A nomogram for predicting intracranial infection in patients with spontaneous intraventricular hemorrhage after external ventricular drainage: a retrospective cohort study

Abstract Background Spontaneous intraventricular hemorrhage (SIVH) necessitating external ventricular drainage (EVD) is a critical condition often accompanied by the risk of intracranial infection (ICI), which significantly affects patient prognosis. This study aimed to develop a predictive model for ICI following EVD in patients with SIVH. Methods We retrospectively analyzed clinical data from 524 patients who underwent ventricular drainage between 2014 and 2024. The cohort was divided into training and validation sets, and variables such as demographic characteristics, underlying conditions, surgical factors, and postoperative management were evaluated. LASSO regression and logistic regression were employed to construct and validate the predictive model. The model was evaluated using calibration plots and the area under the receiver operating characteristic curve (AUC). Additionally, decision curve analysis and clinical impact curves were conducted.To ensure robustness and correct for overfitting, internal validation on the entire dataset was performed, yielding optimism-corrected performance metrics including AUC, calibration slope, and Brier score. Results Multivariate analysis identified a history of diabetes mellitus, duration of EVD use, sterile gauze covering the three-way valve of the drainage system, drainage flushing and catheter loosening as independent predictors of ICI. A nomogram incorporating these risk factors was developed to predict ICI in patients with SIVH. The calibration plot showed strong agreement between predicted and actual values. The AUC for the training cohort was 0.875, and for the validation cohort, it was 0.851. Decision curve analysis indicated robust performance and clinically significant benefits. Additionally, internal validation on the entire dataset further supported the internal validity and stability of our model. Conclusions Our nomogram may be of great value in predicting the ICI risk in patients with SIVH after EVD. This predictive model will assist clinicians and patients in devising effective clinical strategies.