Blood Transfusions Raise Death Risk In Very IllSep 24, 2002 | Healthscout Blood transfusions appear to increase the risk of death for some critically ill patients, a new study finds.
And while a Canadian expert says the results of the study are far from definitive, he says that he and many other Canadian doctors have reduced the number of transfusions they give in such cases, based on the results of a carefully controlled study he reported several years ago.
The study comes as the U.S. government investigates several cases of people who have received the West Nile virus through blood transfusions. At least one has died.
The European study, appearing in tomorrow's issue of the Journal of the American Medical Association, looked at the results of transfusions given to more than 3,500 patients in 146 intensive care units. The death rate for patients who got transfusions was 29 percent, compared to 14.9 percent for those who did not get blood, the researchers report.
"We currently believe that blood transfusions result in immunosuppressive effects, with increased risk of infections," says Dr. Jean-Louis Vincent, the head of intensive care at Erasme University Hospital in Brussels, and the lead author of the report.
The Canadian expert, Dr. Paul C. Hebert, who co-wrote an accompanying editorial, points out an obvious problem in interpreting that fact. "Transfusion is clearly a marker of the severity of illness," says Hebert, who is professor of medicine at the University of Ottawa. And while the European study tried to take the condition of the patients into account, it's hard to tell if they succeeded, he adds.
Hebert's basic criticism of the European study is that it is "observational," meaning that it looked back, using information on all patients who were treated over a given period. Truly accurate data can come only from randomized, controlled trials in which patients are assigned in advance to get certain treatments, Hebert says.
"The inference that can be drawn from this study remain weak and purely in the realm of hypothesis," he says. "It has to be tested again."
But in 1997, Hebert and his colleagues reported a smaller-scale controlled trial that also found an increased risk of death among seriously ill patients who got transfusions. That trial indicated that "restricting transfusions doesn't hurt people and may save lives," he says.
Hebert quickly says that the relatively small number of patients in his trial limits the conclusions that can be drawn from it. "We have been suggesting that our trial needs to be reproduced and that studies should be done in other populations," he says.
There is a fine balance between "the risk of anemia and the risk of transfusion, and we don't know which will win out," Hebert says. But his study tipped the balance slightly, he says. "Most of Canada changed clinical practice because of the results of our trial, and I believe we are transfusing less," he says.
"We tend to transfuse less," Vincent echoes.
"Several controlled trials of the kind needed to resolve the issue are under way", Hebert says.
One of them, testing whether erythropoietin, a hormone that stimulates red blood cell production, works better than transfusing red cells, has been completed and is awaiting publication. Canadian agencies are sponsoring several trials with different groups of patients, and the National Institutes of Health is planning at least one large-scale trial in this country.