Propensity for home death among Taiwanese cancer decedents in 2001-2006, determined by services received at end of life

Context: The discrepancy between patients' preferred and actual place of death highlights the dilemma inherent in achieving their preferences for home death. Research on determinants of home death has been limited largely by focusing on individual-level factors and somewhat on health care resources at the primary hospital and regional levels. Objectives: To investigate factors associated with home death, specifically, services received by cancer patients at the end of life (EOL). Methods: This was a retrospective cohort study using administrative data from 201,201 Taiwanese cancer decedents in the period 2001-2006. Results: Rates of home death decreased significantly over time (from 35.67% to 32.39%). Dying at home was associated with patient demographics (gender, age, and marital status) and disease characteristics (cancer type, metastatic status, postdiagnosis survival time, and comorbidity level). Taiwanese cancer patients were less likely to die at home if they received care from a medical oncologist and in hospitals or regions with abundant health care resources. Furthermore, Taiwanese cancer patients were less likely to die at home if they used life-sustaining treatments (intensive care unit care, cardiopulmonary resuscitation, intubation, and mechanical ventilation) in the last month of life. However, multiple emergency room visits in the last month of life and receiving hospice care increased Taiwanese cancer patients' propensity to die at home. Conclusion: Despite the causal ambiguity in interpreting our research findings, they indicate that using life-sustaining treatments at EOL not only exacts a substantial toll from patients, family members, and society, but also decreases the likelihood of dying at home.