Liver fibrosis risk, missed follow-up, and inequitable specialist referral in the All of Us Research Program: a retrospective cohort study

Background Metabolic dysfunction–associated steatotic liver disease (MASLD) is highly prevalent and frequently undiagnosed, despite liver fibrosis being the strongest predictor of liver-related morbidity and mortality. We aimed to quantify Fibrosis-4 (FIB-4) index-based fibrosis risk and to evaluate guideline-recommended follow-up—FIB-4 documentation, confirmatory testing, and specialist referral—in a large US cohort. Methods We conducted a retrospective study of electronic health record data from the US All of Us Research Program. Among adults aged 35 years or older with sufficient laboratory data, FIB-4 scores were calculated from routinely captured biomarkers, classified as low (<1.3), indeterminate (1.3–2.67; 2.0–2.67 if aged ≥65 years), or high risk (>2.67), and described across sociodemographic and metabolic subgroups; associations with incident major adverse liver outcomes (MALO) were assessed with Cox proportional-hazards regression. We examined FIB-4 documentation, confirmatory testing, and specialty referrals within 365 days of the index FIB-4 date. Findings Among 136,004 participants (mean age 60.8 years [SD 13.5]; 79,365 [58.4%] female; 81,420 [59.9%] White, 21,779 [16.0%] Hispanic or Latino, 20,076 [14.8%] Black), 26,971 (19.8%, 95% CI 19.6–20.0) had indeterminate and 11,217 (8.2%, 8.1–8.4) high FIB-4–based fibrosis risk. Elevated risk was more frequent with cardiovascular disease (41.7% vs 25.6%), hypertension (30.7% vs 23.4%), and type 2 diabetes (29.7% vs 27.4%), rising from 25.6% with no metabolic comorbidities to 36.6% with four. FIB-4 documentation was rare, and fewer than 1% of high-risk participants underwent confirmatory testing or referral; referral was low across risk categories (0.19% high, 0.19% indeterminate, 0.25% low risk). MALO occurred in 1308 (11.7%) high-risk vs 830 (0.8%) low-risk participants (adjusted hazard ratio 19.68, 95% CI 17.88–21.67). Interpretation Routinely collected laboratory data identify substantial liver fibrosis risk in this volunteer US cohort, yet guideline-recommended follow-up is uncommon. Implementation strategies such as automated risk calculation and reflex-based testing pathways could reduce missed opportunities for earlier diagnosis and intervention. Funding No specific funding was received for this study.