This study describes an intervention in one intensive care unit, consisting of a transfusion bundle in which 4 of 5 checkpoints relates to safety (i.e. correct identification of patient and product), a valid haemoglobin (Hb) measurement and informed consent, and 1 checkpoint relates to whether red blood cell (RBC) transfusion is administered according to the individual preset Hb trigger. The authors state that their intervention aimed at reducing unnecessary, inappropriate and unsafe transfusions, but the article only reports on the measured divergence between the preset Hb threshold and the last Hb before transfusion.
The care bundle intervention has a compliance in itself of 37-50%, and the authors observed a decrease of the mean pre-transfusion Hb level by 0.2 g/dL to 7.1 g/dL. No information is given on the dose of RBCs given according to the pre-transfusion Hb (single- or double-unit transfusions) or whether this changed. There was no evaluation of the appropriateness of the chosen Hb threshold, and the patient sample was too small to reliably estimate changes in transfusion safety.
While the actual pre-transfusion Hb approaches the preset individual thresholds and the guideline, it is not certain whether the care bundle was the major cause and whether it will work for others, mainly because of the low compliance and the low external validity. It is suggested to test the external validity of the bundle in a multicentre study. As the majority of the bundle focuses on the technical-administrative aspects of transfusion, it could be very relevant in settings with a low degree of technical-administrative transfusion safety. However, for centres with a high degree of technical-administrative transfusion safety, care bundles with focus on the clinical appropriateness of RBC transfusion and on patient blood management would be more relevant.
– Astrid Norgaard