer population irrespective of care setting. The networks had been operational for many years and were registering a wide variety of health-related problems before end-of-life care was added. Other strengths of the study include the quality of the data collection procedures27 and the fact that the 2008 registration form used was developed simultaneously and in close collaboration between countries specifically for the purpose of cross-national comparability.
The study has several limitations. First, a retrospective study design may induce some bias. GPs might have overestimated some aspects of their caregiving (eg, the level of communication) or underestimated some aspects (eg, symptom distress) because of their awareness of what is socially desirable or of the difficulties of making a retrospective judgment of their own intentions, and these tendencies might differ between countries; however, most of the questions relate to an objective description of the patient's last phase of life. GPs may also have lacked the required information or recalled it incorrectly or did not consult the patient's medical record. The chances of this were reduced by the collection of data on a weekly basis, leaving little time between death and registration, and by instructing the GP to use patient files as much as possible. Second, some bias due to a GP's particular interest cannot be entirely excluded. Third, full nationwide representativeness could not be guaranteed in the Netherlands for nursing home deaths because GPs are not involved in care in these settings. However, given the high involvement of the GP in end-of-life care in both countries,14–16 the largely non-cancer population in Dutch nursing homes,39,59–61 the adjustment for place of death in our analyses, and the representativeness of the sample (if nursing home deaths are excluded), the results provide valuable insights into the manner in which the general population of patients with cancer die in Belgium and the Netherlands.
The results of this study show similarities as well as differences between the end of life of patients with cancer in Belgium and the Netherlands. Notwithstanding that more than 90% of patients were able to live at home for most of the time during the last year of life in both countries, a substantial proportion in both Belgium (29%) and the Netherlands (19%) die in a hospital. As we know, terminally ill patients with cancer prefer to be cared for at home and also to die there62–64; hence, the latter result might be a cause for concern for both practice and policy. However, the problem of death in a hospital seems most obvious in Belgium, and the contrast might be related to government and palliative care policies that, in the Netherlands, are more clearly geared toward encouraging death at home5,60 and also to differences between primary care cultures. In both countries, GPs occupy a prominent position in the provision of palliative care,26,65 but their specific roles at the end of life are different. Being gatekeepers to the health care system, Dutch GPs take on most care tasks themselves, although Belgian GPs seem more often to pursue a coordinating role only.25,26,66 This hypothesis is further supported by the finding that palliative care services are more likely to be used in general practice in Belgium than in the Netherlands, and by the differences in numbers of patient contacts in the last phase of life, especially during the last week.
The higher prevalence of end-of-life patient-GP conversations during the last month of life in the Netherlands compared with Belgium might also be related to the differences in medical culture. Ideas of candor and open communication might be valued more highly in the Netherlands,26 which could also partly explain the variation in the GP's knowledge of their patients' preferences about medical end-of-life treatment (43% in Belgium v 67% in the Netherlands).
Notwithstanding these differences, there are several similarities as well. In both countries, GPs focus more on physical and psychological than on spiritual or social issues; it may be that GPs feel less comfortable in introducing these topics, that other professionals may fulfill this task, or that these issues do not concern patients as much as others do. Options for palliative care were principally discussed during conversations held in the last month of life rather than earlier when the subject was addressed by fewer than half the patients; this is somewhat surprising, given the strong emphasis the palliative care movement places on advocating early communication and advance care planning and the relatively predictable disease trajectory of patients with cancer.2,67,68
From its beginnings, palliative care has predominantly focused on cancer care, and our results show the deep penetration of palliative care into end-of-life care for patients with cancer in both of the countries studied.1 Although a palliative care approach was present for most patients, many experienced symptom distress at the very end of life. Lack of energy and lack of appetite occurred as the most distressing physical symptoms in a majority of patients. Pain was reported far less often as having caused distress, which might indicate that pain is a symptom that can be alleviated by palliative care, although other symptoms are far more challenging. However, pain was still a distressing symptom in more than 20% of patients with cancer in both countries, which suggests that there is potential for improvement in alleviating pain at the end of life.
In conclusion, the use of a nationwide sentinel network of GPs can provide important insights into how patients with cancer in general are dying in a particular country. It can also make comparison between countries possible. Although place of death and communication on end-of-life issues differ substantially between Belgium and the Netherlands, a palliative treatment goal has been adopted for the vast majority of patients in both countries. However, GPs report that the majority of patients with cancer still experience symptom distress at the end of life, which suggests important challenges for improving end-of-life care. Future research should also address other important aspects in the provision of end-of-life care, such as the well-being of families, and it should study more deeply the cause and effect of different health care approaches and their outcomes.
Supported by a grant from the Flemish government agency for Innovation by Science and Technology (Strategic Basic Research No. 050158). The Belgian Sentinel Network of General Practitioners is supported by the Flemish and Walloon Community of Belgium.
The Fund for Scientific Research, the Research Council of the Vrije Universiteit Brussel, the Institute for the Promotion of Innovation by Science and Technology in Flanders, and the Flemish and Walloon Ministry of Welfare, Public Health and Family, did not have any role in the design and conduct of the study, collection, management, analysis, or interpretation of the data, or in the preparation, review, or approval of this manuscript. In Belgium, the study protocol and anonymity procedures were approved by the ethical review board of the University Hospital of the Vrije Universiteit Brussel. An approval from the Ethical Review Board was not required for this study in The Netherlands because of the nature of the data collection (postmortem). In both countries, patient anonymity was preserved, and physician confidentiality was maintained through the registration and data entry processes.
Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
The author(s) indicated no potential conflicts of interest.
Conception and design: Koen Meeussen, Lieve Van den Block, Michael A. Echteld, Johan Bilsen, Viviane Van Casteren, Ebun Abarshi, Gé Donker, Bregje Onwuteaka-Philipsen, Luc Deliens
Financial support: Lieve Van den Block, Viviane Van Casteren, Luc Deliens
Collection and assembly of data: Koen Meeussen, Lieve Van den Block, Michael A. Echteld, Nicole Boffin, Viviane Van Casteren, Ebun Abarshi, Gé Donker
Data analysis and interpretation: All authors
Manuscript writing: All authors
Final approval of manuscript: All authors
| 1. | National Cancer Control Programmes: Policies and Managerial Guidelines (2nd edition) 2002 World Health Organisation Geneva, Switzerland World Health Organisation Google Scholar |
| 2. | C Sepúlveda, A Marlin, T Yoshida , etal: Palliative Care: The World Health Organization's global perspective J Pain Symptom Manage 24: 91– 96,2002 Crossref, Medline, Google Scholar |
| 3. | H Seow, CF Snyder, RA Mularski , etal: A framework for assessing quality indicators for cancer care at the end of life J Pain Symptom Manage 38: 903– 912,2009 Crossref, Medline, Google Scholar |
| 4. | JM Teno, JC Lima, KD Lyons: Cancer patient assessment and reports of excellence: Reliability and validity of advanced cancer patient perceptions of the quality of care J Clin Oncol 27: 1621– 1626,2009 Link, Google Scholar |
| 5. | J Cohen, D Houttekier, B Onwuteaka-Philipsen , etal: Which patients with cancer die at home? A study of six European countries using death certificate data J Clin Oncol 28: 2267– 2273,2010 Link, Google Scholar |
| 6. | SL Mitchell, JM Teno, SC Miller , etal: A national study of the location of death for older persons with dementia J Am Geriatr Soc 53: 299– 305,2005 Crossref, Medline, Google Scholar |
| 7. | VM Johnson, JM Teno, M Bourbonniere , etal: Palliative care needs of cancer patients in U.S. nursing homes J Palliat Med 8: 273– 279,2005 Crossref, Medline, Google Scholar |
| 8. | EP McCarthy, RS Phillips, Z Zhong , etal: Dying with cancer: Patients' function, symptoms, and care preferences as death approaches J Am Geriatr Soc 48: S110– S121,2000 suppl 5 Crossref, Medline, Google Scholar |
| 9. | MT Claessens, J Lynn, Z Zhong , etal: Dying with lung cancer or chronic obstructive pulmonary disease: Insights from SUPPORT—Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments J Am Geriatr Soc 48: S146– S153,2000 suppl 5 Crossref, Medline, Google Scholar |
| 10. | LA Dow, RK Matsuyama, V Ramakrishnan , etal: Paradoxes in advance care planning: The complex relationship of oncology patients, their physicians, and advance medical directives J Clin Oncol 28: 299– 304,2010 Link, Google Scholar |
| 11. | JW Mack, JC Weeks, AA Wright , etal: End-of-life discussions, goal attainment, and distress at the end of life: Predictors and outcomes of receipt of care consistent with preferences J Clin Oncol 28: 1203– 1208,2010 Link, Google Scholar |
| 12. | D Hui, A Con, G Christie , etal: Goals of care and end-of-life decision making for hospitalized patients at a Canadian tertiary care cancer center J Pain Symptom Manage 38: 871– 881,2009 Crossref, Medline, Google Scholar |
| 13. | B Zhang, AA Wright, HA Huskamp , etal: Health care costs in the last week of life: Associations with end-of-life conversations Arch Intern Med 169: 480– 488,2009 Crossref, Medline, Google Scholar |
| 14. | K Bayingana, S Demarest, L Gisle , etal: Health survey interview, Belgium 2004 Scientific Institute of Public Health Belgium, Department of Epidemiology, 2006, Report No. Depotn: D/2006/2505/4, IPH/EPI REPORTS No. 2006-035 Google Scholar |
| 15. | GA Donker: Continue Morbiditeits Registratie Peilstations Nederland [in Dutch] 2007 Utrecht, the Netherlands NIVEL Google Scholar |
| 16. | E de Vogel-Voogt, A van der Heide, AF van Leeuwen , etal: Patient evaluation of end-of-life care Palliat Med 21: 243– 248,2007 Crossref, Medline, Google Scholar |
| 17. | E Abarshi, B Onwuteaka-Philipsen, G Donker , etal: General practitioner awareness of preferred place of death and correlates of dying in a preferred place: A nationwide mortality follow-back study in the Netherlands J Pain Symptom Manage 38: 568– 577,2009 Crossref, Medline, Google Scholar |
| 18. | E Abarshi, M Echteld, L Van den Block , etal: Transitions between care settings at the end of life in the Netherlands: Results from a nationwide study Palliat Med 24: 166– 174,2010 Crossref, Medline, Google Scholar |
| 19. | K Meeussen, L Van den Block, N Bossuyt , etal: GPs' awareness of patients' preference for place of death Br J Gen Pract 59: 665– 670,2009 Crossref, Medline, Google Scholar |
| 20. | L Van den Block, R Deschepper, J Bilsen , etal: Transitions between care settings at the end of life in Belgium JAMA 298: 1638– 1639,2007 Crossref, Medline, Google Scholar |
| 21. | L Van den Block, R Deschepper, N Bossuyt , etal: Care for patients in the last months of life: The Belgian Sentinel Network Monitoring End-of-Life Care study Arch Intern Med 168: 1747– 1754,2008 Crossref, Medline, Google Scholar |
| 22. | L Van den Block, R Deschepper, J Bilsen , etal: Euthanasia and other end of life decisions and care provided in final three months of life: Nationwide retrospective study in Belgium BMJ 339: b2772,2009 Crossref, Medline, Google Scholar |
| 23. | L Van den Block, R Deschepper, J Bilsen , etal: Euthanasia and other end-of-life decisions: A mortality follow-back study in Belgium BMC Public Health 9: 79,2009 Crossref, Medline, Google Scholar |
| 24. | Palliative Care in the European Union, IP/A/ENVI/ST/2007-22, 2008 Policy Department Economic and Scientific Policy http://www.europarl.europa.eu/activities/committees/studies/download.do?file=21421 Google Scholar |
| 25. | E Abarshi, MA Echteld, L Van den Block , etal: Use of palliative care services and general practitioner visits at the end of life in The Netherlands and Belgium J Pain Symptom Manage 41: 436– 448,2011 Crossref, Medline, Google Scholar |
| 26. | J Griffiths, H Weyers, M Adams: Euthanasia and Law in Europe 2008 Portland, OR Hart Publishing Google Scholar |
| 27. | L Van den Block, V Van Casteren, R Deschepper , etal: Nationwide monitoring of end-of-life care via the Sentinel Network of General Practitioners in Belgium: The research protocol of the SENTI-MELC study BMC Palliat Care 6: 6,2007 Crossref, Medline, Google Scholar |
| 28. | D Devroey, V Van Casteren, F Buntinx: Registration of stroke through the Belgian sentinel network and factors influencing stroke mortality Cerebrovasc Dis 16: 272– 279,2003 Crossref, Medline, Google Scholar |
| 29. | GA Donker, DM Fleming, FG Schellevis , etal: Differences in treatment regimes, consultation frequency and referral patterns of diabetes mellitus in general practice in five European countries Fam Pract 21: 364– 369,2004 Crossref, Medline, Google Scholar |
| 30. | DM Fleming, FG Schellevis, V Van Casteren: The prevalence of known diabetes in eight European countries Eur J Public Health 14: 10– 14,2004 Crossref, Medline, Google Scholar |
| 31. | MP Lobet, A Stroobant, R Mertens , etal: Tool for validation of the network of sentinel general practitioners in the Belgian health care system Int J Epidemiol 16: 612– 618,1987 Crossref, Medline, Google Scholar |
| 32. | A Stroobant, V Van Casteren, G Thiers: WJ Eylenbosch D Noah Surveillance systems from primary-care data: Surveillance through a network of sentinel general practitioners Surveillance in Health and Disease 62– 74,1988 Oxford, United Kingdom Oxford University Press Google Scholar |
| 33. | H Wauters, V Van Casteren, F Buntinx: Rectal bleeding and colorectal cancer in general practice: Diagnostic study BMJ 321: 998– 999,2000 Crossref, Medline, Google Scholar |
| 34. | N Bossuyt, V Van Casteren: Epidemiology of suicide and suicide attempts in Belgium: Results from the sentinel network of general practitioners Int J Public Health 52: 153– 157,2007 Crossref, Medline, Google Scholar |
| 35. | J Cohen, J Bilsen, P Hooft , etal: Dying at home or in an institution using death certificates to explore the factors associated with place of death Health Policy 78: 319– 329,2006 Crossref, Medline, Google Scholar |
| 36. | B Gielen, A Remacle, R Mertens: Patterns of health care use and expenditure during the last 6 months of life in Belgium: Differences between age categories in cancer and non-cancer patients Health Policy 97: 53– 61,2010 Crossref, Medline, Google Scholar |
| 37. | GA Donker: Continuous Morbidity Registration at Dutch Sentinel General Practice Network 2008 Utrecht, the Netherlands Netherlands Institute for Health Services Research Google Scholar |
| 38. | N Boffin, N Bossuyt, V Van Casteren: Kenmerken van de peilartsen en hun praktijk; Situatie in 2007 en vergelijking met voorgaande jaren Wetenschappelijk Instituut Volksgezondheid http://www.wiv-isp.be/epidemio/epinl/medvnl/profil_07.pdf Google Scholar |
| 39. | JF Hoek, MW Ribbe, CM Hertogh , etal: The role of the specialist physician in nursing homes: The Netherlands' experience Int J Geriatr Psychiatry 18: 244– 249,2003 Crossref, Medline, Google Scholar |
| 40. | H Seow, CF Snyder, LR Shugarman , etal: Developing quality indicators for cancer end-of-life care: Proceedings from a national symposium Cancer 115: 3820– 3829,2009 Crossref, Medline, Google Scholar |
| 41. | DJ Casarett, J Teno, I Higginson: How should nations measure the quality of end-of-life care for older adults? Recommendations for an international minimum data set J Am Geriatr Soc 54: 1765– 1771,2006 Crossref, Medline, Google Scholar |
| 42. | EJ Emanuel, LL Emanuel: The promise of a good death Lancet 351: SII21– SII29,1998 suppl 2 Crossref, Medline, Google Scholar |
| 43. | JR Lunney, KM Foley, TJ Smith , etal: Describing Death in America: What We Need to Know 2003 Washington, DC The National Academies Press Google Scholar |
| 44. | K Rosenfeld, NS Wenger: Measuring quality in end-of-life care Clin Geriatr Med 16: 387– 400,2000 Crossref, Medline, Google Scholar |
| 45. | PA Singer, DK Martin, M Kelner: Quality end-of-life care: Patients' perspectives JAMA 281: 163– 168,1999 Crossref, Medline, Google Scholar |
| 46. | KE Steinhauser, EC Clipp, M McNeilly , etal: In search of a good death: Observations of patients, families, and providers Ann Intern Med 132: 825– 832,2000 Crossref, Medline, Google Scholar |
| 47. | KE Steinhauser, NA Christakis, EC Clipp , etal: Factors considered important at the end of life by patients, family, physicians, and other care providers JAMA 284: 2476– 2482,2000 Crossref, Medline, Google Scholar |
| 48. | AL Stewart, J Teno, DL Patrick , etal: The concept of quality of life of dying persons in the context of health care J Pain Symptom Manage 17: 93– 108,1999 Crossref, Medline, Google Scholar |
| 49. | JM Teno, BR Clarridge, V Casey , etal: Family perspectives on end-of-life care at the last place of care JAMA 291: 88– 93,2004 Crossref, Medline, Google Scholar |
| 50. | NS Wenger, K Rosenfeld: Quality indicators for end-of-life care in vulnerable elders Ann Intern Med 135: 677– 685,2001 Crossref, Medline, Google Scholar |
| 51. | HR Pasman, HE Brandt, L Deliens , etal: Quality indicators for palliative care: A systematic review J Pain Symptom Manage 38: 145– 156,2009 Crossref, Medline, Google Scholar |
| 52. | L Van den Block, R Deschepper, K Drieskens , etal: Hospitalisations at the end of life: Using a sentinel surveillance network to study hospital use and associated patient, disease and healthcare factors BMC Health Serv Res 7: 69,2007 Crossref, Medline, Google Scholar |
| 53. | C Cartwright, BD Onwuteaka-Philipsen, G Williams , etal: Physician discussions with terminally ill patients: A cross-national comparison Palliat Med 21: 295– 303,2007 Crossref, Medline, Google Scholar |
| 54. | JM Teno: Toolkit of Instruments to Measure End-of-life Care 2010 Center for Gerontology and Health Care Research, Brown Medical School wwwchcrbrownedu/pcoc/Choosing htm Google Scholar |
| 55. | SE Hickman, VP Tilden, SW Tolle: Family reports of dying patients' distress: The adaptation of a research tool to assess global symptom distress in the last week of life J Pain Symptom Manage 22: 565– 574,2001 Crossref, Medline, Google Scholar |
| 56. | MM Oken, RH Creech, DC Tormey , etal: Toxicity and response criteria of the Eastern Cooperative Oncology Group Am J Clin Oncol 5: 649– 655,1982 Crossref, Medline, Google Scholar |
| 57. | J Bilsen, J Cohen, K Chambaere , etal: Medical end-of-life practices under the euthanasia law in Belgium N Engl J Med 361: 1119– 1121,2009 Crossref, Medline, Google Scholar |
| 58. | A van der Heide, BD Onwuteaka-Philipsen, ML Rurup , etal: End-of-life practices in the Netherlands under the Euthanasia Act N Engl J Med 356: 1957– 1965,2007 Crossref, Medline, Google Scholar |
| 59. | HE Brandt, L Deliens, ME Ooms , etal: Symptoms, signs, problems, and diseases of terminally ill nursing home patients: A nationwide observational study in the Netherlands Arch Intern Med 165: 314– 320,2005 Crossref, Medline, Google Scholar |
| 60. | JF Hoek, BW Penninx, GJ Ligthart , etal: Health care for older persons, a country profile: The Netherlands J Am Geriatr Soc 48: 214– 217,2000 Crossref, Medline, Google Scholar |
| 61. | MW Ribbe, G Ljunggren, K Steel , etal: Nursing homes in 10 nations: A comparison between countries and settings Age Ageing 26: 3– 12,1997 suppl 2 Crossref, Medline, Google Scholar |
| 62. | IJ Higginson, GJ Sen-Gupta: Place of care in advanced cancer: A qualitative systematic literature review of patient preferences J Palliat Med 3: 287– 300,2000 Crossref, Medline, Google Scholar |
| 63. | ST Tang: When death is imminent: Where terminally ill patients with cancer prefer to die and why Cancer Nurs 26: 245– 251,2003 Crossref, Medline, Google Scholar |
| 64. | M Beccaro, M Costantini, P Giorgi Rossi , etal: Actual and preferred place of death of cancer patients: Results from the Italian Survey of the Dying of Cancer (ISDOC) J Epidemiol Community Health 60: 412– 416,2006 Crossref, Medline, Google Scholar |
| 65. | J Rocafort, C Centeno: EAPC Review of Palliative Care in Europe 2008 Milan, Italy EAPC Google Scholar |
| 66. | NHG-Standpunt Toekomstvisie Huisartsenzorg 2012 [in Dutch] Huisarts en Palliatieve Zorg 2009 Nederlands Huisartsen Genootschap Google Scholar |
| 67. | JR Lunney, J Lynn, DJ Foley , etal: Patterns of functional decline at the end of life JAMA 289: 2387– 2392,2003 Crossref, Medline, Google Scholar |
| 68. | L Radbruch, S Payne: White Paper on standards and norms for hospice and palliative care in Europe: Part 1. Recommendations from the European Association for Palliative Care Eur J Palliat Care 16: 278– 289,2009 Board of Directors of the European Association for Palliative Care Google Scholar |
Acknowledgment
We thank Nathalie Bossuyt, MA, MD, for support with data collection and statistical advice; Jane Ruthven for editing the English version; and Rita De Boodt, Flemish Ministry of Welfare, Public Health and Family, for encoding the cause of death into International Classification of Diseases, 10th Revision (ICD-10) codes in the Belgian sample. We also thank Marianne Heshusius of the Netherlands Institute for Health Services Research for supervising the data collection process and all participating sentinel general practitioners for providing the study data.
