A Novel Model for Predicting the Clearance of Jaundice in Patients With Biliary Atresia After Kasai Procedure

Frontiers in Pediatrics

article

Yimao Zhang¹  Qi Wang¹  Siyu Pu¹  Junxiang Wang²  Bo Xiang¹  Juxian Liu³  Shuguang Jin¹ 

[1] Department of Pediatric Surgery, West China of Hospital, Sichuan University;Department of Neonatal Surgery, Shenzhen Children's Hospital;Department of Ultrasound, West China of Hospital, Sichuan University

关键词: biliary atresia;    Kasai;    clearance of jaundice;    nomogram;    prediction model;   

DOI  :  10.3389/fped.2022.837247

学科分类：社会科学、人文和艺术（综合）

来源: Frontiers

【 摘 要 】

Background The failed clearance of jaundice (CJ) in patients with biliary atresia (BA) after the Kasai procedure (KP) often leads to a shorter native liver survival (NLS) time and earlier liver transplantation. We aimed to investigate risk factors of failed CJ and establish a novel nomogram model to predict the status of CJ. Methods We retrospectively reviewed institutional medical records from January 2015 to April 2020 and enrolled BA patients post-KP, randomly divided into training and testing cohorts at a ratio of 7:3, and further subdivided into cleared and uncleared jaundice groups. Univariate and multiple logistic regression analyses were used to select risk factors to establish the nomogram in the training cohort. The performance of the nomogram was evaluated by calculating the areas under the receiver operating curve (AUC) in both cohorts. Results This study included 175 BA patients post-KP. After univariate and multiple logistic regression analyses, Cytomegalovirus IgM +ve associated BA (OR = 3.38; 95% CI 1.01–11.32; P = 0.04), ln γ-glutamyl transpeptidase (GGT) (OR = 0.41; 95% CI 0.22–0.80; P = 0.009), thickness of the fibrous portal plate (OR = 0.45; 95% CI 0.27–0.76; P = 0.003), liver stiffness measurement (LSM) (OR = 1.19; 95% CI 1.06–1.34; P = 0.002), and multiple episodes of cholangitis (OR = 1.65; 95% CI 1.13–2.41; P = 0.01) were identified as independent risk factors of unsuccessful CJ to construct the nomogram. The receiver operating characteristic curve (ROC) analysis suggested good nomogram performance in both the training (AUC = 0.96) and testing cohorts (AUC = 0.91). Conclusion Our nomogram model including several risk factors effectively predicts CJ in patients post-KP, which could aid in clinical decision-making.

【 授权许可】

CC BY