PURPOSE:A randomized controlled trial of cancer patients has linked early supportive care with improved hospice use and less-aggressive end-of-life care. In practice, the early use of supportive interventions and potential impact on end-of-life care are poorly understood. We sought to describe early use of medications to treat common breast cancer symptoms (pain, insomnia, anxiety, and depression) and to assess the relationship between early use of these treatments and end-of-life care.METHODS:Secondary analysis of 2006-2012 SEER-Medicare data was performed. Women included had stage IV breast cancer and died within the observation period. We used modified Poisson regression to assess the relationship between supportive medication use in the 90 days post-diagnosis and several end-of-life care measures (hospice use, in-hospital death, chemotherapy receipt within 14 days of death, ICU admission, or >1 hospitalization or emergency department/ED visit 30 days before death).RESULTS:Among the 947 women included, 68 % of women used at least one supportive medication in the 90 days following their diagnosis: 60.3 % used opioid pain medications and 28.3 % received non-opioid psychotropic medications. Early use of any supportive medications was not associated with end-of-life care. Similarly, we found no differences in end-of-life care between opioid pain medication users and non-users. However, we found that non-opioid psychotropic medication users were less likely to receive chemotherapy in the last 14 days of life (aRR 0.33, 95 % CI 0.12-0.88).CONCLUSIONS:Non-opioid psychotropic use was associated with some aspects of end-of-life care. Future research should consider alternative measures of palliative and supportive care use using administrative data sources.