Reliability and predictive validity of a hepatitis-related symptom inventory in HIV-infected individuals referred for Hepatitis C treatment

BACKGROUND: We aimed to determine the reliability and validity of a hepatitis symptom inventory and to identify predictors of hepatitis C (HCV) treatment initiation in a cohort of HIV-infected patients. METHODS: Prospective clinic based study that enrolled patients referred for HCV therapy consideration. A hepatitis symptom inventory and the Center for Epidemiologic Studies Depression Scale (CES-D) were administered to HIV/HCV individuals. The symptom inventory was factor analyzed and subscale reliability estimated with Cronbach's alpha. Predictive validity was evaluated using generalized estimating equations (GEE). Predictors of HCV treatment were identified using logistic regression. RESULTS: Between April 2008 to July 2010, 126 HIV/HCV co-infected patients were enrolled in the study. Factor analysis using data from 126 patients yielded a three-factor structure explaining 60% of the variance for the inventory. Factor 1 (neuropsychiatric symptoms) had 14 items, factor 2 (somatic symptoms) had eleven items, and factor 3 (sleep symptoms) had two items, explaining 28%, 22% and 11% of the variance, respectively. The three factor subscales demonstrated high intrinsic consistency reliability. GEE modeling of the 32 patients who initiated HCV therapy showed that patients developed worsening neuropsychiatric and somatic symptoms following HCV therapy with stable sleep symptoms. Bivariate analyses identified the following as predictors of HCV therapy initiation: lower HIV log(10 )RNA, lower scores for neuropsychiatric, somatic and sleep symptoms, lower CES-D scores and white ethnicity. In stepwise multiple logistic regression analysis, low neuropsychiatric symptom score was the strongest independent predictor of HCV therapy initiation and HIV log(10 )RNA was inversely associated with a decision to initiate HCV treatment. CONCLUSIONS: A 41-item hepatitis-related symptom inventory was found to have a clinically meaningful 3-factor structure with excellent internal consistency reliability and predictive validity. In adjusted analysis, low neuropsychiatric symptom scores and controlled HIV infection were independent predictors of HCV treatment initiation. The usefulness of the HCV symptom inventory in monitoring HCV treatment should be evaluated prospectively.