The Journal of Practical Medicine ›› 2026, Vol. 42 ›› Issue (17): 3111-3118.doi: 10.3969/j.issn.1006-5725.2026.17.009

• Chronic Disease Control • Previous Articles     Next Articles

Construction and evaluation of a predictive model for acute kidney injury in patients with acute heart failure based on the VExUS score and clinical risk factors

Yulong Wei, Yezi Feng, Xinxin Huang, Qianqian Ma, Zhiqiang Dong, Qi Ding()   

  1. Department of Cardiology,Zhengzhou Seventh People's Hospital,Zhengzhou 450016,Henan,China
  • Received:2026-06-23 Online:2026-09-10 Published:2026-10-10
  • Contact: Qi Ding
  • About author:

    Wei Yulong performed conceptualization, data curation, and writing original draft. Feng Yezi performed conceptualization, writing review and editing. Huang Xinxin performed investigation, writing review and editing. Ma Qianqian performed investigation. Dong Zhiqiang performed investigation. Ding Qi performed conceptualization, funding acquisition, writing review and editing. All authors read and approved the final manuscript as submitted.

Abstract:

Objective To develop a combined predictive model incorporating the Venous Excess Ultrasound (VExUS) score and clinical risk indicators for acute kidney injury (AKI) in patients with acute heart failure (AHF) admitted to the cardiac care unit (CCU), and to evaluate the incremental predictive value of the VExUS score. Methods This study included 122 AHF patients hospitalized in the CCU between January 2022 and January 2025. The VExUS score, incorporating Doppler hemodynamic assessments of the maximum inferior vena cava diameter, hepatic vein, portal vein, and renal vein, was evaluated within 24 hours of admission. Patients were categorized into an AKI group (n = 40) and a non-AKI group (n = 82) based on the KDIGO guidelines. A combined predictive model was developed using logistic regression. Model discrimination was evaluated and compared using receiver operating characteristic (ROC) curves and the DeLong test. The net reclassification improvement (NRI) and integrated discrimination improvement (IDI) were calculated to assess the incremental predictive value of the VExUS score. Internal validation was performed using a bootstrapping approach, and calibration curves were plotted to evaluate model calibration. Results Among the 122 patients, the proportion of individuals with a VExUS score ≥ 2 was significantly higher in the AKI group (score of 2: 23%; score of 3: 43%) compared to the non-AKI group (11% and 12%, respectively) (P < 0.001). Multivariate logistic regression analysis revealed that a VExUS score of 2 (OR = 6.44, 95% CI: 1.71 - 24.22, P = 0.006) and a VExUS score of 3 (OR = 7.81, 95% CI: 2.25 - 27.10, P < 0.001) were independent predictors of AKI occurrence. The area under the curve (AUC) for the VExUS score alone in predicting AKI was 0.749. When integrated with a baseline clinical model (including a history of CKD, APACHE Ⅱ score, etc.), the AUC further improved to 0.821, demonstrating superior predictive performance compared to the clinical model alone (P < 0.001). Compared with the VExUS model alone, the combined model yielded a continuous NRI of 0.690 (P < 0.001) and an IDI of 0.063 (P = 0.021). Bootstrap internal validation demonstrated good calibration for the combined model. Conclusion A combined predictive model incorporating the VExUS score and clinical risk factors accurately predicts AKI risk in patients with AHF, outperforming the standalone VExUS score. The VExUS score offers incremental predictive insights beyond traditional clinical risk assessment, helping inform volume management and renal protection strategies in this population.

Key words: venous excess ultrasound score, acute heart failure, acute kidney injury, congestive nephropathy, doppler ultrasound

CLC Number: