Assessing and addressing research-practice gaps in practitioners' provision of decision support for terminally-ill cancer patients considering the place of their end -of -life care: Design and evaluation of a multi-faceted educational intervention

Background. A paradox exists between preferred place of end-of-life cars (home/hospice) and actual place of care (hospital). Patients often need help with decision making yet practitioners generally lack skills in providing quality decision support.

Purpose. To evaluate an intervention designed to address barriers to providing decision support for place of end-of-life care. Objectives were to: (1) describe nurses' perceptions of factors that affect their provision of decision support; and (2) design and evaluate the efficacy of an educational intervention for strengthening the quality of practitioners' decision support.

Design and methods. (1) Systematic review: Synthesis of primary research studies investigating factors influencing place-of-care. (2) Needs assessment: Semi-structured interviews with 22 nurses. (3) Randomized trial: Comparison of intervention and control groups on the quality of decision support provided to simulated patients considering place of care before and after training.

Setting of the Needs Assessment and Randomized Trial. Ontario palliative and oncology practice settings.

Results (1) Systematic Review: Of 735 papers identified, 39 papers (33 studies) met inclusion criteria. Factors related to the illness, the individual and the care and social environment influence place of end-of-life care for cancer patients. Commonly identified factors were patient preference, social supports and health care resources. Stated preference, care-giver presence and contact patterns with health care services influence place of end-of-life cancer care. (2) Needs assessment: Nurses held favourable attitudes towards providing decision support for place-of-care at end-of-life. Nurses considered decision support an important part of patient-centred care yet identified a lack of skills, confidence and tools as barriers. (3) Randomized control trial: The before-after changes in the quality of decision support were greater (p<0.0001) in the intervention group (mean change 3.75, 95%Cl: 2.46, 5.03) compared to the control group (mean change -0.667, 95% Cl: -1.57, 0.24).

Conclusions. Numerous determinants influence terminal cancer patients' decision making about place of care. Nurses believe they have a role in providing decision support, but lack skills, confidence and tools. These barriers can be addressed with an educational program and a patient decision aid.

Broad implementation of the intervention may improve end-of-life care planning. Full implementation of the intervention requires examination of broad system perspectives.

Trial registration: NCT00614003