OBJECTIVE:  The aim of this study was to validate the FibroScan system compared with liver histology and serum markers for the diagnosis of hepatic fibrosis. We also tried to determine the cut‐off levels and assess the feasibility of using FibroScan values to predict the fibrosis stage. METHODS:  In 44 patients with HCV infection, liver stiffness was evaluated by FibroScan, serum fibrosis markers and a liver biopsy. Associations between these indices were also analyzed. RESULTS:  FibroScan values showed a good correlation with serum levels of type IV collagen, hyaluronic acid and procollagen‐III‐peptide, and with the platelet count. Compared with liver histology, the FibroScan values increased proportionally with the progression of the histological fibrosis stage. Advanced fibrosis (F3 or F4) could be efficiently predicted by a FibroScan cut‐off value of 15 kPa. The FibroScan sensitivity, specificity, positive predictive value, negative predictive value and accuracy were 100%, 73.9%, 77.8%, 100%, and 86.4%, respectively. CONCLUSION:  FibroScan values gave a good correlation with various markers of fibrosis and increased proportionally with the progression of the hepatic fibrosis stage. A FibroScan value of 15 kPa was found to be a significant separation limit for differentiating advanced fibrosis stages (F3 and F4) from the milder stages (F0–F2). FibroScan values are clinically useful for predicting the fibrosis stages and helpful in managing interferon therapy in patients with chronic hepatitis C.

To assess the total daily inulin dose (TDD) and contribution of basal insulin to TDD and to identify the predictive factors for insulin requirement profiles in subjects with type 2 diabetes, we retrospectively examined insulin requirement profiles of 275 hospitalized subjects treated with basal-bolus insulin therapy (BBT) (mean age, 60.1 ± 12.9 years; HbA1c, 10.2 ± 4.5%). Target plasma glucose level was set between 80 and 129 mg/dL before breakfast and between 80 and 179 mg/dL at 2-hour after each meal without causing hypoglycemia. We also analyzed the relationship between the insulin requirement profiles (TDD and basal/total daily insulin ratio [B/TD ratio]) and insulin-associated clinical parameters. The mean TDD was 0.463 ± 0.190 unit/kg/day (range, 0.16-1.13 unit/kg/day). The mean B/TD ratio was 0.300 ± 0.099 (range, 0.091-0.667). A positive correlation of TDD with B/TD ratio was revealed by linear regression analysis (r=0.129, p=0.03). Stepwise multiple regression analysis identified post-breakfast glucose levels before titrating insulin as an independent determinant of the insulin requirement profile [Std β (standard regression coefficient) = 0.228, p<0.01 for TDD, Std β = -0.189, p<0.01 for B/TD ratio]. The TDD was <0.6 unit/kg/day and the B/TD ratio was <0.4 in the majority (70.2%) of subjects in the present study. These findings may have relevance in improving glycemic control and decreasing the risk of hypoglycemia and weight gain in subjects with type 2 diabetes treated with BBT.