Study Design The study is designed as a prospective randomised controlled trial. Eligible patients from the departments of oncology, cardiology and respiratory medicine at Aarhus University Hospital will be included in the study and randomised in two groups: one receiving usual care and the other receiving usual care and ACP. Patients will be recruited from one department at a time, allowing the recruitment period to be up to six months for each department. Because patients with terminal cardiac and respiratory diseases are known to have more unpredictable life expectancy than cancer patients, initial recruitment will be from cardiology and respiratory medicine departments, respectively.9 This procedure will enhance the possibility of the estimated number of deaths in all three diagnose groups to be registered within the study period (see power calculation below). The research group has a long and ongoing close cooperation with the departments mentioned and collaboration has been established.
Practical Execution The researcher (the applicant) will be provided with a list of patients who will attend the outpatient clinic the following day. Using an identification tool (see Appendix 1) the researcher will assess the patients, identifying those patients anticipated to be in need of palliative and supportive care. Patients' data are found in Electronic Patient Files (EPJ). After the patients have been seen by staff in the outpatient's clinic, the researcher will ask the doctor in the clinic whether he/she would be surprised if this patient died within the next 6-12 months. (i.e. "The surprise Question"). 10 If the answer is no, the patient will be enrolled in the study. Within two days hereafter the patient will be contacted by mail and later by telephone. Information about ACP will be given written and verbally. If the patient accepts, informed written consent is obtained. Returning consent form blank or not responding after two reminders will be considered a decline. Patients will be randomised by usual standards; non-research staff will carry out the randomisation using sealed envelopes containing allocation cards assigned by random numbers. All patients included in the study will receive a questionnaire concerning quality of care (see questionnaire data). The patients are also asked to point out a close relative who is likely to accept participation. Participating relatives will be approached in order to fill in questionnaires during the patient's trajectory and in the bereavement period (see questionnaire data).
Intervention Patients in the intervention group will be offered a formal ACP discussion with the researcher. This will be conducted as a discussion between the researcher and the patient based on the Respecting Patient Choices model.11 This programme, developed at the Austin Hospital in Melbourne, Australia, is based on the Respecting Choices programme and involves a coordinated approach to ACP whereby facilitators, in collaboration with the treating doctor, assist patients and their families to reflect on the patient's goals, values, and beliefs, and to discuss and document their future choices about health care. Patients are encouraged to appoint a proxy and to document their preferences about end-of-life care; including the preferences for life prolonging treatments and cardiopulmonary resuscitation recorded on an Advance Care Plan (see Appendix 2). Whenever possible, treating doctors will participate in the discussion to ensure that the patients understand their illness, treatment options, and likely prognosis. The discussion will be audio taped for scientific use and the ACP will be documented in EPJ. The discussion will take place at the hospital or at the patient's home. With the acceptance of the patient, the general practitioner and the community nurse will receive a copy of the ACP documents. Patients will be encouraged to include their closest relative in the ACP discussion. The aim is to complete the ACP discussion within two weeks after the identification of patients in the outpatient's clinic. Consistent with usual practice patients in the control group will receive usual medical and psychosocial care but not a formalised ACP discussion.
Data Collection Demographic data: will be collected at enrolment and will include age, sex, education, children living at home, admission diagnosis, the existence of a form requesting no resuscitation, whether the patient has a proxy, and whether the patient already has any EPJ documented preferences on end-of-life care, including life prolonging treatment or cardiopulmonary resuscitation. Data will be collected from EPJ and in the questionnaires, respectively.
Questionnaire Data A thorough literature review of existing, validated tools will be conducted. Accordingly the questionnaires will be composed by validated and well-established questionnaires and supplemented by ad hoc questions when no existing questionnaires are available. Translation and adaption as well as pilot testing in accordance with international guidelines will be performed before use. All patients will be requested to fill in a questionnaire three to four weeks after inclusion and the relatives are similarly asked to evaluate the care at this time.12 In those cases where the patient dies within the study period, the bereaved relatives will receive a questionnaire evaluating the levels of stress, anxiety and depression as well as a questionnaire concerning the quality of the patient's end-of-life care and preferences for place of care and death.13-15
Power Calculation Based on the existing literature, we expect that the proportion of patients whose end-of-life preferences were known will be 15 % and that an increase to 65 % would represent a clinically meaningful effect. To achieve 90 % power to detect a difference in the primary outcome between groups with a certainty of 95 %, we estimated that 22 deaths would be required in each group, i.e. 44 deaths in total. It is therefore planned to continue the inclusion period until at least 22 deaths are observed in each group among patients enrolled. In a similar study, the population included patients had an average lifespan of 3.5 months the time of inclusion until death and 86 % of included patients died during the study period which was 20 months.16 In order to observe 44 deaths, an estimated inclusion of 90 patients is necessary. The estimated inclusion period will be approximately 18 months.
Perspectives The project is an integrated part of a large research program within the palliative field. The project is, with its focus on creating better trajectories and better treatment for patients throughout the end of their terminal illness, highly relevant. The ongoing development of palliative care and the lack of evidence that characterises the area challenge the health care even more while also dealing with the utilisation of limited financial resources. If ACP is effective in a Danish context, it will improve the end-of-life care for patients and their families and reduce the psychological distress in the bereaved relatives. ACP may therefore contribute to a more rational use of health care resources as well as being a tool to help patients have their preferences met.
Ethics Terminally ill patients are a vulnerable group to involve in research. However, the research group has positive experiences with involving patients and relatives in this kind of research. The patients' main motivation to participate has been to help others in similar situation. It should also be noted that no patients are treated less favourably than the current standard procedure in healthcare. It is optional for patients and their relatives in the intervention group, whether they wish to participate in the ACP or not.
