Introduction Non-alcoholic fatty liver disease (NAFLD) can involve advanced fibrosis and has implications with cardiovascular and liver disease related mortality. The definitive diagnosis of NAFLD and non-alcoholic steatohepatitis (NASH) is liver biopsy. We aim to evaluate and compare the accuracy of the fibrosis-4 (Fib-4) index in predicting hepatic fibrosis among young and adult age groups with biopsy proven NAFLD or NASH.

Methods We conducted a retrospective chart review to identify patients who were 18 years of age and over who underwent a liver biopsy between 2020 and 2023. Data was collected on patient demographics, comorbidities, and liver biopsy findings. Stage of fibrosis was determined using the Metavir Scoring System (F0-F4). Exclusion criteria for the study focused on identification of other causes of acute or chronic liver disease. Patients were split into two age groups: young (18-44 years) and older (45 years or more). All statistical analyses were conducted using Stata V.18 to compute means and frequencies of patient data to determine accuracy of Fib-4 index.

Results A total of 254 patients were included with a mean age of 53.1 years, 62.2% females, and 80.3% Caucasians. The majority of patients had metabolic syndrome (50.8%), obesity (75.2%), hypertension (62.6%), and dyslipidemia (64.9%). Fib-4 index was interpretable in 171 out of 254 patients, out of which 28.7% were predicted to have advanced fibrosis. Fib-4 index was indeterminate in the remaining 83 patients out of whom 6% had advanced fibrosis on liver biopsy. Overall, Fib-4 index had low sensitivity (10%) and higher specificity (65.7%) in predicting advanced fibrosis with the best outcomes seen in the young age group (Table 1). Area under the receiver operating characteristic (AUROC) curve analysis showed unsatisfactory accuracy of Fib-4 index in the overall sample and older age group with AUROC of 52.6% and 40.9%, respectively. There was good accuracy for the young age group with AUROC of 81.6%. Overall, the Fib-4 index had a better negative predictive value, indicating better accuracy with predicting non-advanced fibrosis cases.

Conclusion Our study suggests that there is utility of interpretable Fib-4 index scores in increasing the pre-test probability of non-advanced fibrosis (F0-F2) in both the young and older age groups. Liver biopsy may be needed to exclude advanced fibrosis. Those with indeterminate Fib-4 index scores will benefit from further investigation to determine the degree of liver fibrosis.