ABSTRACT
INTRODUCTION AND OBJECTIVES: Assessing fibrosis risk noninvasively is essential. The steatosis-associated fibrosis estimator (SAFE) score shows promise but needs validation. PATIENTS AND METHODS: This was a three-part study. In part 1, we compared the SAFE score with the Fibrosis-4 (FIB-4) and NAFLD fibrosis score (NFS) in the National Health and Nutrition Examination Survey (NHANES) cohort (2017-2020), using transient elastography (TE) as screening reference. In part 2, we examined patients who underwent liver biopsies at an Asian center between 2018 and 2020 to assess these models in various liver diseases. In part 3, the SAFE score was applied to adults in the NHANES cohort (1999-2016) to assess the correlation with mortality. RESULTS: In part 1, we studied 6,677 patients, comprising 595 screening positive (TE ≥8 kPa). SAFE (cutoff 100) displayed a lower proportion of false positives (10.4 %) than FIB-4 (cutoff 1.3) and NFS (cutoff -1.455) (22.1 % and 43.6 %) while retaining a low proportion of false negatives (5.5 %). In part 2, SAFE outperformed FIB-4 (P = 0.04) and NFS (P = 0.04) in staging significant fibrosis (≥S2) in NAFLD and had similar accuracies in other etiologies. In part 3, the FIB-4, NFS, and SAFE score were associated with all-cause mortality in the general population, with c-statistics of 0.738, 0.736, and 0.759, respectively. CONCLUSIONS: The SAFE score reduced futile referrals more effectively than FIB-4 without raising the missed TE ≥ 8 kPa rate. It correlated with all-cause mortality in the general population and excelled in staging significant fibrosis in NAFLD.