The impact of the intensity of single symptoms on the diagnosis and prevalence of major depression

Abstract

ObjectiveDiagnoses are based on algorithms which count symptoms. The question is whether all or only sufficiently severe complaints of patients should qualify as diagnostic criteria and how prevalence rates of mental disorders change when all or only moderate or severe complaints are counted as symptoms in diagnostic algorithms. MethodOne hundred thirty-nine general practice patients were diagnosed as suffering from major depression according to DSM-IV on the basis of the standardized clinical International Neuropsychiatric Diagnostic Interview (MINI). They additionally filled in the self-rating SCL-90, which allows patients to rate the intensity of each symptom on a five-point Likert scale. The diagnostic algorithm for major depression was modeled on the basis of the SCL-90 self-rating. MINI-diagnoses were compared with SCL-diagnoses when symptoms with different intensity were taken into account. ResultsThe prevalence of “SCL-90-major depression” is 77.2% if all symptoms are counted, and 38.7% if only at least moderately severe symptoms, or 5.0% if only extremely severe symptoms are included in the diagnostic algorithm. Sensitivity rates vary from 10.8%, if only extremely severe symptoms are counted, to 94.2% if all complaints were included. Specificity rates vary from 37.4% for all complaints and 100% for extremely severe items. Accuracy is best when “a little bit” of complaint is omitted (78.1%) and lowest if only extremely severe symptoms are counted (58.9%). ConclusionThe data demonstrate the importance of the severity of single symptoms for the diagnosis of mental disorders. Symptom definition, recognition, and evaluation must find greater attention in research, clinical practice, and training of physicians.