ABSTRACT
Objective:To investigate the improved performance of hepatic elastography combined with the serum biomarkers for the diagnosis of biliary atresia.Methods:A total of 193 patients with suspected biliary atresia in Beijing Children′s Hospital from March 2019 to November 2020 were consecutively collected. All patients were randomly divided into the training cohort and validation cohort at a ratio of 7∶3. LASSO regression analysis was used for the selection of the model index based on the data set from the training cohort including the serum biomarkers, demographic features (age and sex) and hepatic elastic measurement, and a diagnostic model for biliary atresia was subsequently developed by weighting on the basis of the dominance ration. The performance of the model was respectively evaluated with respect to the discrimination and calibration in each cohort.Results:Alanine aminotransferase (ALT), glutamyl transferase (GGT) and hepatic elastic measurement were selected to build the model. The area under the ROC curve of the final diagnostic model was 0.943 with a sensitivity of 90.9% and a specificity of 85.7% in the training cohort, and 0.955 in the validation cohort. Hosmer-Lemeshow test ( P=0.292, P=0.951) and calibration curves further validated its satisfactory calibration in both cohorts. As demonstrated by Delong et al.test, employing the model in the training cohort achieved the best diagnostic performance compared with using single model index ( P<0.001, P=0.016, P<0.001). In the validation cohort, the decision curve analysis showed the model had a higher overall net benefit over using hepatic elastography alone in every predicted probability. Conclusions:The diagnostic model for biliary atresia, which incorporates ALT, GGT and hepatic elastic measurement, can improve the performance of hepatic elastography with a higher clinical value.