A diagnostic model of autoimmune hepatitis in unknown liver injury based on noninvasive clinical data

Development and validation of a noninvasive prediction model for significant hepatic liver fibrosis in Chinese patients with autoimmune hepatitis

INTRODUCTION AND OBJECTIVES

Autoimmune hepatitis (AIH) is a prevalent noninfectious liver disease. However, there is currently a lack of noninvasive tests appropriate for evaluating liver fibrosis in AIH patients. The objective of this study was to develop and validate a predictive model for noninvasive assessment of significant liver fibrosis (S ≥ 2) in patients to provide a reliable method for evaluating liver fibrosis in individuals with AIH.

MATERIALS AND METHODS

The clinical data of 374 AIH patients were analyzed. A prediction model was established through logistic regression in the training set, and bootstrap method was used to validate the models internally. In addition, the clinical data of 109 AIH patients were collected for external verification of the model.The model was expressed as a nomogram, and area under the curve (AUC) of the receiver operating characteristic (ROC), calibration curve, and decision curve analysis were used to evaluate the accuracy of the prediction model.

RESULTS

Logistic regression analysis revealed that age, platelet count (PLT), and the A/G ratio were identified as independent risk factors for liver fibrosis in AIH patients (P < 0.05). The diagnostic model that was composed of age, PLT and A/G was superior to APRI and FIB-4 in both the internal validation (0.872, 95%CI: 0.819-0.924) and external validation (0.829, 95%CI: 0.753-0.904).

CONCLUSIONS

Our predictive model can predict significant liver fibrosis in AIH patients more accurately, simply, and noninvasively.