• Louanne M Carabini, Carine Zeeni, Natalie C Moreland, Robert W Gould, Michael J Avram, Laura B Hemmer, John F Bebawy, Patrick A Sugrue, Tyler R Koski, Antoun Koht, and Dhanesh K Gupta.
• Departments of *Anesthesiology ‡Anesthesiology & Neurological Surgery §Neurological Surgery ∥Anesthesiology, Neurology, & Neurological Surgery, Northwestern University Feinberg School of Medicine, Feinberg, Chicago, IL †Department of Anesthesia, American University of Beirut, Beirut, Lebanon.
• J Neurosurg Anesthesiol. 2014 Jul 1; 26 (3): 205-15.

BackgroundSurgery for posterior spine instrumentation often requires major transfusion. The aim of this study was to develop and test the validity of a model for predicting intraoperative major transfusion (>4 U total red blood cells), based on preoperative patient and surgical variables, that was applicable to adult patients undergoing cervical, thoracic, and/or lumbar spine deformity surgery with and without osteotomies.Materials And MethodsThe perioperative data from 548 patients who underwent ≥ 3 levels of posterior spinal fusion with instrumentation between January 1, 2003 and May 30, 2009, were retrospectively collected to create a model for predicting major blood transfusion. The validity of the model was retrospectively tested with a separate data set of 95 patients who underwent surgery from June 1, 2009 through September 30, 2010.ResultsThere was a 59.5% incidence of major transfusion in the derivation set of patients. Independent predictors of major transfusion were operation duration, number of posterior levels instrumented, surgical complexity score, and preincision hemoglobin. This model was able to predict major transfusion significantly better than a previously published model (ROCAUC=0.89; 99% confidence interval, 0.80-0.90; P<0.001).ConclusionsOur model has an increased accuracy for predicting the probability of major transfusion compared with a previously published model. In addition, our model is applicable to all types of spine fusion surgery and accounts for the complexity of surgical instrumentation, the number of levels instrumented, and the predicted duration of surgery as independent variables.

19 keywords...

hide…