Background: There is a dearth of methodological studies critically evaluating reliability, validity and feasibility of measures of common mental disorders (CMD) in low-income countries.
Methods: Test-retest and inter-rater reliability of categorisation of CMD caseness, according to locally agreed criteria using the Comprehensive Psychopathological Rating Scale (CPRS), was measured in 99 women from out-patient clinics (inter-rater) and 99 women from a primary healthcare centre (test-retest) in Ethiopia. The construct validity of CMD as measured with CPRS was assessed with exploratory factor analysis using maximum likelihood with varimax rotation.
Results: Test-retest reliability was fair (kappa = 0.29). Subsequent assessment of inter-rater reliability found excellent agreement (kappa = 0.82). The construct of CMD appeared unidimensional, combining depressive, anxiety and somatic symptoms.
Conclusions: Detection of socioculturally meaningful cases of CMD in Ethiopia can be reliably achieved with local psychiatrist assessment using CPRS, although thorough training is essential.