Analysis of CRRT withdrawal failure in patients with infectious shock complicating AKI based on decision tree algorithm
10.3969/j.issn.1673-9701.2025.03.006
- VernacularTitle:感染性休克并发AKI患者CRRT撤机失败的决策树分析
- Author:
Wei XIONG
1
;
Hao LIAO
1
;
Wenjuan XU
1
;
Yan TU
1
Author Information
1. 中国人民解放军联勤保障部队第九○八医院重症医学科,江西南昌 330001
- Publication Type:Journal Article
- Keywords:
Infectious shock;
Acute kidney injury;
Continuous renal replacement therapy;
Withdrawal;
Decision tree;
Prediction
- From:
China Modern Doctor
2025;63(3):22-26
- CountryChina
- Language:Chinese
-
Abstract:
Objective To establish a risk prediction model of continuous renal replacement therapy(CRRT)withdrawal failure in patients with infectious shock complicating acute kidney injury(AKI)based on the decision tree algorithm,and to explore the influencing factors of CRRT withdrawal failure in patients with infectious shock complicating AKI.Methods 220 patients with infectious shock complicating AKI admitted to our hospital from May 2020 to May 2023 were retrospectively analyzed,and divided into success group and failure group according to the success or failure of the withdrawal,univariate and multivariate Logistic regression analysis were used to screen risk factors of CRRT withdrawal failure in patients with septic shock complicated with AKI,C-reactive protein/albumin(CRP/ALB)at admission and sepsis-related organ failure assessment(SOFA)score at withdrawal,acute physiology and chronic health evaluation(APACHE Ⅱ)score,N-terminal B-type natriuretic peptide(NT-proBNP)level at the beginning of CRRT,mean arterial pressure(MAP)grading,urine volume after withdrawal,and serum creatinine(Scr)level after withdrawal were taken into constructing a decision tree model and validating the model efficacy.Results In this study,there were 41.82％of patients failed to withdrawal.Combination of univariate and multivariate Logistic regression analysis showed that SOFA score at withdrawal,APACHE Ⅱ scores,urine volume after withdrawal,Scr level after withdrawal,NT-proBNP at the beginning of CRRT,and MAP grading,CRP/ALB at the time of admission were an independent risk factor for CRRT withdrawal failure in patients with septic shock complicated with AKI(P＜0.05).The results showed that the higher CRP/ALB was the most important influencing factor on the failure of CRRT evacuation in patients with infectious shock complicating AKI,and the area under the receiver operating characteristic curve was 0.965.Conclusion The decision tree model constructed by CRP/ALB at admission,SOFA score at withdrawal,urine volume after withdrawal,Scr level after withdrawal,APACHE Ⅱ score,and MAP grading has a better predictive efficacy of CRRT withdrawal failure in patients with infectious shock complicating AKI,which is a guideline for patients'prognostic assessment.
