• Lauren Burrage, Henry Zimmerman, Sarah Higgins, Kerrin Param, Clare Orme, Jonathan Mitchell, and James O'Beirne.
• Department of Gastroenterology, Sunshine Coast University Hospital, Sunshine Coast Region, Queensland, Australia.
• Intern Med J. 2021 Apr 1; 51 (4): 533-539.

BackgroundCurrent guidelines suggest using transient elastography (TE) or aspartate aminotransferase to platelet ratio index (APRI) score <1 to exclude cirrhosis prior to commencing treatment for hepatitis C virus (HCV). Recently, fibrosis-4 (FIB-4) <0.93 has been shown to have a high negative predictive value (NPV) for the presence of cirrhosis.AimsTo assess FIB-4 and APRI in a cohort of HCV patients and to validate FIB-4 <0.93 in populations of HCV-infected individuals with differing cirrhosis prevalence, including secondary care, primary care and prisons.MethodsFrom our treatment database, we identified patients with complete data (n = 793). We calculated FIB-4 and APRI and correlated this with the presence of cirrhosis, determined by TE. We analysed the performance of FIB-4 and APRI using area under the receiver operating curve analysis. We calculated sensitivity, specificity, positive predictive value, NPV and number of patients misclassified using published cut-offs in populations with varying cirrhosis prevalence.ResultsFIB-4 was superior to APRI for the diagnosis of cirrhosis (area under the receiver operating curve 0.868 vs 0.802). In secondary care (cirrhosis prevalence 32%), APRI <1 had a NPV of 80% and misclassified 14% of patients. FIB-4 <0.93 had a NPV of 97% and misclassified 1%. In primary care and prison (cirrhosis prevalence 13% and 8%), the NPV for APRI <1 was 93% and 96%, respectively, but 5% of patients with cirrhosis were misclassified. FIB-4 <0.93 had excellent NPV in both primary care (97%) and prisoners (100%).ConclusionsFIB-4 <0.93 is highly efficient at ruling out cirrhosis in HCV patients and allows TE to be appropriately avoided, thereby streamlining treatment algorithms.© 2020 Royal Australasian College of Physicians.

15 keywords...

hide…