FIB-4 and APRI for cirrhosis detection in a privately insured national cohort.

Background and Aims: Non-invasive laboratory-based fibrosis indices have been proposed as a tool for population-based screening for advanced fibrosis. We aimed to examine the performance of fibrosis indices at the time of and prior to cirrhosis diagnosis. Methods: We included adult patients with cirrhosis diagnosis codes in a privately insured database (Optum) from 2010-2018 with 1:4 birth year-matched controls without cirrhosis diagnosis codes. We analyzed aspartate aminotransferase-to-platelet ratio index (APRI), and fibrosis-4 index (FIB-4) up to 30 months prior to the entry of cirrhosis diagnosis codes. Cut-offs of <1 and ≥2 were used for APRI and <1.3 and ≥2.67 were used for FIB-4. Results: We included 10,650 patients with cirrhosis (median age 62 years), who were predominantly white (57.8%) and male (51.9%). The most common etiologies of cirrhosis were non-alcoholic steatohepatitis (23.8%), hepatitis C (23.0%), and alcohol-related liver disease (20.5%). At the time of cirrhosis diagnosis (±3 months), 9.3% of patients with cirrhosis had APRI ≥2 and 41.3% had a FIB-4 ≥2.67 compared to 1.2% and 8.9% in control patients, respectively. In the periods spanning 3-12, 12-21, and 21-30 months prior to cirrhosis diagnosis, APRI was ≥2 in 9.4%, 6.6%, and 6.5% of patients, respectively; FIB-4 was ≥2.67 in 42.1%, 37.1%, and 34.3% of patients, respectively. The sensitivity and specificity of APRI at the time of cirrhosis diagnosis were 9.3% and 98.8%, respectively, while the sensitivity and specificity of FIB-4 were 41.3% and 91.0%, respectively. Lower cut-off values for APRI and FIB-4 showed similar performance. Conclusions: Existing non-invasive fibrosis makers are suboptimal when used for advanced fibrosis identification, missing over half of patients with cirrhosis at the time of and prior to clinical diagnosis. Impact and implications: Commonly available laboratory-based indices, including APRI and FIB-4, have been proposed to rule in or rule out advanced fibrosis in the general population. In a study of a large privately insured cohort from the US, FIB-4 and APRI were not sufficient for screening for advanced fibrosis at the time of or prior to clinical diagnosis. While performance for screening out advanced fibrosis was better, a significant percentage of patients with cirrhosis have lab indices below threshold values. Future studies to develop laboratory-based algorithms to help stratify liver fibrosis for population-based screening are warranted.