Objective Advance Care Planning (ACP) encourages the expression of a person’s values, beliefs and life goals and is a route that all patients with chronic disease can benefit from, especially those at high risk of deterioration (1). A recent randomised trial showed that ACP improved awareness of patients’ end-of-life (EOL) wishes and reduced their families’ distress. ACP is likely to be beneficial for patients with chronic heart failure (HF) in our institution.

Methods In this pilot study, the Heart Failure Service (HFS) team conducted ACP from Feb 2012 to Nov 2012 in Tan Tock Seng Hospital, a tertiary hospital in South-East Asia which receives about 1500 HF admissions annually. The HFS is a multidisciplinary team comprising a physician, advanced practise nurses, case managers and pharmacists. The team received formal training in Basic Advance Care Planning (First Steps) and End-of-life treatment options (Last Step) as per the “Respecting Choices” programme pioneered by the Gundersen Lutheran Medical Foundation. ACP was initiated by the Case Managers or other team members for patients in Stage C or Stage D of HF, which was conducted together with patients’ families, caregivers, and other significant persons as permitted by the patients. Patients who refused, who were mentally incompetent, had a history of self-harm or intent to self-harm, were uncommunicative, or who had other had acute life-threatening comorbidities were excluded from this pilot study. A post-ACP telephone survey of patients’ and their families’ perception of ACP was performed.

Results We initiated ACP for 22 patients, age 70 ± 13 years old, 73% Male, 59% Stage C, 41% Stage D, LVEF 27% ± 15%. Ten patients completed full ACP discussion. The rest were unwilling to complete the discussion. Nine patients died a mean of 53 days after initiation of ACP, of whom only 3 had completed ACP. These three had their end-of-life preferences honoured. In the post-ACP survey 3 patients expressed better understanding of their medical condition, were relieved that their EOL preferences were made known, and felt more in control of their treatment plan and thus less anxiety. The rest of the patients were too ill to do the survey or had died, but their family members felt relieved of guilt as they did not need to decide patients’ EOL preferences.

Conclusions ACP appears to be beneficial in the patients who agreed to participate. This may be due to selection bias as they are likely to be open to the ACP process, having perhaps already had prior exposure to the concept. It is unclear how to identify those ready to proceed with ACP, and this should be studied further.