Screening for Advanced Liver Fibrosis: Comparing Serum Indices With Elastography in a Nationally Representative U.S. Cohort.

INTRODUCTION: Metabolic dysfunction-associated steatotic liver disease is common and often progresses silently to advanced fibrosis, necessitating reliable, scalable screening strategies. We compared widely used serum fibrosis indices against vibration-controlled transient elastography (VCTE) measured liver stiffness in a nationally representative sample of U.S. adults. MATERIALS AND METHODS: We performed a cross-sectional analysis of National Health and Nutrition Examination Survey 2017-2018 participants aged ≥20 years with valid VCTE data. Participants with high alcohol intake (>30 g/day for men; >20 g/day for women), positive for hepatitis C, hepatitis B core antibody, or other chronic liver disease were excluded. We calculated values for aspartate aminotransferase-to-platelet ratio index (APRI), the body mass index, aspartate aminotransferase/alanine aminotransferase ratio, and diabetes score (BARD), Fibrosis-4 (FIB-4), Forns, and NAFLD fibrosis score (NFS) using published formulas and applied standard cutoffs (BARD ≥ 2; NFS ≥ 0.67; FIB-4 > 3.25; APRI > 1.50; Forns > 6.9). Advanced fibrosis was defined as median liver stiffness ≥9.7 kPa. Weighted analyses accounted for the complex survey design. Diagnostic performance metrics included positive and negative predictive values (PPV and NPV, respectively), false negative rate, and diagnostic odds ratio (DOR). RESULTS: Among 3,797 participants (weighted to >180 million U.S. adults), 5.5% (n = 251; 95% CI, 4.4-7.0) had VCTE-consistent advanced fibrosis (F3/4). All serum indices except BARD were associated with advanced fibrosis (BARD P = .22). NPVs for all tests (93.5%-94.8%) approximated the baseline NPV assuming no disease (93.4%). APRI yielded the highest PPV (62.5%) and highest DOR (24), followed by FIB-4 (PPV, 34.9%). BARD produced the lowest PPV (6.8%) and a DOR (1.2). CONCLUSIONS: In a nationally representative U.S. cohort, commonly used serum fibrosis indices demonstrate limited accuracy for detecting liver stiffness associated with advanced fibrosis. Among commonly recommended first-line indices, FIB-4 and NFS offer the best balance of accessibility and performance but have nontrivial false negative rates, and the BARD demonstrated minimal discriminatory value in this population. Further research is needed to develop a more accurate predictive model for advanced fibrosis.