MedPAC Explores Impact of Hospice on Overall Medicare Outlays
May 25, 2016 12:52 PM
At its April meeting, the Medicare Payment Advisory Commission (MedPAC) discussed the findings of a contractor report released last summer, with staff providing an overview of the contractor’s findings and asking for guidance from commission members on what actions might next be taken by the commission. The report “Spending in the Last Year of Life and the Impact of Hospice on Medicare Outlays,” examined national trends, reviewed and replicated recent studies on the topic in literature, and developed a new “market-level” approach to assessing the impact of hospice on overall Medicare expenditures.
The report found that between 2002 and 2012, use of hospice increased (from 26 to 47 percent of decedents) and spending in the last year of life also increased by 1.1 percent. The contractor found that nothing in the national trends suggested that use of hospice decreased Medicare costs for beneficiaries in the last year of life. The study also found that one-third of hospice spending (approx. $5 billion) was spent outside of the final year of life.
In its review and replication of studies in the literature on hospice’s impact on Medicare outlays, the contractor looked at “fixed period studies” and “enrollment/pseudo-enrollment” studies. Fixed period studies compare spending for decedents who do and do not enroll in hospice over a fixed period of time (often the last six to 12 months of life). As a general rule, these studies show low additional costs or small amounts of savings generated by use of hospice services. The contractor’s replication of the fixed period study showed that overall hospice was not associated with an aggregate reduction in Medicare spending but that by diagnosis hospice was associated with significant savings for cancer decedents and significantly higher cost for non-cancer decedents.
In enrollment/pseudo-enrollment studies, researchers examine spending during hospice enrollment as compared with spending for a pseudo-enrollment period created by the researcher for decedents that did not use hospice care. Under enrollment/pseudo-enrollment studies, the contractor and historical studies found a substantial reduction in spending associated with hospice for a wide range of patients.
The contractor then examined the contradictory findings resulting from the two different approaches and concluded that the discrepancies between the findings might be due to the way in which the pseudo-enrollment periods were created, since they force the researcher to make a “call” as to when the patient would have elected hospice care so that comparisons could be made about spending with or without hospice. But when spending for both populations (hospice and non-hospice) over the last year of life is examined, spending was actually greater for patients admitted to hospice. This led the contractor to conclude that savings found in the pseudo-enrollment approach to studying hospice impact on Medicare outlays are likely due to a flawed methodology rather than representative of the actual effect of hospice.
Next, the contractor used a market-level approach based on the assumption that if hospice reduces aggregate Medicare expenditures then greater hospice market penetration would be associated with lower end-of-life costs. Using this approach the contractor found that higher hospice penetration in a market is associated with slightly higher costs per decedent, and these higher costs were connected to hospice use by patients with non-cancer diagnoses and (mostly) to patients with very long stays on hospice. The contractor ultimately concluded that use of hospice does not appear to result in a reduction in aggregate Medicare spending relative to use of conventional care at the end-of-life.
Additional findings that were part of the MedPAC discussion included statistics showing that in 2013 about 35 percent of hospice payments were for care prior to the final year of life, and about 38 percent of spending on routine home care (RHC) was spent prior to the final year of life. In looking at diagnosis-related spending, approximately 16 percent of hospice payments for beneficiaries with cancer were for care outside the final year of life as compared with 40 percent of spending for non-cancer patients. Further, a much lower share of payments for general inpatient care (GIP) and continuous home care (CHC) were expended prior to the final year of life. About 29 percent of payments to nonprofit hospices were for care prior to the last year of life as compared to 40 percent of payments to for-profits. Hospice providers that began participating in Medicare before 2000 had a smaller proportion of their payments from care provided outside the final year of life as compared with newer providers.
MedPAC staff suggested that the findings discussed could be used to help focus program integrity efforts (such as conducting medical review of hospices that receive a high proportion of their payments for care provided outside the final year of life) and/or to develop a payment adjuster applicable to such providers, but sought input from Commission members on this.
One Commissioner noted that the materials indicated that 39 percent of beneficiaries who received hospice outside of the final year of life were less than 65 years of age, and it might be worthwhile to examine the characteristics of the under-65 population in hospice. Another suggested it might be worthwhile to look at differential utilization of hospice in the last year of life based on different health plans, and staff suggested they could look at finer detail related to care provided to Medicare Advantage (MA) patients that opt for hospice services.
Commissioners noted that in the early years hospice was elected primarily by cancer patients, and prognosis was more predictable, but that has changed so it may be advisable to look at diagnoses and to also monitor whether enough has been done relative to payment changes that were implemented early in 2016. Staff agreed that variation in diagnoses under hospice has contributed to the variation in lengths of stay, but the influx of different types of providers has contributed, as well. Commission members noted that analysis should not be focused exclusively on cost; instead MedPAC should focus on where constructive changes can be made -- possibly concentrating on the for-profit vs. nonprofit angle, but also on the issue of who decides that the patient is eligible for care, and whether the certifying physician has actually seen the patient.
A Commissioner raised the issue of informed discussions about end of life care, and how to encourage that these discussions take place sooner; another expressed concern that the studies examined for the report used some older data that may not be applicable at this time, and that it may also be advisable to look at what happens with patients that “get the benefit as it was designed.” Another noted that perhaps with conditions other than cancer it might be appropriate to look at lower payment rates or some type of value-based adjustor applied after the fact. Another noted that many patients are dying very shortly after electing hospice care and that real savings could come from patients electing hospice earlier rather than coming in very close to death, and it makes sense to look at the value that payment for advance care panning services might bring to the equation. Further, it may be worthwhile to look at hospices with a high proportion of long-stay patients and examine their quality programs to see if they are reviewing patients for continuing eligibility or are these patients getting hospice services that are more appropriately classified as a chronic disease management program.
Another Commissioner mentioned the possibility of a for-profit payment differential to address the differences in the behaviors relative to seeking out long-stay patients, while another discussed state regulatory underpinnings related to Physician Orders for Life-Sustaining Treatment (POLST) that could have an impact on how much of a role the patient’s declaration of desired interventions play in the services that are ultimately provided. Yet another Commissioner indicated that he was hesitant to recommend the standard policy recommendations that MedPAC might normally pursue because they could have a seriously negative impact, but would like to see if increased accountability by referring clinicians can be pursued in some way, and one element of getting there may be better alignment of hospice with the rest of the health care system, including looking at an MA carve in. Another Commissioner commented on cultural and other issues that must be considered in relationship to utilization of hospice services. In a final comment, one of the Commissioners mentioned that he examined hospice utilization by individuals in their health system and found that patients they bring onto service represent a significant savings to the Medicare program over patients that remain on conventional curative care.
It is anticipated that MedPAC will continue discussions on these issues as part of future meetings.