Red blood cell transfusion practice in elective orthopedic surgery: a multicenter cohort study

The indications for red blood cell (RBC) transfusions remain unclear despite published guidelines. Our hypothesis was that the transfusion practice varies inside the Centre hospitalier de l'Université de Montréal (CHUM).

A total of 701 charts of patients who underwent a knee or hip arthroplasty or prosthesis revision in three hospitals of the CHUM were reviewed. Demography, hemoglobin (Hb) concentrations, details on transfusions, and postoperative adverse events (AEs) were collected up until discharge. The primary outcome was the presence or absence of RBC transfusion. Secondary outcomes were the nadir Hb, number of units transfused, discharge Hb, blood losses, and postoperative AEs.

The rate of postoperative transfusion was 29%. We found no significant difference between odds ratios of each site for sex, coronary artery disease, chronic heart failure, type of procedure, American Society of Anesthesiologists physical status, weight, height, body mass index, body surface area, and estimated blood volume. Overall, patients were transfused at a Hb between 75 and 80g/L. Eighty-five percent of postoperative transfusions could be predicted using only nadir Hb and adding patient characteristics did not substantially improve the model (86.1%). Discharge Hb was below 100g/L in 66% of patients.

There was no difference among hospitals regarding the way RBC transfusions are used. Our data suggest that physicians mainly based their decision to transfuse on a single variable, the Hb concentration, with the use of a restrictive strategy. Future trials should focus on the optimal transfusion trigger to adopt in major orthopedic surgery.