New Evidence on End-of-Life Hospital Utilization for Enhanced Health Policy and Services Planning

Abstract

Background:

Long-standing concern exists over hospital use by people near or at the end of life (EOL) related to the appropriateness, quality, and cost of care in hospital. It is widely believed that most people die in hospital after an escalation in hospital use over the last year of life. As most deaths in high-income countries are not sudden or unexpected, opportunities exist for planning compassionate, effective, and evidence-based EOL care.

Objective:

Gain current population-based evidence for EOL health policy and services planning.

Design:

Retrospective study of population-based hospital utilization data.

Setting/Subjects:

All hospital patients in every Canadian province and territory except Quebec. All decedents with hospital separations in 2014–2015.

Measures:

Descriptive–comparative and logical regression analysis tests.

Results:

In 2014–2015, 3.5% of hospital episodes ended in death and 43.7% of all deaths in Canada (excluding Quebec) took place in hospital. 95.2% of those dying in hospital were only admitted once or twice during their last 365 days of life. 3.6% of those dying in hospital had been living in the community and receiving publicly funded home care before the hospital admission that ended in death, while 67.0% had been living at home without home care. 79.0% of hospital deaths followed an unplanned admission through the emergency room, with 70.5% arriving by ambulance. The hospital care provided in the last stay was largely noninterventionist.

Conclusions:

These findings reveal the need for a major reconceptualization of death, dying, and EOL care to ensure sufficient capacity of palliative home care and other services to support dying people and prevent the health and family caregiver crises that lead to hospital-based EOL care and death.

Introduction

For many years, concern has existed over the use of hospitals by people who are near or at the end of life (EOL).^1–5 Research has often revealed inappropriate and costly “curative” care, and thus low quality care provided to terminally ill and dying people in hospitals.^6–9 With annual deaths increasing in number with population aging,^10–12 research is needed for compassionate and effective EOL care.

Many studies have documented EOL hospital utilization, although these have often used hospital data collected in previous decades on samples.^5,13,14 Hospital utilization patterns have changed considerably in recent years. Among these is a major shift to ambulatory and community-based care, with most tests, treatments, and surgeries performed now on an outpatient or day care basis.^15–17 Given these developments and pressures, current population-based evidence is essential for enhanced EOL health policy and services planning. This article reports population-based research on the use of hospitals by terminally ill and dying Canadians. Eight questions were addressed using 2014–2015 data primarily:

1. What proportion of hospitalizations ended in death?

2. What proportion of deaths in Canada took place in hospital?

3. How did the average length of stay for inpatient decedents compare with the average length of other stays?

4. What are the sociodemographic and other characteristics of inpatient decedents?

5. Among patients who died in hospital, how many times were they hospitalized in the last 365 days of life?

6. What proportion of admissions to hospital in the last 365 days of life was through the emergency room (ER) and/or involved an ambulance?

7. What was the location before the hospital admission that ended in death and what locations were patients discharged to following previous hospital admissions in the last 365 days of life?

8. How did inpatient decedents who were admitted once to hospital in the last year of life differ from decedents admitted more than once?

Design

Retrospective hospital data were obtained for analysis from the discharge abstracts database (DAD) of the Canadian Institute for Health Information (CIHI). It contains sociodemographic, administrative, and clinical data on all hospital episodes in every hospital in all provinces and territories except Quebec (Quebec does not provide data to CIHI). Complete individual-anonymous hospital utilization data (excluding stillbirths) were obtained for 2013–2014 and 2014–2015, the two most recent data years available for this mid-2016 study. Every individual patient had been given a unique number; this number was used to create a third dataset containing all DAD information collected over the last 365 days of life on every inpatient who died in the 2014–2015 year. The SAS program was used to analyze the 2014–2015 and 365-day data, through descriptive–comparative and logistic regression methods.

The University of Alberta's Health Research Ethics Board (#Pro00063626) provided ethics approval.

Results

There were 2,525,987 hospital admissions in 2014–2015, with 88,100 (3.5%) ending in death. These deaths in hospital represented 43.7% of all deaths that year in Canada outside of Quebec.¹² The mean length of hospital stay was 16.4 days (median = 7, mode = 1, and standard deviation [SD] = 40.6) for stays ending in death, compared to 6.3 days (median = 3, mode = 1, SD = 16.07) for stays not ending in death. The mean length of stay for individuals who died in a special care unit such as an intensive care unit (ICU) or coronary care unit (CCU) was 7.5 days. Individuals who died in hospital spent 23.8 days on average in hospital over their last 365 days of life (median = 13, mode = 1, and SD = 45.17). In comparison, individuals who were hospitalized one or more times and discharged alive in the 2014–2015 year had a mean of 9.9 days in hospital that year (median = 3, mode = 1, and SD = 23.77).

Decedent Characteristics

As shown in Table 1, patients dying in hospital were more likely to be male and aged 65 or older (mean = 75.1 years, median = 79.0, mode = 84, and SD = 16.24) compared to patients who did not die. Moreover, 29.2% of all deaths took place in a special care unit, with 23.4% of all 141,620 admissions by the 88,100 decedents in the last year of life involving special care units. Only 724 or 0.8% of all deaths occurred during an intervention (such as intubation, cardiopulmonary resuscitation [CPR], surgery, and angioplasty). The most responsible diagnosis for the last hospital stay varied considerably, but nearly one in four (24.6%) had the diagnosis “factors influencing health status and contact with health services.”

Table 1.

Patients Who Died in Canadian Inpatient Hospital Beds in the 2014–2015 Year

	Number	Percentage
Total	88,100	100
Age
0–64	17,894	20.3
65+	70,206	79.7
Gender
Male	46,629	52.93
Female	41,468	47.47
Province or territory of residence
Alberta	10,521	11.96
British Columbia	14,177	16.12
Manitoba	5403	6.14
New Brunswick	3921	4.46
Newfoundland and Labrador	2523	2.87
Nova Scotia	4775	5.43
Northwest Territories	19	0.02
Nunavut	26	0.03
Ontario	41,304	46.97
Prince Edward Island	665	0.76
Saskatchewan	4427	5.03
Yukon Territories	15	0.02
Special care unit (ICU/CCU) use
Last hospital stay	25,740	29.2
In last 365 days of life	33,140	23.4
Emergency room admission
Last hospital stay	62,077	79.0
In last 365 days of life	111,515	78.7
Ambulance use
Last hospital stay	62,077	70.5
In last 365 days of life	88,729	62.7
Most responsible diagnosis (last hospital stay)
Factors influencing health status and contact with health services	21,704	24.6
Diseases of the circulatory system	16,385	18.6
Diseases of the respiratory system	13,411	15.2
Infectious and parasitic diseases	6285	7.1
Neoplasms	8474	9.6
Diseases of the digestive system	5578	6.3

ICU, intensive care unit; CCU, coronary care unit.

ER and Ambulance Use

Most (70.5%) patients arrived for the hospital stay that ended in death by ambulance and most (79.0%) were admitted through the ER. Similarly, over the last year of life, 78.7% of all admissions were through the ER and 62.7% of admissions involved an ambulance. In comparison, for patients admitted and discharged alive in the 2014–2015 year, 47.7% of their admissions that year were through the ER and 26.3% of these involved an ambulance. Similarly, 42.9% of their last admissions to hospital in the 2014–2015 year were through the ER and 22.7% of these admissions involved an ambulance.

Hospital Admissions and Transfers

Inpatient decedents were admitted to hospital a mean of 1.6 times during the last year of life (median = 2, mode = 1, range = 1–19, and SD = 0.81), with 49.1% having only one hospital admission. Another 46.1% were admitted once more to hospital in the last 365 days of life. The remaining 4.9% were admitted multiple times in the year before death in hospital, with 24 individuals (0.001%) admitted 12 or more times, the equivalent of at least once a month.

As shown in Table 2, 67.0% of those who died in hospital were admitted from a private residence where publicly funded home care services were not being provided; another 3.6% were admitted from a home where they were receiving publicly funded home care. Similarly, over the last 365 days of life, 70.6% of all admissions were from a home without publicly funded home care services. Discharges from hospital in the last 365 days of life were most often to a home not receiving home care services (46.9%), followed by a home with home care services (22.5%).

Table 2.

Locations of End-of-Life Care in Relation to Inpatient Hospitalizations (Percentages)

Location	Arrived from (last hospital stay)	Arrived from (last 365 days of life)	Discharged to (last 365 days of life)
Home (with no home care services)	67.0	70.6	46.9
Home (with home care services)	3.6	3.5	22.5
Another hospital	9.3	8.0	15.4
Nursing home	6.2	5.1	4.7
Aged care home	5.1	4.7	3.1
Ambulatory care clinic	4.8	4.7	2.4
Chronic care facility	1.5	1.2	2.3
All other	2.5	2.2	2.7

Hospital Admission Differences

As illustrated in Table 3, a multivariate logistic regression model designed to explain differences between those persons admitted once to hospital (reference number) and those admitted more often to hospital in the last 365 days of life revealed that females and younger (<65) people were more likely to have multiple admissions. The mean and median ages of inpatient decedents admitted only once to hospital in the last 365 days of life were 76 and 80, respectively, compared to 74 and 78 for those admitted more often. The mean and median ages of decedent cohorts declined steadily with each additional admission; by the 17th admission to hospital, the mean and median ages of this cohort were 42.6 and 53.0, respectively.

Table 3.

Multivariable Logistic Regression Model Findings for Single versus Multiple Admissions in Last Year of Life (1 as Reference)

Characteristics	Adjusted odds ratio	95% confidence interval	p
Age	0.995	(0.994–0.996)	<0.0001
Gender
Female	1
Male	0.937	(0.912–0.963)	<0.0001
Place of residence
Ontario	1
Alberta	1.070	(1.023–1.119)	0.0029
British Columbia	1.086	(1.044–1.129)	<0.0001
Manitoba	0.823	(0.776–0.873)	<0.0001
New Brunswick	1.010	(0.944–1.081)	0.7654
Newfoundland and Labrador	0.931	(0.857–1.011)	0.894
Nova Scotia	0.788	(0.741–0.839)	<0.0001
Northwest Territories	1.278	(0.490–3.331)	0.6157
Nunavut	1.945	(0.859–4.402)	0.1104
Prince Edward Island	1.013	(0.864–1.186)	0.8759
Saskatchewan	1.213	(1.137–1.294)	0.0001
Yukon Territories	3.924	(1.075–14.317)	0.0385
Emergency use (last stay)
No	1
Yes	0.388	(0.374–0.403)	<0.0001
Ambulance use (last stay)
No	1
Yes	1.227	(1.189–1.266)	<0.0001
Special care unit use (last stay)
No	1
Yes	0.722	(0.698–0.747)	<0.0001
Most responsible diagnosis (last stay)
Infectious and parasitic diseases	1
Circulatory system diseases	0.681	(0.642–0.723)	<0.0001
Respiratory system diseases	0.855	(0.804–0.909)	<0.0001
Factors influencing health status and contact with health services	0.965	(0.909–1.024)	0.2382
Neoplasms	0.742	(0.693–0.794)	<0.0001
Digestive system diseases	0.933	(0.867–1.004)	0.0629

Province of residence was also relevant. Compared to patients from Ontario, patients in British Columbia and Saskatchewan were more likely to be hospitalized two or more times in the last 365 days of life. In contrast, Manitoba, Yukon Territory, and Nova Scotia patients were less likely to be hospitalized more than once compared to Ontario patients.

Patients who died in a special care unit were less likely to have been admitted multiple times. Similarly, those dying in hospital after being admitted through the ER were less likely to have multiple hospital admissions compared to those with planned admissions to hospital. In contrast, those arriving by ambulance were more likely to have multiple hospitalizations.

Primary diagnosis was also relevant (using infectious and parasitic diseases as the reference). Patients with circulatory system diseases, respiratory system diseases, and cancer were less likely to have been admitted more than once to hospital in the last 365 days of life.

Discussion

This study revealed low use of hospitals by dying Canadians, as only a small proportion (3.5%) of hospitalizations ended in death and less than half of all deaths (43.7%) in Canada (excluding Quebec) occurred in hospital. Moreover, the vast majority (95.2%) of hospital decedents had one or two hospital admissions in the last year of life. The final hospital stay (16.4 day mean) was not exceptionally long compared to hospitalizations not ending in death (6.3 day mean) and over the last year of life, inpatient decedent days averaged 23.8, ∼2 days per month. These findings vary considerably from the common belief that hospital use is high in the last year of life, and it escalates as death draws near.^5,18,19 These findings also vary considerably from hospital utilization studies using older Canadian data.^13,20,21,22

As 78% of all deaths in Canada in the mid-1990s took place in hospitals,²² the findings from this study reveal a major change in location of death and by implication EOL care. Similar changes may have occurred in other countries, with current evidence essential for health policymakers and providers. This evidence is also needed to inform dying people and their families who may believe that a hospital is the only appropriate place for death and dying, as better care will be provided there over homes and nursing homes.²³

Other findings are also important to note, including those demonstrating that the hospital care provided was mostly palliative in type. Finding under 1% of deaths took place during an intervention indicates that CPR and other lifesaving efforts were not contemplated or considered appropriate. In these cases, impending death was likely obvious and may have been for some time. Previous studies have already identified a low rate of CPR in hospital.^8,9,24 More studies of EOL care are needed to determine if noninterventionist care is commonly provided now in hospitals to people who are terminally ill or dying.⁸

However, 29.2% died in a special care unit. A sudden illness or rapid decline in health is likely in these cases, with ICU or CCU care thus appropriate. Regardless, supportive nursing care appears to have been the most common type of care provided in hospitals to most inpatient decedents. Given the availability of life-sustaining therapies in Canadian hospitals and the propensity to use them without financial or other restrictions to “save” lives when reasonably possible, impending inevitable death appears to have been clearly evident in many cases.

Inevitable death is also likely illustrated by the most common diagnosis “factors influencing health status and contact with health services” for admissions ending in death; this ICD-10 Chapter XXI diagnosis is used whenever palliative or comfort-oriented care alone will be provided. Many hospital deaths could therefore potentially have occurred in other places, such as a home or nursing home. Alternatively, these deaths could have taken place in a hospice, although Canada lags behind other countries in residential hospice accessibility.^25,26

Although this study focused on hospital utilization, the predominance of nonhospital EOL care was revealed. More than half (56.3%) of all deaths in Canada outside Quebec did not take place in hospital and 95.2% of inpatient decedents were only admitted once or twice to hospital for an average of around 3 weeks of care in the last 365 days of life. Moreover, home was the location of residence before most final admissions and most admissions in the last year of life. As terminal illnesses typically last weeks, months, or even years,^9,19,27 nonhospital settings such as private residences, nursing homes, and aged care facilities like lodges or assisted living facilities clearly are common places of EOL care now.

Of great concern is finding that only 3.6% of all patients who died in hospitals arrived from a home where they were receiving publicly funded home care services. Moreover, only 22.5% of hospital discharges in the last year of life were to a home receiving publicly funded home care services. It is possible that private home care services had been purchased to enable these people to remain at home throughout much of their dying trajectory.^5,28,29 Moreover, people with life-limiting illnesses may not always be in need of home care, and it is also possible that those needing EOL care are well served by family members or other care providers in homes, nursing homes, and other aged care facilities.^9,30,31 Unfortunately, the needs of terminally-ill and dying people have not been adequately mapped to determine who needs EOL care and when, and what types of EOL services are needed. Regardless, these findings are highly suggestive of a need to increase palliative home care services. Not only is this care important for humanitarian purposes, but it could be essential to prevent the health and family caregiver crises that can lead to ambulance calls, ER visits, and hospitalizations. Family caregiver burden has been identified as a major issue in Canada and many other countries.^28,32–34

The high rate of admissions to hospital that involved an ambulance and the ER indicate a crisis occurred, one that could potentially be prevented by an improved quantity, diversity, and quality of palliative home care services.^21,35,36 Not only have home care services been shown to reduce EOL hospital stays and hospitalizations^37–39 but also these services are important as home deaths are greatly preferred in Canada and elsewhere.^40,41 Home care services are often needed as many dying persons do not have family members or friends who can provide much, if any, home-based EOL care.^42,43

Regardless, hospitals remain appropriate places for EOL care when emergency treatment is failing to save the lives of individuals suffering from sudden serious health problems. This situation may explain why 9.3% of patients were transferred from one hospital to receive care and die in another hospital. The findings that younger people, females, and those dying of uncommon diseases were more often admitted to hospital in the last year of life demonstrates that the EOL phase can be unpredictable, with hospitalizations potentially occurring because curative treatment efforts are indicated, pain and other symptoms need intervention, or because EOL care has not been foreseen and planned for. Hospitals will therefore always provide some EOL care. However, as inappropriate EOL care in hospital is a long-standing concern,^9,44,45 the care provided in hospital to people who are dying or died needs to be monitored routinely for quality improvement purposes. Accessible high-quality nonhospital EOL care must also be assured and routinely monitored to enable good deaths outside of hospital.

Some European countries are already reporting very low hospital death rates, such as the Netherlands (25%), as an outcome of deliberative EOL planning.⁴⁶ Effective and accessible nonhospital EOL care must become more available in Canada if the avoidable use of hospitals is to be reduced below the current rate of 43.7%. Moreover, it is critical to realize that nonhospital EOL care services will be increasingly needed in the future. With population aging, Canada and most other countries will have twice as many annual deaths occurring in as little as 15 years.¹²

Planning for a reduction in EOL hospital utilization will need to be based on the evidence that nearly 80% of all inpatient decedents were admitted through the ER and 2/3 arrived by ambulance. Unmet EOL care needs or sudden changes in health likely led to these unplanned hospitalizations. Research studies have consistently pointed out that care needs change as death nears.^30,47,48 Given the issue that EOL care needs can be expected to escalate and may be highly volatile as death approaches, one of the most essential developments will be a planned reduction in the need for ambulance calls and ER visits. Much can be done, with some measures like interdisciplinary home palliative care teams having gained clear and irrefutable evidence to support widespread implementation.

Another solution with a growing evidence base is case managers (sometimes called care coordinators or patient navigators) to help dying people remain comfortably and safely in place.^49–51 With 69.4% of discharges from hospital in the last year of life being to a home not receiving home care services, most terminally ill Canadians are discharged into the care of one or more family members. As family members are not often skilled in EOL care nor have a good working knowledge of the healthcare system, individualized care from a palliative case manager has the potential to successfully anticipate and address EOL care needs outside of hospital.⁴⁹ Gerontological case managers may also be helpful as this study found that many people dying in hospitals were older. Older people often have multiple chronic illnesses and senescence with advanced aging.¹⁰ Case managers may be particularly important for identifying those needing EOL care in a residential hospice or nursing home,⁴⁹ with this care arranged in advance of ambulance calls, ER visits, and hospitalizations.

Another option with a growing body of supportive evidence is EOL care in nursing homes and other aged care facilities, places where a growing proportion of deaths are occurring.^39,52 This study found that 10.8% of decedent inpatients had been transferred from these facilities, which indicates that these facilities could be encouraged or required, such as through legislative measures, to provide onsite palliative care so their residents can remain comfortably in place and die in place.^39,53 Funding could be mandated, for instance, to hire onsite nurse practitioners or physicians as they can rapidly assess sudden or escalating symptoms onsite and prescribe pain and other medications to negate the need for hospital transfers. Extra staff for onsite EOL care may also be needed as dying residents typically require more care.⁵⁴ Ideally, these care facilities should become recognized as places of excellence in palliative care.

Conclusions

Only 43.7% of all deaths in Canada (excluding Quebec) are taking place in hospitals now, demonstrating a major, but unrecognized, shift of death, dying, and EOL care out of hospital. Deliberative efforts to prevent the need for hospitalized EOL care are needed now, as every ER visit, every hospital admission, and every hospital discharge in the last year of life is a care setting transition that introduces the risk of care mishaps and other EOL quality concerns such as disrupted care teams. Much of this EOL care could be shifted, as the care provided in hospitals was largely noninterventionist or palliative in orientation. As such, another major change in EOL care was revealed. These changes have not been recognized yet, nor did they occur because of deliberative policy and health services planning. Regardless, this shift to community-based EOL care should be sustained, not only because it is the right thing to do but also because the need for compassionate and effective EOL care is growing.

Footnotes

Acknowledgments

We appreciate the funding support for this study from the Law Commission of Ontario. The research plan, findings, and conclusions are the sole responsibility of the researchers.

The data analysis was undertaken at the University of Alberta, and this statistical analysis checked by the third author at McMaster University.

Author Disclosure Statement

No competing financial interests exist.

References

Bekelman

, Halpern

, Blankart

, et al.: Comparison of site of death, health care utilization, and hospital expenditures for patients dying with cancer in 7 developed countries. J Am Med Assoc, 2016; 315:272–283.

Blinderman

, Billings

: Comfort care for patients dying in the hospital. N Engl J Med, 2015; 373:2549–2561.

Fowler

, Hammer

: End-of-life care in Canada. Clin Investig Med, 2103; 36:E127–E132.

Gao

, Ho

, Verne

, et al.: Changing patterns in place of cancer death in England: A population-based study. PLoS Med, 2013; 10:e01410.

Scitovsky

: “The high cost of dying”: What do the data show?. Milbank Q, 2005; 83:825–841.

Cohen

, Pivodic

, Miccinesi

, et al.: International study of the place of death of people with cancer: A population-level comparison of 14 countries across 4 continents using death certificate data. Br J Cancer, 2015; 113:1397–1404.

Earle

, Landrum

, Souza

, et al.: Aggressiveness of cancer care near the end of life: Is it a quality-of-care issue?. J Clin Oncol, 2008; 26:3860–3866.

Thurston

, Wilson

, Hewitt

: Current end-of-life care needs and care practices in acute care hospitals. Nurs Res Practice, 2012, Article ID 869303, 8 pages.

Wilson

: A report of an investigation of end-of-life patient care practices in health care facilities, and the influences for those practices. J Palliat Care, 1997; 13:34–40.

10.

Christensen

, Doblhammer

, Rau

, et al.: Ageing populations: The challenges ahead. Lancet, 2009; 374:1196–1208.

11.

Mladovsky

, Allin

, Masseria

, et al.: Health in the European Union. Trends and analysis, 2009. www.euro.who.int/__data/assets/pdf_file/0003/98391/E93348.pdf (last accessed March 6, 2017).

12.

Statistics Canada: Deaths, estimates, by province and territory, 2015. www.statcan.gc.ca/tables-tableaux/sum-som/l01/cst01/demo07a-eng.htm (last accessed March 6, 2017).

13.

Guerriere

, Zagorski

, Fassbender

, et al.: Cost variations in ambulatory and home-based palliative care. Palliat Med, 2010; 24:523–532.

14.

Warren

, Barbera

, Bremner

, et al.: End-of-life care for lung cancer patients in the United States and Ontario. J Nat Cancer Inst, 2011; 103:853–862.

15.

Canadian Institute for Health Information: Hospital trends in Canada, 2005. https://secure.cihi.ca/free_products/Hospital_Trends_in_Canada_e.pdf (last accessed March 6, 2017).

16.

Organisation for Economic Co-operation and Development: Health at a glance: Europe. 2012. OECDiLibrary. www.oecd-ilibrary.org/sites/9789264183896-en/05/04/index.html?itemId=/content/chapter/9789264183896-54-en (last accessed March 6, 2017).

17.

York

, Kaufman

, Grube

: Where have all the inpatients gone? A regional study with national implications. Health Affairs Blog, 2014. http://healthaffairs.org/blog/2014/01/06/where-have-all-the-inpatients-gone-a-regional-study-with-national-implications (last accessed March 6, 2017).

18.

Hamel

, Lynn

, Teno

, et al.: Age-related differences in care preferences, treatment decisions, and clinical outcomes of seriously ill hospitalized adults: Lessons from SUPPORT. J Am Geriatr Soc, 2000; 48(5 Suppl):S176–S182.

19.

Teno

, Gozalo

, Bynum

, 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. J Am Med Assoc, 2013; 309:470–477.

20.

Menec

, Lix

, Nowicki

, et al.: Health care use at the end of life among older adults: Does it vary by age?. J Gerontol A Biol Sci Med Sci, 2007; 62A:400–407.

21.

Seow

, Brazil

, Sussman

, et al.: Impact of community based, specialist palliative care teams on hospitalisations and emergency department visits late in life and hospital deaths: A pooled analysis. Br Med J, 2014; 348:g3496.

22.

Wilson

, Truman

, Thomas

, et al.: The rapidly changing location of death in Canada, 1994–2004. Soc Sci Med, 2009; 68:1752–1758.

23.

Thompson

, McClement

, Menec

, et al.: Understanding bereaved family members' dissatisfaction with end-of-life care in nursing homes. J Gerontol Nurs, 2012; 38:49–60.

24.

Ehlenbach

, Barnato

, Curtis

, et al.: Epidemiologic study of in-hospital cardiopulmonary resuscitation in the elderly. N Engl J Med, 2009; 361:22–31.

25.

Canadian Hospice Palliative Care Association: Fact sheet: Hospice palliative care in Canada, 2014. www.chpca.net/media/330558/Fact_Sheet_HPC_in_Canada%20Spring%202014%20Final.pdf (last accessed March 6, 2017).

26.

World Health Organization: Global atlas of palliative care at the end of life, 2014. www.who.int/nmh/Global_Atlas_of_Palliative_Care.pdf (last accessed March 6, 2017).

27.

Reed

, Corner

: Defining the illness trajectory of metastatic breast cancer. BMJ Support Palliat Care, 2015; 5:358–365.

28.

Pivodic

, Van den Block

, Pardon

, et al.: Burden on family caregivers and care-related financial strain at the end of life: A cross-national population-based study. Eur J Public Health, 2014; 24:819–826.

29.

Raphael

, Ahrens

, Fowler

: Financing end-of-life care in the USA. J R Soc Med, 2001; 94:458–461.

30.

Wilson

: The duration and degree of end-of-life dependency of home care clients and hospital inpatients. Appl Nurs Res, 2002; 15:81–86.

31.

Wilson

, Truman

: Addressing myths about end-of-life care: Research into the use of acute care hospitals over the last five years of life. J Palliat Care, 2002; 18:29–38.

32.

De Korte-Verhoef

, Pasman

, Schweitzer

, et al.: Burden of family carers at the end of life; a mixed-method study of the perspectives of family carers and GPs. BMC Palliat Care, 2014; 13:16.

33.

Stajduhar

: Burdens of family caregiving at the end of life. Clin Invest Med, 2013; 36:E121–E126.

34.

Turcotte

: Family caregiving: What are the consequences?. 2015. www.statcan.gc.ca/pub/75-006-x/2013001/article/11858-eng.htm (last accessed March 6, 2017).

35.

Lustbader

, Mudra

, Romano

, et al.: The impact of a home-based care program in an accountable care organization. J Palliat Med, August, 2016 online ahead of print; [Epub ahead of print].

36.

Morris

, King

, Turner

, et al.: Family carers providing support to a person dying in the home setting: A narrative literature review. Palliat Med, 2015; 29:487–495.

37.

Bainbridge

, Seow

, Sussman

: Common components of efficacious in-home end-of-life care programs: A review of systematic reviews. J Am Geriatr Soc, 2016; 64:632–639.

38.

Gomes

, Calanzani

, Curiale

, et al.: Effectiveness and cost-effectiveness of home palliative care services for adults with advanced illness and their caregivers. Cochrane Database Syst Rev, 2013; 6:(CD007760).

39.

Costa

: The determinants of place of death: An evidence-based analysis. Ont Health Technol Assess Ser, 2014; 14:1–78.

40.

Tang

: When death is imminent: Where terminally ill patients with cancer prefer to die and why. Cancer Nurs, 2003; 26:254–251.

41.

Wilson

, Cohen

, Deliens

, et al.: The preferred place of last days: Results of a representative population-based public survey. J Palliat Med, 2013; 16:502–508.

42.

McKee

, Kelley

, Guirguis-Younger

: So no one dies alone: A study of hospice volunteering with rural seniors. J Palliat Care, 2007; 23:163–172.

43.

Wilson

, Truman

, Huang

, et al.: The possibilities and realities of home care. Can J Public Health, 2005; 96:385–389.

44.

Downar

, You

, Bagshaw

, et al.: Nonbeneficial treatment Canada: Definitions, causes, and potential solutions from the perspective of healthcare practitioners. Crit Care Med, 2015; 43:270–281.

45.

Palda

, Bowman

, McLean

, et al.: “Futile” care: Do we provide it? Why? A semistructured, Canada-wide survey of intensive care unit doctors and nurses. J Crit Care, 2005; 20:207–213.

46.

Pivodic

, Pardon

, Morin

, et al.: Place of death in the population dying from diseases indicative of palliative care need: A cross-national population-level study in 14 countries. J Epidemiol Commun Health, 2016; 70:17–24.

47.

Chochinov

, Johnston

, McClement

, et al.: Dignity and distress towards the end of life across four non-cancer populations. PLoS One, 2016; 11:e0147607.

48.

Wodchis

, Austin

, Henry

: A 3-year study of high-cost users of health care. Can Med Assoc J, 2016; 188:182–188.

49.

Thomas

, Wilson

, Woytowich

: Examining end-of-life care management: Systematic review. Nurs Res Pract, 2014; 2014:1–14.

50.

Johnson

: Systematic review of oncology nurse practitioner navigation metrics. Clin J Oncol Nurs, 2015; 19:308–313.

51.

Tan

, Wilson

, McConigley

: Experiences of cancer patients in a patient navigation program: A qualitative systematic review. JBI Database System Rev Implement Rep, 2015; 13:136–168.

52.

US. Department of Health and Human Services: Long-term care providers and services users in the United States: Data from the National Study of Long-term Care Providers, 2013–2014. www.cdc.gov/nchs/data/series/sr_03/sr03_038.pdf (last accessed March 6, 2017).

53.

Cimino

, McPherson

: Evaluating the impact of palliative or hospice care in nursing homes. J Gero Nsg, 2014; 40:10–14.

54.

Cable-Williams

, Wilson

: Dying and death within the culture of long-term care facilities in Canada. Int J Older People Nurs, 2016; 9:169–179.