Preliminary medication adherence data over one year in a longitudinal cohort of post-menopausal women with early-stage breast cancer initiating aromatase inhibitors

BACKGROUND: In 2005, aromatase inhibitors (AIs) became standard care for postmenopausal early-stage breast cancer patients to prevent recurrence. AIs are taken orally daily for 3-5 years. There is concern about nonadherence to AIs in clinical populations. Our objectives are to measure AI adherence prospectively and better understand AI adherence predictors. METHODS: To date we collected data from 72 postmenopausal female oncology outpatients within 10 days of AI initiation, via self-report by paper surveys, completed every 2 weeks for 12 weeks, and at 52 weeks. We assess also clinical and sociodemographic characteristics, side effects, physical function, depression, and menopausal symptoms. Adherence is assessed via the Morisky Medication Adherence-Hormonal scale with scores ranging from 0-8. Scores <6 denote low adherence and =8 denote high adherence. We conducted univariate analyses and multivariable logistic regression modeling with robust standard errors. RESULTS: Seven women (10%) were classified as having low adherence, 17(24%) medium adherence, and 48(67%) high adherence over the 52-week period. In univariate analyses, we observed no association between adherence and any other factor measured. In the multivariable analysis, baseline income (p<0.01), physical function (p=0.04), and depression (p=0.04), and longitudinal out-of-pocket costs (p<.01) and menopausal symptoms (p=0.04) were associated with low adherence. DISCUSSION: If only two thirds of women taking AIs are adherent, reduced clinical effectiveness of AIs could lead to recurrence and mortality in cancer populations. In this cohort, we plan in future analyses to further investigate factors associated with adherence and to compare self-reported adherence with electronic monitoring and pill count data.