Stepwise combination of simple noninvasive fibrosis scoring systems increases diagnostic accuracy in nonalcoholic fatty liver disease

Abstract

Objective: Nonalcoholic fatty liver disease (NAFLD) is a common chronic liver disease ranging from simple fatty liver to steatohepatitis, fibrosis, and cirrhosis. We aimed to analyze the diagnostic performance and clinical utility of simple noninvasive tests alone or in combination for the detection of advanced fibrosis in patients with NAFLD.

Design and subjects: Data from 323 patients with biopsy-proven NAFLD/NASH who presented to the Clinic for Gastroenterology and Hepatology, University Hospital of Cologne between July 1998 and November 2009, were analyzed retrospectively. Sensitivity, specificity, positive predictive values, and negative predictive values were determined along with the area under receiver operating characteristic curves (AUROC) using published formulas for NAFLD, FIB-4, and BARD fibrosis scores.

Results: The area under receiver operating characteristic curves were as follows: NAFLD fibrosis score 0.96 [95% confidence interval (CI), 0.92-0.99], FIB-4 0.95 (95% CI, 0.91-1.00), BARD 0.82 (95% CI, 0.71-0.92) with negative predictive values for advanced fibrosis of 96%, 98%, and 96%, respectively. When applying the NAFLD, FIB-4, or BARD scoring systems 25%, 15%, or 26% of cases with advanced fibrosis would have been missed. Combining FIB-4 and BARD in a stepwise fashion, patients would have been correctly classified without biopsy in 67% of cases without missing a single case of advanced fibrosis.

Conclusions: The FIB-4 and NAFLD fibrosis scores perform better than the BARD scoring system. Liver biopsy can securely be replaced only with a stepwise combination of simple noninvasive tests, otherwise the assessment of risk due to advanced fibrosis may be misleading in a clinically meaningful proportion of patients.