Hospital Readmissions
May 19, 2017

Hospice Patients Have Lower Hospitalization Rates, Reduced Medical Costs and Are Five Times More Likely to Die at Home

Hospice worker helps patient up off bed

Medicare patients with poor-prognosis cancer who enroll in hospice are significantly less likely than nonhospice patients to be hospitalized, admitted to an intensive care unit (ICU) or to receive invasive procedures near the end of life, according to a report published in JAMA Internal Medicine. Further, hospice patients have lower healthcare expenditures in the last year of life and are five times more likely than nonhospice patients with similar characteristics to die at home.

“Our study shows very clearly that hospice matters,” says lead author Ziad Obermeyer, MD, MPhil, of Brigham and Women’s Hospital and Harvard Medical School, Boston. “Hospice and nonhospice patients had very similar patterns of healthcare utilization right up until the week of hospice enrollment—then, the care started to look very different.”

“Patients who didn’t enroll in hospice ended up with far more aggressive care in their last year of life, most of it related to acute complications like infections and organ failure, and not directly related to their cancer diagnosis.” Such care is unlikely to match the end-of-life wishes of most patients, point out the authors.

Investigators analyzed data from a nationally representative 20% sample of 86,851 Medicare beneficiaries newly diagnosed with poor-prognosis cancer (e.g., lung, pancreatic, brain or any metastatic malignancy) who died in 2011. Patients who enrolled in and died while under hospice care were matched to a control group of those who died without hospice services, creating 18,165 matched pairs (mean age, 80 years) for the matched cohort study. 

Overall:

  • Median length of time from first poor-prognosis diagnosis to death was 13 months (interquartile range, three to 34 months).
  • 60% of patients were enrolled in hospice.
  • Median length of hospice stay was 11 days, with less than 6% of stays exceeding six months.

Key Findings

Compared with those in hospice, nonhospice patients had higher rates of:

  • Hospitalizations (65.1% vs 42.3%; risk ratio [RR], 1.5; 95% confidence interval [CI], 1.5 to 1.6)
  • ICU admissions (35.8% vs 14.8%; RR, 2.4; 95% CI, 2.3 to 2.5)
  • Invasive procedures such as intubation, dialysis or resuscitation (53.0% vs 26.7%; RR, 1.9; 95% CI, 1.9 to 2.0)
  • Death in a hospital or nursing facility (74.1% vs 14.0%; RR, 5.3; 95% CI, 5.1 to 5.5)
  • Healthcare expenditures during the last year of life ($71,517 vs $62,819; difference, $8,697; 95% CI, $7,560 to $9,835)

Nearly three-quarters of nonhospice patients died in hospitals or nursing homes, compared with just 14% of hospice patients. “These findings highlight the importance of honest discussion between doctors and patients about our patients’ goals of their care at the end of life relating to treatment decisions and quality of life,” says Obermeyer. This is particularly pertinent in light of the recent addition of reimbursement codes for advance care planning discussions by the Centers for Medicare and Medicaid, he notes.

Lower Costs

Enrollment in hospice was associated with a decrease in overall healthcare costs of $8,697 per patient during the last year of life. Patients enrolled for five to eight weeks represented the greatest cost difference ($17,903) compared with nonhospice patients, although shorter stays also resulted in lower, but still statistically significant, cost differences. Only among the 2% of patients who stayed in hospice longer than one year did expenditures exceed those of the nonhospice group.

“Cost trajectories began to diverge in the week after hospice enrollment, implying that baseline differences between hospice and nonhospice beneficiaries were not responsible for cost differences,” observe the authors. “Hospice enrollment of five to eight weeks produced the greatest savings; shorter stays produced fewer savings, likely because of both hospice initiation costs and need for intensive symptom palliation in the days before death.”

Hospice Stays Too Short

Median length of stay among hospice patients in the study was 11 days. As the number of patients with cancer who use hospice continues to rise, the length of hospice stays grows shorter, while care intensity outside of hospice is increasing, note the authors.

“Patients with cancer, the single largest group of hospice users, have both the highest rates of hospice enrollment and the highest rates of hospice stays of less than three days,” they point out. Since the initiation of the Medicare hospice benefit in 1982, the number of people of all diagnoses receiving its services has increased, yet the median length of stay has decreased over the same time period.

“We know that many people access hospice care too late to fully take advantage of all this team-based, patient and family-focused model of care can offer,” says J. Donald Schumacher, PsyD, former president and CEO of the National Hospice and Palliative Care Organization.

“While patients with cancer still make up more than a third of all those cared for by hospice providers, their lengths of stay in hospice are among the shortest. This points to the desperate need for clinicians treating cancer to have conversations about palliative care and hospice.”

Source: “Association between the Medicare Hospice Benefit and Health Care Utilization and Costs for Patients with Poor-Prognosis Cancer,” JAMA Internal Medicine; November 12, 2014; 312(18):1888-1896. Obermeyer Z, Makar M, Abujaber S, Dominici F, Block S, Cutter DM; Departments of Emergency Medicine and Health Care Policy, Harvard Medical School; Departments of Emergency Medicine and Psychiatry, Brigham and Women’s Hospital; Ariadne Labs; Department of Biostatistics, Harvard School of Public Health; Department of Psychosocial Oncology and Palliative Care, Dana-Farber Cancer Institute; Department of Economics, Harvard University, all in Boston.

CLINICIANS: SIGN UP FOR OUR EMAILS

Join our email list for webinars, hospice care news & more.