Reliability of the Camberwell Assessment of Need for South African mental health service users

Background : The Camberwell Assessment of Need (CAN) is a widely used instrument to assess the service needs for people with a severe mental illness. No reliability data are available for its use in South Africa. Method : Interrater and test-retest reliability were documented for a sub-sample of 194 patients participating in a needs assessment study among people receiving mental health services. Initially, the CAN was administered by one field worker, while another observed and also made ratings. The participant was then re-interviewed approximately one week after the initial interview. Results : With the exception of ‘psychotic symptoms’ ( κ = -0.4) and ‘looking after home’ ( κ = 0.008), Cohen’s κ coefficients for interrater reliability ranged from 0.150 to 0.760 for Section 1 of the CAN. Test-retest reliabilities for Section 1 resulted in moderate to substantial agreement, with Cohen’s κ coefficient ranging from 0.25 to 0.81. High correlations were reported for the test-retest reliabilities of the remaining sections of the CAN ( r = 0.719–0.845). However, the figures for the interrater reliabilities were lower and more variable ( r = −0.082–0.408). Conclusion : The results of this study suggest that with additional interviewer training, the CAN is a relatively reliable instrument for assessing the needs of youths with severe mental illness.