Background Established legal mandates andhigh expectations for psychiatric advance directives are not matchedbyempirical evidence

Declaration of interest None. Psychiatric advance directives are legal instruments that allow competent individuals to appoint proxies and specify how treatment decisions should be made in the event they become incompetent. These directives have been hailed as instruments for honouring patient autonomy, improving communication and ensuring participation in treatment decisions, and providing pragmatic alternatives to coercive measures (Appelbaum, 1991; Swanson et al, 2000; Srebnik et al, 2003). Progress towards implementing advance directives as routine planning options has not kept pace with such enthusiasm. The only randomised effectiveness study of psychiatric advance directives showed no impact on outcome of care with regard to compulsory admissions, hospital days or patient satisfaction (Papageorgiou et al, 2002). However, as with other complex community-based interventions (Schoenwald & Hoagwood, 2001; Ablon & Jones, 2002; Hohmann & Shear, 2002), questions have been raised about actual operational specifics and how they might have influenced the impact of local efforts to implement such directives (Thomas, 2003). No scientific consensus exists on what the ‘intervention’ is. It is unclear what it means for people with severe and persistent mental illness to undertake the anticipatory planning culminating in an advance directive. This study empirically explores the deliberations of informed mental health service users contemplating this option.