Kessler et al1 err in stating that “ . . . theRegier-proposed restriction of cases in DSM-V isan attempt to declare that mild cases do not exist.”1(p1118) Our earlier article had nothing to do with DSM-V; rather, we merely presented a method for updating DSM-III and DSM-III-R prevalence rates byapplying previously unused information embedded in the Epidemiologic CatchmentArea Study and National Comorbidity Survey data sets to recalculate ratesof more stringently defined DSM-IV disorders.2 We applied to the scoring algorithms for an entiresyndrome or diagnosis those rules contained in the original Diagnostic InterviewSchedule and Composite International Diagnostic Interview assessment instrumentsthat were used to assess the “clinical significance” of reportedsymptoms in the Epidemiologic Catchment Area Study and National ComorbiditySurvey. We did not, as these authors assert, require service use in the past12 months. If our approach to assessing the clinical significance of syndromesis flawed, it follows that the determination of the clinical significanceof all symptoms that have been used to determine prevalence rates in almostall mental disorder epidemiological studies since 1980 would be similarlyflawed.
Regier DA, Narrow WE, Rae DS. For DSM-V, It’s the “DisorderThreshold,” Stupid. Arch Gen Psychiatry. 2004;61(10):1051. doi:10.1001/archpsyc.61.10.1051-a