The diagnosis of non-alcoholic steatohepatitis (NASH) requires liver biopsy. Patients with NASH are at risk of progression to advanced fibrosis and hepatocellular carcinoma. A reliable non-invasive tool for the detection of NASH is needed. We aimed at developing a tool to diagnose NASH based on a predictive model including routine clinical and transient hepatic elastography (TE) data. All subjects undergoing elective cholecystectomy in our center were invited to participate, if alcohol intake was < 30 g/d for men and < 15 g/d for women. TE with controlled attenuation parameter (CAP) was obtained before surgery. A liver biopsy was taken during surgery. Multivariate logistic regression models to predict NASH were constructed with the first 100 patients, the elaboration group, and the results were validated in the next pre-planned 50 patients. Overall, 155 patients underwent liver biopsy. In the elaboration group, independent predictors of NASH were CAP value [adjusted OR (AOR) 1.024, 95% confidence interval (95% CI) 1.002-1.046, p = 0.030] and HOMA value (AOR 1.847, 95% CI 1.203-2.835, p < 0.001). An index derived from the logistic regression equation to identify NASH was designated as the CAP-insulin resistance (CIR) score. The area under the receiver operating characteristic curve (95%CI) of the CIR score was 0.93 (0.87-0.99). Positive (PPV) and negative predictive values (NPV) of the CIR score were 82% and 91%, respectively. In the validation set, PPV was 83% and NPV was 88%. In conclusion, the CIR score, a simple index based on CAP and HOMA, can reliably identify patients with and without NASH.
© 2022. The Author(s).