• Risha Gidwani, Nina Joyce, Bruce Kinosian, Katherine Faricy-Anderson, Cari Levy, Susan C Miller, Mary Ersek, Todd Wagner, and Vincent Mor.
• 1 Health Economics Resource Center (HERC), Center for Innovation to Implementation (Ci2i) , VA Palo Alto Health Care System, Menlo Park, California.
• J Palliat Med. 2016 Sep 1; 19 (9): 957-63.

BackgroundSpecialty societies recommend patients with advanced cancer receive early exposure to palliative care and exposure to hospice care.ObjectiveWe sought to understand real-world practice of care, specifically, the timing of palliative care, and how timing and duration of hospice care varied across Medicare, VA, and VA-Purchased care.DesignWe conducted a retrospective analysis of administrative data for veterans aged 65 years or older who died with cancer in 2012. Multilevel logistic regression was used to evaluate the likelihood of receiving palliative care, receiving hospice care, and receiving hospice care for at least three days.SettingMedicare, VA, and VA-Purchased care environments.MeasurementsThe receipt and timing of palliative care within VA and the receipt and timing of hospice care across three healthcare environments.ResultsMost veterans received hospice care (71%), whereas fewer received palliative care (52%). Among all cancer decedents, 59% received hospice care for their last three days of life. Patients who received hospice care did so a median of 20 days before death (interquartile range [IQR]: 7-46). Patients who received palliative care did so a median of 38 days before death (IQR: 13-94). Adjusted analyses revealed significant differences in receipt of palliative care across cancer type, and significant differences in receipt of hospice care across cancer type. After adjusting for age and cancer type, patients who received VA hospice care were significantly less likely to receive it for at least three days compared with patients who received VA-Purchased or Medicare hospice care.ConclusionsThere remains a gap between recommended timing of supportive services and real-world practice of care. Results suggest that difficulties in prognosticating death are not fully responsible for underexposure to hospice.

8 keywords...

hide…