ADVERTISEMENT
Implementing Palliative Care in the Nursing Home
Background
In a recent column in The New York Times, Jane Gross described her family’s experience of death in the nursing home setting. In the face of progressive illness and catastrophic debility, the columnist’s mother chose to refuse food and fluid. She died 13 days later after a very difficult time. Ms. Gross states, “For days my brother and I badgered the doctors and nurses to medicate her sufficiently to stay ahead of the discomfort. We cried. We screamed. We threatened to take her home...what good was medication on demand if she had to suffer to get it?” The author concludes, “Rage at their inability to make her comfortable from day 1 has softened into an understanding of the challenges of dying this way in a nursing home.”1
Approximately 20% (nearly 500,000) of people who die each year do so in our nation’s nursing homes or shortly after a transfer from the nursing home to an acute care hospital.2 While many nursing home residents have chronic, life-limiting conditions for which palliative care may be most appropriate, up to 40% of nursing home residents are transferred to the hospital in the 30 days prior to their deaths.3 Many dying nursing home residents have limited access to palliative care, largely because the incentives for reimbursing such care are sorely lacking.4 Thus, most nursing home residents die in need of palliative care, or with palliation delayed until the last few days of life.5
The limited literature on palliative care in nursing homes highlights major problems regarding advance care planning, pain management, and psychosocial support. Inadequate pain management is a major problem in the nursing home setting. Pain is a common feature of life for nursing home residents.6,7 Teno and colleagues8 reported that over 40% of residents in pain at an initial assessment were in severe pain 60-180 days later. Bernebei and colleagues9 reported that 24-38% of elderly residents with cancer had daily pain, and of these, 26% received no analgesics. A subsequent study by Won and colleagues10 found that 26% of residents without cancer reported daily pain, and that one-quarter of these residents also went untreated. Very old residents, members of minority populations, and those with severe cognitive impairment were at greatest risk for underevaluation and management of pain.9,10
Possible reasons for the poor pain management include inadequate assessment by staff with limited training11 and problems assessing pain in cognitively impaired residents.12 In a setting where death is common, the careful completion of advance directives (living wills and durable power of attorneys for health care) and conversations about preferences at the end-of-life are uncommon.5,13 Studies show that while many residents may have a “do not resuscitate” order in their health care record, few families and even fewer residents have discussions about life-sustaining treatment with their physicians.14 When discussions about life-sustaining treatment are postponed until a medical crisis, the decisions made may not reflect the resident’s long-standing goals and values.
It is widely believed that psychosocial support to families and residents is lacking in the nursing home, despite growing recognition of the needs of family members with loved ones in long-term care settings. In one study, interviews with families identified caring behaviors by the staff, participation in the dying process, and the provision of spiritual support as improving residents’ and their families’ experience at the end of life.15 Elsewhere, families who experienced nursing home hospice care believed that symptoms were better managed, hospitalizations were reduced, and quality of care was improved.16 Unfortunately, there are few grief and bereavement services for families of residents in long-term care.17
Efforts to overcome deficits in palliative care services in nursing homes have resulted in a variety of educational strategies,18,19 but until now, no systematic studies have been undertaken to examine how palliative care can be integrated into routine nursing home practice. In this article we describe our experiences with such a program. With funding from the Robert Wood Johnson Foundation through an initiative, “Promoting Excellence in End of Life Care,” the palliative care program was carried out as a collaboration between Genesis ElderCare and the University of Pennsylvania School of Nursing.20 It specifically addressed advance care planning, pain and symptom management, and psychosocial support, and introduced an integrated Palliative Care Delivery Process. Population and Setting Six community nursing homes in Maryland, their residents, and staff, participated.
The homes were part of Genesis ElderCare, a large geriatric health care company managing over 300 nursing homes. Genesis recommended participating homes on the basis of location, relative stability of administrative staff, and willingness to implement the palliative care program. Three homes were located in suburban Baltimore and three on the rural Eastern Shore of Maryland. The average number of beds per home was 135 (range 100-200). Residents were elderly (average age, 78 years), predominantly female (68%), and Caucasian (78%). Approximately 85% of residents were admitted directly from an acute care hospital, and nearly 45% of those discharged from the facility were returned to the community. Most residents were extremely frail, had significant limitations in physical and cognitive function, and an average length of stay of 202 days. On average, three to five deaths occurred per home per month. The homes’ major payer sources were Medicare (55%) and Medicaid (15.5%), with the remainder being assorted private insurances and private pay. None of the homes offered subacute care.
Key Elements of the Palliative Care Delivery Process
We defined palliative care as care focused on quality of life and delivered to persons with progressive, far advanced disease with a limited prognosis for recovery. The palliative care program targeted those residents facing terminal or end-stage disease, or those who had experienced significant functional decline, recent multiple hospitalizations, or major alterations in nutritional status. A Palliative Care Delivery Process was used to guide the sequence of activities for the program (Figure 1). Assessment focused on advance care planning, pain and symptom management, and psychosocial concerns.
Advance care planning was introduced at time of admission when the goals of palliative care were briefly described by the admissions coordinator to new residents and their families. Advance care planning was approached as an ongoing discussion rather than simply filling out a Living Will. Decisional capacity, knowledge of the illness, expectations for care (curative versus restorative), expected functional outcomes, life prolongation versus quality of remaining life, and preferences for any specific intervention (eg, hospitalization, artificial nutrition) were routinely addressed as each resident’s status was reviewed and ongoing discussions held with patients and family members to ensure thoughtful decision-making.
A “Pain Assessment Tool” and “Pain and Symptom Assessment Summary Sheet” guided the staff in better pain and symptom management. Residents were screened for pain and other symptoms on a schedule concurrent with regular Minimum Data Set assessments: on admission and quarterly, or whenever a significant clinical change occurred. Frequent inquiries by all staff into the presence of pain and other discomforts were emphasized, with special attention to behavioral and facial cues, in cognitively impaired residents. Staff followed a variety of care planning protocols to manage pain and other symptoms (eg, nausea, constipation, anorexia, dyspnea). An assessment of family and community support, resident needs for spiritual support, and potential bereavement needs of families guided care planning. This allowed staff to individualize the care plan to include personal preferences of the resident (eg, chosen foods, diversional activities, hospice services [if appropriate and desired]), and interactions with family, clergy, and others. Needs of family members, significant others, as well as other staff and residents, were also accounted for in order to provide appropriate memorial services, bedside rituals, bereavement materials, and access to community referral services.
Care planning and evaluation were based on problem identification and reviewed with the physician, other team members, resident, and family. Care plans were tailored to residents’ specific needs, using a series of AIM protocols (Assess, Intervene, Monitor), which were available to assist staff in addressing such problems as pain, dyspnea, or need for bereavement support (Figure 2). Through regularly scheduled care planning, team and/or quarterly quality improvement meetings, gaps in care were identified and problems addressed. A systematic audit was put in place to evaluate critical outcomes (eg, advance care planning, pain and symptom management, psychosocial concerns) following all deaths. Specific criteria included documentation, appropriateness of assessments, completeness of care plans, and consistency between resident goals/wishes and interventions (Figure 3).
Program Implementation
Among the six participating homes, four (two suburban and two rural) served as the intervention homes and received training in the Palliative Care Delivery Process as outlined above. The two remaining homes (one suburban and one rural) did not receive any specific palliative care training and served as control homes. In the four intervention homes, the first step was to train key staff in the concepts of palliative care. This training included administrators, directors of nursing, social workers, and nurses. The first day of training focused on five modules:
Introduction to Palliative Care;
Advance Care Planning;
Pain and Symptom Management;
Psychosocial Support;
and the Palliative Care Delivery Process.
Copies of the training materials were provided in loose-leaf binder format. Following the training for key staff, the nursing home administrator designated one or two palliative care coordinators (nurses or social workers) in each home. The coordinators delivered inservice training for their staff over a 3-month period, using the materials as a guide. At the conclusion of the formal inservice training for all disciplines and care providers, staff were instructed on how to integrate palliative care approaches into the existing systems of care, specifically admissions, nursing assessment, care planning, and quality improvement. From this point onward, training in palliative care occurred as part of new staff orientations on site as described below, and as needed.
To aid in the training for palliative care, and to facilitate implementation of the program, all intervention homes also received ongoing guidance from a nurse with expertise in palliative care. The nurse consultant was available to the centers on a weekly basis (4-5 hours per week per home averaging in total about 20 hours per week for the four homes). The nurse consultant assisted the nursing home staff in identifying residents appropriate for palliative care, utilizing relevant assessment tools, and resolving complex care problems (eg, pain and symptom management). In so doing, she participated in relevant care planning and quality improvement meetings, provided formal and informal training of staff, interacted with all team members (including attending physicians, pharmacy providers, hospice where available, and pastoral care), and communicated with the project directors regarding the progress of program implementation. The nurse consultant was critical in assisting the homes to become proficient in the palliative care delivery process. As staff became more comfortable with palliative care, their reliance on the consultant decreased, allowing for greater ownership and sustainability of the program. A corporate representative from Genesis ElderCare was also available to the consultant for assistance in implementing the program. The representative was familiar with all aspects of the project and visited participating homes on a regular basis.
In addition to implementing the palliative care process with nurse consultation, two of the four intervention homes also developed a dedicated interdisciplinary palliative care team. The team, led by the medical director, consisted of members from nursing, social work, activities, dietary, other therapies, hospice (where available), other professionals and community resources as indicated. The team met weekly to discuss those residents deemed appropriate for the palliative care approach, taking into account advance directives and current wishes of the residents and families, pain and symptom management, and other physical and psychosocial needs. The control homes followed existing regulatory and Genesis ElderCare guidelines for care, but received no additional training or consultative support for palliative care. During this time, however, new corporate policies relating to pain management and advance directives were implemented in all Genesis ElderCare homes, including the control homes. Implementation was designed to evaluate two levels of palliative care programming (education/consultation or education/consultation/team) versus conventional care (control).
For this article, we are focusing on program implementation and evaluation. However, we did collect data on outcomes of interest including events at time of death, staff attitudes and knowledge, documentation of assessment and treatment of physical and emotional symptoms, and observation of process implementation and culture change. Data collection instruments included tools for screening for appropriateness of palliative care, tracking documentation of physical and emotional symptoms, recording events at the time of death (eg, pain, hospital transfer, family present, etc), recording advance directives, surveying family satisfaction, and determining staff knowledge and attitudes. Results have been reported elsewhere and are described below.
In an analysis of the events near the time of death, intervention homes had more residents with advance care plans (P < 0.01), more residents with pain managed by narcotics (P < 0.01), higher rates of symptoms in the last weeks of life (P < 0.05), and higher rates of physician visits and x-rays in the last weeks of life (P < 0.01).21 The most common symptoms reported at the homes were pain (77%), constipation (77%), weakness (71%), cough (46%), depression (36%), and anxiety (24%). There was a significant interaction between location effects and intervention group. The homes on the Eastern Shore did have a significantly higher percentage of physical symptoms with documented assessment and treatment than those in Baltimore (P < 0.01). There was no difference between intervention and control homes on the Eastern Shore or Baltimore on assessment and treatment of emotional symptoms. There was also no difference between intervention and control groups from the Eastern Shore on assessment and treatment of physical symptoms.
In the Baltimore homes, the education home had more assessment and treatment of physical symptoms than either the team or control homes (P < 0.01).22 Staff from both intervention and control homes reported a basic comfort level with caring for the dying and the principles of palliative care. Staff in palliative care homes were able to identify two problems with caring for the dying (loneliness and difficulty communicating with family) and one way to improve care of the dying (greater emphasis on spiritual care) significantly more often than staff in nonpalliative care homes (P < 0.05).
At the time of the second survey, staff in all homes identified Regulatory Concerns, Difficulty in Communication with Family, Concern for Other Residents, and Inability to Meet Spiritual Needs as problems with caring for the dying, and Greater Family Involvement and Use of a Palliative Care Team as ways to improve care significantly more often (P < 0.05) than staff during the first survey.23
Program Evolution
The chief barriers to progress included problems that are general to long-term care, as well as challenges to team building and development of a palliative care perspective. In the former category, implementation was hampered by turnover in leadership staff (especially among critically important core staff such as the Director of Nursing/Nursing Home Administrator, unit managers, and social worker), reconciling workload issues, and recruiting a Medical Director able to lead one of the Palliative Care Teams. Challenges to building a strong palliative care team included minimizing frustrating communications with those physicians less sure of the goals of a palliative care approach, and achieving greater ownership, decision-making and follow-up on the part of all staff, not just the designated palliative care coordinators and advocates.
Our successes were greatest where staff were committed and stable from the outset. Our sites were not immune, however, to the staffing problems that challenge the entire long-term care industry: only one home had a consistent administrator, director of nursing, and assistant director of nursing, who also served as the palliative care coordinator. Success also demanded an integration of palliative care concepts into the ongoing care processes of the center; for example, integrated discussions of palliative care occurred as part of the weekly care planning conferences or quarterly quality improvement sessions, and included all levels of staff, especially nursing assistants. The program flourished in sites when most staff articulated (although not always documented) individualized care planning, reported comfort in talking with residents and families about end-of-life care, focused on promoting the quality of remaining life, and expressed empowerment and pride with their engagement in palliative care (usually bolstered by one or more significant success stories). By the end of the project, all of the homes had accepted the idea of palliative care and made considerable progress in integrating it into daily practice. Each had in place a process for discussing and reviewing advance directives, identifying persons appropriate for palliative care, assessing and treating pain and other troublesome symptoms, and attending to the psychosocial needs of families and residents at the end of life.
Nevertheless, staff still became frustrated when messages to families by physicians and nurses were inconsistent and interfered with good end-of-life decision making. Moreover, staff remained particularly challenged by fragmentation occurring between hospital discharge and nursing home admission, especially when it failed to account realistically for patient prognosis and family wishes. Limited resources that severely constrained the provision of psychosocial support for residents or bereavement support to families were also a source of frustration for staff.
Conclusions
Long-term care emphasizes rehabilitation and maintenance of function. Due to the nature of regulated documentation processes designed to detect changes in health status, most functional declines are targeted for intervention. In this context, death is often viewed as a negative outcome or the result of inadequate care. Only when palliative care is a routine part of nursing home care, can care be more appropriate to resident needs at the end of life. Currently, the full integration of palliative care into the life of nursing homes is hampered by staff turnover, shortages of qualified staff, regulatory requirements, competing administrative directives, health policies, and reimbursement. Nonetheless, we have seen that it is possible to promote positive change in end-of-life care in nursing homes. Addressing critical processes, implementing ongoing and directed training programs, integrating palliative care principles into existing systems, and providing ongoing support and consultation are key to overcoming barriers and sustaining change. Among the most important lessons learned were these defining characteristics for program success, which were articulated in a summary report by Jeanne Twohig at the National Program Office for “Promoting Excellence in End-of-Life Care”:
• Strong and committed leadership at both the facility and corporate level.
• Active involvement of the medical director as a link to community providers and enhancement of physician communication and buy-in.
• Availability of a nurse with expertise in hospice and palliative care to guide and support the implementation.
• Availability of an onsite champion to provide cohesion to the effort.
• An approach to organizational change that is both “bottom up” and “top down.”
• Implementing palliative care in small, manageable increments.
• Inclusion of all staff in the effort, from housekeeping to administration.
• Regular meetings where discussions of palliative care occur, keeping staff united in their mission and focused on the importance of the care planning process.
• Maintenance of palliative care residents within the existing unit structures, so that care is truly woven into daily practice.
• Capacity to use the palliative care program to improve marketability to residents, families, and referring institutions, enhance recruitment and retention of staff, and as a benchmark of quality care and regulatory compliance.
In summary, a model such as ours can be implemented in nursing homes, recognizing that it takes time to achieve such cultural change and that administrative and corporate leadership are essential. Long-term sustainability, however, will also depend on health care policies that encourage palliative care, reimbursement incentives recognizing that palliative care requires skill and resources, and public demand for palliative care as a reasonable expectation in every nursing home.
Acknowledgments
Without the support of corporate leaders at Genesis HealthCare and the staff at six participating Genesis facilities, this project would not have been possible.
Our appreciation goes as well to the staff of the National Program Office for “Promoting Excellence in End-of Life Care” at the Practical Ethics Center, University of Montana, and to the Robert Wood Johnson Foundation for 3 years of funding for the project.
Ms. Morrison reported that she is employed by and has received research support from Genesis HealthCare.
Ms. Parrish reported that she is employed by Genesis HealthCare and has received research support from the Robert Wood Johnson Foundation, the University of Pennsylvania School of Nursing, and Genesis HealthCare.
