Validation of a New Method to Automatically Select Cases With Intraoperative Red Blood Cell Transfusion for Audit.

BACKGROUND: Hospitals review allogeneic red blood cell (RBC) transfusions for appropriateness. Audit criteria have been published that apply to 5 common procedures. We expanded on this work to study the management decision of selecting which cases involving transfusion of at least 1 RBC unit to audit (review) among all surgical procedures, including those previously studied.

METHODS: This retrospective, observational study included 400,000 cases among 1891 different procedures over an 11-year period. There were 12,616 cases with RBC transfusion. We studied the proportions of cases that would be audited based on criteria of nadir hemoglobin (Hb) greater than the hospital's selected transfusion threshold, or absent Hb or missing estimated blood loss (EBL) among procedures with median EBL <500 mL. This threshold EBL was selected because it is approximately the volume removed during the donation of a single unit of whole blood at a blood bank. Missing EBL is important to the audit decision for cases in which the procedures' median EBL is <500 mL because, without an indication of the extent of bleeding, there are insufficient data to assume that there was sufficient blood loss to justify the transfusion.

RESULTS: Most cases (>50%) that would be audited and most cases (>50%) with transfusion were among procedures with median EBL <500 mL (P < .0001). Among cases with transfusion and nadir Hb >9 g/dL, the procedure's median EBL was <500 mL for 3.0 times more cases than for procedures having a median EBL ≥500 mL. A greater percentage of cases would be recommended for audit based on missing values for Hb and/or EBL than based on exceeding the Hb threshold among cases of procedures with median EBL ≥500 mL (P < .0001). There were 3.7 times as many cases with transfusion that had missing values for Hb and/or EBL than had a nadir Hb >9 g/dL and median EBL for the procedure ≥500 mL.

CONCLUSIONS: An automated process to select cases for audit of intraoperative transfusion of RBC needs to consider the median EBL of the procedure, whether the nadir Hb is below the hospital's Hb transfusion threshold for surgical cases, and the absence of either a Hb or entry of the EBL for the case. This conclusion applies to all surgical cases and procedures.