The relationship between adherence to adjuvant hormonal therapy and survival among low-income, insured women with primary breast cancer

Abstract

e11522 Background: Clinical trials and meta-analyses show that adjuvant hormonal therapy for hormone receptor positive breast cancer significantly decreases risk of death. We explored the relationship between adherence to adjuvant hormonal therapy and death in a low-income, Medicaid-insured population. Methods: Using a Medicaid claims-tumor registry linked database and National Death Index data (NDI), we evaluated adherence to adjuvant hormonal therapy [defined as >80% Medication Possession Ratio (MPR)] and mean six-year overall and cancer-specific survival by local versus regional stage for all female breast cancer diagnosed in years 2000–2002, in North Carolina. The Kaplan-Meier and Cox Proportional Hazards models were used to determine the role of adherence on cancer-specific survival. Models were adjusted for age, race, Charlson comorbidity score, number of prescription medications, type of surgery, use of radiation therapy, prior chemotherapy, hormone receptor status (positive or unknown). Results: The final sample consisted of 1,042 cases [ages range 29–97 years (mean 65.9 years; 56% Caucasian; mean Charlson comorbidity score 4.1 (SD 2.9); 680 local and 362 regional stage], of which 732 filled a prescription for adjuvant hormonal therapy within the year after breast cancer diagnosis. Filling a prescription for adjuvant hormonal therapy, versus not, was not significantly associated with cancer-related death: HR 1.04 (95% CI 0.66 - 1.64) overall; HR 0.75 (95% CI 0.39 - 1.43) for local stage and HR 1.01 (95% CI 0.51 - 2.00) for regional stage. However, adherence in the highest quartile (MPR>95) is associated with an increase in mortality risk. Conclusions: In this low income insured group of breast cancer patients, no statistically significant association was found between death rates and use of adjuvant hormonal therapy. However, an unexpected association between very high adherence and increase in mortality was found. This may reflect methodological limitations of claims data involving bias and unidentified patient risk. More research is needed to explore reasons for higher mortality among low-income women with high medication adherence. [Table: see text]