Beyond generalized anxiety disorder: Psychometric properties of the GAD-7 in a heterogeneous psychiatric sample

Background: Although developed as a screener for Generalized Anxiety Disorder (GAD) in primary care, the GAD-7 is now commonly used as a measure of general anxiety symptoms across various settings and populations. However, little is known about its psychometric properties when used in such heterogeneous samples. We examined the internal consistency, convergent validity, sensitivity and specificity, sensitivity to change, and structure of the GAD-7 in patients receiving brief, intensive CBT treatment in a partial hospital setting. We also examined the properties of a modified version that assessed symptoms over the past 24-h. Methods: Participants (n = 1082) completed the GAD-7 upon admission and discharge from a partial hospital program. They also completed measures of worry, depression, and well being and a structured diagnostic interview. We examined psychometric properties in the total sample and separately for patients with GAD, post-traumatic stress disorder, Social Anxiety Disorder (SAD), and panic disorder. Results: Internal consistency and convergent validity were good for the total sample and each anxiety disorder group. The GAD-7 demonstrated poor specificity and a high false positive rate for all anxiety disorders. Sensitivity to change was generally good. Factor analysis revealed that a one-factor structure did not fit the data well. The 24-h version performed similarly to the original version. Conclusions: The GAD-7 performed well as a measure of anxiety symptom severity, but not as a screener in this psychiatric sample. It is a useful outcome measure for hetereogenous samples, but it may not perform as well specifically for individuals with SAD. A modified version of the GAD-7 that assessed anxiety symptoms over the past 24-h appears to be a reliable and valid modification. (C) 2014 Elsevier Ltd. All rights reserved.