Utility of Four Non-invasive Scores in Predicting Advanced Fibrosis in a Predominantly Hispanic Population

• DOI: 10.14309/00000434-201610001-00852
• Publication Date: 2019-09-30T23:38:05Z
• Authors: Aradhna Seth; Maya Balakrishnan; Aaron P. Thrift; Gagan K. Sood
• Keywords: Liver disease; Chronic liver disease

Introduction: Non-alcoholic fatty liver disease (NAFLD) is a leading cause of chronic in the United States. NAFLD prevalence highest among Hispanics. Several non-invasive scores for predicting fibrosis patients have been well validated; however these were derived and validated majority Caucasian cohorts. Evidence Hispanics lacking. The aim this study was to assess accuracy four advanced our predominantly Hispanic patient population. Methods: This retrospective cross-sectional 137 adults with who underwent biopsies between 2010-2014 at county hospital Houston,TX. Exclusion criteria were: evidence heavy alcohol use (>30 g/day men;>20 women), secondary disease, drug induced injury, or HIV infection.The score (NFS), FIB-4, BARD, aspartate aminotransferase platelet ratio (APRI) calculated from labs obtained within six months biopsy compared histologic stage fibrosis. Microsoft Excel Stata used calculate descriptive statistics, area under receiver operating curve (AUROC), sensitivity, specificity, negative predictive value (NPV), positive (PPV). Results: overall mean age BMI 47± 11.9 years 32 ± 6.7kg/m2; 77% (n=105) 78% (n=107) women; 50% diabetic (n=69) hemoglobin A1c 7.5%. 40% had (n=55). Advanced older (p=0.026), lower platelets (p< 0.001), albumin 0.001). NFS, APRI NPV's of: 85%, 78.2%, 78.3%, 73.4%, respectively. FIB-4 PPV 92.6% followed by NFS (81.5%), (58.5%), BARD (54.5%). diagnostic accuracies relatively similar having (AUROC 0.80). Using scores, could avoided 63% Subgroup analyses only revealed results. Conclusion: Our results original literature suggest that may be safely Overall, especially appear more reliable excluding versus it. Albeit quite high suggesting they best models prediction fibrosis.Table 1: Statistical Analysis