Consensus on red cell transfusion

BMJ 1995;311:962-963 (14 October)

Editorials

Consensus on red cell transfusion

Critically ill patients are often short of blood

Perioperative blood and red cell transfusion in elective surgeryvaries, apparently without justification, among clinical units.¹Guidelines have not been agreed about when to give a transfusionand how much blood to give. Evidence has not been gathered tojustify the heterogeneity of transfusion practice, which exposesrecipients to numerous hazards, albeit a small risk per "donorexposure."²

Previous attempts have been made to establish guidelines: in1988 the US National Institutes of Health's consensus conferenceon perioperative transfusion proposed that a haemoglobin concentrationof <70 g/l justified transfusion. The recommendations, however,were so hedged around with the need to consider patients' clinicalproblems individually that the suggested haemoglobin concentrationwas only a small component in the decision.³ This is entirelyappropriate as the haemoglobin concentration or packed cellvolume alone gives insufficient information on which to basea decision to give . . . [Full text of this article]