Development of a preoperative risk score for predicting blood transfusion in pediatric scoliosis surgery: a two-center retrospective cohort study.

Background: Despite recent advancements in blood conservation strategies, perioperative allogeneic red blood cell (RBC) transfusion remains common in pediatric patients undergoing scoliosis surgery. This study aimed to develop a simplified preoperative risk score to predict allogeneic RBC transfusion requirements in this population. Methods: We conducted a retrospective cohort study of 1,992 pediatric patients (<18 years) who underwent scoliosis surgery at two tertiary care centers between January 2018 and May 2023. The primary outcome was perioperative allogeneic RBC transfusion. Missing data were addressed via multiple imputation, generating 10 imputed datasets. Predictor variables were screened through univariate and least absolute shrinkage and selection operator (LASSO) regression. Multivariable logistic regression was used to develop the prediction model, followed by stepwise variable selection to identify a parsimonious set of predictors without compromising model performance. A simplified risk score and corresponding risk stratification system were derived from the final model coefficients. Results: The overall perioperative RBC transfusion rate was 32.9% (655/1,992). After variable selection and model optimization, the final prediction model included four preoperative variables: American Society of Anesthesiologists (ASA) physical status grade, Cobb angle, diagnosis, and weight. The resulting risk score ranged from 0 to 20 points and exhibited strong discriminative ability, with an area under the receiver operating characteristic curve (AUC) of 0.818 [95% confidence interval (CI): 0.793-0.843] in the training set and 0.814 (95% CI: 0.764-0.864) in the validation set. Three risk categories were established: low risk (0-4 points), medium risk (5-10 points), and high risk (≥11 points). Transfusion rates exhibited a progressive increase across risk categories in both training (9.6, 37.3, 76.3%) and validation (7.3, 32.1, 74.1%) sets. Conclusion: We developed a simplified risk score based on four readily available preoperative variables (diagnosis, weight, Cobb angle, and ASA grade) to predict perioperative allogeneic RBC transfusion in pediatric scoliosis surgery. This tool facilitates preoperative transfusion risk assessment and stratification, serving as a practical clinical aid that may contribute to optimized perioperative blood management strategies.