A systematic review of clinician-rated instruments to assess adults’ levels of functioning in specialised public sector mental health services

Philip M Burgess, Meredith G Harris, Tim Coombs, Jane E Pirkis

Functioning is one of the key domains emphasised in the routine assessment of outcomes that has been occurring in specialised public sector mental health services across Australia since 2002, via the National Outcomes and Casemix Collection. For adult consumers (aged 18-64), the 16-item Life Skills Profile (LSP-16) has been the instrument of choice to measure functioning. However, review of the National Outcomes and Casemix Collection protocol has highlighted some limitations to the current approach to measuring functioning. A systematic review was conducted to identify, against a set of pre-determined criteria, the most suitable existing clinician-rated instruments for the routine measurement of functioning for adult consumers. We used two existing reviews of functioning measures as our starting point and conducted a search of MEDLINE and PsycINFO to identify articles relating to additional clinician-rated instruments. We evaluated identified instruments using a hierarchical, criterion-based approach. The criteria were as follows: (1) is brief (<50 items) and simple to score, (2) is not made redundant by more recent instruments, (3) relevant version has been scientifically scrutinised, (4) considers functioning in a contemporary way and (5) demonstrates sound psychometric properties. We identified 20 relevant instruments, 5 of which met our criteria: the LSP-16, the Health of the Nation Outcome Scales, the Illness Management and Recovery Scale-Clinician Version, the Multnomah Community Ability Scale and the Personal and Social Performance Scale. Further work is required to determine which, if any, of these instruments satisfy further criteria relating to their appropriateness for assessing functioning within relevant service contexts, acceptability to clinicians and consumers, and feasibility in routine practice. This should involve seeking stakeholders' opinions (e.g. about the specific domains of functioning covered by each instrument and the language used in individual items) and testing completion rates in busy service settings.