Generalized anxiety disorder (GAD) is one of the most prevalent psychiatric presentations; however, GAD has the lowest diagnostic reliability of the anxiety disorders and is poorly recognized in clinical practice. A more reliable assessment of GAD could lead to earlier detection and treatment of the disorder, which has an otherwise debilitating course and significant associated impairment. The 7-item GAD Scale (GAD-7) has shown promise as a measure with good clinical utility and strong psychometric properties in primary care and community settings but has yet to be assessed in acute psychiatric populations. This study examined the validity of the GAD-7 in a sample of 232 patients enrolled in a partial hospital programme. Patients completed a diagnostic interview and a battery of self-report measures before and after treatment. Findings suggest that the GAD-7 has good internal consistency and good convergent validity with worry, anxiety, depression and stress, and the measure was sensitive to change over the course of a short intensive cognitive–behavioural therapy partial hospital programme. However, the confirmatory analysis failed to support the hypothesized unidimensional factor structure; and although the GAD-7 demonstrated good sensitivity (.83), specificity was poor (.46) in identifying patients with GAD. Overall, the GAD-7 appears to be a valid measure of generalized anxiety symptoms in this sample, on the basis of good internal consistency, convergent validity and sensitivity to change, but does not perform well as a screener for GAD. Copyright © 2012 John Wiley & Sons, Ltd.
Key Practitioner Message
- The GAD-7 Scale is an easy-to-score, self-report measure of core generalized anxiety disorder symptoms.
- The GAD-7 Scale has good internal consistency and convergent validity with depression, anxiety, stress and worry, and is sensitive to change.
- The GAD-7 Scale appears to be a good measure of generalized anxiety symptoms in an acute psychiatric sample.
- The GAD-7 Scale does not perform well as a screener for GAD and should not be used to identify cases of GAD in acute psychiatric samples.