gms | German Medical Science

Background: Patients presenting to Centres for Rare Diseases (CRD) often suffer from undiagnosed mental disorders [1]. Standardised screening scales for mental health such as the GAD-7 [2] and the PHQ-9 [3] might support mental health experts (MHE) in the process of diagnosing mental disorders. The aim of this analysis was to test the psychometric properties of the GAD-7 and the PHQ-9 in patients presenting to a CRD and to examine their discriminatory power for a diagnosis of anxiety or depressive disorder.

Methods: Data were derived from 1300 patients (control group (CG) n=636, intervention group (IG) n=664) participating in ZSE-DUO (Dual guidance structure in Centres for Rare Diseases), a multicentre, prospective controlled trial in 11 CRD (funding by G-BA, Grant 01NVF17031), evaluating the benefit of involving a mental health expert (MHE) in the diagnostic process [4]. Patients in the CG received standard care. Patients in the IG received standard care, complemented by consultation of the MHE in the entire diagnostic process. The MHE documented diagnoses of mental disorders within 12 months after patients’ first visit at a CRD. Several diagnoses could be documented for one patient. Both, in CG and IG, depressive and anxiety symptoms were assessed at patients’ first visit at a CRD using the GAD-7 [2], and the PHQ-9 [3]). Psychometric properties of the questionnaires were analysed in exploratory factor analyses (EFA) to evaluate theoretically assumed factor structures, including CG and IG patients. The internal reliability was reported as Cronbach’s α. In the IG, the discriminatory power of the established cut-off score of ≥10 for the GAD-7 and the PHQ-9 summary score for the diagnosis of anxiety (ICD-10: F40-F41) and depressive disorders (ICD-10: F32-F39), respectively, was tested in ROC analyses. In addition, the Youden-Index was used to analyse appropriate cut-off scores for our study population.

Results: The results of the EFA confirmed the theoretically assumed factor structures for both scales (GAD-7: 1 factor Cronbach’s α=0.87; PHQ-9: 2 factors; Cronbach’s α=0.82). The discriminatory power of the GAD-7 and the PHQ-9 for a diagnosis of anxiety or depressive disorders in the IG, was moderate to low (GAD-7: AUC=0.68 [95%CI=0.63-0.74], sensitivity=45%, specificity=79%; PHQ-9: AUC=0.78, [95%CI=0.74-0.82], sensitivity=71%, specificity=69%). According to the Youden-Index optimal cut-off scores were ≥9 for the GAD-7 (sensitivity=51%, specificity=73%) and ≥11 for the PHQ-9 (sensitivity=68%, specificity=75%).

Conclusion: The underlying theoretical constructs for the GAD-7 and the PHQ-9 were confirmed in patient presenting to a CRD. However, the established cut-offs for the GAD-7 and PHQ-9 summary scores seem to lack discriminatory power in this population. Hence, further research is needed to adapt these established scales to screen for patients who need consultation by MHE in CRDs.

The authors declare that they have no competing interests.

The authors declare that a positive ethics committee vote has been obtained.