Differences Between For-profit and Non-profit Hospice Agencies in the US Medicare Population

BACKGROUND

Hospice is a program that provides end-of-life care for patients with an expected survival of 6 months typically at their homes. In addition, hospice also provides caregiver support of loved ones, and bereavement services after the patient’s demise. The Medicare Hospice Benefit was created by Congress in 1982 but the recent proliferation of for-profit hospices, now accounting for a majority of hospice agencies, has raised concerns about the quality of care received by patients enrolled in hospice [1].

OBJECTIVE

Our objective was to assess differences between for-profit and non-profit hospices regarding patient characteristics, costs, and patterns of care provided in the entire US Medicare population. We hypothesized that care would be costlier in for-profit agencies.

METHODS AND FINDINGS

We combined the CMS Fee-For-Service Post-Acute Care Provider Public Use Files for calendar year 2017 [2] and the CMS dataset for general information on hospice providers with details on characteristics and ownership of hospice agencies [3], categorized as for-profit, non-profit/government, and combined government/non-profit ownership status. Hospice agencies with other ownership status and those that did not have a match in the second dataset were excluded. To assess differences, analysis of variance and logistic regression were used.

In 2017, most hospice agencies were for-profit (n=2533, 59.5%) vs non-profit (n=1048, 24.6%) or other (n=673, 15.8%); similarly, most Medicare beneficiaries were enrolled in for-profit agencies (n=707,864, 45.9%) vs non-profit (n=623,276, 40.4%) or other (n=213,163, 13.8%) (Table 1). Compared to non-profits, for-profit agencies received 51.6% more total Medicare payment ($9.4 billion USD vs $6.2 billion USD) as well as 33.6% more payment per beneficiary ($13,246.15 vs $9915.15) (p<0.0001 for all).

Table 1.

Differences In Medicare Beneficiaries in Hospice by Agency Ownership Type in 2017

	Total	For-profit (n=2533 agencies, 707,864 patients)	Non-profit (n=1048 agencies, 623,276 patients)	Odds ratio (95% CI)
Demographics
Average age (years)	82.4	82.6	82.2	−0.34 (−0.54, −0.26)
Female beneficiaries	58.2%	59.2%	57.3%	0.92 (0.92, 0.93)
White beneficiaries	82.3%	78.3%	86.7%	1.80 (1.78, 1.82)
Black beneficiaries	7.8%	8.9%	6.3%	0.69 (0.68, 0.70)
Asian beneficiaries	1.3%	1.3%	1.5%	1.16 (1.13, 1.20)
Hispanic beneficiaries	5.9%	8.2%	3.6%	0.43 (0.42, 0.43)
Medicare advantage	34.7%	35.8%	33.9%	0.92 (0.91, 0.93)
Medicaid	25.2%	28.2%	21.9%	0.71 (0.71, 0.72)
Hospice use patterns
Routine home care days	98.1%	98.8%	97.1%	0.42 (0.42, 0.42)
Total live discharges	11.3%	13.7%	8.4%	0.58 (0.57, 0.58)
Beneficiaries w/ ≤7 hospice days	25.5%	20.4%	30.6%	1.71 (1.70, 1.73)
Beneficiaries w/ >60 hospice days	32.5%	38.1%	26.7%	0.59 (0.59, 0.60)
Beneficiaries w/ >180 hospice days	12.5%	15.9%	9.1%	0.5 (0.52, 0.53)
Primary diagnosis
Cancer	26.1%	21.7%	30.7%	1.60 (1.58, 1.61)
Dementia	17.9%	20.7%	14.8%	0.67 (0.66, 0.67)
Circulatory/heart disease	17.3%	18.0%	16.5%	0.90 (0.90, 0.91)
Respiratory disease*	8.7%	8.7%	8.7%	1.00 (0.99, 1.01)
Stroke	9.2%	10.3%	8.4%	0.80 (0.79, 0.81)

Units are percentages unless specified otherwise in parentheses. All p values for differences < 0.001 except the data indicated with asterisk (*) which was non-significant

Beneficiaries enrolled in for-profit agencies spent more mean days in hospice than those in non-profit agencies (81.7 vs 55.7) and were more likely to have extended hospice stays spanning >180 days (15.9% vs 9.1%) and less likely to spend <7 days. Beneficiaries in for-profit agencies received more routine home care days of service and were also more likely to experience live discharge from hospice.

Beneficiaries in for-profit agencies were more likely to be slightly older, female, racial/ethnic minority group, and enrolled in Medicare Advantage and Medicaid. For-profit agencies enrolled lower proportions of patients with cancer and greater proportions of patients with dementia, circulatory/heart disease, and stroke.

DISCUSSION

Our data show most Medicare beneficiaries are enrolled in for-profit vs non-profit hospice agencies, which cost Medicare 34% more per beneficiary. Higher total costs of care directly relate to longer lengths of stay in for-profit agencies, even when accounting for lower reimbursement rates for days 61 and greater. This difference could be explained by for-profit agencies enrolling more patients with dementia, who live longer in hospice than cancer patients, who have greater acuity yet shorter survival in hospice [1]. Cost of care was higher in for-profit agencies even though they provided higher proportion of routine home care days. While previous studies have evaluated association of hospice ownership status with setting and timing of hospice care, and there have been investigative and federal criminal inquiries into enrollment of low-acuity patients by for-profit hospice agencies, this is the largest and most representative study to date [4, 5]. Given that for-profit hospices enroll a greater relative proportion of patients who are racial and ethnic minorities, likely because of their growth in urban areas [6], understanding the differences in care they provide is particularly important from a racial equity perspective as well. One limitation is that we lacked data regarding hospice location. Further work should attempt to incorporate patient and family-based measures of quality or satisfaction, which these data sources lacked.

In conclusion, our study highlights the need to carefully monitor for-profit hospices to ensure that the patient-centered mission of hospice is retained, and patients receive high-quality care to the very end. Hospice is an essentially moral program and care needs to be taken to retain its original vision. Hospice payment reform tying reimbursement to clinical quality outcomes should be considered. Further research studying quality and transitions of care and enrollment practices of patients from for-profit nursing homes is needed.

Declarations

Disclosures

Diane Meier receives grant funding from the Arthur Vining Davis Foundations, the Berkman Charitable Trust, Trudy Elbaum Gottesman and Robert W. Gottesman, the Y.C. Ho/Helen and Michael Chiang Foundation, the Gordon and Betty Moore Foundation, the John A. Hartford Foundation, the Gary and Mary West Foundation, the West Health Institute, the Patty and Jay Baker Foundation, Thelma Lyon, the Brookdale Foundation, the Cameron and Hayden Lord Foundation, the Milbank Foundation, the Stavros Niarchos Foundation, the American Academy Of Hospice and Palliative Medicine, the Weiner Family Fund, Joy Schein, and the United States Cancer Pain Relief Committee. All other authors have no relevant disclosures.