Psychometric evaluation of a patient-rated troubling symptom scale for generalized anxiety disorder clinical trials

The majority of clinical outcome assessments developed for generalized anxiety disorder (GAD) may not efficiently and sensitively reflect heterogeneous symptom clusters from a patient perspective. The Patient-Rated Troubling Symptoms Scale for Anxiety (PaRTS-A) instrument was developed to provide an individualized assessment of patient relevant GAD symptoms.

The objective of this analysis was to evaluate the psychometric characteristics of the PaRTS-A.

Data were taken from a 6-week clinical trial evaluating the efficacy of adjunctive risperidone therapy in GAD patients undergoing standard treatment. Patients completed the PaRTS-A, the Quality of Life Enjoyment and Satisfaction Questionnaire (Q-LES-Q), and the Sheehan Disability Scale (SDS) as well as impressions of improvement. Physicians completed the Hamilton Anxiety Rating Scale (HAM-A). A instrument characteristics were assessed through statistical tests of reliability, reproducibility, construct and discriminant validity, and responsiveness.

In the item response theory (IRT) analysis, the PaRTS-A items fit into a single construct of anxiety. Internal consistency reliability exceeded 0.70. The PaRTS-A total score was moderately correlated with the Q-LES-Q and the SDS. The PaRTS-A distinguished between patients with high and low HAM-A scores (p<0.001). The PaRTS-A was responsive to changes in clinical status and the minimal important difference was identified.

The PaRTS-A provides a unique patient-rated assessment of anxiety symptoms. The instrument may provide clinicians with useful information regarding patient's self-rated anxiety disorder symptoms and the hierarchy of symptoms that they considered most troubling. Our analysis suggests that the PaRTS-A is an internally consistent, valid, and responsive instrument that may be a beneficial adjunct to other instruments in clinical trials.