Less Could Be More in Blood Transfusions

The discovery that it is safe--and sometimes better--to cut back on blood transfusions for critically ill patients may have effects far beyond the obvious ones of blood and money.

It could open the way for more studies of whether common sense makes medical sense, said Dr. Gordon R. Barnard, chief of critical care service at Vanderbilt Medical Center in Nashville, Tenn.

Common sense would say that a normal red-cell count is better than a low one, but the work directed by Dr. Paul C. Hebert shows that isn’t always so, Barnard said.

Hebert, of the University of Ottawa, found that critical care patients are as likely to recover if they get transfusions only when they become severely anemic as they are when mild anemia is the trigger for transfusion.

And, it found, the more restrictive strategy is better for patients under the age of 55 and those who are less critically ill. Among those patients, those who got more transfusions were more likely to die.

“If common sense is not reliable, we need clinical trials,” Barnard said. “I think this paves the way for clinical trials that seem to fly in the face of logic.”

He said Hebert’s study asked a simple question in everyday practice: What hemoglobin level is best in the ICU?

“There are a number of questions like that which remain unanswered or inadequately answered, not only in general medicine but in critical care particularly,” he said.

Some of those questions: What is the best oxygen level to maintain? When drugs have to be used to raise a patient’s blood pressure, what is the best blood pressure to maintain?

“We don’t know if we should try to achieve normal levels, 10% less than normal, or where. It’s probably different depending on what physiological breakdown we’re talking about,” Barnard said.

Red-cell transfusions are routine to fight anemia. However, critical-care doctors disagree on when they are needed. Some give them when a patient becomes slightly anemic, with 10 grams of hemoglobin per deciliter of blood, compared to the usual 11.5 to 12.5 grams. Others wait until a patient is severely anemic, at 7 grams of hemoglobin per deciliter.

Hebert and the Canadian Critical Care Trials Group randomly divided 838 critically ill and anemic patients into two groups, one for each treatment strategy.

He said he expected the two approaches to produce similar results, a finding that by itself could save millions of dollars a year. And, overall, that was the result, he reported in the Feb. 22 edition of the New England Journal of Medicine.

The 420 patients in the liberal strategy group got an average of 5.6 units of blood apiece, while the 418 in the restrictive group averaged 2.6 units--about 54% less.

In addition, one-third of the patients in the restrictive group did not get any transfusions at all.

And in two groups, less was better: Younger and less critically ill patients were more likely to die if they got more transfusions.

“This is a landmark study. This is big-time,” said Dr. Stephen Cohn, chief of trauma and surgical critical care at Jackson Memorial Hospital-University of Miami School of Medicine.

He and Dr. Robert Taylor, president of the Society for Critical Care Medicine, said they plan to use the more restrictive guidelines in their ICUs.

Taylor, of St. John’s Mercy Medical Center in St. Louis, emphasized that the findings don’t apply to several groups of patients. He said that includes patients who are bleeding, whether from ulcers or wounds; those suffering from heart attacks or reduced flow of blood to the heart; and those with emphysema and cardiovascular disease.

And, he said, “This is a large study. A well-done study. But there needs to be other studies that confirm this.”