This study aimed to explore the variation in transfusion practice amongst critical care physicians using a cross-sectional online survey of critical care providers identified through the European Society of Intensive Care Medicine (ESICM).

Participants were presented with four clinical scenarios in which each patient had a haemoglobin of 7.3 g/dL and one of the following four diagnoses: acute myocardial infarction (AMI), abdominal sepsis, traumatic brain injury (TBI) or post-surgical complications after gastrointestinal surgery. For each clinical scenario, participants were asked if they would transfuse or not as well as to list the perceived probabilities of benefits and harms to transfusion and to the rank the characteristics that had most influenced their decision-making surrounding transfusion. 

The greatest variation in transfusion practice was observed in the abdominal sepsis scenario, in which 49% of participants chose to transfuse while 51% chose not to transfuse. The least variation was observed in the acute myocardial infarction scenario, where 75% decided to transfuse while 25% decided not to transfuse. The patient characteristics that participants ranked as the most important factor in their decision to transfuse was the haemoglobin level, followed by patient diagnosis and haemodynamics. 

Although utilizing a restrictive transfusion strategy has been shown to be non-inferior to a liberal transfusion strategy, critical care physicians weighed patient characteristics differently when making decisions regarding transfusion. Willems and colleagues have demonstrated that variation in transfusion practice exists amongst critical care physicians. Further research is warranted to explore these differences.

– Stacey Valentine (SABM reviewer)