A Good Death: Facilitating End-of-life Care Decisions That Honor Patients' Wishes

Cancer has been the leading cause of death in Taiwan for almost 3 decades, but the quality of care received by terminally ill cancer patients at end of life (EOL) is poorly understood due to insufficient and limited-scope research. To improve patient care, the Institute of Medicine (IOM) calls for health systems to be more responsive to patients’ wishes and preferences. Our previous research demonstrated great discrepancies between patients’ preferred prognostic information and actual prognostic disclosure as well as discrepancies between preferences for EOL care and that received by Taiwanese cancer decedents. Such great discrepancies are detrimental to the well-being of both terminally ill cancer patients and their family caregivers as shown by our own research results and the literature. However, our previous nationwide survey used a cross-sectional design, which cannot capture dynamic changes in prognostic awareness and preferences for EOL care in response to worsening physical symptoms and possibly diminished family resources as the disease progresses and death approaches. Therefore, for this 3-year, study the applicant proposes to: (1) longitudinally investigate prognostic awareness and preferences for EOL care in dyads of terminally ill cancer patients and their primary family caregivers, (2) examine the congruence between patients’ preferred and actual EOL care, and (3) evaluate the impact of prognostic awareness, EOL care preferences, and congruence between patients’ preferred and actual EOL care on patients’ psychological well-being, family caregivers’ caregiving burden and bereavement outcomes, and healthcare-resources utilization at the end of life. Aims 1-3 will be addressed by surveying a convenience sample of 298 dyads of terminally ill cancer patients and their family caregivers. Well-established instruments will be used to measure preferences for prognostic information and EOL care, confounding factors (symptom distress, functional dependency, and social support), and outcome variables (QOL, depressive symptoms, satisfaction with care, caregiving burden, and bereavement grief). Use of healthcare services identified as indicators of aggressive treatments will be recorded. Data will be collected at 2-week intervals until the patient’s death. Bereavement interviews will be conducted 1, 3, 6, and 13 months after the patient’s death to avoid contamination from anniversary grief reactions. Kappa coefficients will be computed to correct for agreement between the preferred and actual EOL care that can be expected to occur by chance alone. The generalized estimating equation (GEE) will be used to examine the impact of prognostic awareness and the congruence between patient’s preferred and actual EOL care on outcome variables. Translating the results from the proposed study into clinical care may improve the psychological well-being and QOL of terminally ill cancer patients, lift the caregiving burden of families, which may decrease their bereavement morbidity, avoid futile life-sustaining treatments, and lead to a good death in accordance with the patients’ and families’ wishes as the IOM suggested.

Project ID:PG9912-0277
External Project ID:NHRI-EX100-9906PI