Background ACDs enable individuals to document future treatment preferences. Documented wishes need to be ‘clinically meaningful’ and able to inform medical treatment plans during acute hospitalisation.

Aim To assess preferences indicated on ACDs for content, consistency within the document and concordance with medical orders on the Resuscitation Plan (RP) (hospital POLST form).

Methods Retrospective analysis of ACDs completed during facilitated advance care planning in 2011; including assessing consistency between the specific treatment wishes (CPR/Life-Prolonging Treatment (LPT)) and documented outcome preferences, and evaluating concordance of RP orders and ACD requests.

Results 153 patients completed ACDs, 88% also appointed a Substitute-Decision-Maker and 90% documented CPR/LPT preferences. Of these 51% chose ‘No CPR’, 40% chose ‘CPR if medically beneficial’ and 9% accepted medical advice that ‘CPR would be of no benefit’. 48% indicated they didn't want LPT, 50% wanted LPT if doctors anticipated a reasonable outcome and only 2% wanted LPT for as long as possible. 71% identified other undesired treatment(s) (38% ventilation, 37% artificial nutrition, 12% surgery, 10% dialysis). 84% documented their interpretation of unacceptable outcomes (inability to eat—38%, or communicate—43%, impaired cognition 33%, bedbound 33%).CPR and LPT requests were internally consistent in 88% and 85% of cases. A RP was completed during admission in 70 patients; 89% were concordant with ACD requests, 10% limited treatment more and 1% recommended more treatment.

Discussion Most patients clearly documented preferences and undesired outcomes that were interpretable into actionable medical orders which were followed.

Conclusion Facilitated ACD completion ensures patients' wishes are known and respected.