Hepatitis B validity of drug users' self-report.

Abstract

Although drug users' self-report has provided data for much of the published literature about drug use, little is known about self-report validity when participants are asked about diseases that are associated with drug abuse, such as hepatitis. Injecting drug users and crack cocaine smokers (N = 659) were recruited in Anchorage, Alaska, and asked whether they had been diagnosed previously with hepatitis B. These self-report data were compared to various hepatitis B and C seromarkers as measures of validity of self-report expressed as sensitivity and specificity. Results indicate that although test-retest reliability for self-report is high (.905) and specificity is high (96.06% for hepatitis C virus, or HCV), sensitivity is low (23.74% for HCV). Thus, because of its low sensitivity, self-report of hepatitis should be used only as a prevalence estimate lower bound. More than half of the drug users who had contracted hepatitis had never been told that they were infected.