Canadian researchers have found that giving extra blood, despite its essential role in maintaining life, can cause more harm than good in those who have been thought in greatest need of it.
The unexpected finding, from a study of more than 800 critically ill patients, will deliver a serious blow to the idea of accident victims being stretchered into hospital with a bag of life-giving blood suspended above them.
Doctors had assumed that any means of boosting oxygen supply to the body's tissues in those who were seriously ill would improve their chances of survival. But the Canadian research shows that critically ill patients who received more blood were up to twice as likely to die as those who received less.
In Britain, an estimated 800,000 patients receive blood transfusions each year, most after routine surgery, and demand for blood is rising. Using less blood would also save costs and conserve a scarce resource.
The aim of a transfusion is to restore levels of haemoglobin, the oxygen- carrying constituent of red blood cells. Normal levels of haemoglobin range from 14 to 17 grams per decilitre. In the Canadian study, half of the 838 seriously unwell patients were given enough blood to raise their haemoglobin over 10 grams per decilitre, while the other half were restricted to only 7-9 grams per decilitre, half the normal level. One third of the restricted group received no blood at all.On average the restricted group received three units of blood (one and a half litres) less than the unrestricted group - but more of them survived. Among patients under 55 and the less severely ill, those who received less blood (or no blood at all) were half as likely to die as those who received more. The only exceptions were patients with heart attacks and unstable angina (chest pain).
Dr Paul Hebert and his colleagues from the University of Ottawa say in the New England Journal of Medicine: "Our findings indicate that the use of a threshold ... as low as 7 grams of haemoglobin per decilitre ... was at least as effective as and possibly superior to a liberal transfusion strategy ... in critically ill patients."
It is well-known that giving too much blood or other infusions can cause heart attacks or water on the lungs because of the effort required to pump the extra liquid round the body. But the researchers found these factors could not explain the differences they found.
In a commentary on the findings, Wesley Ely and Gordon Bernard of the Vanderbilt School of Medicine, Nashville, say it is difficult to explain the harm caused by liberal transfusion. One possibility is that a transfusion, like a transplant, delivers a shock to the immune system. "Harmless" viruses in the blood may also be implicated.
They say more trials are needed to determine the best treatment for critically ill patients. "With such knowledge, more physicians will be able to adhere to the dictum `first do no harm', and we will have a surplus of blood for transfusion rather than a shortage."