| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
Correspondence: Address correspondence to Deborah P. Waldrop, MSW, PhD, Assistant Professor, University at Buffalo School of Social Work, 633 Baldy Hall, Box 601050, Buffalo, NY 14260. E-mail: dwaldrop{at}buffalo.edu
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Key Words: End-of-life care • Cancer • Qualitative research • Hospice • Terminal illness
The quality of physicians' communication with family members about a terminal illness directly influences their level of understanding about disease progression. Suboptimal communication leads to limited understanding of terminality and can contribute to delayed discussion of end-of-life options such as hospice care (Christakis & Iwashyna, 1998). However, because prognostication is difficult, physicians are often uncomfortable discussing terminality with patients, and some never mention that the illness will end in death (Drought & Koenig, 2002; Gage & Dao, 1996). Physicians are sometimes concerned that by introducing hospice care, they will be perceived as giving up on the patient or that hospice will be interpreted as a cost-saving measure (Brickner, Scannell, Marquet, & Ackerson, 2004; Johnson & Slaninka, 1999). Such miscommunication about the end of life can result in medical care that is contradictory to patient preferences (Institute of Medicine, 1997).
The length of hospice utilization consistently remains lower than the minimum benefit period of 6 months. Between 1995 and 2002, the mean length of stay decreased from 64 to 50.6 days and the median length of stay was between 20.9 and 29 days (National Hospice and Palliative Care Organization [NHPCO], 2004). Hospice professionals tend to view the prevalence of short stays as an underutilization of benefits and as problematic because they cause discontinuity in care during the often-turbulent final phase of an illness. Short stays allow little time to establish rapport or implement comprehensive care plans; they also may indicate that terminally ill people have experienced unmanaged symptoms such as pain or nausea before admission (Brickner et al., 2004; Miller, Weitzen, & Kinzbrunner, 2003). However, short periods of hospice utilization also may reflect knowledge of available benefits and thus be an indicator of personal preferences about the desired timing of formal end-of-life care. The purpose of this study was to explore the psychosocial dynamics that accompanied limited hospice utilization in people over the age of 65 who died from cancer. This article presents a typology of the illness-related psychosocial dynamics that preceded hospice stays that were shorter in length than the national median.
Psychosocial Dynamics and Hospice Utilization: An Ecological Framework
The reciprocity between an individual and the social environment is central at life's end. From the perspective of the ecological framework, a dying person can be seen within ever-widening social networks: family, social support, and health care (Bronfenbrenner, 1986). This study was guided by concepts from this theoretical framework, including microsocial (individual), mezzosocial (family, social support, and provider) and macrosocial (health care programs and policies) social environments.
Individual or microsocial factors such as age, race, gender, marital status, financial resources, religion, and culture shape perspectives about illness and death (Barriers to Hospice Care, 2000). The type and trajectory of an illness, previous experience with hospitals and hospice, family roles and relationships, the complexity of care needs, the amount of burden, and the availability of caregivers are also influential (Bernard & Guarnaccia, 2003). Emotional responses often result from fear of managing death at home, stopping curative treatment, or being abandoned by providers (Barriers to Hospice Care). Fear can lead to perceptions that hospice means "giving up" or that "nothing" will be done (Huskamp, Buntin, Wang, & Newhouse, 2001). Coexisting emotions, such as anticipatory grief, avoidance, or denial, also influence caregivers' understanding and acceptance of a terminal diagnosis (Jennings, Ryndes, D'Onofrio, & Bailey, 2003).
Admission to hospice can be viewed as both a microsocial and mezzosocial issue; it requires that patients, families, and physicians establish a shared understanding that death will likely result within 6 months and that treatment will shift from "cure" to "care" (Myers, 2002). Because hospice care is family focused, it may prepare caregivers for approaching loss and bereavement experiences (Institute of Medicine, 1997; Schacter & Coyle, 1998). This is an important feature of hospice, because research shows that caregivers of patients enrolled with hospice for short stays are likely to have major depressive disorders (Bradley et al., 2004).
Health care policy and benefits, regional availability, differential interpretation, and cost of care are macrosocial factors that influence utilization. Because hospice regulations stipulate eligibility as fewer than 6 months to live, end-of-life care planning often occurs only during the final phase of an illness rather than being integrated throughout the course of treatment (Christakis & Escarce, 1996; Huskamp et al., 2001). Varying interpretation of regulations can lead to restrictive policies that may deter potential admissions and contribute to uncertainty about the appropriate timing for hospice (Lorenz, Asch, Rosenfeld, Liu, & Ettner, 2004). For those who do enroll in hospice, even for short stays, the mean of total expenses in the last month of life has been found to be significantly less than that in other settings (Miller et al., 2004).
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
The overall project employed mixed methods (qualitative–quantitative) in an equivalent status sequential design (Tashakkori & Teddlie, 1998). The study aimed to investigate the views of hospice professionals and caregivers about psychosocial factors that influence cancer patients' utilization of hospice. Phase I was an ethnography of seven hospice teams, which focused on team communication about limited utilization. Phase II involved focus groups to gather professionals' perspectives about short stays. The results of the first two phases were used to develop the questions that guided Phase III interviews with caregivers. Phase III interviews had four sections: (a) medical history, (b) social support, (c) terminal decline, and (d) end-of-life policies.
This article presents the analysis of the medical history segment of the interviews and a typology of the psychosocial factors that preceded short stays. The medical history segment had three sections: (a) open-ended questions about the progression of the illness (history), (b) a timeline of illness events (timeline), and (c) a subsequent review of the patient's hospice record (hospice record).
The medical history section was guided by 23 open-ended questions grouped under eight domains that represent components of the illness: baseline, symptoms, diagnosis, treatment, progression of the illness, hospice decision, and death. Appendix A includes the questions in each of the eight domains.
|
Hospice records were systematically reviewed to collect supplementary descriptive medical data that was compiled on a medical records review sheet. The summary sheet included categories for the diagnosis, activities of daily living, the Edmonton Symptom Assessment Scale (e.g., anxiety, confusion, shortness of breath, appetite), Palliative Performance Scale (a 10–100 score of patient's functional decline), pain intensity rating, and the patient's goals.
Sampling
Purposeful or theoretical sampling was used to identify and contact potential respondents who had experienced short hospice utilization (Padgett, 1998; Strauss & Corbin, 1998). During the recruitment phase, an introductory letter explaining the study's purpose was included in all bereavement packets. An invitation letter was mailed to caregivers who met inclusion criteria: the hospice patient was 65 years of age or older and died from cancer within 2 weeks of admission.
A sample size of 50 patients was established to obtain maximum breadth and heterogeneity in caregiver experience and to test the statistical significance of quantitative measures in subsequent analyses. Letters were sent to 151 potential participants 6 weeks after the patient's death. When follow-up phone calls to invite participation were made 1 week later, 34 people declined. An additional 67 participants were called repeatedly, but the response rate was only 33%. Nine participants suggested including relatives or friends who helped with caregiving. The final sample consisted of 59 caregivers.
Eighty-one percent (48) of the participants were female. Caregivers ranged in age from 26 to 88, with a mean age of 61; 88% (53) were White and 12% (6) were Black. Forty-six percent (23) of the patients were female; they received hospice care from 3 hr to 14 days (M = 6 days). At the time of their hospice admission, 48% (24) were diagnosed with multiple types of cancer, 36% (18) with one type of cancer, and 16% (8) with both cancer and coexisting medical conditions. Table 1 presents patients' primary diagnoses, place of death, and their relationship with the caregiver.
|
A research team (a research assistant and a qualitative methods consultant) and I conducted the data analysis. A research assistant and I independently coded the interviews and later compared them for congruence. We determined interrater reliability by dividing the number of agreements by the sum of the number of agreements and disagreements (Miles & Huberman, 1994; p. (64). Initially, interrater reliability was 80%, but it subsequently improved to 92% as a result of regular meetings to compare coding results, refine and document the analytic decision making, address discordance in the analysis, and achieve consensus on the final coding structure. The qualitative methods consultant, who assisted with development of the typology, reviewed an audit trail or log of analytic decision making.
Coding
Coding procedures aimed to identify manifest content, or the objective, quantitative elements, and latent content, or the underlying meaning of the narrative (Tashakkori & Teddlie, 1998). First, we used systematic or axial coding to incorporate an a priori or start list of 47 codes based on concepts from the ecological model (Miles & Huberman, 1994). Examples of these codes include diagnosis, provider communication, family dynamics, and death experience.
Next, we used open coding or line-by-line analysis for the purpose of discovering the naturally occurring themes within the groups (Padgett, 1998). We incorporated "free codes" or phrases that capture the meaning of a text passage in the coding scheme and used them to reexamine previously coded transcripts. Examples of free codes are "patient avoidance" and "comorbidities."
A distinct pattern emerged that distinguished caregivers whose relative had been diagnosed with cancer late in the illness from those whose relative had lived with cancer for months to years. We used timing of the cancer diagnosis to separate the sample into two subgroups: (a) late diagnosis (a time period of less than 1 month before death) and (b) known diagnosis (months to years before death).
Matrices
After the coding was completed, data analysis using within-case and cross-case analysis continued. We conducted within-case analysis by creating a matrix of individual, family, and provider variables that contributed to limited hospice utilization. Cross-case analysis involved separating individual, family and provider factors into two matrices: (a) late diagnosis and (b) known diagnosis (Miles & Huberman, 1994).
Typology
We used constant comparative analysis to construct a typology. This process involved breaking the text into units of information, which we used to create mutually exclusive and internally consistent definitions of the categories (Lincoln & Guba, 1985, pp. 347–351). We used negative case analysis to exclude three situations that did not fit within the typology (e.g., death from postsurgical complications; caregiver could not recall details about hospice; earlier admission was declined by hospice).
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
Late: Missed Diagnosis
The caregivers in this group (n = 8) reported that the cancer was unknown to them and undetected by a provider, even though they noticed significant functional decline in the ill person and knew something was wrong. Caregivers generally attributed the missed diagnosis to fragmented medical care (e.g., missed lab work) or lack of follow-up. When the diagnosis was made, shock reverberated through these families. One patient's daughter described her response to learning of her mother's terminal condition 10 days before her death by saying, "How do you go from thinking it's a sinus infection to terminal thyroid cancer?"
Another caregiver described how her mother was not eating well, but the doctor's initial diagnosis did not detect stomach cancer:
Mom wasn't eating well. Her blood count was low, so they gave her a couple transfusions. She was going to have an endoscopy; when we picked her up she was vomiting so we took her to the hospital. They gave her blood, did the test and found cancer in 75% of her stomach. She had seen her doctor of 20 years the week before—he told her that she was in great health and she would live to be 107. I don't understand how it was missed when it was that big.
Her mother died during the second day of hospice care.
Late: Masked by Comorbidities
The caregivers in this group (n = 8), aware of their relative's multiple medical problems and marked physical deterioration, knew that the person was ill. The primary working diagnosis was a coexistent medical condition that masked the cancer; for example, increasing aches and pains were attributed to arthritis or age (masking bone cancer), a "smoker's cough" to bronchitis or pneumonia (masking lung cancer), diarrhea and weight loss to Crohn's disease (masking colon cancer). In these situations, the patient was being treated for medical conditions without improvement.
Some patients had improved after facing previous life-limiting situations. These situations set up a zigzag trajectory in which the person approached death but recovered. Mrs. V's son described how her coexisting cardiovascular disease masked the detection of her cancer: "She was an incredible fighter, and overcame health problems that many people don't. She was in and out of the hospital, but each time we felt that things were getting worse, she recovered." Her diagnosis of ovarian cancer was made 2 weeks before death.
Mrs. M describes how dementia masked an earlier diagnosis of cancer. Her aunt lived in an assisted living facility (ALF):
My aunt was 91 and there was some evidence of dementia. She started having some swelling of her feet so her doctor put her on Lasix and gave her TED hose but it didn't help. She wasn't eating much and her belly was getting big. He did a couple x-rays. The ALF called and they said she was having trouble breathing. When I called he said, "Your aunt has lung cancer that has spread to her liver; she is full of cancer."
When the diagnosis was masked by comorbidities, some caregivers initially responded with shock and anger. Others were resigned to the decline and were not surprised by the final diagnosis.
Late: Patient Avoidance
Some patients lived with multiple medical problems and marked functional decline, but actively resisted medical care. Caregivers (n = 6) rationalized that their relatives were afraid to see a physician, despite the caregivers' ongoing and sometimes intense efforts to encourage the ill person to seek medical attention. One patient's daughter recalled that since her father's death some years before, her mother had repeatedly told her, "If I have cancer, I don't want to know about it." Physical decline, decreasing independence, and poor health were described by the caregivers. Caregivers believed that their relative "knew" something was terribly wrong, but was "hiding" discomfort, not wanting to complain, demonstrating stoicism, or avoiding any discussion of their health. The central theme identified in these interviews was avoidance, which was a barrier to both treatment and a longer hospice stay. Mrs. L described her father's resistance:
He couldn't use one leg anymore and didn't have a doctor so we took him to the ER. They sent him home with a quad cane and said it's probably arthritis. Soon he couldn't walk on the quad cane. I got an appointment but he said, "I don't want to go," so I cancelled it. Then he had to use a urinal during the night because he couldn't get to the bathroom; and he was coughing up blood. He didn't want to go to the ER and be sent home again. I made another appointment but he said "No, I'm not going."
After Mrs. L called an ambulance and admitted him to the hospital, metastatic lung cancer was discovered. He was transferred to a hospital-based hospice bed and died 10 days later.
Age-Related Attributions
This theme emerged from analysis of the late diagnosis groups. Some patients attributed their symptoms to aging. Mr. R remembered his mother saying, "I'm 74 years old, I've been here a long time. Whatever is going to be will be." Beliefs that discomfort and functional decline were caused by advancing age diminished urgency about obtaining a diagnosis. In other instances, providers attributed the ill person's decline to aging. Ms. S described her interactions with her father's primary care doctor:
I think the MD thought that because Dad was 75 he was old, he'd survived two bouts of cancer and should be thankful to have survived. We weren't asking for organ transplants, but he could have lived to 80 or 85 if they'd done screening.
According to caregivers, these age-related attributions influenced both the patient's understanding of the illness and his or her relationship with the provider.
The families of patients in the late diagnosis group moved quickly from diagnosis to hospice admission and death. In some situations, the diagnosis was made in the hospital and the person never went home; he or she was transferred to a hospice bed in the hospital or to a nursing home.
Known Diagnosis
The caregivers for 25 (50%) patients reported that they had known about the diagnosis of cancer, for a time period ranging from several months to years. An exploration of these transcripts yielded two additional subgroups: those who received ongoing treatment and those whose families managed the care without hospice until the very last days.
Ongoing Treatment
Some patients (n = 15) continued treatment until the final days before death. They knew the diagnosis was terminal, but they continued chemo, radiation, alternative therapy, hyperalimentation, antibiotics, or rehabilitation. They reported believing that treatment would cure the cancer; "treatment was hope" and it meant that the person was "buying time" or "fighting it." In some situations, the patient's doctor had supported ongoing treatment by saying, "This cancer is curable" or "We can treat this disease." When patients moved directly from treatment to hospice care, it precluded the mental and emotional shift that typically accompanies the end of treatment. One patient's wife described her difficulty with this abrupt transition: "When I put him in the hospital on that Saturday I just figured they'd clear up the infection, he'd go for rehab and he would be home. I did not expect my husband to die."
Mrs. N describes the juxtaposition of ongoing curative treatment and the dying process:
Mom's lymphoma came back and she had chemo. After she developed an obstruction, her doctor said, "Looking at you, your spirit and strength, I think you can get through the surgery." Her oncologist said, "I want to give her chemo." They planned hyperal to increase her albumin so they could bypass the lymphoma, she could eat, then she could go through Ritoxan; she was absolutely game to do this. They showed us how to do the hyperal at home. We thought she would beat it.
Two days after beginning hyperalimentation Mrs. N's mother became nonresponsive and her family contacted hospice; she received care for less than 24 hr before her death. Some in this group focused intensely on treatment and were surprised by the person's death. Others recognized that the treatment was not working and shifted their focus from cure to care for a brief but peaceful time before death.
Family Managed Care
In these situations, caregivers (n = 10) reported understanding that the cancer was terminal. When the ill person had reached the end of treatment and learned that nothing more was available, his or her caregivers recognized the decline and knew the illness would end in death. However, either the patient refused hospice care or the family did not want outside help. Some caregivers explained that hospice meant that the person was "giving in" and that hospice admission might cause the person to succumb to death. Caregivers described wanting to retain normalcy for as long as possible. Some feared that admission to hospice would change their focus from living to dying; they were not ready to make this shift until the final stage of life. In other situations, the ill person refused to talk about the approaching death and struggled to retain independence. Other caregivers wished to retain control of the situation and expressed the fear that the transition to hospice would bring a new care regimen with strangers and a loss of control. Mrs. E described how she and her husband made the decision to contact hospice:
I approached him about hospice but he didn't want anything to do with it, because it was too final. He thought we were rushing him. When it got harder for us to get medicine and there was nothing more they could do for him, then he felt hospice could help us by delivering the meds, so we called.
In the family-managed situations, a shared decision was initiated at particular turning points.
Turning Points
The caregivers in the known diagnosis group reported that they cared for the terminally ill person until a turning point made care impossible to manage. Caregivers in both the ongoing treatment and family managed care groups described this point of no return as when illness-related changes made it impossible for them to manage care without help (e.g., the person could not stand, became incontinent, or was in intense pain). Medical turning points occurred when a physician stated unequivocally that treatment was no longer working and nothing else was available. Functional turning points took place when patients could not walk, stand, or get out of bed and the physical burden of care became overwhelming. Pragmatic turning points occurred when the person's care needs became intense and the caregiver either could not leave the home even to pick up medicines, or the need for equipment, information, respite, or support became central. Emotional turning points occurred when the ill person and caregiver felt overwhelming anxiety or fear.
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
The explicit knowledge of a terminal diagnosis was linked to deferred hospice admission for half of the sample. Hospice admission is a critical juncture on the trajectory of a terminal illness, but crossing this threshold requires changing perspective and routine as well as integrating emotional, psychosocial, and cognitive factors that can cause resistance to the transition. Families cannot control the disease process, but they may maintain feelings of control by avoiding hospice admission, because (a) treatment holds a promise not for a cure but for the extension of time shared, or (b) hospice represents a "savings account" of benefits and services that are available when necessary.
Reflecting carefully on the meaning of these findings raises several critical issues. First, it is crucial to consider whether short stays are problematic and, if so, how. The appropriate length of hospice care varies and is itself situation specific. Hospice professionals' expertise can potentially enhance the end-of-life experience when they become involved earlier; alternatively, hospice professionals do experience frustration with short stays, which limit the time available to get to know the patient–family dyad and to intervene before death. However, caregivers who knew about the terminal diagnosis often wanted to feel some control over the advancing condition, either by continuing treatment or by managing the illness on their own until assistance became essential. These findings both build on and extend the work of Christakis (Barriers to Hospice Care, 2000), who identified psychosocial dynamics as a barrier to earlier admission. Extending the work of Lackan, Ostir, Freeman, Mahnken, and Goodwin (2004), this study reveals additional factors that contribute to diverse patterns of health care use at the end of life. It also underscores the importance of understanding decisions about and transitions between types of care in long-term illness (Iwashyna, Chang, Zhang, & Christakis, 2002).
Policy Implications
This study's results point to the need to examine the link between large numbers of short stays and the long-term financial viability of hospice organizations. Hospice benefits are paid at per diem rates that depend on the acuity of care (Centers for Medicare and Medicaid Services, 2004). Because most services are provided during the first and last days of a stay, the charges on these days often include multiple services that exceed the reimbursement rate (Buntin & Huskamp, 2002). Because short stays allow insufficient time to recoup the cost, large numbers of them strain the financial viability of hospice organizations (Medicare Payment Advisory Commission, 2002).
Practice Implications
To minimize late diagnoses, practitioners should attend carefully to functional and physical changes reported by older people and not attribute them to age. This subtle form of ageism may be a barrier both to earlier diagnosis as well as to open discussions about end-of-life care and hospice. However, practitioners should also respect older adults' rights to self-determination. The right to refuse a provider's advice is paramount even when people avoid medical diagnoses or hospice utilization. Forcibly confronting avoidance is neither helpful nor advisable. Given different patterns of readiness to acknowledge an approaching death and to prepare for it, these results illustrate the importance of understanding and respecting unique and distinctive patient–family dynamics.
This study has several limitations. First, it sought only family caregivers' perspectives; the views of providers and patients would have deepened our understanding of the dynamics. Second, learning from caregivers' experiences is critical to understanding consumers' perspectives. However, this study provides only retrospective perceptions, which means that the extent of participants' prior knowledge and perceptions of hospice and how they changed with experience are unknown. Third, the sample demographics present limitations. Despite the wide geographic area and large daily census of Hospice Buffalo, the sample was drawn from only one hospice. In addition, the sample was primarily White. The sample statistics reflect national hospice utilization demographics (81% White, 9% Black, 4.3% Hispanic, and 4.6% multiracial; see NHPCO, 2004), but they also illustrate the need for greater understanding of variance based on racial, ethnic, and geographic factors. Finally, the study focused only on cancer patients, who represent just 50% of all hospice patients. Future studies should explore whether differences exist between hospice patients with cancer and those with other terminal conditions.
This study does have implications for future research. The hospice patient population continues to change, gradually evolving since the program's establishment (Gage & Dao, 1996). As service options, consumer populations, and patterns of utilization shift, it is import to define subgroups of users and to understand the dynamics that characterize them. In addition, as life expectancy lengthens, how hospice is perceived and utilized will continue to evolve. Integrating caregivers' perspectives is a central component of enhanced care for the dying (George, 2002). Short stays are problematic when people do not understand the terminality of their illness and have no options for earlier discussion of end-stage care. For others, short stays are a reflection of how they lived while dying and an intentional way of meeting their final needs. Understanding the differential characteristics and needs of subgroups of terminally ill people is central to providing good care at life's end.
|
Footnotes |
|---|
1 University at Buffalo School of Social Work, NY.
Decision Editor: Linda S. Noelker, PhD
Received for publication January 13, 2005. Accepted for publication July 15, 2005.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  N. E. Schoenberg, D. Shenk, and C. S. Kart Food for Thought: Nourishing the Publication of Qualitative Research Journal of Applied Gerontology, February 1, 2007; 26(1): 4 - 16. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
