TITLE:
Development and Validation of a Clinical Prediction Model for Postoperative Wound Infection in Patients Undergoing Spinal Tumor Surgery
AUTHORS:
Hongying Zhang, Xinyi Liang
KEYWORDS:
Spinal Tumors, Wound Infection, Spinal Surgery, Risk Factors, Risk Prediction Model
JOURNAL NAME:
Health,
Vol.18 No.7,
July
23,
2026
ABSTRACT: Objective: To investigate risk factors for postoperative incisional infection after spinal tumor surgery, develop a risk prediction model, and evaluate its predictive performance. Methods: Using a retrospective design, 223 patients who underwent spinal tumor surgery at our hospital between December 2024 and March 2026 were enrolled. According to whether surgical wound infection occurred, patients were assigned to a surgical site infection-positive group and a surgical site infection-negative group. Baseline clinical data were collected, and univariate and multivariate analyses were performed; Firth penalized likelihood Logistic regression was adopted to avoid overfitting. A risk prediction model was then constructed based on the results of logistic regression analysis. Results: Among the 223 patients who underwent spinal tumor surgery, 11 developed postoperative surgical wound infection, yielding an overall postoperative incisional infection rate of 4.93%. Multivariate analysis indicated that age and surgical approach were independent risk factors for postoperative incisional infection after spinal tumor surgery. The area under the receiver operating characteristic (ROC) curve (AUC) of the prediction model was 0.9419, After correction via internal Bootstrap validation, the area under the curve (AUC) was 0.921. The calibration curve demonstrated satisfactory model fitting (Hosmer-Lemeshow test, P = 0.437); with a sensitivity of 100% and a specificity of 88.37%. Conclusion: The risk of incisional infection after spinal tumor surgery is relatively high. Influencing factors include age and surgical approach. The prediction model constructed based on Logistic regression exhibited good discrimination and calibration, with certain clinical application value.