To develop and validate a nomogram predicting the postoperative adding-on (AO) phenomenon in Lenke 1/2 adolescent idiopathic scoliosis (AIS) patients, providing a quantitative tool for therapeutic benefit assessment and personalized risk prediction. A retrospective analysis was conducted using clinical data from 213 patients. The least absolute shrinkage selection operator (Lasso) was applied to identify the best predictors. Multivariate logistic regression was used to construct a nomogram with a training set (80%), while a validation set (20%) was used for internal validation. The receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA) assessed the model's clinical utility. The model was constructed using the four best predictor variables: the number of vertebrae in the main thoracic curve (NV-MTC), apical vertebra translation (AVT), last instrumented vertebra rotation (LIVR), and Risser sign (RS). The model's C-index was 0.932 (95% confidence interval, 0.8858-0.9775). DCA indicated a clinically meaningful high-risk threshold between 0 and 1, and the calibration curve demonstrated satisfactory accuracy. The model proved useful for predicting the occurrence of AO in Lenke 1/2 AIS patients after surgery.
Read full abstract