Saving money isn’t really the point of hospice care. Helping dying patients have the best possible remaining life, followed by a good death, is really the purpose.
But whether hospice care saves money has prompted debate for years. Most hospice patients die at home, which is what the great majority of Americans say they prefer, or in nursing homes. Wouldn’t that save money compared with the cycle of 911 calls and hospitalizations that characterize so many American deaths?
Studies of this question have been small and have reported contradictory results. “There hasn’t been a lot of consensus,” said Dr. Ziad Obermeyer, an emergency physician and health policy specialist at Harvard Medical School. “There are people in the policy establishment who are still skeptical about whether the costs are lower.”
He hopes the large study he and his colleagues at Brigham and Women’s Hospital in Boston published recently in JAMA will finally put that concern to rest. It matched two groups, each containing more than 18,000 older Americans with metastatic cancer, comparing patients who enrolled in hospice with those who had the same poor prognosis but didn’t use hospice. The researchers tracked participants in both groups until their deaths.
The hospice patients had far fewer hospitalizations and less than half as many intensive care unit stays, their Medicare records showed. They endured half as many invasive procedures. They were five times less likely to die in a hospital or nursing home.
And yes, care in the last year of their lives cost an average of about $8,700 less than for the nonhospice patients. Costs for patients in the nonhospice group were $71,517 on average in the last year of life, compared with $62,819 for those in hospice. If extended nationwide, the difference could save Medicare billions of dollars.
“I hope this will be the last time someone asks the question,” Dr. Obermeyer said. “For this population, enrollment in hospice is associated with a large difference in costs.”
His own experiences in a busy urban emergency room led to this research. “We see so many people with end-stage diseases, and they’re still getting very aggressive treatment,” he told me in an interview. “I’m often the first person to discuss end-of-life care with them, even though they’ve seen a whole list of doctors.”
Such conversations often prove difficult, but they’re particularly tough in an emergency room, he pointed out, when “you might have a five-minute window in which to make an important decision” — like whether to put a patient on a ventilator.
In this study, all the patients had the kind of “poor-prognosis” cancers — primarily of the breast, colon and lungs — that made them eligible for hospice, a decision that generally does involve that kind of conversation. Hospice and nonhospice patients with the same prognoses were matched by age (the average was 80), sex, length of survival and region, because studies have repeatedly shown that geography is destiny when it comes to the amounts and kinds of medical care people receive.
From diagnosis to death, both groups lived a median seven months. The hospice patients had heart-sinkingly short stays: a median of just 11 days, lower than the national median, and a pity, since hospice could have supported them and their families for much longer.
Still, the differences were stark. During the period they received hospice care, 42 percent of hospice patients were hospitalized compared with 65 percent of nonhospice patients during the same period. More than a third of the nonhospice patients spent time in an I.C.U., versus fewer than 15 percent of the hospice enrollees.
About a quarter of the hospice patients underwent “invasive procedures,” though often those were related to comfort, like a having catheter inserted to administer pain-relieving drugs. More than half of the nonhospice group had invasive procedures, which were more likely to include intubation, dialysis and C.P.R.
A key finding: 74 percent of the nonhospice group died in a hospital or nursing home, versus 14 percent of hospice patients.
Costs varied depending on length of hospice stay: Those enrolled for five to eight weeks represented the greatest savings ($17,903), compared with nonhospice patients. Shorter stays meant lower, but still statistically significant, savings. Only among the 2 percent who spent a year or more in hospice did costs exceed those incurred by nonhospice patients.
Because you can never randomly assign dying people to hospice care — it will always be their choice — this study measured associations, not causes. People who opt for hospice are different from those who decline it, Dr. Obermeyer acknowledged: “Hospice is the way people express their preferences about what kind of care they get.”
And the study looked only at people with cancer. That’s still a leading hospice diagnosis, but a greater proportion of hospice patients have other conditions, and that could influence the costs.
Still, greater hospice use – and for longer periods — looks like a money-saver.
As the health policy specialist Joan M. Teno pointed out in an accompanying editorial, that really shouldn’t matter. “We’ve set up this notion that it must be cost-neutral or save money,” she told me in an interview. “Why can’t it cost a little more if it provides high quality care? Nobody questions whether I.C.U. care saves money.”
But perhaps seeing hospice as a way to save Medicare dollars will lead to policies that encourage its use. And this might, too: Dr. Obermeyer’s next study will look at whether hospice reduces out-of-pocket expenses, “not just for the system, but for patients and families,” he said. “We talk a lot less about costs for individuals.”
True. And wouldn’t that be interesting to know?