Predicting Operative Blood Loss During Spinal Fusion for Adolescent Idiopathic Scoliosis

Background: Patient and surgical factors are known to influence operative blood loss in spinal fusion for adolescent idiopathic scoliosis (AIS), but have only been loosely identified. To date, there are no established recommendations to guide decisions to predonate autologous blood, and the current practice is based primarily on surgeon preference. This study is designed to determine which patient and surgical factors are correlated with, and predictive of, blood loss during spinal fusion for AIS. Methods: Retrospective analysis of 340 (81 males, 259 females; mean age, 15.2 y) consecutive AIS patients treated by a single surgeon from 2000 to 2008. Demographic (sex, age, height, weight, and associated comorbidities), laboratory (hematocrit, platelet, PT/PTT/INR), standard radiographic, and perioperative data including complications were analyzed with a linear stepwise regression to develop a predictive model of blood loss. Results: Estimated blood loss was 907 +/- 775 mL for posterior spinal fusion (PSF, n = 188), 323 +/- 171 mL for anterior spinal fusion (ASF, n = 124), and 1277 +/- 821 mL for combined procedures (n= 28). For patients undergoing PSF, stepwise analysis identified sex, preoperative kyphosis, and operative time to be the most important predictors of increased blood loss (P < 0.05). For ASF, the mean arterial pressure at incision and the operative time were predictive (P < 0.05). The following formula was developed to estimate blood loss in PSF: blood loss (mL) = C + Op-time (min) x (6.4) - pre-op T2-T12 kyphosis (degrees) x (8.7), C = 233 if male and - 270 if female. Conclusion: We find sex, operative time, and preoperative kyphosis to be the most important predictors of increased blood loss in PSF for AIS. Mean arterial pressure and operative time were predictive of estimated blood loss in ASF. For posterior fusions, we also present a model that estimates blood loss pre-operatively and can be used to guide decisions regarding pre-donation of blood and the use of antifibrinolytic agents. Level of Evidence: Retrospective study: Level II.