Objective: Psychological distress questionnaires are often used as screening instruments for mental disorders in clinical and epidemiologic settings. Poor physical health may affect the screening properties of a questionnaire. We evaluate the effect of self-perceived health status on the screening performance of the Kessler K10 and K6 scales in a community sample.
Methods: We used data from the Canadian Community Health Survey: Mental Health and Well-Being (CCHS 1.2). Psychological distress was measured by the 6-item (K6) and the 10-item (K10) Kessler instrument. Depression and anxiety disorders were assessed using the World Mental Health Composite International Diagnostic Interview (1-month estimates). Optimal cut-off points regarding health status were determined by finding the K6 and K10 values that allowed for the best balance between sensitivity and specificity. Stratum-specific likelihood ratios (SSLRs) were computed to define strata with discriminating power.
Results: There was a strong association between the screening performance of the K6 and K10 scales and self-perceived health status: for the K10 scale, a cut-off point of 5/6 yielded the best balance between sensitivity and specificity for subjects with excellent or very good health status, while a cut-off point of 14/15 yielded the best balance between sensitivity and specificity for subjects with poor health status.
Conclusions: The combination of the K6 and K10 scales, with a self-rated health status item, may improve screening properties of the 2 scales.