Craniosynostosis is defined as a premature closure of one or more calvarial sutures. Surgical intervention is indicated to improve appearance and prevent complications of increased intracranial pressure such as cognitive impairment, auditory and visual loss. Correction of craniosynostosis includes various surgical techniques, some of which are associated with substantial blood loss. Cranial vault remodeling is the historical gold standard for treating this condition and yields predictable results in terms of cranial shape and volume expansion. The technique involves wide scalp dissections with multiple osteotomies followed by reshaping and stabilization of these cranial bones. Consequently, these procedures are lengthy (4-6 …show more content…
Among the 38 patients who underwent cranial vault remodeling, we applied multivariable logistic regression to identify independent predictors of substantial intraoperative blood loss (i.e., > 60 mL/kg) and the need for coagulation products with the c-index used to judge the quality of model prediction and the Youden J-index in receiver operating characteristic (ROC) curve analysis to determine the optimal cut-off values for TEG parameters and the platelet fibrinogen product.{Katz, 2011 #408} We developed a ROC curve based on a simplified predictive algorithm according to the number of predictors (using their respective cut-off values) to characterize the accuracy (i.e., area under the curve), sensitivity and specificity in classifying patients with and without substantial blood loss as well as those who required and did not require coagulation products.{Zhou, 2011 #407} Based on significant independent multivariable predictors, we derived the probability of substantial intraoperative blood loss with 95% confidence intervals. Box-and-whisker plots were constructed to illustrate median values and interquartile ranges for TEG parameters and platelet fibrinogen product for comparing cranial vault remodeling patients according to blood loss and need for coagulation products. Statistical analysis was performed using IBM SPSS Statistics (version 23.0, IBM, Armonk, NY). Two-tailed values of P < 0.05 were considered statistically