Feasibility of the development and psychometric properties of a standardized screening instrument for mental disorders in patients with suspected rare diseases: results of the ZSE-DUO study.

Introduction: Patients presenting in centers for rare diseases (CRDs) show complex symptoms, requiring multiprofessional expertise. Many patients suffer from unnoticed mental disorders. Currently, mental health experts (MHEs) are rarely involved in the diagnostic process. The aim of this study was to test the feasibility of developing a new short screening instrument for mental disorders and to test its predictive value. Methods: Data were derived from 1,300 adult patients participating in ZSE-DUO (dual guidance structure in centers for rare diseases), a multicenter study in 11 CRDs (funded by G-BA, Grant 01NVF17031), evaluating the benefit of involving an MHE in the diagnostic process. Patients completed standardized questionnaires on anxiety [the 7-item Generalized Anxiety Disorder (GAD-7)], depression [the 9-item Patient Health Questionnaire (PHQ-9)], dissociation [4-item Dissociation-Tension Scale (DSS-4)], psychopathology [Symptom-Checklist K-9 (SCL-K-9)], and quality of life [12-item Short Form Health Survey (SF-12) and EQ-5D-5L] prior to and during their first CRD visit as part of the routine assessment necessary for a first contact at a CRD. Exploratory factor analysis (EFA) was performed for item reduction. The reliability of the factor structure was assessed using Cronbach's α. Model fit was tested using confirmatory factor analysis (CFA). The predictive value of the new screening instrument was tested by calculating a receiver operating characteristic (ROC) curve using the diagnoses from the MHE. Results: The extracted 18-item model had a four-factor structure with acceptable to high reliability. The extracted mental health dimensions were 1) anxiety and depression, 2) mobility and activities of daily living, 3) energy or fatigue, and 4) dissociation. Excluding the dimension "dissociation" (well assessed using the DSS-4 separately), a summary score was calculated ranging from 0 to 45, with the lowest score representing the best mental health. In the CFA, the model fit indices for the extracted factor structure did meet the established thresholds. The predictive value of the new screening instrument on any mental disorder was moderate [area under the curve (AUC)=0.68; 95% CI=0.64-0.73]. Conclusion: The design of a new short screening instrument for mental disorders in patients presenting at CRDs was feasible. Furthermore, the new short screening instrument may help MHEs to identify patients in need of a more thorough screening and adapted care, particularly in identifying patients with potential depressive disorders. However, due to the heterogeneity of the mental disorders in this patient group, in addition, individual assessment of patients by MHEs is necessary.