Enormous Variation in Late-Life Care Documented Once Again

The latest report from the Dartmouth Atlas Project concludes that the quality of end-of-life cancer care and use of hospice and palliative care services in a cancer patient's last month of life are enormously variable, depending on the community and the specific hospital or health system where the patient sought care. The mainstream media have picked up on this story (read here and here) as have  publications serving  the health care industry  (read here). Dale Lupu has an insightful post on Pallimed (read here) on the study's conclusions. Some newspapers have taken a closer look at study results within  their communities (read here for a report on Chicago).

A recent post on this blog (read here) commented on variation in late-life care (as identified in an earlier Dartmouth Atlas report), and how we should be looking at top decile performers as beacons whose practices should be emulated,  and another post (read here) wondered if a time will arrive when people evaluating communities in which to retire will  use a community's performance in providing late-life care as a factor in their final choice.
David Goodman, MD,   the director of the Center for Health Policy Research at the Dartmouth Institute for Health Policy and Clinical Practice and lead author of this study, put it best when he said , "Doctors and health systems need to take a look at themselves... to examine where they need to make their investments, and to make sure that they're fairly investing in what I think is one of the most undertreated (services) that we have in Medicare today, and that is undertreatment with palliative care and hospices services for those with advanced disease."

Among the striking examples of  variation, consider these:

•On average, 6% of cancer patients received chemotherapy during the last two weeks of life nationally, but in Olympia, WA it was 12.6% and in San Antonio, TX, 10.8, but only 3% in Worcester, MA and Baton Rouge, LA.

•Nationally, about 24% of Medicare patients with cancer were admitted to intensive care at least once during their last month of life. But variation across the country was seven-fold. In  Los Angeles and Miami it was 40.3%.  But in Mason City, Iowa, it was only 6%.

•Nationally, about 55% of cancer patients used hospice services in the last month of life. But again, variation was nearly fourfold across the country depending on hospital referral region. .

•About 9% of patients with end stage cancer nationally received life sustaining treatments, such as endotracheal intubation, feeding tube placement and cardiopulmonary resuscitation. But rates varied by more than six across the nation, from 18.2 in Manhattan and 17.5% in Los Angeles to 3.9% in Minneapolis.

Nancy Foster, vice president for quality and patient safety with the American Hospital Association, said studies like the Dartmouth Atlas Project's "help shine a light on the importance of end-of-life care." She went on to say,"Unfortunately, hospice care is not readily available in all parts of the United States." A curious comment, when one considers that use of hospice is below national average in some cities, regions, and states which have a relatively higher number of hospices per capita.

How much of this care is futile? Hard to say, of course. Living well has a different meaning for each patient, and it is the responsibility of clinicians and health care systems to help patients articulate their goals for living and for their medical care.

But we do know that there is a huge gap between patient preferences and the care they receive. In a post titled " Marginal Treatments at a High Emotional Cost from Families" on the KevinMD.com site, Dr. Marya Zilberberg wrote about her father's battle with advanced cancer and was left wondering  "what are we prolonging in many of the cases that we treat at the end of life — life or death?"
 
As Dale Lupu wrote in her post, perhaps this study will  be the impetus for performance improvement projects in some of these regions with poor results. I'm optimistic it will be.  Progress will be slow, however, if these performance improvement projects are isolated to individual hospitals and hospices - what are needed, as our (DAI Palliative Care Group) studies of exemplar palliative care communities have shown, are community-wide initiatives, or networks, whose "members" span settings of care and adopt "tools" like POLST to "spread the science" of HPM.

What a wonderful opportunity for HPM practitioners to assume leadership positions in the development of such networks. Much more on this subject to appear in future posts.