Preference and actual place-of-death in advanced cancer: prospective longitudinal study.

Objectives: To evaluate the preferred place-of-death (PPoD) among patients with advanced cancer over time, and the concordance between preferred and actual place-of-death.

Methods: Prospective cohort study. A total of 190 patients with advanced cancer and their caregivers (n=190) were interviewed every 3 months, from study enrolment to 12 months (M0, M1, M2, M3, M4). PPoD data were obtained under four different end-of-life scenarios: (1) severe clinical deterioration without further specification; (2) clinical deterioration suffering from severe symptoms; (3) clinical deterioration receiving home-based visits; and (4) clinical deterioration receiving home-based visits and suffering from severe symptoms.

Results: Home was the most common PPoD over time among patients in scenarios 1 (n=121, 63.7%; n=77, 68.8%; n=39, 57.4%; n=30, 62.5%; n=23, 60.5%) and 3 (n=147, 77.4%; n=87, 77.7%; n=48, 70.6%; n=36, 75.0%; n=30, 78.9%). PPoD in palliative care unit (PCU) and hospital were most frequent at baseline in scenario 2 (n=79, 41.6%; n=78, 41.1%), followed by hospital over time (n=61, 54.5%; n=45, 66.2%; n=35, 72.9%; n=28, 73.7%). During the curse of illness, 6.3% of patients change their PPoD in at least one of end-of-life scenario. About 49.7%, 30.6% and 19.7% of patients died in PCU, hospital and home, respectively. Living in rural area (OR=4.21), poor health self-perception (OR=4.49) and pain at the last days of life (OR=2.77) were associated with death in PPoD. The overall agreement between last preference and actual place-of-death was 51.0% (k=0.252).

Conclusion: Home death was not the preferred place for a large number of patients when this option was presented within a clinical context scenario. The PPoD and actual place-of-death were depending on the clinical situation.