AIHF-nomogram

AIHF-nomogram predicts the probability of advanced liver fibrosis in patients with autoimmune hepatitis using a multivariable nomogram derived from clinical and laboratory predictors.

Key Features:

• Development and Validation: Constructed from data on 235 autoimmune hepatitis patients who underwent liver biopsy and divided into training and validation sets.
• Predictive Methodology: Employed least absolute shrinkage and selection operator (LASSO) regression to identify six independent predictors—sex, age, red cell distribution width, platelets, alkaline phosphatase, and prothrombin time—and integrated them into the nomogram.
• Performance Evaluation: Assessed by receiver operating characteristic (ROC) curve analysis with area under the ROC curves of 0.804 (training) and 0.781 (validation), outperforming fibrosis-4 and aminotransferase-to-platelet ratio scores, with calibration curves showing agreement between predicted and observed probabilities and decision curve analysis indicating clinical utility.

Scientific Applications:

• Clinical Utility: Identifies patients with autoimmune hepatitis at high risk for advanced liver fibrosis to support non-invasive risk stratification without sole reliance on liver biopsy.
• Research and Development: Provides a statistical foundation for further research into non-invasive diagnostic tools in hepatology and potential broader applications beyond AIH.

Methodology:

Data from 235 biopsied autoimmune hepatitis patients were split into training and validation sets; LASSO regression selected six predictors that were combined into a predictive nomogram, and performance was evaluated using ROC curve analysis, calibration curves, and decision curve analysis.