Nonprofit hospices disproportionately care for costly patients, researchers conclude

The findings appear in the Feb.2 issue of the Journal of the American Medical Association (JAMA).

"There was a big increase in the number of for-profit hospice agencies from 2000 to 2007, and previous work has shown that those agencies tended to have significantly higher profit margins than their nonprofit counterparts," said lead author, Melissa W. Wachterman, MD, MPH, a palliative care physician and research fellow in BIDMC's Division of General Medicine and Primary Care. "The Medicare per diem payment rate is the same, regardless of patient diagnosis, location in which care is received (for example, private home versus nursing home), or length of stay, and we wanted to know whether for-profit and nonprofit hospices were responding differently to potential financial incentives inherent in the Medicare Hospice Benefit."

Researchers examined a nationally representative sample of patients discharged from hospice, primarily due to death (85 percent) in 2007. In all, data from 4,705 patients, representing an estimated 1.03 million patients discharged from hospice nationwide, were analyzed, looking at diagnosis, location of care, length of stay, and number of visits per day by different hospice care providers.

The data showed that nonprofit hospice agencies had a higher proportion of the types of patients who required more visits from skilled care providers than for-profit agencies. For example, nonprofit agencies had a higher proportion of cancer patients, while for-profit agencies had a higher proportion of dementia patients. Cancer patients required more visits per day from skilled personnel such as nurses and social workers than patients with dementia. Wachterman explains that patient selection of this nature has important policy implications because caring for dementia patients rather than cancer patients "could be financially advantageous for hospices under the current capitated reimbursement system."

However, it is important to note that "clinicians caring for patients considering hospice can be reassured that for-profit hospices provide as many nursing visits to patients with a given diagnosis as nonprofit hospices," adds senior author Ellen McCarthy, PhD, MPH, an epidemiologist at BIDMC and a professor of medicine at Harvard Medical School.

The study also looked at length of stay and found that the median LOS for patients in for-profit hospice agencies was four days longer when compared with nonprofit hospice agencies. Because there are considerable fixed costs at the time of enrollment in hospice and again at the time of death, longer stays are "thought to be more profitable," the study noted.

The significant differences between for-profit and nonprofit hospice agencies mean that hospices serving the neediest patients "may face difficult financial obstacles to providing appropriate care in this fixed per-diem payment system," explains Wachterman.

These findings may "have potentially important implications both for clinicians taking care of patients at the end of life and for policymakers in the area of Medicare hospice payment." The study may help inform current debate around payment reform in the Medicare Hospice Benefit.

The Medicare Payment Advisory Committee (MedPAC) has recommended a U-shaped reimbursement plan that considers the intensity of care required at the beginning and end of a hospice stay. The plan also recommends that a higher per diem rate be paid for the first 30 days of enrollment and a standard payout be made at the time of death.

Study coauthors include BIDMC investigators Edward R. Marcantonio, MD, SM and Roger B. Davis, ScD. This study was supported through funding from the Health Resources and Services Administration of the Department of Health and Human Services to support the Harvard Medical School Fellowship Program in General Medicine and Primary Care.

The study was deemed exempt by the Beth Israel Deaconess Medical Center institutional review board because the researchers used publicly available deidentified data.