Factors associated with poor adherence to antiviral treatment for hepatitis B

Antiviral therapy for hepatitis B is effective and reduces the risk of progression to cirrhosis and liver cancer but is often required for an indefinite duration. Treatment adherence is important to prevent the development of resistance and optimize outcomes. Pharmacy adherence measures can be used to assess treatment adherence, with the medication possession ratio being less susceptible to bias than physician- or self-reported adherence. The aim of this study was to measure adherence in public hospital outpatients over a 3-year period and to examine factors associated with nonadherence. A retrospective study of pharmacy records of patients dispensed antiviral therapy for hepatitis B from four major hospitals in Melbourne between 2010 and 2013. Hospital record numbers were linked with and de-identified demographic information including age, sex, Indigenous status, country of birth, interpreter requirement, spoken language and postcode of residence. The medication possession ratio was the outcome measure with poor adherence defined <.90. Univariate logistic regression and multivariate logistic regression were performed to examine associations with nonadherence. Records of 1026 patients were included in the analysis. Twenty per cent of all participants met the definition of poor adherence. Significant factors affecting adherence included age <35 years (P=.002), hospital site and treatment by multiple doctors within shorter time periods. This is the largest study examining detailed factors associated with adherence to hepatitis B treatment. Understanding poor adherence in clinical settings, and the factors associated with lower adherence, is important to inform efforts towards promoting treatment adherence for hepatitis B.