Despite efforts to reduce aggressive and expensive care for the terminally ill, a greater percentage of dying Medicare patients underwent hospitalization in their last 90 days—not fewer—and more patients spent time in intensive care units in their last month in 2009 than in 2000, says a surprising report.
The study, published in the Journal of the American Medical Association, found that the percentage of patients hospitalized in their last 90 days of life went from 62.9% in 2000 to 69.3% in 2009. And the percentage of patients who spent time in the intensive care unit in their last 30 days went from 24.3% to 29.2%.
Joan Teno, MD, lead author of the report and a health policy and practice professor at Brown University in RI, says such patterns of more aggressive care "is really only ethically defensible if that's what the patient wanted," and she doubts that more patients are wanting that level of intensity.
Rather, she says, her own research and that of others, shows that patients and their families are not getting the opportunity to make such choices.
Her report did find that a larger percentage of patients are dying at home, and far fewer are dying in acute care hospitals. But more patients are being moved around from one care setting to another in the last 90 days of life as well as in the last three.
Additionally, a slightly higher percentage of patients had at least three hospitalizations in the last 90 days between 2000 and 2009.
Teno says the findings surprised her and her team.
"I've spent my entire career as a physician essentially practicing palliative care to improve quality for the seriously ill and dying, but when I saw these results, I said, 'Wow.' I was quite surprised," she says.
Asked if it was possible that families and patients just want providers to do everything possible until the end, Teno says, "There are some families where that may be the explanation. But I think the more predominant thing that's going on here is a kind of 'Don't ask don't tell.' We're not openly communicating with the dying patient and their family about what the treatment options and goals of care are. And I think we can do a better job of that."
Teno adds that the study results are "not surprising when you look at the financial incentives within our fee-for-service healthcare system: We pay (doctors and hospitals) for doing more and we pay for another day in the ICU."
"We don't pay for doctors to have in-depth discussions about the goals of care, or to educate patients about their prognosis, or help them to arrive at treatment goals or develop a plan to honor a dying patients wishes."
The report found some positive trends toward less aggressive end of life care: The number of median days in hospice care during the last 30 days of a person's life went from 3.3 in 2000 to 6.6 in 2009, and many more patients were enrolled in a hospice program at their time of death, from 21.6% to 42.2%. Also, a slightly higher percentage of patients had continuous hospice care in their last month of life.
It appears, however, that more patients are spending fewer than four days in hospice settings before their death, indicating that the hospice option may be offered too late for some.
Another finding is that more patients spent time in hospice units within general inpatient hospitals in 2009 than in 2000. "I guess that's good," Teno says, "but my concern is that the use of these services really come only after an aggressive pattern of care. Hospice has become sort of an add-on with these people who spend three days or less on hospice services."
The researchers' sample used a cohort of a random 20% sample of fee-for-service Medicare beneficiaries who died in 2000, 20005 or 2009 and who were 66 years or older, and did not have health insurance coverage.
The Residential History File that assigns Medicare beneficiaries to a given location each day was used to determine place of death, number of healthcare transitions, and places of care.
An accompanying JAMA editorial by Grace Jenq, MD and Mary Tinetti, MD, of the Yale School of Medicine and the Yale School of Health suggest that a solution might be a policy in which physicians and others set "criteria for ICU admission as for other sites of healthcare (which) could reduce inappropriate and costly ICU stays that deprive many patients of the end-of-life care they would prefer if asked.
"Perhaps there should be a threshold of likely benefit and life expectancy or an ICU admission," they wrote. "Elicitation and documentation of goals of care and a plan detailing how intensive care is the optimal method for meeting those goals should be required to receive reimbursement or ICU care."