Dying patients are more likely to receive less intense medical care than in the recent past. The cost of that healthcare, however, provided by hospitals, physicians, and hospice, continues to rise, says a Dartmouth Atlas report.
Terminally ill patients received less aggressive interventions and were more likely enrolled in hospice in their last six months if they died in 2010 than if they died in 2007, according to the latest Dartmouth Atlas report on end-of-life care.
Patients dying in 2010 also were less likely to spend a portion of their last days in the hospital, and were less likely to die in one than in 2007. For example, in 2007, 28.1% of patients died in an acute care setting; in 2010, 25% were in a hospital when they expired. And fewer episodes of terminal care involved time spent in an intensive care unit, from 17.6% in 2007 to 16.7% in 2010.
"The surprise was the rapidity of change that we saw nationally and in individual healthcare systems," says David Goodman, MD, co-principal investigator of the Dartmouth Atlas Project.
For example, the University of Pittsburgh Medical Center "in a three-year –period had a 20% reduction in inpatient days in the last six months life, and 16% reduction in intensive care unit days," he says. And Fletcher Allen Health Care in Burlington, VT, "which already had a very low intensity of care in 2007, one of the lowest in the country, still reduced [its] inpatient days by 22%."
At Hershey Medical Center in Hershey, PA, "there was a "26% reduction in ICU days over a three year period and an 18% reduction in physician visits. This does not happen by accident," Goodman says.
More disappointing, however, was the report's failure to find a decline in cost of care at the end of life. The institute published its first end-of-life care report on 2007 care practices in 2010. Since then, costs have risen.
The average spending for a terminally ill Medicare patient during the last two years of life increased 15.2%, from $60,694 in 2007 to $69,947 in 2010. Those costs were not adjusted for inflation, but the increase was much higher than the consumer price index of 5.2%, Goodman says.
"Whether we're beating the cost of living index or lagging behind, the norm is what's killing us," he explains. "These are unsustainable increases in Medicare program payments." More important, he says, is that there's strong evidence that lengthy end-of-life stays in the hospital and "receiving uncomfortable procedures" are not what patients want.
"The important takeaway," from the latest report, Goodman says, is that hospitals that are spending high amounts on end-of-life care "need to look very closely at the care they deliver. There's going to be expectations that all places, particularly those that are spending a lot more money without being able to demonstrate comparable better outcomes, will have to explain that. And when there are opportunities to improve care, they will need to improve care."
The latest report finds that while intensity of care declined overall, dying patients were more likely to have more than 10 outpatient physician visits during the last six months of life in any setting, including in the hospital, a sign that care at the end of life is increasingly fragmented among multiple clinicians who may not be communicating well with each other, Goodman says.
However, patients' chance of spending time in the intensive care did not change. Patients dying in 2010 spent on average nearly three days longer in hospice care, 21 days, compared with 18.3 in 2007.
Also disappointing is the report's finding that regional trends persist, because at hospitals in about 18 states, practices to encourage more appropriate care for terminally ill patients have stagnated or become even more intense rather than less.
That's shown by a measurement the Dartmouth Atlas project refers to as the "hospital care intensity index" for each hospital referral region, state, and hospital. The Atlas is based on 1.1 million Medicare patients who died in 2010, and 1.16 who died in 2007.
The index is a composite of various healthcare services at the end of life such as time spent in a hospital or an ICU, number of physician visits, enrollment in hospice, use of imaging, home health services, ambulance services, durable medical equipment, and whether the patient died during a hospital stay.
Among the nation's 306 hospital referral regions, 2,400 hospitals, and among all states, practice patterns show wide variation in this index. For example, hospitals in New Jersey, New York, Florida, California, and Nevada had the highest end-of-life intensity scores: Hospitals in Utah, Idaho, Oregon, Montana, and Vermont had the lowest.
In 2007, New Jersey, New York, Nevada, Delaware and California were the highest utilizers of intensive end-of-life care, while Idaho, Utah, Oregon, Montana and Vermont were the lowest.
Variations extend to specific cities and towns. Hospitals in Los Angeles in 2010 had the highest intensity index, followed by Miami, Manhattan, and McAllen, TX. And Medicare's costs for those patients were correspondingly higher than other metropolitan regions or towns, ranging from $112,263 to $109,557 for these four cities.
On the other hand, Minot, ND; Bismarck, ND; La Crosse, WI; and Dubuque, IA had the lowest cost of care and among the lowest hospital intensity scores. Medicare spent on average between $47,620 and $46,563 per patient in those areas.
A separate section of the Dartmouth update analyzes changes in end-of-life service delivery among the nation's academic medical centers. The report said that while "striking variation" has been noted among these teaching hospitals, "but that care is changing quickly in many centers.
"For example, while patients cared for by New York University's Langone Medical Center who died in 2007 spent a relatively high number of days in the hospital in the last six months of life, NYU led academic medical centers with a 29% decrease (26.9 to 19.1 days) by 2010," the report says.
"Some hospitals with a relatively low number of hospital days reduced them even further," for example, at Fletcher Allen Health Care, hospital days for dying patients "fell about 22%, from 11 to 8.6 days."
Goodman puts some blame for high intensity of end-of-life care on clinicians who "are not very good at assessing patient preferences. We don't receive the training to do it. It's not enough for physicians to be well-meaning, and not sufficient for physicians to assume that everyone wants to live one more day."
Doctors need to "provide accurate information about uncertainty, choices and possible outcomes" of aggressive care… and what treatment entails," and that making good on a pledge to extend life a week or a month longer, saying "there's hope for you," may mean the patient will spend most of that time in the hospital.
"Too often those discussions never occur, and when they do, physicians paint an overly optimistic scenario of patients' chances. It's rare that a patient would be seeking those additional days if they knew they would all be spent on a ventilator in an ICU."
See Also:
Hospitals' Fear of 30-Day Penalties May Speed Hospice Admissions
Aggressive Care for Dying Cancer Patients Futile, Study Finds