Attempt to discriminate patients of non-A, non-B hepatitis as to their responsiveness to interferon treatment using multivariate analysis

Abstract

We analyzed the characteristics of the 25 patients of non-A, non-B hepatitis according to their responsiveness to interferon treatment using multivariate analysis. Patients were initially analyzed by univariate analysis (t-test, chi-square test) as to 15 variables. From the results of the univariate analysis, we applied 5 variables; age, pretreatment serum of 2-5AS activity, peak 2-5AS after initiation of interferon treatment, increment ratio of 2-5AS and pretreatment serum level of GPT for discriminant analysis. We performed discriminant analysis and computed the discriminant function to discriminate the patients. 18 out of 19 patients (94.7%) were discriminated properly. Discriminant analysis were considered useful to assess the interferon treatment. Next to Age, 2-5AS increment ratio contributed to the discrimination. So patients' reactivity to extermally administered interferon was considered important.