Self-reported hepatitis B and C virus infections had low sensitivity among HIV-infected patients

Abstract Objective To determine the validity of self-reported hepatitis B virus (HBV) and hepatitis C virus (HCV) in HIV-infected injection drug users (IDUs) vs. nonIDUs. Methods A cross-sectional study was performed among HIV-infected IDUs and nonIDUs in the Penn Center for AIDS Research (CFAR) Database. Self-reported past HBV, HCV, and serostatus were obtained from the CFAR Database. Results Among 970 subjects (798 nonIDUs; 172 IDUs), there was no difference in sensitivity of self-reported HBV between nonIDUs (27% [95/346]; 95% CI, 23%–32%) and IDUs (26% [31/117]; 95% CI, 19%–35%; P > 0.5), but specificity was greater among nonIDUs (96% [360/374; 95% CI, 94%–98%] vs. 78% [28/36; 95% CI, 61%–90%]; P < 0.001). Sensitivity of self-reported HCV was greater among IDUs (78% [101/130; 95% CI, 70%–85%] vs. 62% [47/76; 95% CI, 50%–73%]; P = 0.02), but there was no difference in specificity (97% [626/643]; 95% CI, 96%–98% for nonIDUs vs. 93% [26/28]; 95% CI, 76%–99%] for IDUs; P = 0.2). Conclusions The sensitivity of self-reported HBV and HCV compared to actual serostatus are not sufficiently high enough to warrant their use to estimate the prevalence and incidence of these infections.