Almost one of every five patients in the intensive care units of a major teaching hospital got treatment that was futile  or “probably” futile, according to the doctors who treated them.  And older patients—especially those admitted from a nursing facility—were most likely to get care that does nothing to improve their quality of life, or even keep them alive for more than a brief period of time.

Those are the conclusions of a new study by a group of UCLA researchers published this week in JAMA’s Journal of Internal Medicine. This study will be extremely controversial but should not be ignored. It raises important questions about the hospital care given to very sick older patients.

First, take a look at the results, which were based on a survey of 36 critical care physicians who treated 1136 patients over a three-month period in 2011-2012. Of those ICU patients, 11 percent got care that their doctors deemed futile and another 9 percent received treatment that physicians considered probably futile. These patients accounted for about 7 percent of all ICU patient days. The cost of futile treatment in one hospital’s ICUs: about $2.6 million over 3 months.

While the researchers found that older patients and those admitted from long-term care facilities were more likely to get futile care, they saw no differences by race or ethnicity.

Importantly, the survey was done at the time treatment was being administered and not in retrospect. Thus, the ICU docs couldn’t know for sure what the outcomes would be for these patients.

However, the researchers were able to follow the patients and what they learned seemed to confirm the judgments of the physicians.  Almost no patients had what most of us would consider a good quality of life six months after their ICU stay.

More than two-thirds (68 percent) of those receiving what was deemed futile treatment died before they could be discharged. An additional 16 percent died within 6 months. Of the remaining 16 percent, nearly all required ventilators, feeding tubes, or other life-sustaining devices to keep them alive.

Still, as the authors readily admit, some of this gets pretty murky. The researchers did not attempt to set objective criteria for futile or probably futile treatment. Instead, they relied on the subjective judgment of the surveyed doctors. And, of course, if ICUs provide less aggressive treatment to those patients whom doctors think are likely to die, their predictions could become self-fulfilling.

Did the patients or their families think the care was futile? The study did not say. If the intensive care docs thought treatment was a waste of time (or worse), why did they do it?  The authors didn’t ask and can only speculate: Lack of agreement on whether to treat by the family or among the care team, or failure to address end-of-life issues (always a good guess). And of course money can’t be ruled out.

This study leaves a lot of important unanswered questions. But it highlights an issue we often prefer to ignore:  Hospitals do over-treat patients who are near death. The only real question is how often. Many of these patients needlessly suffer even as the health system spends a significant amount of money that could be put to better use.

We know that at least some of this can be alleviated by good hospice or palliative care. This study helps us understand the consequences of failing to provide that care.