Community mental health services in China are in need of a brief, valid instrument to identify mental disorders. Many studies support the Self-Reporting Questionnaire 20 (SRQ-20) as a cost-effective instrument with which to measure community mental health, but there have been no previous studies conducted with the Chinese version of SRQ-20. In this paper, the psychometric properties of the Chinese version of SRQ-20 in community settings were investigated.
SRQ-20 was used to evaluate 959 participants from 10 communities of Hangzhou City and 60 participants from two community primary care centres for the presence of mental disorders. Discriminative, concurrent, and construct validity as well as internal consistency and screening properties of SRQ-20 were evaluated.
Cronbach's alpha for SRQ-20 was 0.90 for primary care and 0.91 for the community. Item-total correlation coefficients ranged from 0.40 to 0.73 in primary care, and from 0.51 to 0.69 in the community. The test-retest correlation coefficient was 0.93 in primary care and 0.94 in the community. Varimax-rotated principal component analysis of the SRQ-20 items yielded three factors: depressive symptoms, anxiety symptoms and somatic symptoms. The area under the ROC curve was 0.83 (SE =0.04, 95% CI 0.75-0.90) for the community and primary care samples. The optimal cut-off point for SRQ-20 was 6/7 yielding sensitivity of 93% and specificity of 62%.
SRQ-20 appears to be a reliable and valid measure of mental disorders in the community in China.