Impact of advance care planning on the place of death for cancer patients: A retrospective cohort study.

e24007 Background: Place of death is an important indicator of quality of end-of-life care (EOLC). Most patients do not want to die in hospital and hospital deaths are costly to healthcare organisations. Advance care planning (ACP) includes discussions about where patients want to die. Cancer patients often have a distinct illness trajectory with a clear terminal phase, enabling opportunities for EOLC planning. This work aims to explore the impact of ACP on the outcomes and place of death for cancer patients. Methods: A retrospective decedent cohort study of 10,727 cancer patients with a Coordinate My Care (CMC) record created 2011-2019. CMC is a digital platform that enables patients to record, revise and share their ACP and EOLC preferences with primary, secondary and urgent healthcare providers in London. Logistic regression analysis was applied to examine the relationship between in-hospital death and a range of individual characteristics and EOLC preferences. Results: Only 1% chose hospital as their preferred place of death. Overall, 71% died in their place of preference. Home (42%) was the most common place of death, while 20% died in hospital. The likelihood in-hospital death is positively associated with being older (OR = 1.12, p < 0.1), hospital being the preferred place of care (OR = 2.56, p < 0.001), hospital being the preferred place of death (OR = 2.85, p < 0.001) and if the preferred place of death is not recorded (OR = 1.32, p < 0.001). By contrast, patients who are severely frail (OR = 0.61, p < 0.001), who chose not to be resuscitated (OR = 0.56, p < 0.001) and who have a ceiling of treatment for symptomatic rather than intensive hospital based treatments (OR = 0.40, p < 0.001) are less likely to die in hospital. Conclusions: Most cancer patients prefer out-of-hospital death. Discussing, recording and sharing individual’s EOL preferences leads to better quality EOLC. Patient characteristics influence in-hospital deaths, but EOLC preferences exert stronger impact. Enabling cancer patients to die outside of hospital has positive implications for patients and their families and supports cost-effective use of healthcare resources.