ADVERTISEMENT
End-of-Life Nursing Home Care: Resource Allocation and Ethics
Many individuals will receive their end-of-life care in long-term care settings. Nursing home residents with life-limiting illness may or may not be formally enrolled in a hospice program, but in either case are entitled to appropriate care at the end of life.1 The broader proposition is that it is or should be a societal norm to value providing compassionate end-of life care to persons across the continuum of care settings. An important question to consider is how well or poorly the current economic and regulatory frameworks support the quality of end-of-life care that nursing home residents are entitled to receive.
It is often problematic to apply traditional concepts of cost-benefit analysis to the value of geriatric care in general, and to end-of-life care in particular. The economic analyses that traditionally are done under the cost-benefit rubric tend to weigh strongly against care of the older individual.2 The argument often offered is that in a world of limited resources, hard choices have to be made in healthcare resource allocation, and that those choices should be guided by the cost per quality-adjusted life-year (QALY) saved. In this metric, one might undertake those interventions that cost no more than, for example, $100,000 per life-year saved in good health, and proportionately reduce the value assigned if the lengthening of life is of reduced quality because of comorbidity.
The limits of this kind of analysis are starkest when one considers end-of-life issues. High-quality end-of-life care is not likely to extend lifespan much, or at all, but rather is aimed at improving the quality of a brief span of life. At the extreme, one might make a very effective multimodality palliative intervention (with effective pain control, emotional support, and other compassionate measures), but do so very late in the course, and thereby make the patient’s last few days of life much improved in quality. Most patients and families would value such a quality-of-care intervention greatly, yet the brevity of lifespan affected means that on the QALY metric it might only be worth a few hundred dollars, since it only yields a fraction of a percent of a quality-adjusted life-year.
Alternatively, consider a rather different metric for placing a value on end-of-life services. It’s not one that is as commonly used, but it illustrates how much the conclusions reached are influenced by the criteria used. Focus not on the cost to society of publicly provided benefits, but rather on the rational choices of individuals for themselves, as they may allocate their own resources. This would reflect the kind of care people would choose for themselves if they had the choice to make, and acted as strictly rational and self-interested economic beings.
Persons with rational economic thinking—in this case, the rational decisionmakers at end of life—would spend almost as much as they could on anything that could improve the quality of their end of life by even a little. The more assets they have, and the shorter their anticipated survival, the more they’d be willing to pay per day on end-of-life care.
Even for those individuals who are without means, one would presumably want to provide them with end-of-life healthcare that is some reasonable approximation of the choices that those who are not impoverished would make for themselves at end of life. On this model, therefore, you would accord a very high social value to the provision of high-quality end-of life care.
The fundamental point is that while models are interesting, they should be approached with some caution. This is particularly true when one can set up two plausible models, QALY and the economic decisionmaker, which are polar opposites in how they value end-of-life care. It seems reasonable when two models disagree so fundamentally to treat both with a healthy skepticism.
Accordingly, one might want to look at an analytic model that is grounded in less abstract concepts and that is more centered on patients, or at least on patients’ families. Of interest in that regard is the report on a survey of the families of North Carolina patients who were simultaneously getting hospice and nursing home care, and who died between December 1997 and May 1998.3 The median value the family members placed on the hospice care, above that for the nursing home care, was $75 a day. It seems likely that a decade later now, the figure might well be higher.
When it comes to a subjective measure of value, of course, a lot depends on the eye of the beholder. Another study from North Carolina with one of the same investigators asked the administrators of nursing homes and of hospices their opinion on the value of hospice in nursing homes.4 The nursing home administrators’ questions were only from those homes that actually had gone so far as to have a current contract set up for hospice services (as 83% did), and so the sample likely omitted administrators from homes that were most inclined to negative attitudes about hospice in the nursing home. It nevertheless is unsurprising that about a quarter of nursing home administrators thought nursing homes could provide good care without any hospice for end-of-life care, while only 1% of the hospice administrators felt that way. Of even more interest, perhaps, is the fact that, although the nursing homes in the sample set all had hospice contracts, the attitudes of nursing home administrators about the value of hospice was correlated with how often hospice was actually used in their nursing home.
It is a fair question to ask (whether one is a nursing home administrator, a payor, or a patient) what is the empiric evidence that a hospice program objectively improves end-of-life care in nursing homes. The evidence here is limited, but leaves some room for concern. One study looked at pain control for residents in two proprietary skilled nursing facilities (SNFs) of a single nursing home chain, and found substantial undertreatment of pain in dying patients; this was true whether or not hospice was involved in the particular nursing home resident’s care.5 Another study looking at a sample across multiple nursing home and assisted living facilities suggested that most residents did receive adequate pain treatment, and the differences between hospice and nonhospice were small and largely explicable by differences in case mix in hospice versus outside hospice.6 Either way, it appears that the overall bundle of care provided may be more important, in many cases, than whether or not it is formally denoted as hospice.
A recent randomized study gives even a better idea of how patients and families value the additional contribution of hospice in a nursing home setting. Casarett and colleagues7 randomized residents at three nursing homes who met clinical criteria for hospice, either to usual care or to a very modest intervention, which was an informational visit to tell them about hospice. The informational visit was associated with roughly a quadrupling of the rate of those enrolling in hospice in the succeeding 30 days; there was only a 6% rate in the usual care group, but a 25% enrollment rate in the informational visit group. Although only a minority of the patients apparently were followed until date of death, it appears that the main effect of an informational visit was at least as much to hasten the time of referral to hospice as to increase the frequency of referral; patients in the usual care group tended to have hospice much closer to death than did those who received the informational visit.
If one chooses to try to look from a cost-benefit analysis point of view at hospice, then obviously one also has to consider the cost of hospice as compared to nonhospice care. The data here suggest that the net-cost impact of hospice is probably not huge in the overall scheme of things. It seems plausible that a good integration of palliative care into the continuum of care will certainly improve quality of care. It is likewise plausible that it could save money, such as by enabling providers to more quickly and efficiently match patients at end of life with the bundle of services that they may wish to receive at that point in their course.
The evidence for net savings is mixed, and appears to vary depending on both the methods by which palliative care is delivered and by the patient characteristics. A single institution report from Virginia Commonwealth University Medical Center found that major cost savings could be achieved by setting up a Palliative Care Unit in the facility, in part because it was able to accept appropriate patients as direct admissions from the emergency department.8
On the other hand, another report looked at Medicare data from the late 1990s, comparing Medicare cost in the last year of life for patients who enrolled in hospice versus those who did not.9 Across all hospice diagnoses as a whole, hospice was associated with a 4% higher cost, but among the patients with cancer there was actually a net savings for hospice as compared to nonhospice. In this nonrandomized study, although the investigators made an attempt to correct for differences in factors such as comorbid conditions, it is not clear how much of the difference was related particularly to use of hospice. In their data, as one would expect, very late referrals to hospice were common, and especially in those admitted for a diagnosis other than cancer. A large fraction were admitted to hospice within a week of death: 23% for those with cancer, 36% for those with other diagnoses. A very late hospice admission suggests that both the economic and the quality-of care impact of hospice on that patient’s course are likely to be limited.
On one level, the issue of quality of life is one that raises the concern of trying to make it easier for people to spend their last days in whatever venue they prefer, which will often be their home. On the other hand, many of the palliative care issues that need to be addressed are similar across the various venues.10 The less extensive the array of baseline services provided (such as the lower level of services in assisted living as compared to SNFs), the more the addition of hospice is likely to add to the mix of services that will enable a patient to experience end of life in place.
As a result, the simplest way to think of the societal task is to make high-quality end-of-life care available across all venues of care, so that patients’ choice of venue can be based on where they prefer to be, not restricted by any need to be in particular settings to have access to palliative care services. The focus could be less on how much of the bundle of services come from each component (hospice, nursing home, or other), and more on the issue of whether the bundle of services as a whole is appropriate to meet patient care needs.
The author serves as Deputy Director on Grant Number 5 P20 MD 524-2, from the National Center of Minority Health and Health Disparities, National Institutes of Health. The contents of this article, however, are solely the responsibility of the author, and do not necessarily represent the official views of the National Institutes of Health nor Annals of Long-Term Care: Clinical Care and Aging. Ethical challenges dealing with residents with cognitive impairment is a common occurrence in our Long-Term Care practices.
We would welcome your comments, experiences, and reflections on this article. Please send comments to Fred Feinsod, MD, MPH, CMD, at feinsod@fmfmd.com.
