The Gerontologist Copyright 2008 by The Gerontological Society of America Vol. 48, No. 4, 477–484
Referral and Timing of Referral to Hospice
Care in Nursing Homes: The Significant
Role of Staff Members
Lisa C. Welch, PhD,
1Susan C. Miller, PhD,
2Edward W. Martin, MD,
3and Aman Nanda, MD
4Purpose: Given concerns about end-of-life care for many nursing home (NH) residents, this study sought to understand factors influencing hospice referral or nonreferral as well as timing of referral. Design and Methods: We conducted semistructured interviews with personnel from seven participating NHs and two hospices. We interviewed NH directors of nursing regarding facility referral practices and conducted interviews with 34 NH nurses, 30 NH aides, and 17 hospice nurses knowledgeable about the factors that led to the hospice status of 32 NH decedents. Selected decedents varied by diagnosis and hospice status (received hospice for.7 days,7 days, or not at all). We audiotaped, transcribed, and coded interviews for themes. Results: NH staff members’ recognition of terminal decline, beliefs about hospice, and initiative significantly influenced hospice referral and timing of referral. Staff members’ recognition of familiar signs of decline facilitated hospice referral; in contrast, a perception that death was unexpected impeded referral, and a perception of uncertain prognosis delayed referral. Staff members’ beliefs that hospice does not add value to NH care or is for crises only impeded referral, and a belief that hospice is only for the ‘‘very end’’ delayed referral. Residents received hospice for longer periods when staff believed that hospice complemented NH care and
when staff took the initiative in raising the option of hospice. Implications: Enhanced training about rec-ognizing terminal decline, hospice services in NHs, and the role of staff initiative would support NH staff in raising the option of hospice when appropriate. Key Words: End-of-life care, Dying, Recognizing terminal decline, Staff beliefs, Continuing education
Although nursing homes (NHs) increasingly are sites of end-of-life care (Mitchell, Teno, Miller, & Mor, 2005), concerns have been raised about the quality of that care for many residents. Pain assess-ment and manageassess-ment often is inadequate (Cohen-Mansfield & Lipson, 2002; S. C. Miller, Mor, & Teno, 2003; Teno et al., 2004), and family members of dying residents report stress due to limited physician visits and insufficient staffing (Shield, Wetle, Teno, Miller, & Welch, 2005). The utilization of hospice to augment end-of-life care in NHs is one viable option for addressing such concerns.
Hospice is an interdisciplinary team approach to the care of terminally ill patients and their families that includes management of physical symptoms, social and spiritual care, emotional and psycholog-ical support, and bereavement counseling and support (Moon & Boccuti, 2002). Documented benefits of hospice care for NH patients include better pain management (S. C. Miller, Mor, Wu, Gozalo, & Lapane, 2002), fewer hospitalizations (Gozalo & Miller, 2007), greater family satisfaction with end-of-life care (Baer & Hanson, 2000; Facts on Dying, 2004), and lower costs (Gozalo et al., 2008; S. C. Miller et al., 2004). In addition, evidence has shown a spillover effect in that nonhospice residents living in NHs with a higher proportion of residents enrolled in hospice are less frequently hospitalized at the end of life and more frequently have a pain assessment performed (S. C. Miller, Gozalo, & Mor, 2000, 2001).
We would like to thank Sharon Bayha, RN, and Nancy Grossman for their research assistance. This research was funded by the Project on Death in America, Open Society Institute, and, in part, by the Robert Wood Johnson Foundation (Grant 049891).
Address correspondence to Lisa C. Welch, PhD, Department of Sociology and Criminal Justice Studies, Southern Illinois University Edwardsville, Box 1455, Edwardsville, IL 62026. E-mail: Lwelch@ SIUE.edu
1
Department of Sociology and Criminal Justice Studies, Southern Illinois University Edwardsville.
2Department of Community Health and Center for Gerontology and
Health Care Research, Brown University, Providence, RI.
3Home and Hospice Care of Rhode Island, Pawtucket. 4Rhode Island Hospital Division of Geriatrics, Providence.
Although hospice is one documented option for enhancing end-of-life care in NHs, it is not the only means of providing quality end-of-life care to residents and families. A recent study (Munn, Hanson, Zimmerman, Sloane, & Mitchell, 2006) found that, among NH and residential care/assisted living residents and their families, differences in end-of-life care provided by facility staff and hospice largely disappeared when death was expected (as reported in retrospective accounts). More research is needed to disentangle whether expectation of death has an effect independent of hospice utilization or presence in an NH, but this study does bring evidence that some NH residents receive quality end-of-life care without hospice.
Still, the documented benefits of hospice care for many members of this vulnerable population show that hospice is a viable option that residents and families can use to access a broad array of end-of-life services. As such, the literature supports the value of attaining a clearer understanding of the factors that impact referral and timing of referral to hospice in NHs.
Although hospice utilization in U.S. NHs has increased substantially (Han, Remsburg, McAuley, Keay, & Travis, 2006; S. C. Miller, Intrator, Cang, & Mor, 2006), hospice use across states, NHs, and NH residents varies considerably (S. C. Miller et al., 2006; S. C. Miller & Mor, 2004). In 2000, the proportion of NHs using hospice ranged from 36% in Wyoming to 96% in Florida, and the proportion of NH decedents enrolled in hospice varied from 9% in Vermont to 42% in Oklahoma (S. C. Miller et al., 2006). Among NH residents, factors associated with a lower likeli-hood of hospice enrollment include residence in facilities with a higher proportion of non-White residents as well as individual-level characteristics of male gender, non-White race, the presence of conges-tive heart failure, and higher physical and cogniconges-tive impairment (Gozalo & Miller, 2007).
Length of stay in hospice is an important factor for achieving its full benefit. Nationally, hospice lengths of stay have decreased (General Accounting Office, 2000; G. W. Miller, Williams, English, & Keyserling, 2002). Between 1991 and 2000, the proportion of adult hospice patients with stays of 7 days or less increased from 23% to 37% (Han et al., 2006). Shorter lengths of hospice stay concern policy makers and providers (Health Care Financing Administration, 2001; G. W. Miller et al., 2002). Later referrals are unlikely to allow sufficient time for hospices to accomplish their goals of self-determined life closure, safe and comfortable dying, and effective grieving (Emanuel, von Gunten, & Ferris, 2000; S. C. Miller et al., 2003; National Hospice and Palliative Care Organization, 1997; Schockett, Teno, Miller, & Stuart, 2005). Short lengths of stay allow little time to control symptoms effectively or to deal with complex emotional and spiritual issues that may accompany a dying process.
Research has documented multiple patient bar-riers to hospice referral. Such barbar-riers include misunderstandings about end-of-life care (Silveira, DiPiero, Gerrity, & Feudtner, 2000) and reluctance to discuss end-of-life issues with doctors (Pfeifer, Mitchell, & Chamberlain, 2003). Additionally, patients are often overly optimistic in estimating their prognoses, leading to preferences for more aggressive treatments if they believe they have a life expectancy of greater than 6 months (Weeks et al., 1998).
The process of hospice referral also depends on physicians. In the NH setting, though, staff members provide more direct care to residents (Shield et al., 2005). A recent intervention that assessed NH residents for their hospice appropriateness and then shared this determination with physicians along with requests to authorize hospice informational visits re-sulted in significantly longer hospice stays (Casarett et al., 2005). This finding suggests the importance of understanding more about the everyday NH pro-cesses associated with hospice referral or nonreferral and with timing of referral, which was the goal of the present study.
Methods
A qualitative research design allowed for respon-dents to explain from their perspectives the factors that led to a resident receiving hospice or not as well as the circumstances around the timing of referral when it occurred. We conducted 88 semistructured interviews and analyzed them for main themes.
The sample design sought data from NH and hospice personnel about hospice referral practices in participating NHs and the factors that impacted the hospice status of particular decedents. To achieve this, sample selection followed a multistage process: (a) recruitment of hospices and NHs, (b) selection of decedents within NHs, and (c) identification of NH and hospice providers who were familiar with the end-of-life care of the selected decedents.
First, we recruited two hospices in the same state; these hospices provided information about referrals from local NHs. Based on this information, we recruited two NHs within each of three categories: (a) lower hospice referral (below a hospice’s median number of referrals per NH in a 1-year period), (b) higher hospice referral (above a hospice’s median number of referrals per NH in a 1-year period) with a higher proportion (40%) of short hospice stays (7 days or fewer), and (c) higher hospice referral with a lower proportion (,40%) of short hospice stays. We chose these NH categories to increase the likelihood of identifying specific decedents within NHs who had various hospice statuses (no hospice, shorter hospice stay, or longer hospice stay). We recruited a seventh NH to replace one participating NH that closed during the study period (see Table 1).
At the time an NH agreed to participate, the principal investigator (Susan C. Miller) and a trained inter-viewer interviewed the director of nursing in person regarding that NH’s hospice referral practices.
Second, we identified a stratified sample of NH decedents from weekly lists of decedents sent by participating NHs and hospices over an 8-month period during 2003–2004. The sampling goal was to identify 24 NH decedents. We sought six decedents from each of four diagnosis categories: cancer with Alzheimer’s disease (AD)/other dementia, cancer without AD/other dementia, AD/other dementia without cancer, and other diagnoses. We chose these categories based on research that shows different patterns of care and costs across these diagnosis groups (S. C. Miller et al., 2001, 2002, 2004; S. C. Miller & Mor, 2001). Within each diagnosis category, we sought two decedents for each of three hospice subcategories: without hospice care, with hospice for 7 days or less, and with hospice for more than 7 days. To lessen the respondent burden within any one hospice or NH, we chose no more than two decedents from one hospice or NH in a 1-week period.
In seeking to fulfill the decedent sample goals as well as limit respondent burden within a given facility, we chose a total sample of 32 NH decedents (see Table 2). Most decedent sample goals (by diagnosis and hospice status) were met or exceeded; however, even with expanding the number of decedents, only one resident with cancer (with or without AD/other dementia) who received hospice for 7 days or less and no residents with ‘‘other’’ diagnoses who received hospice more than 7 days died during the study period.
Third, as each decedent was selected from the weekly lists, interviews were conducted with at least one NH nurse, NH aide, and hospice nurse (if applicable) who was familiar with that decedent’s end-of-life care. The trained interviewer conducted telephone interviews with a total of 34 NH nurses, 30 NH certified nursing assistants, and 17 hospice nurses about the factors that led to those in the decedent sample receiving hospice or not and the timing of referral (if applicable). Data collection ended when most decedent sampling goals were met (or exceeded) and additional interviews no longer provided new information.
To conduct the interviews, we developed three semistructured interview guides (one for directors of
nursing, one for NH staff, and one for hospice nurses), piloted them with non-study providers, and modified them accordingly (see Table 3). We did not include pilot interviews in the analysis. In addition to specific interview guide questions, the interviewer used probes to elicit detailed explanations in respondents’ own words.
We audiotaped all interviews and transcribed them verbatim. An interdisciplinary team composed of two physicians (a hospice physician and an NH medical director), a nurse research assistant, a health services researcher with gerontology and epidemiol-ogy expertise, and a sociologist coded interviews to identify major themes. Each team member brought a particular base of knowledge and experience to the analysis. The team began with the two broad a priori categories of ‘‘impediments’’ and ‘‘facilitators’’ to referral and earlier referral; within these broad categories, specific themes emerged from the data.
As analysis began, each team member reviewed transcripts individually and then met as a group to discuss and reach consensus about codes. When coding disagreements arose, team researchers typi-cally deferred to the experience of team physicians in NH settings to reach consensus. When new inter-views were not yielding additional themes, two team members (Lisa C. Welch and Susan C. Miller) coded subsequent interviews, team physicians reviewed these codes for validity, and codes were modified accordingly. We identified as major findings the themes that reappeared most often across interviews. We used ATLAS.ti qualitative data software (http:// www.atlasti.com/) to facilitate analysis.
Results
According to the directors of nursing, none of the study NHs had written procedures for regularly assessing residents’ terminal status or eligibility for hospice care, and none had formal protocols in place for communicating with physicians about hospice eligibility. Instead, NH staff members shaped hospice referral and timing of referral according to their abilities to recognize decline, beliefs about hospice services, and initiative in raising hospice as an option.
Table 1. Nursing Home Facility Characteristics (N= 7)
Characteristic n M(Range) Years certified 16.9 (4–28) Medicare participation 7 Medicaid participation 7 Certified beds 167.7 (61–260) For-profit ownership 5 Freestanding 7
Table 2. Decedent Sample
Group No Hospice Hospice 7 Days Hospice .7 Days Total
Cancer and AD/dementia 2 0 4 6
Cancer, no AD/dementia 2 1 5 8
AD/dementia, no cancer 2 5 4 11
Other diagnosis 4 3 0 7
Total 10 9 13 32
Recognizing Decline
NH staff members’ abilities to recognize terminal decline played a central role in whether a resident received hospice and for how long. The most common facilitator of referral was staff recognition of familiar signs of terminal decline, especially a diagnosis of cancer with metastasis or hospital admission. Regarding an 83-year-old resident with cancer who received hospice for 37 days, an NH staff member said, ‘‘We knew that she had cancer. We knew it was terminal. We had spoken to her nephew who is her next of kin. And [decedent] and her nephew made the decision to go with hospice before she became very ill.’’ Similarly, speaking of a 97-year-old resident with AD/dementia who received hospice for 81 days, an NH staff member reported, ‘‘As I recall, this little lady had gone into the hospital
with an exacerbation of a CHF [congestive heart failure], and I think that’s what brought the family to decide not to hospitalize any further.’’
When staff members did not recognize terminal decline, referrals were impeded. When speaking of the care of residents who did not receive hospice care, staff members frequently reported their perception that death was unexpected due to a rapid decline. In describing an 80-year-old resident with a diagnosis other than AD/dementia or cancer who was not cared for by hospice, a registered nurse (RN) stated, ‘‘. . .we didn’t think she was terminal. I mean she wasn’t supposed to die.’’ An NH staff member described another 85-year-old decedent with a diagnosis other than AD/dementia or cancer who also received no hospice services as ‘‘calm. . .comfortable and. . .very responsive and. . .maybe a little more sleepy.’’ Table 3. Main Interview Questions by Interviewee Type
Interviewee: Nursing Home Nurse and Aide 1. Can you discuss whether you were surprised when (resident) died? Why?
2.For nurses only:Were orders limiting terminal interventions such as resuscitation or hospitalization in place prior to (resident’s) death?
If yes:
a. For approximately how long?
b. Why were orders limiting terminal interventions initiated? c. Who initiated discussion regarding these orders?
3.For nurses only:Was a hospice referral discussed? Who initiated this discussion?
4. Would you discuss what factors you feel led to (resident) (being/not being) cared for by hospice? 5.If resident received hospice:
a. Can you discuss in more detail why you think (resident) was referred to hospice at the time (he/she) was?
b.If hospice stay7 days:What do you think would have needed to happen for the hospice referral to have occurred earlier? c. Can you describe what happened once the hospice was contacted?
d. Do you feel hospice benefited (resident)? the family?
e. Would you discuss why you feel that hospice care was (beneficial/not beneficial)? 6.If resident did not receive hospice:
a. Do you feel that hospice care would have benefited (resident)? the family?
b. Would you discuss why you feel that hospice care would (have been/not have been) beneficial? Interviewee: Hospice Nurse
1. Were orders limiting terminal interventions such as resuscitation or hospitalization in place when (resident) was admitted to hospice?
2. Were legal (planning) documents in place such as an advance directive or durable power of attorney?
3. From your perspective, can you describe what factors may have led (resident) or (his/her) family to choose to be cared for by hospice?
4. Can you discuss why you think (resident) was referred to hospice at the time (he/she) was?
5.If hospice stay7 days:Can you cite a specific nursing home or hospice policy or practice that if added or improved could possibly have resulted in (resident) receiving hospice for a longer period of time?
6. Do you feel that (resident) and (his/her) family received the full benefits of hospice care?
7. What do you think would have allowed hospice staff to be of greater benefit to (resident) and to his/her family? Interviewee: Nursing Home Director of Nursing
1. Can you discuss how hospice referral occurs? 2. What residents get referred to hospice?
3. What factors do you feel foster hospice referrals? What factors are barriers to referrals? Are Medicare Part A skilled nursing home residents likely to access hospice care prior to death?
4. In your opinion, are referrals generally being made at the appropriate time given the nature of the patient’s condition? Explain. a.If no:What is inhibiting more timely referrals?
A death perceived as unexpected was an imped-iment to referral even among staff who believed that a resident and family could have benefited from hospice services. In describing a 76-year-old resident with cancer and dementia who received no hospice services, an RN stated, ‘‘I think it would have been a benefit to [the resident] and his wife and family if that would have been the case. . .But like I said, he died suddenly.’’
Staff member perceptions of uncertain prognosis delayed the timing of referral for decedents with less dramatic signs of decline (e.g., changed eating habits or recurring pneumonia). Although none of the participating NHs had a formal procedure for assessing terminal decline, it may be of even more significance that many NH staff believed that recognizing a terminal decline cannot be done with much accuracy in an NH setting. One NH staff member reported a commonly held view when she said, ‘‘Unfortunately, it’s very difficult in a nursing home to judge when end of life is coming. It’s just very difficult. . .’’
Beliefs About Hospice Services
NH staff members’ beliefs about hospice services also shaped hospice referral and timing of referral. A belief that hospice does not add value to NH end-of-life care or that hospice is only for crises impeded referrals. Among those who received hospice ser-vices, a belief that hospice is appropriate only for the ‘‘very end’’ impeded earlier referral to hospice. In contrast, when NH staff believed that hospice complemented NH services, residents received hos-pice for longer periods.
A major impediment to referral to hospice was a belief among staff that hospice does not add value to NH end-of-life care. Some staff believed that NHs provide essentially the same services for dying residents as do hospices. An NH staff member stated:
To be quite honest with you, what hospice does and what we do, there really isn’t much of a differ-ence.. . .We have dying patients all the time that we care for, and it’s pretty much keep them comfort-able, you know, and address their needs. . .
When services were viewed as similar, hospice was considered competition, and NH aides in particular expressed concern that they would ‘‘lose residents to hospice.’’
Other staff suggested that the NH could provide better end-of-life care than hospice because dying residents already had established relationships with the NH staff: ‘‘I think she was more comfortable with the [NH] staff. Like I said, we knew what she liked, and when she wanted something we would call the doctors.. . . I think she received excellent care at
[the NH].’’ In these responses, NH staff typically focused on the quality of care that the NH provided and were either unaware of or unwilling to acknowledge the breadth of end-of-life services that hospice can offer beyond typical NH services.
Among staff who viewed hospice as potentially adding value to NH end-of-life care, a belief that hospice was appropriate only for crises impeded the referral of residents and families for whom a crisis was not recognized. In all, 21 of the 34 NH nurses interviewed described hospice as a path that was not taken when residents and family members appeared to be ‘‘comfortable.’’ For example, an NH staff member stated that she did not think that an 89-year-old resident with AD/dementia would have benefited from hospice because ‘‘. . . she was comfortable. We’re usually [referring to hospice when], you know, the patient’s uncomfortable, we can’t well manage the pain, or the family also [needs] support, hospice support, you know.’’
In fact, a resident’s uncontrolled or chronic pain was the condition most commonly cited as rising to the level of a crisis that required additional services: ‘‘The only difference is with hospice if we have someone in chronic pain; they have an excellent pain protocol . . .’’ In the absence of a recognizable crisis such as uncontrolled pain, these staff members did not consider hospice to be an end-of-life care option for residents.
The belief that hospice is appropriate only for crises extended to staff views of support for family members. Although hospice offers a range of services for family members, some staff assumed that hospice was unnecessary for family members unless the staff recognized a family crisis. For example, the family of an 89-year-old resident with AD/dementia who received no hospice services was described as ‘‘very realistic that, you know, he lived a long life. . . was well taken care of. . .[and should not have] anything extraordinary.’’ In this view, hospice was considered as an option only for those families who staff members perceived to be experiencing a crisis.
Among those NH residents who did receive hospice, the timing of referral was delayed when staff members believed that hospice is appropriate only for the very end of life. Typical of this case was a discussion about a 92-year-old resident with AD/ dementia who received 2 days of hospice services, even though both the NH staff and the family apparently recognized earlier that the resident’s death was likely. According to the hospice RN, ‘‘. . . it was a matter of when she started having symptoms that were troublesome; that’s when they called us in.’’ NH staff reported that referrals occurred when ‘‘it got to the point where he couldn’t swallow anymore’’ or ‘‘he had a great deal of trouble breathing. . ..’’
In fact, a pattern of referral gaps recurred in the data. We defined referral gaps as instances when more than 7 days passed between the staff reporting
that they had identified terminal decline and the time that the hospice referral was made. There was often a time discrepancy of weeks or months between when staff reported recognizing terminal decline and when referral to hospice occurred. For example, a resident who received 1 day of hospice was reported as being ‘‘more inactive. . . . She wasn’t getting out of bed, she was becoming weaker, she wasn’t eating. . . .’’ For another resident who re-ceived 1 day of hospice care, an NH nurse reported that she had known for ‘‘a couple months’’ that the resident was in the last stages of life because he was ‘‘just not able to recover from the pneumonia.’’
NH staff often reported that they were not surprised when a resident died because he or she had displayed signs of decline; however, despite not being surprised that a resident died, some NH staff did not make a connection between recognizing signs of decline and taking the opportunity to refer to hospice at that point. According to hospice nurses, these referral gaps can be understood in the context of the belief that hospice is only for the last days of life. In describing the referral gap of an 89-year-old resident with AD/dementia who had hospice services for 1 day, a hospice nurse explained, ‘‘That’s what I get a lot working in nursing homes through hospice.. . . people believe that hospice is for the end of life care only. . .[but] it is not just out there for the last 24 hours or 72 hours of life.’’ By waiting until the last days of life before suggesting a referral to hospice, the opportunity was missed for the resident and his or her family members to receive the full benefit of hospice.
In contrast, staff members who cared for dece-dents with more timely referrals described their belief that dying residents benefit from a good working relationship between NH and hospice staff. NH staff with this view saw benefits from the partnership in terms of providing additional atten-tion to residents; managing residents’ pain; and offering needed support for residents, family mem-bers, and NH staff.
NH staff members who believed hospice care complements NH care reported valuing the addi-tional one-on-one attention that hospice can provide dying residents, particularly given the workload of NH staff. For example, a staff member said the following:
Hospice can come in up to six months prior to something being at a real imminent situation, and we thought what was good about it is the fact that they had somebody come in who was actually able to stay with her for a two-hour period and just concentrate on her and turn her and, you know, having that extra attention that would be good for her to have.
In addition, NH staff who believed that hospice services complement NH care reported that hospice
can help with routine pain management, perhaps pre-venting a crisis due to uncontrolled pain. One staff member stated that a decedent ‘‘got good care . . .
I’m a believer that no one should have to suffer on the way out, and [hospice staff] do things that we really can’t.’’ Help with pain management also comforted NH staff. An NH aide said the following:
I have found that hospice care is very beneficial for end of life . . . It gives us and nurses . . . the reassurance that we have someone we can call at any time . . . to maybe increase medications if the person’s. . .discomfort level is going up.
These staff members also reported that hospice provides additional emotional support for residents and NH staff: ‘‘Hospice helped us with the meds, helped us with some spiritual help too. . .was there all the time for her, and it helps the staff get through it, too.’’ Finally, NH staff who saw the benefits of a partnership believed that hospice improves care by providing support for family. In describing a referral for a resident with cancer who received more than a month of hospice services, an RN stated, ‘‘The deciding factor was for the social support for the family as well as for herself. . .’’
These staff members had constructed a view of services for dying residents that was not bounded by the walls of the NH. They were open to the idea that cooperation and collaboration with people who were employed by an outside organization could help residents, family members, and even NH staff to experience a dying process with less distress and more emotional support. For these respondents, involvement of both NH and hospice staff facilitated high-quality care at the end of life.
Taking Initiative
For residents who received timely referrals to hospice, NH staff played an important role in raising and discussing the option of hospice with residents and families. In these cases, NH RNs understood that they often could recognize a patient’s declining condition earlier than family members and that families may not be knowledgeable about the option of hospice. In speaking about a resident with cancer who received hospice services for 1 month prior to death, an RN stated, ‘‘I called her family member, her daughter, after speaking with the physician.. . .
The family didn’t really know too much about it.’’ Staff members often made decisions to suggest hospice care after a conversation with another staff member about the appropriateness of a referral. For example, an RN described the following exchange with a coworker about a resident with cancer and AD/ dementia who received 6 months of hospice services:
‘‘You know what we really need to do with [this resident]?’’ And she said: ‘‘Oh my God! I can’t
believe we thought along the same lines. Are you gonna say hospice?’’ And I said: ‘‘Yeah, it’s time.’’ And that’s when we did it.
For several cases in which residents received hospice for longer periods, NH staff alluded to informal norms surrounding care of dying patients that guided action in situations when terminal decline was recognized. As an RN stated when speaking about a resident with AD/dementia who received hospice services for 3 or 4 weeks: ‘‘During the process when we know that death is coming up, we have to invite hospice.’’ In this way, they viewed hospice as a routine option for end-of-life care.
Discussion
In an NH setting, nurses and aides play a crucial role in whether residents receive hospice care at the end of their lives and for how long that care is received. Results show that skills in recognizing terminal decline, beliefs about hospice services, and initiative in raising the option of hospice are decisive factors. Although residents and families may choose not to elect hospice, NH staff can play a key role in ensuring that the option of hospice is raised and that this occurs early enough in a decline process for residents and families who choose hospice to receive its full benefits.
These results call for increased training for all NH staff in recognizing terminal decline and understand-ing the breadth of hospice services. First, NH staff members need more training to recognize less dramatic signs of terminal decline. Although recog-nizing familiar signs of decline facilitates referrals, NH staff members frequently report deaths among NH residents as being unexpected or sudden. Although prognosis is difficult in many cases, the issue raised here is whether it is more likely that NH residents, many of whom have chronic illnesses, will experience a sudden death or whether less dramatic signs of a dying process are not being recognized. More widespread training incorporating resources for assessing terminal decline, such as the Flacker Mortality Score (Henderson, Hanson, & Reynolds, 2003), into routine care plans would support NH staff in recognizing a dying process and communicating with physicians about the probability of mortality.
Second, misunderstandings about the type and breadth of services hospice provides need to be addressed. The misunderstanding that hospice is only for crises decreases the likelihood that dying patients and their family members whom NH staff do not perceive to be in crisis will be presented with the option of hospice. In addition, it is important that staff members have a clear understanding that hospice services in NH settings are intended to complement the care provided by facility staff. Addressing these misunderstandings requires that
the same type of education about hospice that so far has been targeted mainly toward physicians be extended to NH staff members.
In addition, effective interventions need to be developed and funded to encourage NH staff to take the initiative in raising the option of hospice and to do so at earlier stages in a resident’s illness. The finding that staff initiative is key for raising the option of hospice suggests a need for more research to better understand how organizational patterns may impact whether an aide or nurse is likely to begin a discussion about a terminal decline and/or the option of hospice with a physician, resident, or family member.
The limitations of this study also provide useful avenues for further research on the role NH staff members play in shaping whether dying residents and their families receive hospice care and for how long. A prospective observational study would remove the potential effect of memory on retrospec-tive reports like those used in this study. In addition, although the findings from this qualitative study are not generalizable, future research can build on these results by examining the impact of staff recognition of terminal decline, staff beliefs about hospice, and staff initiative in a larger sample. A larger sample size also would allow for a systematic examination of how facility characteristics (e.g., size and number of deaths per year), provider and resident character-istics (e.g., race/ethnicity, gender, and age), and resident’s diagnosis may mediate the impact of NH staff members on hospice referral and timing of referral. Identifying variation by organizational and individual-level characteristics may add valuable specificity to the practice and policy implications of this study.
References
Baer, W. M., & Hanson, L. C. (2000). Families’ perception of the added value of hospice in the nursing home. Journal of the American Geriatrics Society, 48,879–882.
Casarett, D., Karlawish, J., Morales, K., Crowley, R., Mirsch, R., & Asch, D. A. (2005). Improving the use of hospice services in nursing homes: A randomized controlled trial.Journal of the American Medical Association, 294, 211–217.
Cohen-Mansfield, J., & Lipson, S. (2002). Pain in cognitively impaired nursing home residents: How well are physicians diagnosing it?Journal of the American Geriatrics Society, 50,1039–1044.
Emanuel, L. L., von Gunten, C. F., & Ferris, F. D. (2000). Gaps in end-of-life care.Archives of Family Medicine, 9,1176–1180.
Facts on dying: Policy relevant data on care at the end of life.(2004). Retrieved May 10, 2007, from Brown University, Center for Gerontology and Health Care Research Web site: http://www.chcr.brown.edu/dying/ factsondying.htm
General Accounting Office. (2000). Medicare: More beneficiaries use hospice but for fewer days of care (Publication No. HEHS-00-182). Washington, DC: U.S. Government Printing Office.
Gozalo, P., & Miller, S. C. (2007). Hospice enrollment and evaluation of its causal effect on hospitalization of dying nursing home patients.Health Services Research, 42,587–610.
Gozalo, P., Miller, S. C., Intrator, O., Roy, J., Barber, J., & Mor, V. (2008). Hospice effect on government expenditures among nursing home residents.Health Services Research, 43,134–153.
Han, B., Remsburg, R. E., McAuley, W. J., Keay, T. J., & Travis, S. S. (2006). National trends in adult hospice use: 1991–1992 to 1999–2000. Health Affairs, 25,792–799.
Health Care Financing Administration. (2001). HCFA responds to HAA Medicare hospice benefit questions.Caring, 20,34–35.
Henderson, M. E., Hanson, L. C., & Reynolds, K. S. (2003).Improving nursing home care of the dying: A training manual for nursing home staff.New York: Springer.
Miller, G. W., Williams, J. R., Jr., English, D. J., & Keyserling, J. (2002). Delivering quality care and cost-effectiveness at the end of life: Building on the 20-year success of the Medicare hospice benefit. Alexandria, VA: National Hospice and Palliative Care Organization. Miller, S. C., Gozalo, P., & Mor, V. (2000). Outcomes and utilization for
hospice and non-hospice nursing facility decedents. In Synthesis and analysis of Medicare’s hospice benefit. Washington, DC: Office of Disability, Aging, and Long Term Care Policy, Office of the Assistant Secretary for Planning and Evaluation, Department of Health and Human Services. Retrieved May 10, 2007, from http://aspe.hhs.gov/ daltcp/reports/oututil.htm
Miller, S. C., Gozalo, P., & Mor, V. (2001). Hospice enrollment and hospitalization of dying nursing home patients. American Journal of Medicine, 111,38–44.
Miller, S. C., Intrator, O., Cang, Y., & Mor, V. (2006). State policy and practice and end-of-life care in nursing homes [Abstract]. The Gerontologist, 46(Special Issue 1), 127.
Miller, S. C., Intrator, O., Gozalo, P., Roy, J., Barber, J., & Mor, V. (2004). Government expenditures at the end-of-life for short and long-stay nursing home residents: Differences by hospice enrollment.Journal of the American Geriatrics Society, 52,1284–1292.
Miller, S. C., & Mor, V. (2001). The emergence of Medicare hospice care in U.S. nursing homes.Palliative Medicine, 15,471–480.
Miller, S. C., & Mor, V. (2004). The opportunity for collaborative care provision: Nursing homes contracting with hospices in U.S. states. Journal of Pain and Symptom Management, 28,537–547.
Miller, S. C., Mor, V., & Teno, J. (2003). Hospice enrollment and pain assessment and management in nursing homes. Journal of Pain and Symptom Management, 26,791–799.
Miller, S. C., Mor, V., Wu, N., Gozalo, P., & Lapane, K. (2002). Does receipt of hospice care in nursing homes improve the management of pain at the end-of-life?Journal of the American Geriatrics Society, 50, 507–515.
Mitchell, S. L., Teno, J. M., Miller, S. C., & Mor, V. (2005). A national study of the location of death for older persons with dementia.Journal of the American Geriatrics Society, 53,299–305.
Moon, M., & Boccuti, C. (2002).Medicare and end-of-life care.Princeton, NJ: Last Acts, Robert Wood Johnson Foundation.
Munn, J. C., Hanson, L. C., Zimmerman, S., Sloane, P. D., & Mitchell, C. M. (2006). Is hospice associated with improved end-of-life care in nursing homes and assisted living facilities?Journal of the American Geriatrics Society, 54,490–495.
National Hospice and Palliative Care Organization. (1997).A pathway for patients and families facing terminal illness.Arlington, VA: National Hospice and Palliative Care Organization Standards and Accreditation Committee.
Pfeifer, M. P., Mitchell, C. K., & Chamberlain, L. (2003). The value of disease severity in predicting patient readiness to address end-of-life issues.Archives of Internal Medicine, 163,609–612.
Schockett, E. R., Teno, J. M., Miller, S. C., & Stuart, B. (2005). Late referral to hospice and bereaved family member perception of quality of end of life care.Journal of Pain and Symptom Management, 30,400–407. Shield, R. R., Wetle, T., Teno, J., Miller, S. C., & Welch, L. (2005).
Physicians ‘‘missing in action’’: Family perspectives on physician and staffing problems in end-of-life care in the nursing home.Journal of the American Geriatrics Society, 53,1651–1657.
Silveira, M. J., DiPiero, A., Gerrity, M. S., & Feudtner, C. (2000). Patients’ knowledge of options at the end of life: Ignorance in the face of death. Journal of the American Medical Association, 284,2483–2488. Teno, J. M., Clarridge, B. R., Casey, V., Welch, L. C., Wetle, T., Shield, R.,
et al. (2004). Family perspectives on end-of-life care at the last place of care.Journal of the American Medical Association, 291,88–93. Weeks, J. C., Cook, E. F., O’Day, S. J., Peterson, L. M., Wenger, N., Reding,
D., et al. (1998). Relationship between cancer patients’ predictions of prognosis and their treatment preferences. Journal of the American Medical Association, 279,1709–1714.
Received June 11, 2007 Accepted October 26, 2007
