Abstract
The Panic Disorder Severity Scale (PDSS) is a seven-item scale designed to assess overall severity of panic disorder symptoms. Although the PDSS is widely used, there have been limited independent studies confirming reliability and validity, particularly of the self-report version of the PDSS, and even fewer studies directly comparing the two versions. Given the need for psychometrically sound measures that comprehensively assess the main features of panic disorder, this study examined the psychometric properties of both the clinician-administered and self-report versions of the PDSS. As the PDSS targets symptom severity across a number of specific domains, PDSS items were compared to several scales designed to measure similar constructs, including the Anxiety Sensitivity Index, Illness Intrusiveness Ratings Scale, Mobility Inventory for Agoraphobia and Agoraphobic Cognitions Questionnaire. Results indicated acceptable reliability for both the clinician-administered and self-report versions. Results indicated acceptable validity for the clinician-administered PDSS and promising validity for the self-report. However, correlations between both versions of the PDSS and comparison measures were lower than predicted. Results also indicated that scores on the self-report and clinician-administered versions were significantly associated, although correlations between specific PDSS items on the two versions were also lower than expected. Results suggest that total scores on each of the PDSS versions provide a useful indication of panic-related severity. However, clinicians and researchers are cautioned against relying on scores from individual items to measure specific constructs due to their poor specificity. Copyright © 2010 John Wiley & Sons, Ltd.
Key Practitioner Message:
• Results indicated acceptable validity for the clinician-administered PDSS and promising validity for the self-report, and suggest that total scores on each version provide a useful indication of panic-related severity. However, clinicians and researchers are cautioned against relying on scores from individual items to measure specific constructs due to their poor specificity.