Negative affectivity as a transdiagnostic factor in patients with common mental disorders

Background: Screening and monitoring systems are increasingly used in psychotherapy, but it has been questioned whether outcome measurement using multiple questionnaires is warranted. Arguably, type and number of assessment instruments should be determined by empirical research. This study investigated the latent factor structure of a multi-dimensional outcome measurement strategy used in English services aligned to the Improving Access to Psychological Therapies (IAPT) programme.

Methods: Factor analyses and structural equation models were performed on 11,939 intake assessments of outpatients accessing an IAPT service between 2008 and 2010. We examined whether three routinely employed instruments (PHQ-9 for depression, GAD-7 for anxiety, WSAS for functional impairment) assess empirically different dimensions.

Results: The instruments were found to assess mainly one general dimension and only some items of the GAD-7 and WSAS assess unique variance beyond this general dimension. In a structural equation model the disorder-specific factor scores were predicted by patients׳ diagnostic categories.

Limitations: Since a large naturalistic data base was used, missing data for diagnoses and scale items were encountered. Diagnoses were obtained with brief case-finding measures rather than structured diagnostic interviews.

Conclusion: Although the items seem to address mostly one dimension, some variance is due to differences between individuals in anxiety and impairment. While this generally supports multi-dimensional assessment in a primary care population, the clinical upshot of the study is to concentrate attention on transdiagnostic factors as a target for treatment.