Study of Trends and Outcomes of Prescription Medications Continued During Hospice Care (STOP Med)

Hospice was founded with the primary goal of eliminating aggressive treatment measures and providing holistic, end of life palliative care aimed at relieving pain and symptoms and maximizing quality of life. This includes the discontinuation of limited benefit medications (LBMs) which may no longer provide a patient benefit in the context of limited life expectancy. The continued use of LBMs in the hospice population was poorly characterized in the literature, and evidence was needed to inform providers and policy makers regarding the scope of suboptimal medication use in this vulnerable patient group. This dissertation aimed to broadly characterize LBM use and continuation and identify factors associated with LBM continuation after admission among patients admitted to hospice for cancer and non-cancer causes. Our research aims were accomplished via a series of retrospective cross-sectional and cohort studies in the Medicare hospice population using patient-level administrative claims data from the Surveillance, Epidemiology and End Results (SEER)-Medicare linked database between 2007 and 2013. LBMs for the purposes of this research consisted of anti-hyperlipidemics, antihypertensives, oral antidiabetics, antiplatelets, anti-dementia and anti-osteoporotic medications, and proton pump inhibitors, with medication use measured via Medicare Part D dispensing claims. We found that the prevalence of LBM use after hospice admission was high, differed by primary hospice admission diagnosis and LBM class, and was relatively stable among patient cohorts defined by hospice admission year. After hospice enrollment, approximately 30% of patients continued at least one LBM that had been used prior to hospice admission. Anti-dementia medications were the most frequently continued LBM class while anti-osteoporotic medications were continued least often. In adjusted analysis, the likelihood of continuing at least one LBM after hospice admission was significantly greater in older patients, patients admitted to hospice in a nursing or assisted-living facility (compared to home hospice), and patients with longer hospice stays. Patients experiencing a care transition from a hospice to a non-hospice setting after hospice admission, particularly those experiencing hospitalizations, were also at an increased risk of continuing LBMs. Despite hospice care principles, use and continuation of LBMs in the hospice population is common. This work provides critical evidence on the scope of problematic medication use at the end of life and will ideally be used to promote the development of tailored medication management strategies that minimize burden without impacting quality of life during the patient’s final weeks and months. Until then, a more careful review of patients’ drug regimens at the time of hospice enrollment is warranted, including thoughtful consideration of the need to continue each medication in the context of its risks, benefits and the patient’s preferences and care goals.