Purpose: To assess the factors influencing adherence in long-term medication use as exemplified by statins.
Methods: Data from an in-depth survey of Australians aged 45 years and over were linked to national prescription reimbursement data to assess medication possession ratio (MPR) for statins for the middle two years of a four-year period of statin possession. Poisson regression was used to calculate the relative risk (RR) for adherence (MPR ≥ 80%) for patient characteristics and factors related to access to and need for health care services. Separate models were fit for patients receiving healthcare concession subsidies and those who do not ('general beneficiaries').
Results: In the analysis, 42 492 concessional and 16 110 general beneficiary patients were included, with 80.1% and 56.7% showing MPR ≥ 80%, respectively. In both models, RR for adherence was significantly elevated for older (age 65+) and less healthy (worse self-rated health, pre-existing heart condition or obese) individuals, and for those who had private health insurance. Significantly lower RR (i.e. more non-adherence) was found for individuals reporting speaking a language other than English at home, who were smokers, were employed, and had higher levels of education, and for those who reported psychological distress. Income had no significant relationship with adherence, and the pattern of adherence by remoteness of area of residence was inconsistent.
Conclusions: Poor adherence in long-term use of statins is commonplace, but a number of key predictors-including age, language other than English spoken at home, smoking status and psychological distress-are readily assessable by the prescribing practice.
Keywords: epidemiologic determinants; medication adherence; pharmacoepidemiology; statins.
Copyright © 2013 John Wiley & Sons, Ltd.