Modelling the cost effectiveness of non-alcoholic fatty liver disease risk stratification strategies in the community setting

Background Non-alcoholic fatty liver disease (NAFLD) is highly prevalent worldwide. Identifying high-risk patients is critical to best utilize limited health care resources. We established a community-based care pathway using 2D ultrasound shear wave elastography (SWE) to identify high risk patients with NAFLD. Our objective was to assess the cost-effectiveness of various non-invasive strategies to correctly identify high-risk patients. Methods A decision-analytic model was created using a payer's perspective for a hypothetical patient with NAFLD. FIB-4 [>= 1.3], NAFLD fibrosis score (NFS) [>= -1.455], SWE [>= 8 kPa], transient elastography (TE) [>= 8 kPa], and sequential strategies with FIB-4 or NFS followed by either SWE or TE were compared to identify patients with either significant (>= F2) or advanced fibrosis (>= F3). Model inputs were obtained from local data and published literature. The cost/correct diagnosis of advanced NAFLD was obtained and univariate sensitivity analysis was performed. Results For >= F2 fibrosis, FIB-4/SWE cost $148.75/correct diagnosis while SWE cost $276.42/correct diagnosis, identifying 84% of patients correctly. For >= F3 fibrosis, using FIB-4/SWE correctly identified 92% of diagnoses and dominated all other strategies. The ranking of strategies was unchanged when stratified by normal or abnormal ALT. For >= F3 fibrosis, the cost/correct diagnosis was less in the normal ALT group. Conclusions SWE based strategies were the most cost effective for diagnosing >= F2 fibrosis. For >= F3 fibrosis, FIB-4 followed by SWE was the most effective and least costly strategy. Further evaluation of the timing of repeating non-invasive strategies are required to enhance the cost-effective management of NAFLD.