ABSTRACT
BACKGROUND/AIMS: Several simple tests for hepatic fibrosis employ indirect markers. However, the efficacy of using direct and indirect serum markers to predict significant fibrosis in clinical practice is inconclusive. We analyzed the efficacy of a previously reported indirect marker of hepatic fibrosis - the aspartate aminotransferase to platelet ratio index (APRI) - in patients with nonalcoholic chronic liver diseases (CLDs). METHODS: A total of 134 patients who underwent a percutaneous liver biopsy with a final diagnosis of chronic hepatitis B (n=93), chronic hepatitis C (n=18), or nonalcoholic fatty liver disease (n=23) were enrolled. A single-blinded pathologist staged fibrosis from F0 to F4 according to the METAVIR system, with significant hepatic fibrosis defined as a METAVIR fibrosis score of > or =2. RESULTS: The mean area under the receiver operating characteristic curve (AUROC) of APRI for predicting significant fibrosis in nonalcoholic CLDs was 0.84 [95% confidence interval (CI), 0.78-0.91]. APRI yielded the highest mean AUROC in the patients with chronic hepatitis B (0.85; 95% CI, 0.771-0.926). The positive predictive value of APRI > or =1.5 for predicting significant fibrosis was 89%. The negative predictive value of APRI <0.5 for excluding significant fibrosis was 80%. CONCLUSIONS: APRI might be a simple and noninvasive index for predicting significant fibrosis in nonalcoholic CLDs.