External Validation of Four Point-of-Care Noninvasive Scores for Predicting Advanced Hepatic Fibrosis in a Predominantly Hispanic NAFLD Population

Background The development of point-of-care biomarkers of disease has become a major focus of interest in nonalcoholic fatty liver disease (NAFLD). The NAFLD fibrosis score (NFS), BARD, FIB-4, and aspartate aminotransferase-to-platelet ratio index (APRI) are commonly used for advanced NAFLD fibrosis prediction. However, the performance of these scores among in a predominantly Hispanic patient population, a population with the highest prevalence of NAFLD, has not been examined. Methods We performed a retrospective study among patients with histologically confirmed and staged NAFLD at the Ben Taub General Hospital, Houston, Texas, to externally validate four noninvasive advanced fibrosis prediction scores. Their discriminatory ability was assessed using the area under the receiver operating characteristic curve (AUROC). Sensitivity, specificity, positive, and negative predictive values were calculated. Results We included 99 NAFLD patients, of whom 37 (37.4%) had advanced fibrosis. The cohort was predominantly Hispanic (73.7%). The AUROC for detecting advanced fibrosis were: NFS 0.79 (95% confidence interval, 0.69-0.88), BARD 0.76 (0.67-0.86), FIB-4 0.77 (0.68-0.87), and APRI 0.70 (0.59-0.81). Using the low cutoff for the NFS (- 1.455) had the highest sensitivity (81.1%) and the highest negative predictive value (85.4%) among the overall study population. Conclusions Noninvasive scores for advanced NAFLD fibrosis have moderate discriminatory ability in Hispanic patients with NFS having a small advantage. The AUROCs of these scores were similar to those reported in Caucasian populations. However, they had uniformly lower negative predictive values among our predominantly Hispanic study population, suggesting that they are not reliable for ruling out advanced fibrosis among this high-risk population.