Development and Validation of a Preoperative Prediction Model for Major Complications After Elective Craniotomy in Brain Tumor Patients: Integrating Nursing Assessments into Risk Stratification

Background Elective craniotomy for brain tumor resection carries a substantial risk of major postoperative complications. Existing prediction models are limited by methodological weaknesses and often fail to incorporate comprehensive preoperative data, including nursing assessments. This study aimed to develop and validate a novel prediction model integrating multidimensional variables to individually estimate the risk of major complications. Methods This single-center retrospective cohort study analyzed electronic medical records from 1,263 adult patients undergoing elective craniotomy for brain tumor resection between 2018 and 2024. The cohort was randomly allocated into a training set (70%) and a validation set (30%). Variable selection was performed using Least Absolute Shrinkage and Selection Operator regression, after which a multivariable logistic regression model was developed. The primary outcome was defined as a composite of major complications occurring within 30 days after surgery. Model performance was evaluated based on discrimination (area under the curve, AUC), calibration (calibration plots and the Hosmer–Lemeshow test), and clinical utility (decision curve analysis). Results The final model included nine predictive variables: age, ASA grade, preoperative use of anticoagulants or antiplatelets, maximum tumor diameter, high-grade glioma pathology, cavernous sinus invasion, preoperative platelet count, albumin level, and Caprini score. The model showed strong discriminatory ability, with an AUC of 0.852 (95% CI: 0.825–0.879) in the training set and 0.831 (95% CI: 0.789–0.873) in the validation set. Calibration was satisfactory, and decision curve analysis indicated clinical utility over a broad range of threshold probabilities (10%–65%). Subgroup analyses revealed consistent performance across different patient strata. Conclusion We developed and internally validated a robust prediction model that integrates preoperative nursing assessments to accurately estimate the risk of major complications (such as postoperative hematoma, thromboembolism, and infection) following elective craniotomy for brain tumors. The nomogram provides a practical tool for personalized risk stratification, supporting clinical decision-making and targeted preventive interventions.