Interactive network dynamic nomogram for predicting poor neurological outcomes of post-cardiac arrest brain injury patients
10.3760/cma.j.issn.1671-0282.2025.05.012
- VernacularTitle:心脏骤停后脑损伤不良神经预后交互式网络动态列线图
- Author:
Guowu XU
1
;
Jinxiang WANG
;
Heng JIN
;
Lijun WANG
;
Muming YU
Author Information
1. 天津医科大学总医院急诊医学科，天津 300052
- Keywords:
Post-cardiac arrest brain injury;
Poor neurological prognosis;
Interactive network dynamic nomogram
- From:
Chinese Journal of Emergency Medicine
2025;34(5):684-691
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To develop and validate an interactive network dynamic nomogram for early prediction of poor neurological prognosis in patients with post-cardiac arrest brain injury (PCABI).Methods:A retrospective study was conducted on hospitalized patients who achieved return of spontaneous circulation after cardiac arrest at Tianjin Medical University General Hospital between January 2020 and April 2024. Patients were classified into favorable and poor prognosis groups based on the Glasgow-Pittsburgh Cerebral Performance Category at discharge. Eligible patients were randomly assigned to a training cohort and an internal validation cohort in a 7:3 ratio. Univariate and multivariate logistic regression analyses were performed to identify independent predictors of poor neurological outcomes in PCABI, which were subsequently used to develop a nomogram prediction model. The predictive performance of the nomogram was evaluated by comparing its area under the curve (AUC) of receiver operating characteristic with those of individual predictors using the DeLong test. Model calibration and clinical utility were assessed using calibration curves and decision curve analysis, respectively. Internal validation was conducted, and an interactive dynamic nomogram was developed using web-based visualization techniques.Results:A total of 276 PCABI patients were enrolled (training set: 196; validation set: 80), with 82 cases (29.7%) classified as poor prognosis. Multivariate logistic regression analysis identified age ( OR=1.071, 95% CI: 1.021-1.124, P=0.005), APACHEⅡ score ( OR=1.746, 95% CI: 1.393-2.190, P<0.001), initial shockable rhythm ( OR=0.142, 95% CI: 0.025-0.819, P=0.029), defibrillation ( OR=0.228, 95% CI: 0.060-0.869, P=0.030), cardiopulmonary resuscitation duration ( OR=2.116, 95% CI: 1.487-3.010, P<0.001), and lactate level ( OR=1.392, 95% CI: 1.005-1.927, P=0.047) as independent predictors of poor neurological outcomes in PCABI. A nomogram prediction model was developed based on these factors, achieving an AUC of 0.965 (95% CI: 0.939-0.989) in the training cohort and 0.987 (95% CI: 0.967-1.000) in the internal validation cohort. The nomogram demonstrated significantly superior predictive performance compared to individual predictors ( P<0.001) and exhibited excellent discrimination, calibration, and clinical net benefit. The interactive dynamic nomogram, developed through web-based visualization, further enhanced its applicability in clinical practice. Conclusions:The interactive network dynamic nomogram, developed based on age, APACHEⅡ score, initial shockable rhythm, defibrillation, cardiopulmonary resuscitation duration, and lactate level, demonstrated favorable predictive value for poor neurological outcomes in PCABI. This tool facilitates clinical application and offers a novel strategy for early identification and targeted interventions in high-risk patients.
