Article Text
Abstract
Cross-sectional and case-control studies were conducted in a US Army unit which had experienced a protracted outbreak of viral hepatitis. Serological, demographic, and exposure data were collected. The cross-sectional study found that there was no association between the prevalence of hepatitis B virus (HBV) infection and ethnicity, education, and rank. There was an association with age and duration of assignment to the unit. Having social contact with an identified hepatitis patient and the sharing of personal hygiene items with a case were significant risk factors for HBV infection. Through the use of anonymous questionnaires, the case-control study evaluated various behavioural factors. Univariate analysis indicated moderate but not significant increases in risk associated with a history of multiple sex partners and a history of disciplinary problems while in the military. Six drug-use related risk factors were significantly associated with HBV infection on univariate analysis. When evaluated with multivariate analysis, the only risk factor that was significantly associated with HBV infection was injection of drugs while assigned to the study unit. Two sources of misclassification bias were identified, both of which acted to underestimate the true risk associated with identified risk factors.