More people over 65 are dying in hospice care and fewer are dying in hospitals. But this good news is tempered by a very different story. People are also being hospitalized more frequently in the last three months of their lives, are more likely to spend time in intensive care units, and are often receiving hospice care for just a few days before they die.

In a new study in the Journal of the American Medical Association (JAMA), Joan Teno and colleagues painted a nuanced picture of end-of-life care in the U.S.  A key finding: Simply knowing where someone dies may not say much about the care she received at the end of life.  

As Teno and her colleagues note, one patient may spend her final week at home, but her last day in the hospital for pain control. Another may spend her last week being moved from home to hospital to nursing home and back to the hospital. Both die in the same place but their experiences are very different.

A recent report by the Centers for Disease Control and Prevention found more people are dying at home. And Medicare is considering site of death as a quality measure for end-of-life care, on the theory that dying at home may be less stressful than dying in an institution.

But Teno tells a much more complex tale. She and her co-authors found that about 33.5 percent of Medicare beneficiaries died at home in 2009, 10 percent more than in 2000. Only about 24.6 percent died in the hospital in 2009, down one-quarter from 2000, while the percentage of people dying in nursing homes was little changed. At the same time, about 42 percent died in hospice care in 2009, nearly twice the percentage in 2000.

For those of us who believe in the importance of palliative care at the end of life, this is all good. But it isn’t the whole story. And the rest is less positive.

Transitions from one care setting to another in the last 90 days of life increased by 50 percent, from an average of two moves in 2000 to three in 2009. Ten percent faced a care transition in their last three days in 2000, but 14 percent were moved in 2009. While 24 percent of those who died used the ICU in 2000, 29 percent received intensive care in 2009. More than 11 percent had three or more hospitalizations in the last 90 days of their life.

While these patients spent a bit less time in the hospital, they also spent more days in the ICU. Most disturbingly, dementia patients spent more time in the intensive care in 2009 than in 2000.

Finally, while hospice use increase, more than 28 percent of hospice users were enrolled for three days or less.  Medicare’s own attitude towards hospice is…complicated.

There are some big caveats to this study. The first is that it includes only patients in traditional fee-for-service Medicare, not Medicare Advantage. It will be interesting to learn whether experiences are any different for managed care patients.

Second, the study looks retrospectively at all patients who have died though in some cases nobody would expect them to receive end of life care. Imagine, for instance, an active 65-year-old who breaks a leg skiing. She has the leg repaired in the hospital, goes to a skilled nursing facility for rehab, and has a completely unexpected heart attack. She is returned to the hospital, where she dies in the ICU. She’d appear as someone with multiple care transitions at the end of life, though such care may have been completely appropriate.  

But many people don’t die that way. They suffer from chronic diseases, such as heart or lung disease or dementia, and death is often not unexpected (by them or their physicians). Teno’s study tells us there is still a lot more we need to learn about how to deliver care at the end of life.