The purpose of this study was to determine which patient- or surgery-related factors are predictive of need for perioperative transfusion to avoid obtaining unnecessary pre-operative type and screens (T&S). We conducted an observational retrospective cohort study of 1200 women ≥ 18 years old undergoing gynecologic surgery for benign, possibly benign, or malignant indications on a gynecologic oncology service at a university medical center from 2009-2016. A logistic regression model was used to examine patient-related and surgery-related variables predictive of outcome of transfusion. Independent variables included patient demographics, comorbidities, and surgical indication surgical route, and surgical type. Dependent variable was transfusion outcome (T&S only, conversion to type and cross (T&C), or transfusion). Eight hundred ninety-nine (74.9%) women underwent pre-operative T&S, of which 118 (9.8%) were converted to T&C, and 80 (6.7%) received a transfusion of blood or blood products. Cancer indication, major surgery, and preoperative hematocrit less than 36% were significantly associated with need for transfusion (P = 0.002, P < 0.0001, P < 0.0001, respectively). Patients with a benign indication undergoing minor procedures and with normal preoperative hematocrit are least likely to require transfusion.
Keywords: Blood loss; Cost savings; Evidence-based practice; Hemorrhage; Minimally-invasive surgery.