An examination of the efficiency of the 12-item SPHERE questionnaire as a screening instrument for common mental disorders in primary care


David M. Clarke, Associate Professor (Correspondence)


Objective: To further examine the screening capabilities (sensitivity, specificity, etc.) of the 12-item Somatic and Psychological Health Report (SPHERE) questionnaire.

Method: A more extensive examination was carried out on data presented in a recent report.

Results: The sensitivities of the questionnaire, scored in three different ways, ranged from 100% to 47%. Specificities ranged from 18% to 72%. As a global screen for mental disorder it had a very high false-positive rate, with, in one sample, 83% of patients screening positive while only 27% had a current psychiatric diagnosis, and in the other sample 55% screened positive with only 13% having a current psychiatric diagnosis. The PSYCH-6 scale used by itself had similar properties to the 30-item General Health Questionnaire from which it was derived. The addition of the scale measuring fatigue (SOMA-6) increased or decreased specificity depending on how it was combined with the PSYCH-6 scale.

Conclusion: The evidence is insufficient to recommend the 12-item SPHERE, in its current form, as a screening instrument for DSM-IV mental disorders in general practice, as the specificity is inadequate. Ways of raising the threshold for caseness need to be explored. The argument for adding a measure of fatigue to a general screening measure is not supported.