In routine psychiatric practice, clinicians, patients and carers are all involved in the care process and their perspectives should all be considered when assessing mental health outcome. The fact that their views may markedly differ (Gunkel & Priebe, 1993; Ellwood, 1998; Walter et al, 1998; Priebe et al, 2007; Brown et al, 2008) is confirmation that taking into account multiple perspectives and integrating these views are necessary steps when evaluating mental health outcomes.

The subjective appraisals made by patients who are mentally ill had been neglected for a long time, in both clinical practice and research, due to the view (or the prejudice) that they were unable, by the nature of their illness, to appreciate their need for treatment or to make the ‘correct’ decisions about how to conduct their lives. Therefore, for many years, the mental health system took a paternalistic stance and controlled the lives of people diagnosed with severe mental disorder, and the only form of ‘involvement’ of service users was as passive recipients of decisions made by others (Chamberlain, 2005).

The idea of patients’ involvement in the planning and evaluation of mental healthcare has been growing over the last 20 years, especially in those countries where institutional service provision has been transformed into a community-oriented model of care (Beresford, 2005; Chamberlain, 2005; Thornicroft & Tansella, 2005). This represents a major challenge for mental health services, since the so-called ‘users’ movement’ advocates a shift towards a modern ‘shared model’ or ‘partnership model’ between health professionals and service users, based on mutual respect for each other's skills and competencies and recognition of the advantages of combining these resources to achieve beneficial outcomes (Simpson & House, 2002; Crawford et al, 2003; Bramesfeld et al 2007; Brimblecombe et al, 2007; Killaspy et al, 2008; Stringer et al, 2008).

Within this methodological framework, it is particularly important not only to combine optimal measures of both the ‘service level’ and the ‘patient level’ (Thornicroft & Tansella 1999) but also to pay attention to (in addition to the traditional outcome dimensions such as clinician-rated symptoms and disability) patients’ self-perceived outcomes, such as self-rated needs for care, subjective quality of life, carer burden and satisfaction with care (Jenkins, 1990; Mirin & Namerow, 1991; Attkisson et al, 1992; Ruggeri & Tansella, 1995; Smith et al, 1997).