# Analysis of risk factors for delayed intracranial hemorrhage following ventriculoperitoneal shunt surgery and construction of a nomogram model

## TL;DR

This study identifies risk factors for delayed brain bleeding after shunt surgery in hydrocephalus patients and builds a predictive model.

## Contribution

The study introduces a nomogram model to predict delayed intracranial hemorrhage after ventriculoperitoneal shunt surgery.

## Key findings

- Age, previous craniotomy, and elevated neutrophil-to-lymphocyte ratio are independent risk factors for delayed intracranial hemorrhage.
- The nomogram model achieved an AUC of 0.80 with high sensitivity and good calibration.
- The model suggests increased postoperative monitoring for patients with identified risk factors.

## Abstract

The objective of this study is to identify the independent risk factors associated with delayed intracranial hemorrhage (DICH) subsequent to ventriculoperitoneal shunt (VPS) surgery in patients diagnosed with hydrocephalus and to construct a nomogram model for this condition.

We conducted a retrospective analysis of clinical data from 266 patients who underwent VPS procedures at Ningbo University Affiliated Lihuili Hospital between January 2015 and December 2024. The patients were stratified into two groups: those with postoperative DICH (DICH group) and those without (non-DICH group). Initially, univariate analysis was used to evaluate differences in clinical data between the two groups. Subsequently, multivariate logistic regression analysis was employed to identify independent risk factors for DICH (p < 0.05); R software was used to construct a nomogram model for predicting the occurrence of delayed intracranial hemorrhage (DICH) following ventriculoperitoneal shunt (VPS) surgery. Finally, the receiver operating characteristic (ROC) curve and calibration curve were employed to comprehensively evaluate the predictive performance of the nomogram model.

The incidence of DICH was observed to be 17.3% (46/266). Statistical analysis revealed that the DICH group had significantly higher age, a greater prevalence of previous craniotomy, and an elevated postoperative-to-preoperative neutrophil-to-lymphocyte ratio ratio (NLRR) compared to the non-DICH group (p < 0.05). Multivariate logistic regression analysis identified age (OR = 1.061, 95% CI: 1.021–1.103), a history of craniotomy (OR = 2.676, 95% CI: 1.196–5.989), and NLRR (OR = 1.931, 95% CI: 1.373–2.717) as independent risk factors for DICH (p < 0.05). The results of ROC curve analysis showed that the Area Under the Curve (AUC) was 0.80 (95% CI: 0.73–0.87). The sensitivity was calculated to be 87.7%, the specificity was 60.9%, the positive predictive value (PPV) was 91.5%, the negative predictive value (NPV) was 50.9%, and the cutoff value was 0.254. Additionally, the Hosmer-Lemeshow (HL) test for goodness-of-fit yielded a chi-square value (χ2) of 10.145 and a p-value of 0.255, indicating that the model has good calibration.

Advanced age, a history of craniotomy, and elevated NLRR are independent risk factors for DICH following VPS in hydrocephalus patients. The constructed nomogram model for predicting DICH occurrence after VPS surgery in hydrocephalus patients exhibits favorable predictive performance. Postoperatively, enhanced vigilance for DICH is recommend when any of these risk factors are present postoperatively. Increasing the frequency of postoperative head CT scans may improve detection sensitivity for asymptomatic patients with minor bleeding, thereby facilitating early intervention.