Medicare Part D low-income subsidy and disparities in breast cancer treatment.

Cost, Value, and Policy in Quality
Session Type and Session Title: 
General Poster Session A: Science of Quality and Cost, Value, and Policy in Quality
Oral Abstract Session: Plenary Abstracts
Abstract Number: 
J Clin Oncol 32, 2014 (suppl 30; abstr 2)
Alana Biggers, Joan Neuner, Elizabeth Smith, John A. Charlson, Liliana Pezzin, Purushottam Laud; University of Illinois at Chicago, Chicago, IL; Medical College of Wisconsin, Milwaukee, WI

Abstract Disclosures


Background: Breast cancer outcomes are worse among black than white women, but the role of income and out-of-pocket costs (OOPCs) in these disparities is understudied. The Medicare D program provided medication insurance for older women and also included a low-income subsidy (LIS) which eliminated or reduced OOPCs among women with low assets and limited income (based on federal poverty level). We examined differences in adherence to HT by race/ethnicity among a Medicare D population, hypothesizing that LIS might reduce racial disparities in HT adherence. Methods: With data collected from a national sample of women enrolled in Medicare Parts A, B, and D, we identified Medicare Part D enrollees ≥65 years diagnosed with breast cancer who underwent mastectomy or breast conserving surgery in 2006-07 and received either tamoxifen or an AI (anastrozole, letrozole, or exemestane) within one year of surgery. Nonadherence rates (medication possession rate of >0.80) were calculated by race and LIS status for each year after first fill up through December 2011. The association of race with HT adherence was examined in unadjusted Chi-square analyses and in regression models adjusted for age, comorbidity, chemotherapy use, and zip code level- income and education. All models utilized GEE to account for within-patient clustering. Results: Among a sample of 23,299 women (50.6% age 65-74, 40.9% age 75-84), 27.2% received LIS. LIS (but not AI use) varied substantially by race, so that 20.6% of white women and 69.7% of black women received the subsidy. In the first year of therapy, differences in adherence by race were statistically significant, but small (64.2% for white, 63.2% for black and 66.7% for Hispanic). Adherence dropped during years 2-3 of the study, but reductions were much smaller among LIS recipients. Results were confirmed in adjusted models. Conclusions: Enrollment in the Medicare D LIS was high among black and Hispanic breast cancer patients, and disparities in adherence to breast cancer HT among these women were small and remained so over three years. Our study offers important information about the role of medication subsidies and SES in adherence, and suggests their potential to reduce the breast cancer outcomes gap by race.