But it's not clear whether that means patients die any sooner, researchers report in the Archives of Internal Medicine.

Even if they do, that's not necessarily a bad thing, said Dr. Scott Halpern, a critical care expert at the University of Pennsylvania in Philadelphia, who wrote a commentary on the findings.

"A lot of ICU beds in this country are filled with patients that are either too sick to benefit or too well to benefit," Halpern told Reuters Health.

"The present study is interesting in that it raises the possibility that scarcity may in fact be the mother of expedited end-of-life decision-making," he added. "It's much easier to transfer a patient to an intensive care unit whether or not they will benefit from it than it is to have a difficult discussion about the end of life."

Just last month, a study from France showed patients who were denied ICU access because of bed shortages had a higher risk of dying over the next couple months.

In the new study, Dr. Henry Stelfox at the University of Calgary and colleagues used data on nearly 3,500 hospitalized patients who had suddenly gotten very ill, launching the hospital's emergency team into action.

The emergency team was called much less often when there weren't any free ICU beds compared with when at least three were available.

When no ICU beds were free, 12 percent of patients were admitted to the ICU within two hours, compared to 21 percent when more than two beds were free.

The goals of care changed from resuscitation to medical or comfort care 15 percent of the time when the ICU was full, compared to nine percent of the time when three or more beds were available.

Whether that's appropriate is unclear, said Halpern, and depends on individual values.

For example, a woman with fatal cancer whose blood pressure drops suddenly might be kept alive for a little longer if she gets aggressive treatment in the ICU, Halpern said. But she might have a more peaceful death if she is given comfort care instead.

"She would die either way, but in this case in a more palliative setting," he said.

There was no difference in death rates at the hospital, although Halpern warned that result could be misleading because patients who got comfort were often sent home.

"We don't know whether this study documents rationing or the elimination of waste," he said. "Work needs to be done to better understand how commonly ICU beds are not available, and how often that affects patients who could benefit from ICU admission."

SOURCE: Archives of Internal Medicine, online March 12, 2012.

• A Permanent Economic Emergency