Performance of the Enhanced Liver Fibrosis Score, Comparison with Vibration-controlled Transient Elastography Data, and Development of a Simple Algorithm to Predict Significant Liver Fibrosis in a Community-based Liver Service: A Retrospective Evaluation

Liver fibrosis is a key risk factor for cirrhosis, hepatocellular carcinoma and end stage liver failure. The National Institute for Health and Care Excellence guidelines for assessment for advanced (≥F3) liver fibrosis in people with nonalcoholic fatty liver disease recommend the use of enhanced liver fibrosis (ELF) test, followed by vibration-controlled transient elastography (VCTE). Performance of ELF at predicting significant (≥F2) fibrosis in real-world practice is uncertain. To assess the accuracy of ELF using VCTE; investigate the optimum ELF cutoff value to identify ≥F2 and ≥F3; and develop a simple algorithm, with and without ELF score, for detecting ≥F2.Retrospective evaluation of patients referred to a Community Liver Service for VCTE, Jan-Dec 2020. Assessment included: body mass index (BMI), diabetes status, alanine aminotransferase (ALT) levels, ELF score and biopsy-validated fibrosis stages according to VCTE.Data from 273 patients were available. n=110 patients had diabetes. ELF showed fair performance for ≥F2 and ≥F3, area under the curve (AUC) = 0.70, 95% confidence interval (CI) 0.64-0.76 and AUC=0.72, 95% CI: 0.65-0.79 respectively. For ≥F2 Youden's index for ELF=9.85 and for ≥F3, ELF=9.95. Combining ALT, BMI, and HbA1c (ALBA algorithm) to predict ≥F2 showed good performance (AUC=0.80, 95% CI: 0.69-0.92), adding ALBA to ELF improved performance (AUC=0.82, 95% CI: 0.77-0.88). Results were independently validated.Optimal ELF cutoff for ≥F2 is 9.85 and 9.95 for ≥F3. ALT, BMI, and HbA1c (ALBA algorithm) can stratify patients at risk of ≥F2. ELF performance is improved by adding ALBA.