Improving the Quality of End-of-Life Care in Long-Term Care Institutions

Abstract

Purpose:

A knowledge transfer program was carried out to improve knowledge in end-of-life care staff at all levels in residential care homes for the elderly, using a model similar to that developed for a non-acute care hospital setting.

Methods:

The program consisted of a series of seminars and on-site sharing sessions held in the hospital providing outreach support to residential care homes for the elderly (RCHEs), as well as case discussions in the RCHEs. Evaluation was carried out using a knowledge assessment questionnaire before and after the initiative, as well as recording RCHE staff feedback and in-depth interviews with selected residents and their family members.

Results:

Knowledge gaps among RCHE staff existed in the areas of mortality relating to chronic diseases, pain and use of analgesics, feeding tubes, dysphagia, sputum management, and attitudes towards end-of-life care issues, which improved after the program. From the qualitative study, RCHE staff highlighted knowledge and service gaps, issues relating to use of feeding tubes and refusal to eat, lack of confidence in managing the dying process, application of Advance Care Plan (ACP) in the RCHE setting, and the need for training in these areas. Residents and family members highlighted the preference for death over suffering, planning for death, misconceptions about life-sustaining treatments and the advance directive (AD) document, and service gaps in advance care planning.

Conclusion:

Considerable knowledge and service gaps exist among staff and residents of RCHEs, which can be improved by the hospital geriatric team providing services to RCHEs.

Introduction

As in other developed countries, the Hong Kong population is aging. Currently approximately 7% of the population is cared for in residential care homes for the elderly (RCHEs),¹ where on-site nursing care is available during office hours and personal care is available throughout the day. These residents tend to be frail, with multimorbidity, are dependent, and have cognitive impairment,² with a rather high five-year mortality of 54%.³ A survey of these institutions showed that among residents who had normal cognition, 88% expressed a wish for an AD regarding future medical treatment, and approximately 35% would prefer to die in their place of long-term care.⁴ Although up to half of long-term care residents may die in their residential care home in many countries,^5,6 in Hong Kong, the legislation has made dying in place the extraordinary rather than the habitual, and nearly all RCHE residents died in hospitals. As a result, as death approaches, and the physical condition becomes unstable, the resident undergoes frequent transitions between the hospital and their long-term care facility, similar to nursing home residents elsewhere not enrolled in hospice services.⁷

Well-known barriers to good end-of-life care in long-term care settings include inadequate staffing, lack of supervision, inadequate knowledge of supervisor or frontline staff, and interference from the staff's personal attitudes and beliefs on death and dying.^8,9 Studies on long-term care staff of different ranks have showed marked differences in familiarity with the concept of palliative care between different groups of staff and different educational backgrounds.^10,11 Since most frontline nonprofessional staff will have difficulty arranging extra time to receive training, on-site education may be a way to improve attendance and engagement.

Among older people who died, chronic disease patients suffered symptoms comparable to those of their cancer counterparts, yet often received less support and fewer resources due to lower priority and lack of service provision. Both patients and caregivers also had to cope with psychological symptoms such as depressive mood, grief, and anxiety regarding death.¹² Using an incremental approach of promoting the key principles of end-of-life care, we have conducted a continuous quality improvement initiative in a non-acute care hospital involving staff of all levels and disciplines;¹³ subsequent evaluation showed that symptoms were better controlled, fewer unnecessary investigations were carried out, and patients were able to spend less time in hospital.¹⁴

Since older people with end-stage chronic diseases tend to be concentrated in long-term residential care homes for the elderly, and a significant proportion of these patients in non-acute care hospitals reside in residential care homes for the elderly, adoption of this care model through a knowledge transfer process to improve the quality of care would be the next logical step in the diffusion of good practice. It has been shown that educating nursing home staff may improve family satisfaction¹⁵ in addition to improving end-of-life knowledge and attitudes among staff members.¹⁶ We therefore carried out a knowledge transfer program to improve knowledge in end-of-life care staff at all levels in residential care homes for the elderly, using a model similar to one developed for the non-acute care hospital, and we evaluated the results.

Methods

The project consists of disseminating knowledge and experience gained from the non-acute hospital quality improvement initiative¹⁴ to residential care homes for the elderly in the same region in Hong Kong through a series of seminars and on-site sharing sessions conducted in the hospital and each RCHE. Topics included the following:

• Identifying patients near the end of life but who have not reached the terminal stage.

• Assessing readiness to talk about dying; promoting awareness of dying in end-stage diseases; helping to overcome denial of dying and learning about disease progression (symptoms and signs along the dying trajectory).

• Addressing information needs; discussing options and preferences regarding treatment, place of care, and advance care plan; providing education to staff to support patients and caregivers to make informed choices and decisions at the end of life.

• Case conferences: focused on probable end-of-life residents in each home, their prognosis and symptoms, family and staff concerns, and expected end-of-life decisions.

Invitations explaining the purpose of the program were sent out to RCHEs with at least 50 residents. After consent was given by the manager, pre-training questionnaires were sent out to all staff in each home. After the return of these questionnaires, management and senior nursing staff from participating homes were invited to attend three identical introductory seminars focusing on the principles of end-of-life care, the importance of chronic diseases other than cancer in relation to death in older people, and relevant documents (Do-Not-Resuscitate order, advance care planning, ADs). One on-site educational talk and interactive session was conducted in each RCHE for frontline care staff. These were conducted by the project geriatrician, with either the visiting medical officer or the geriatrics outreach nurse who usually worked in that home. The outreach doctors had direct admission rights to the hospital unit, bypassing the emergency department.

These sessions focused on the monitoring and handling of common end-of-life symptoms (e.g., feeding problems, pain, and agitation); use of analgesics; and psychosocial or spiritual issues regarding good death. A case conference was held in each home, led by the project geriatrician, with the visiting medical officers and outreach nurses, to guide nursing home staff in the identification of potential end-of-life residents, to clarify concepts, to determine care focus, and to improve problem solving techniques. After all training, nursing home staff was asked to complete a post-training questionnaire, given to the home manager on the last training session and collected within four weeks from that day.

Questionnaire

The questionnaire was divided into (1) demographic data of the respondent: gender, current work place, discipline, and years of working in a geriatric field; (2) knowledge regarding pain and analgesics; (3) knowledge regarding dysphagia management; and (4) attitude towards end-of-life care choices in cancer and noncancer diseases such as dementia or organ failures. Comparison was made between the correct response rates in the pre-training and post-training questionnaires. Student t test was used to compare variables with continuous data and χ² test was used to compare categorical data between staff. A pvalue of<0.05 was taken as statistically significant.

Qualitative assessment and analysis

Comments of staff regarding end-of-life care in the long-term residential care setting were elicited during on-site seminar and case conference discussions. In-depth interviews were also conducted with selected patients and their family members. These narratives were recorded in the form of detailed field notes compiled from an audiorecording of the interviews by two research assistants, who ensured that idiosyncratic expressions and choice of words were captured in its original form. The notes were organized into categories according to the topic. By means of content analysis, narrative data consisting of respondents' own words and expressions were interpreted and inductively coded. The codes were aggregated to form themes. The process was repeated in the field notes for each respondent until a number of overarching themes were generated and the data became saturated.

Results

Twenty RCHEs were invited to participate. Nineteen homes (16 nonprofit, 3 for-profit) responded. Staff training was conducted between June 1, 2011 and February 19, 2012. After collection of the completed pre-training questionnaires, 47 senior management and nursing staff attended three identical introductory seminars. Out of a total of 739 care staff, 320 attended 22 on-site educational talks and interactive sessions, at least one in each home. One case conference was held in each home, with one home having an extra session. Out of 2925 residents, 745 were estimated to be at the end-of-life stage as defined by having two hospitalizations within the past year and estimated life expectancy of approximately six months. The prevalence per home ranged from 2.2% in the smallest home of 50 residents, to 44% in one of the largest homes, with over 200 residents.

Questionnaire survey

Of 739 staff at the start of the training, 660 (89.3%) participated in the pre-training and 632 in the post-training questionnaire. Among the latter, 328 (52%) also completed the pre-training questionnaire. (Table 1) shows the profile of the participants.

Table 1.

Characteristics of Participants

	Pre-training(n=660)	Post-training(n=632)
Workplace
Subsidized home	503 (80.5)	495 (81.4)
For-profit home	122 (19.5)	113 (18.6)
Female gender	586 (92.9)	567 (95)
Post
Registered nurse	52 (8.3)	34 (5.7)
Enrolled nurse	68 (10.9)	72 (12.1)
Health worker	93 (14.9)	97 (16.3)
Personal care worker	300 (48.1)	314 (52.9)
Physiotherapist	9 (1.4)	6 (1)
Occupational therapist	10 (1.6)	7 (1.2)
Social worker	18 (2.9)	20 (3.4)
Assistant social worker	16 (2.6)	17 (2.9)
Administrative staff	23 (3.7)	14 (2.4)
Others	35 (5.6)	13 (2.2)
Years of working in geriatric field, years, mean (SD)	7.9 (5.6)	7.6 (5.5)

Data are presented as n(%) unless otherwise specified.

Knowledge of chronic diseases in relation to mortality

Before training, 27.9% and 31.5% of participants were able to answer correctly the proportion of deaths attributed to chronic diseases (such as heart disease, chronic lung disease, and chronic kidney disease) and dementia and its complications. There was an underestimation of the proportion of deaths related to these diseases. These figures improved after training (27.9% versus 40.3%, p<0.000, for chronic diseases; 31.5% versus 42.7%, p<0.000, for dementia and related complications).

Knowledge regarding pain and analgesics

There were six questions on knowledge of pain and eight questions on analgesics (see Table 2). For the question “What is the proportion of patients likely to suffer from pain in cancer?” only 11% answered correctly, with 88% overestimating the proportion. For the question “What is the proportion of patients likely to suffer from pain in end-stage dementia?” only 23% answered correctly, with 60% underestimating the proportion. Seventy-seven percent of participants thought the effects of analgesics will decrease with repeated use. Over 60% of participants believed narcotic analgesics will induce gastrointestinal bleeding or ulcers, and that end-stage dementia patients have reduced pain sensations.

Table 2.

Comparisons on Knowledge on Pain and Analgesics

	% Answered correctly
	Pre-training (n=660)	Post-training (n=632)	p value
Do you think morphine can be used on noncancer patients?	78.1	83	NS
Do you think the following painkillers might cause stomach ache or stomach bleeding?
Paracetamol	59.8	66.9	p<0.010
Nonsteroidal antiinflammatory agents (e.g., naprosyn, brufen)	70.9	76.3	p<0.050
Morphine/tramadol	34.5	53	p<0.000
What is the proportion of patients likely to suffer from pain in the following conditions?
Cancer	10.5	24.7	p<0.000
Chronic diseases: coronary heart disease, chronic obstructive pulmonary disease, kidney disease	34.2	40.0	p<0.050
End-stage dementia	23.3	37.8	p<0.000
Do you think existing painkillers might lead to the following side effects and should not be used?
Stomach ache/stomach bleeding	69.7	88.3	p<0.000
Kidney disease	43.9	64.7	p<0.000
Nausea and vomiting	53.2	72.8	p<0.000
Reduced effectiveness with increased use	22.7	40.7	p<0.000
Do you think end-stage dementia patients:
Have degenerated / less acute pain sensations?	39.1	50.0	p<0.000
Do not have pain if they have no wounds?	69.5	83.5	p<0.000
May be in pain when showing shortness of breath?	70.8	78.8	p<0.010

Data were presented as %. Some questions have missing responses; hence not all percentages were based on the same number of responses. Responses as “uncertain” were counted as missing responses.

NS, not significant.

While the percentage of correct responses to most questions increased significantly after training, the proportion of correct responses remained below or equal to 50% for questions about the proportion of patients with pain in various conditions, analgesic effect with repeated use, and diminished pain sensation for patients with dementia.

Knowledge about feeding tubes, dysphagia, and sputum management

The lowest correct responses were for the following: “Using a small or large spoon will not affect the risk of pneumonia in an end-stage dysphagic patient with sputum” (38%), “Sputum in an end-stage dysphagic patient is definitely caused by pneumonia” (47.3%), and “Sputum in an end-stage dysphagic patient can be improved by turning or changing the person's posture” (50.3%). The proportion of correct answers to these improved significantly with training, except for the question regarding pneumonia risk and spoon size, which did not improve.

Regarding tube feeding, the lowest correct response rate in the pre-training survey was for the question “Can feeding tubes lead to longevity?” (29.2%), which improved significantly to 52% after training.

Attitude toward end-of-life care issues

Prior to training, 76.3% of participants answered “yes” to the question whether cardiopulmonary resuscitation (CPR) is appropriate for someone with end-stage dementia. This number was higher than that for end-stage cancer patients (47.5%). Both numbers were reduced significantly after training. More participants considered tube feeding appropriate in end-stage dementia than in end-stage cancer (70.2% versus 45.3%, p<0.000), the difference persisting after training although the proportion choosing artificial feeding for both groups was reduced. With training, the proportion of participants increased significantly who opted for assisted hand feeding rather than tube feeding or nothing-by-mouth with intravenous fluids for end-stage patients who had eating difficulties, eating refusal, or a chest infection. More participants were inclined to relax therapeutic diets and continue with oral feeding, even though food intake may be reduced.

End-of-life care (EOLC) knowledge differed significantly between the grades of frontline workers: registered nurses, enrolled nurses, health workers, and personal care workers. Workers with more health education were more likely to overestimated the proportion of cancer deaths and underestimate that of dementia-related deaths (p for trend<0.001 for both).

Qualitative outcomes

The views of RCHE staff

Six qualitative themes on RCHE staff views towards EOLC were generated from a content analysis of audiorecorded narrative data from training sessions.

Support for EOLC in a care home setting

The clinical conditions of some of the residents were similar to those of patients being cared for in the non-acute hospital. Better palliative day and hospice support and improved interface between RCHEs and hospitals are desirable. More support was needed in medical, psychological, and spiritual domens. Medical support included hospital admissions, medical and nursing outreach teams to help with day to day management as well as liaison with family members. Psychological support included anticipatory grief and burnout among carers. Psychological support include clinical psychologist or psychiatric support, particularly for managing psychosocial symptoms such as anxiety accompanying chronic obstructive pulmonary disease.

Knowledge and service gaps

Gaps in knowledge and service include proactive assessment of end-of-life needs especially for pain; providing regular pain medications especially in dementia and in patients incapable of verbal communications; assessment and management of symptoms such as dehydration, excessive sputum, dyspnea, stiffness, constipation, and poor appetite; eliciting patient preferences in order to improve medication adherence and the effects and side effects of morphine use; suitable techniques for bathing the patient and for pressure sore prevention; and identification of residents with high fall risk. The indications and side effects of morphine need to be understood by staff, for this drug to help alleviate symptoms. If staff thinks that morphine is addictive and should be avoided, then this is a serious knowledge as well as service gap.

Misconceptions regarding tube-feeding and refusal to eat

Misconceptions include that ‘feeding caused diarrhea and hastened death;’ that ‘tube feeding prevents aspiration pneumonia;’ that having nothing to eat means ‘starving to death’ and having ‘poor quality of life;’ that there were legal implications for withholding the feeding tube; and management of feeding problems for those with dementia; that dietary restrictions should be maintained at all cost for diabetes patients.

Managing the dying process

Dying usually occurred in hospitals, as a result of a lack of confidence in managing the dying process, involving discussion with relatives and residents; a lack of skills in providing comfort at the end of life; and a concern that relatives may complain if the elderly person should die in the home. There was a tendency among staff to hide terminal prognosis (e.g., cancer) from residents: ‘Not knowing the diagnosis will help them to maintain the will to live.’

Advance care planning in the home care setting

Staff recognized a pressing need for training to help residents in making ACPs, especially concerning invasive procedures, and to help with the problem with disagreement between a resident and relatives.

Staff self-perceived training needs

Self-perceived training needs include how to discuss with residents and relatives coping with disease progression and the end of life; how to comfort residents' fear of dying; and care and engagement of family members. The staff felt that the program had satisfied their needs and had helped them improve their self-efficacy in introducing end-of-life care services to residents.

Table 3.

Comparison on Knowledge about Sputum and Dysphagia Management

	% Answered correctly
	Pre-training (n=660)	Post-training (n=632)	p value
If an end-stage patient with dysphagia has sputum, it is…
Because of food stuck in the throat. Withholding oral feeding prevents sputum.	78.8	83.7	p<0.050
Definitely caused by pneumonia.	47.3	59.3	p<0.000
Going to improve by turning or changing the person's posture.	50.3	63.4	p<0.000
Preventable by inserting a feeding tube.	83.9	88.9	p<0.010
A normal phenomenon. One should remind the patient to swallow or cough out the sputum.	63.6	73.7	p<0.000
Not going to affect the risk of pneumonia whether a small or large spoon is used for feeding.	38.0	32.8	NS
Do you think feeding tubes can lead to the following effects?
Fewer pneumonias	51.7	66. 6	p<0.000
Fewer bed sores	78.5	87.2	p<0.000
Easier care	41.4	33.5	p<0.050
Longevity	29.2	52.1	p<0.000
Patient discomfort	72.3	79.3	p<0.010
Patient frustration and agitation	65.8	77.2	p<0.000
Restraints	71.1	77.4	p<0.050
Less suffering	61.1	76.9	p<0.000

Data were presented as %. Some questions have missing responses; hence not all percentages were based on the same number of responses. Responses as “uncertain” were counted as missing responses.

NS, not significant.

The views of residents and their family members

Views on ADs, advance care planning, and psychosocial concerns were explored in individual interviews of five residents with chronic organ failure and stroke. Six family caregivers from two families were interviewed for their views on advance care planning and end-of-life care. Both groups expressed the need for more public education in the ACP and the AD. Seven qualitative themes were generated from a content analysis of narrative data from RCHE residents and family caregivers.

Death wish of the older persons

Most of the interviewed residents expressed a wish to die sooner due to suffering. One resident (A) suffered constant itch and numbness: ‘I'm just forcing myself to smile; I'd rather die sooner and end this suffering.’ Another resident (B) said, ‘What I want is not to prolong life; the sooner the better.…I will definitely choose ‘euthanasia.’

Planning for death

Although most of the elderly interviewees had planned for their after-death arrangements, none had stated their preferences for treatment in end-stage illnesses. One lady (C) who survived a cardiac arrest said, ‘I've signed an agreement; I've picked my coffin and the funeral home. I want cremation. Burial grounds are too expensive, a place costs more than a flat.’

Perceptions of life-sustaining treatments

Some elderly opted for cardiopulmonary resuscitation (CPR) at all costs; on the other hand, (C) said, ‘I scolded my children when I came round [from the CPR].…They shouldn't have asked for it. There was so much suffering.’ A man with stroke (D) said he would not want a feeding tube inserted in the last stage, but he would opt for CPR because he would like to ‘be saved anyhow.’

Most family members preferred oral feeding over tube feeding, since eating was highly meaningful as the elderly person's last chance to savour life: Family member (F) said ‘She can no longer recognize her family members, but we can still see her enjoyment and comfort from eating her favorite food.’ And feeding provided a chance to interact and maintain a caring relationship: ‘We take turns to accompany mother and feed her by each mouthful; this patient and intimate care can only be provided by us, that is what RCHE and hospital staff cannot offer.’

Consensus and discrepancy in patient and family perceptions

Most elderly and family members believe that medical treatment should continue until the condition becomes critical. However, they have different attitude towards certain measures such as CPR and restraint. When the harmful effects and futility of CPR were explained to family members, they indicated that they do not want the elderly to endure the suffering. While family members believed that restraint is sometimes necessary, one resident (E) reacted strongly to being restrained: ‘I once told the hospital staff I'd bite my tongue to kill myself if they dare try to restrain me.’

Perceptions of existing AD documents

Elderly and family members did not understand the medical procedures listed on the AD form (cardiac resuscitation, artifical ventilation, nutrition, and hydration) and hence found it difficult to make informed choices, as there ‘has not been enough publicity and discussions to make issues clear for everyone.’

Perceptions of ACP

Elderly patients and family members with previous experience of family members dying in suffering especially value comfort care and welcome having an AD or ACP in place. The elderly valued AD/ACP for its ‘convenience’ and help to relieve the burden of decision making for their family members. Both the elderly and family members highly valued the arrangement of direct admissions to hospital, as they preferred to stay in the RCHE during the end stage and avoid the suffering caused by Accident and Emergency Department attendances.

ACP service gap

Most elderly wanted to make end-of-life choices, but were not confident in communicating their choices to family members or health professionals. AD and ACP were especially valued by elderly residents who had no family members to advocate for them.

Changes to practice after program implementation

The program has impacted service provision in the following areas: flexibility in ambulance service handling of these patients to bypass the accident and emergency department; involvement of speech therapists and dietitians regarding food texture and tastes to promote oral rather than tube feeding; increase in the number of residents adopting ACPs (24 in participating homes; 31 at home); and adoption of training material by other geriatric teams and RCHEs in Hong Kong.

Discussion

This study shows that among RCHEs in Hong Kong, knowledge regarding end-of-life care issues among staff, residents, and their family members is inadequate, which may hamper the provision of good end-of-life care services in RCHEs. This in turn is likely to be a result of the avoidance of discussion regarding death and dying, not only among the lay public but also among health care professional staff. The need to raise health literacy among staff, residents, and members of the public is highlighted, and is an essential first step in the effort to improve quality of care.

The initiative described represents a first step in diffusing good practice from hospital to RCHEs resulting in a closer link between hospital support and RCHEs in improving end-of-life care. While hospital support is essential in providing specialist support and advice, changes in staff culture and practices are also necessary. In this regard, the Hong Kong Hospital Authority already supports outreach geriatric services for RCHEs, and future improvement in care quality could be build around this basic support structure, with input from palliative care specialists based in the hospital.

Inadequate knowledge regarding pain and analgesics use in end-of-life patients was similar to that found in other surveys, and pain in advanced dementia patients was largely underrated. Education in this aspect is particularly important, since up to a third of long-term care home residents suffer from dementia. Misconceptions regarding dysphagia during feeding existed among staff, similar to that among the lay public, suggesting that public education on this topic is also indicated. Training improved knowledge of end-of-life care issues. As with other reports,⁹ qualitative interviews identified areas of deficits such as lack of knowledge and confidence in assessing symptoms and providing comfort for end-of-life residents, resulting in difficulties in communications between the staff, the resident, and the family. Residents and family members expressed variable expectations concerning the dying process, depending on their background and personal experience of death and dying, comparable to findings among Western populations.¹⁷ Among the older persons there was little understanding of the concepts of futility or autonomy, and an assumption that doctors should make decisions. The views on end-of-life treatments in this group of residents were more aggressive than those usually reported in Western literature.¹⁸ This finding accords with other authors who have suggested that cultural adaptation of end-of-life decisions may be needed.¹⁹ A passive attitude among the elderly has been observed in both Asian^20,21 and Western cultures.²²

Limitations and strengths

It was difficult to ascertain the response rate of the same staff to both surveys due to their frequent movements during the training period. Due to rapid staff turnover, the pre- and post- test population is not exactly the same. The increase in knowledge may not be solely attributed to the program, due to variable attendance and to preexisting knowledge levels and educational background among RCHE staff. Standardized measures of knowledge were not used, as applicability to the local service setting was uncertain. Knowledge attrition with time was not tested. The strengths of this study include the large number of RCHEs and the use of quantitative as well as qualitative assessments to attempt to include all the issues relating to improving end-of-life care quality. Future studies can be designed to evaluate in greater detail the impact on use of hospital services and on quality care indicators of residents receiving end-of-life care.

Conclusion

It is possible to improve end-of-life care knowledge, change attitudes, and ultimately improve end-of-life care in RCHEs using a knowledge transfer model involving hospital geriatricians, outreach geriatrics teams, and all levels of residential care home staff.

Footnotes

Acknowledgment

This study was partially supported by the Knowledge Transfer Fund of the Chinese University of Hong Kong.

Author Disclosure Statement

No competing financial interests exist.

References

Department CaS: Thematic Household Survey Report No. 40. Hong Kong. Census and Statistics Department, the Government of the Hong Kong Special Administrative Region. 2009.

Woo

, Chau

. Aging in Hong Kong: The institutional population. J Am Med Dir Assoc, 2009; 10,7:478–485.

Lee

, Chau

, Hui

, Chan

, Woo

. Survival prediction in nursing home residents using the Minimum Data Set subscales: ADL self-performance hierarchy, cognitive performance and the changes in health, end-stage disease and symptoms and signs scales. Eur J Public Health, 2009; 19,3:308–312.

Chu

, Luk

, Hui

et al. Advance directive and end-of-life care preferences among Chinese nursing home residents in Hong Kong. J Am Med Dir Assoc, 2011; 12,2:143–152.

Flory

, Yinong

, Gurol

, Levinsky

, Ash

, Emanuel

. Place of death: U.S. trends since 1980. Health Aff (Millwood), 2004; 23,3:194–200.

Mor

, Papandonatos

, Miller

. End-of-life hospitalization for African American and non-Latino white nursing home residents: Variation by race and a facility's racial composition. J Palliat Med, 2005; 8,1:58–68.

Gozalo

, Miller

. Hospice enrollment and evaluation of its causal effect on hospitalization of dying nursing home patients. Health Services Res, 2007; 42,2:587–610.

Kayser-Jones

, Schell

, Lyons

, Kris

, Chan

, Beard

. Factors that influence end-of-life care in nursing homes: The physical environment, inadequate staffing, and lack of supervision. Gerontologist, 2003; 43,Spec 2:76–84.

Wowchuk

, McClement

, Bond

Jr.

The challenge of providing palliative care in the nursing home Part II: Internal factors. Int J Palliat Nurs, 2007; 13,7:345–350.

10.

, Kwan

, Lau

, Kwan

, Lam

, Woo

. The needs, current knowledge, and attitudes of care staff toward the implementation of palliative care in old age homes. Am J Hosp Palliat Care, 2010; 27,4:266–271.

11.

McDonnell

, McGuigan

, McElhinney

, McTeggart

, McClure

. An analysis of the palliative care education needs of RGNs and HCAs in nursing homes in Ireland. Int J Palliat Nurs, 2009; 15,9:446–455.

12.

Woo

, Lo

, Cheng

, Wong

, Mak

. Quality of end-of-life care for non-cancer patients in a non-acute hospital. J Clin Nurs, 2011; 20,13–14:1834–1841.

13.

Woo

, Lo

, Lee

et al. Improving end-of-life care for non-cancer patients in hospitals: Description of a continuous quality improvement initiative. J Nurs Healthc Chronic Illn, 2009; 1:237–244.

14.

Woo

, Cheng

, Lee

et al. Evaluation of a continuous quality improvement initiative for end-of-life care for older noncancer patients. J Am Med Dir Assoc, 2011; 12,2:105–113.

15.

Arcand

, Monette

et al. Educating nursing home staff about the progression of dementia and the comfort care option: Impact on family satisfaction with end-of-life care. J Am Med Dir Assoc, 2009; 10,1:50–55.

16.

Parks

, Haines

, Foreman

, McKinstry

, Maxwell

. Evaluation of an educational program for long-term care nursing assistants. J Am Med Dir Assoc, 2005; 6,1:61–65.

17.

Moorman

. Older adults' preferences for independent or delegated end-of-life medical decision making. J Aging Health, 2011; 23,1:135–157.

18.

Kwak

, Haley

. Current research findings on end-of-life decision making among racially or ethnically diverse groups. Gerontologist, 2005; 45,5:634–641.

19.

Winzelberg

, Hanson

, Tulsky

. Beyond autonomy: Diversifying end-of-life decision-making approaches to serve patients and families. J Am Geriatr Soc, 2005; 53,6:1046–1050.

20.

Htut

, Shahrul

, Poi

. The views of older Malaysians on advanced directive and advanced care planning: A qualitative study. Asia Pac J Public Health, 2007; 19,3:58–67.

21.

Low

, Ng

, Yap

, Chan

. End-of-life issues: Preferences and choices of a group of elderly Chinese subjects attending a day care centre in Singapore. Ann Acad Med Singapore, 2000; 29,1:50–56.

22.

Gott

, Small

, Barnes

, Payne

, Seamark

. Older people's views of a good death in heart failure: Implications for palliative care provision. Soc Sci Med, 2008; 67,7:1113–1121.