P0947 - Investigating Target Populations for NAFLD Screening: What Does a Diabetes-Focused Strategy Miss?

Brenna Sanders, MD, Danielle Lefebvre, MD, Andrew Schreiner, MD
Medical University of South Carolina, Charleston, SC

Introduction: Future NAFLD screening recommendations will need to ensure detection of advanced fibrosis, the best predictor of future severe liver outcomes. Recently proposed models put forth diabetes as an indicator for NAFLD screening given diabetes’ association with NAFLD and advanced fibrosis. We compared non-invasive high-risk advanced fibrosis assessments using the Fibrosis-4 Index (FIB-4) and the AST:Platelet Ratio Index (APRI) between patients with and without diabetes in a sample of primary care patients with NAFLD to determine if a diabetes-focused NAFLD diagnostic strategy would miss high-risk disease.

Methods: In this retrospective cohort study of primary care patients with NAFLD from 2012 to 2018, we calculated FIB-4 and APRI scores for each included patient and categorized scores by advanced fibrosis risk (low, indeterminate, and high). High-risk scores for advanced fibrosis were the primary outcome and the proportion of the overall cohort with a high-risk assessment was calculated. Univariate analysis of patient characteristics for the overall sample was performed, and variables were compared by a diagnosis of diabetes. Multivariable logistic regression models evaluated the association between diabetes and high-risk FIB-4 (> 2.67) and APRI (> 1.0) results, while adjusting for other patient variables. SAS 9.4 was used for statistical analyses.

Results: Of the 652 patients included in the cohort, 300 (46%) had a diagnosis of diabetes, 75 (11.5%) had a high-risk FIB-4, 67 (10.3%) had a high-risk APRI, and 49 (7.5%) patients had high-risk FIB-4 and APRI assessments. Similar proportions of patients with diabetes had high-risk FIB-4 (12.3% vs. 10.8%, p=0.54), APRI (9.7% vs. 10.8%, p=0.64), or both (7.7% vs. 7.4%, p=0.89) as patients without diabetes. The unadjusted logistic regression model demonstrated no significant association with diabetes and high-risk FIB-4 (OR 1.30, 95% CI 0.75-2.26), high-risk APRI (OR 0.83, 95% CI 0.48-1.44), or both (OR 1.01, 95%CI 0.53-1.91). After adjusting for demographic and clinical variables, diabetes had no significant association with high-risk FIB-4 or APRI outcomes (Table 1).

Discussion: A similar of proportion of patients with and without diabetes had high-risk advanced fibrosis assessments. Our data demonstrates that a diabetes-focused NAFLD screening strategy may miss patients with NAFLD at high-risk for advanced fibrosis.



Disclosures:


Brenna Sanders, MD, Danielle Lefebvre, MD, Andrew Schreiner, MD. P0947 - Investigating Target Populations for NAFLD Screening: What Does a Diabetes-Focused Strategy Miss?, ACG 2023 Annual Scientific Meeting Abstracts. Vancouver, BC, Canada: American College of Gastroenterology.