Routine-Data Triage for MASLD Positive on Vibration-Controlled Transient Elastography after FIB-4: An Exploratory NHANES Pathway Comparison Study with Temporal Validation.

BACKGROUND/AIMS: Contemporary metabolic dysfunction-associated steatotic liver disease (MASLD) pathways use the fibrosis-4 index (FIB-4) before vibration-controlled transient elastography (VCTE), but FIB-4-based rules may miss adults with VCTE-defined increased liver stiffness. We quantified VCTE-positive burden across low-risk and indeterminate FIB-4 groups and evaluated an exploratory routine-data second-stage triage strategy in the National Health and Nutrition Examination Survey (NHANES).

METHODS

Adults with MASLD from NHANES 2017 to March 2020 formed the development cohort, and those from August 2021 to August 2023 formed the temporal validation cohort. MASLD required controlled attenuation parameter-defined steatosis, cardiometabolic risk, and exclusion of significant alcohol use and viral hepatitis. The primary endpoint was liver stiffness measurement (LSM) ≥8 kPa; LSM thresholds of 7, 10, and 12 kPa were tested. A regularized logistic model was developed among adults with FIB-4 ≤2.67. The enhanced hybrid pathway referred adults with FIB-4 >2.67 or, among those with FIB-4 ≤2.67, model probability ≥0.159561.

RESULTS

The development cohort included 3,004 adults, 503 (16.7%) with LSM ≥8 kPa; validation included 1,959 adults, 408 (20.8%) with LSM ≥8 kPa. In validation, LSM ≥8 kPa occurred in 221/1,228 with FIB-4 <1.3 and 111/473 with FIB-4 1.3 to 2.67. Standard FIB-4 referral detected 137/408 cases while referring 528 of 1,959 adults; the enhanced pathway detected 233 of 408 while referring 621 of 1,959. In low/intermediate FIB-4 validation adults, the model area under the receiver operating characteristic curve was 0.742.

CONCLUSIONS

Low-risk and indeterminate FIB-4 groups contained substantial VCTE-positive burden. This exploratory enhanced pathway identified more VCTE-positive adults with modestly higher referral burden, supporting prospective validation.