OBJECTIVE To analyze the risk factors for intracranial infection in patients after craniotomy, and to establish and verify a nomogram model for predicting risk.

METHODS A retrospective analysis was conducted on 1 185 patients who underwent craniotomy in the Department of Neurosurgery of Quanzhou First Hospital from Jan. 2022 to Dec. 2024. These patients were divided into the infection group (n=106) and the non-infection group (n=1 079) based on the occurrence of intracranial infection during hospitalization. Logistic regression analysis was used to screen for risk factors of postoperative intracranial infection, and a nomogram prediction model was constructed.

RESULTS Logistic regression analysis results showed that prolonged operation time, preoperative diabetes, lumbar drainage catheterization, ventricular drainage >7 days, artificial dura mater application and intraoperative blood loss ≥300 ml were risk factors for postoperative intracranial infection (P < 0.05). Based on these six indicators, a nomogram prediction model was established. The nomogram model demonstrated good predictive performance upon validation: consistency index (C-index) and the area under the receiver operating characteristic curve (AUC) were 0.804, indicating excellent discrimination. The Hosmer-Lemeshow test (χ²=3.218, P=0.781) and the calibration curve mean absolute error(MAE)=0.009 both indicated precise calibration performance. The validation by the decision curve analysis (DCA) suggested that when the prediction probability threshold of the nomogram model was set between 0.050 and 0.800, the clinical net benefit reached maximum, indicating optimal clinical application value.

CONCLUSIONS The nomogram prediction model for intracranial infection risk after craniotomy constructed in this study demonstrates good prediction performance and clinical applicability. The relevant predictive indicators are scientific and easy to obtain and can provide an effective reference for the diagnosis of intracranial infection after craniotomy.