Deconstructing the 'black box' of the Camberwell assessment of need score in mental health services evaluation.

Objective: The aim of the study was to examine an alternative way of scoring the Camberwell Assessment of Need (CAN) for the purpose of service evaluation, using the by us defined Social Services (SI) and Psychiatric Services (PI) subindices.

Methods: CAN assessments in 1997 and 1999 of 262 outpatients (mean age 45 years, 77.1% psychotic disorders) were reanalysed to fit the SI and the PI, which were compared to the full CAN.

Results: The mean total needs on the full CAN decreased from 6.65 to 6.22 (P = 0.007), as did the mean unmet needs (1.55-1.81, P = 0.049). The mean total needs on the PI decreased from 2.42 to 2.22 (P = 0.006), as did the mean unmet needs (1.66-0.57, P < 0.001). No changes in mean needs occurred on the SI.

Conclusions: All significant changes occurred on the PI, indicating a more beneficial outcome of the psychiatric care than the social care in terms of meeting needs, a result impossible to discern from the total scores of the CAN. Thus, output scores on subindices of the CAN might be useful as outcome measures in service evaluation.