Breast cancer treatment and survival outcomes in relation to age and care settings

Abstract

6583 Background: Breast cancer treatment and outcomes have been examined by age, but not by care setting - a construct that reflects patients' level of dependence that may not be captured through comorbidity measures. This study aims at filling this knowledge gap. Methods: Using the Cancer-Aging Linked Database (CALD), developed by linking the Ohio Cancer Incidence Surveillance System, Medicare and Medicaid files, home health care Outcome Assessment Information Set, long-term care Minimum Data Set, death certificates, and U.S. Census data, we identified women ≥65 years old diagnosed with local or regional stage breast cancer from July 2000 through December 2001, with deaths through December 2005. Multivariable logistic regression and Cox-proportional hazards models were developed to assess the association between age (age ≥ 80 versus younger), care settings (community dwelling (CD), home health care (HHC), or nursing home care (NHC)), and outcomes of interest after adjusting for patient covariates. Results: Of 3,050 patients, 33% were ≥80 years, 6% received HHC, and 3% received NHC. Patients who were less likely to receive lumpectomy with radiotherapy (versus mastectomy) were older, (versus younger, Adjusted odds ratio (AOR) = 0.54, 95% confidence interval (CI) =0.44–0.65); received NHC (versus CD, AOR = 0.24, CI=0.10–0.58); had regional disease (versus local, AOR=0.25, CI=0.20–0.31); and had negative hormone-receptor status (versus positive, AOR=0.50, CI=0.36–0.67). Older patients and those who received NHC were nearly twice as likely as their younger and community dwelling counterparts, respectively, to undergo lumpectomy without radiotherapy. Receipt of lumpectomy with radiotherapy was associated with favorable breast cancer survival. Conclusions: Careful evaluation to identify the oldest old, and nursing home patients who can and do receive lumpectomy with radiotherapy may improve breast cancer survival in this vulnerable population. No significant financial relationships to disclose.