Detecting depressive and anxiety disorders in distressed patients in primary care; comparative diagnostic accuracy of the Four-Dimensional Symptom Questionnaire (4DSQ) and the Hospital Anxiety and Depression Scale (HADS)

BACKGROUND Depressive and anxiety disorders often go unrecognized in distressed primary care patients, despite the overtly psychosocial nature of their demand for help. This is especially problematic in more severe disorders needing specific treatment (e.g. antidepressant pharmacotherapy or specialized cognitive behavioural therapy). The use of a screening tool to detect (more severe) depressive and anxiety disorders may be useful not to overlook such disorders. We examined the accuracy with which the Four-Dimensional Symptom Questionnaire (4DSQ) and the Hospital Anxiety and Depression Scale (HADS) are able to detect (more severe) depressive and anxiety disorders in distressed patients, and which cut-off points should be used. METHODS Seventy general practitioners (GPs) included 295 patients on sick leave due to psychological problems. They excluded patients with recognized depressive or anxiety disorders. Patients completed the 4DSQ and HADS. Standardized diagnoses of DSM-IV defined depressive and anxiety disorders were established with the Composite International Diagnostic Interview (CIDI). Receiver Operating Characteristic (ROC) analyses were performed to obtain sensitivity and specificity values for a range of scores, and area under the curve (AUC) values as a measure of diagnostic accuracy. RESULTS With respect to the detection of any depressive or anxiety disorder (180 patients, 61%), the 4DSQ and HADS scales yielded comparable results with AUC values between 0.745 and 0.815. Also with respect to the detection of moderate or severe depressive disorder, the 4DSQ and HADS depression scales performed comparably (AUC 0.780 and 0.739, p 0.165). With respect to the detection of panic disorder, agoraphobia and social phobia, the 4DSQ anxiety scale performed significantly better than the HADS anxiety scale (AUC 0.852 versus 0.757, p 0.001). The recommended cut-off points of both HADS scales appeared to be too low while those of the 4DSQ anxiety scale appeared to be too high. CONCLUSION In general practice patients on sick leave because of psychological problems, the 4DSQ and the HADS are equally able to detect depressive and anxiety disorders. However, for the detection of cases severe enough to warrant specific treatment, the 4DSQ may have some advantages over the HADS, specifically for the detection of panic disorder, agoraphobia and social phobia.