Using data from a prospective blood management intelligence portal, Lucas and co-authors performed an analysis of 2 905 patients who underwent major abdominal surgery at Johns Hopkins Hospital betweeen 2010 and 2013. Among the 895 patients (30.8%) who received a red blood cell transfusion intra- and/or postoperatively, overtransfusion (defined as a target haemoglobin of 9 g/dL or higher) was identified in more than half (57%) of the cases. The risk of overtransfusion was 1.46 times higher with a transfusion trigger of 8 g/dL compared with a transfusion trigger below 7 g/dL. Transfusion of more than one unit increased the risk for overtransfusion by 1.54 times.
Remarkably, patients discharged at higher haemoglobin levels did not experience less in-hospital complications and had a decreased length of stay (1.4 vs. 2.9 days). Among the 26 surgeons who transfused at least 5 patients, the
proportion of overtransfused patients by surgeon ranged from 0% to 80% and the proportion of patients who were transfused at a haemoglobin trigger of 8 g/dL ranged from 0% to 39%.
A considerable weakness of this study and of a system that considers haemoglobin levels alone is that clinical symptoms of anaemia remain unrecognised. Another limitation is that complications were identified using an administrative database based on coding and were only identified until discharge (the median hospital length of stay is not provided). Thus, the authors’ conclusion that strict adherence to appropriate transfusion targets could have saved 0.8 units per transfused patient, or 20% of all units given, and the suggestion that the observed variability of transfusion rates reflects a lack of standardised care are not well grounded on the reported evidence. However, the results of this study clearly indicate the need for more appropriate blood use and more guideline adherence, especially by avoiding the routine administration of double units.
– Thomas Frietsch
