This study reports on variability in red blood cell transfusion rates among US hospitals using a large dataset of 228 316 patients. Information was taken from discharge records from 2009-2011.

The authors conclude that there still is a large variability in transfusion rates among hospitals, which is inversely associated with hospital procedure volume and with length of stay. This is not a new finding; however, it remains disappointing that such a wide variability still exists.

The authors do not provide an explanation (no information on the use of, and adherence to, restrictive transfusion policies, no information on the use of transfusion alternatives), so for the reader it is not clear how this can be further addressed. A possible case of co-linearity could be present regarding the association found between transfusion rates and length of stay (high transfusion rate was associated with longer hospital stay), since the highest transfusion rates were found in a relatively high proportion of government hospitals that might focus less on length of stay than private hospitals.

In their discussion, the authors suggest to include the transfusion rate among quality indicators for hospitals, which I find a very relevant solution. In the Netherlands, red blood cell use has been adopted as a quality indicator for years and has decreased dramatically, partly due to increased awareness of the need to minimise blood loss among orthopaedic surgeons but also due to improved adherence to restrictive transfusion triggers.

– Cynthia So-Osman