Test performance characteristics of the AIR, GAD-7, and HADS-anxiety screening questionnaires for anxiety in chronic obstructive pulmonary disease

Abstract

Rationale: Anxiety is a common comorbidity of chronic obstructive pulmonary disease (COPD) that is associated with higher morbidity and mortality. We evaluated three anxiety screening questionnaires: the Generalized Anxiety Disorder 7-Item Scale (GAD-7), the Hospital Anxiety and Depression Scale Anxiety subscale (HADS-A), and the Anxiety Inventory for Respiratory Disease (AIR). Objectives: To evaluate and compare the test performance characteristics of three anxiety screening questionnaires, using the Mini-International Neuropsychiatric Interview (MINI), version 7.0, as the “gold standard.” Methods: Individuals with COPD were recruited at 16 centers. The MINI and questionnaires were administered by trained research coordinators at an in-person visit and readministered by telephone 2–4 weeks later. A composite score for the presence of any Diagnostic and Statistical Manual of Mental Disorders, 5th edition (DSM-V) anxiety disorder was computed, based on the MINI as the gold standard, compared with a participant screening positive on self-report measures for these analyses. Results: Two hundred and twenty eligible individuals with COPD were enrolled; 219 completed the study. Eleven percent were identified as having a DSM-V anxiety disorder, based on the MINI. Elevated anxiety symptoms based on questionnaires were 38% for the AIR, 30% for the GAD-7, and 20% for the HADS-A. Area under the receiver operating characteristic curve (AUC) was highest for the GAD-7 (0.78; 95% confidence interval [CI], 0.69–0.87), followed by the HADS-A (0.74; 95% CI, 0.64–0.84) and the AIR (0.66; 95% CI, 0.56–0.76). The AUC for the GAD-7 was significantly greater than for the AIR (P = 0.014). Sensitivity was not statistically different among the questionnaires: 77% for the GAD-7, 63% for the HADS-A, and 66% for the AIR. The HADS-A had the highest specificity, 85%, which was significantly higher than that of the GAD-7 (77%; P, 0.001) and the AIR (65%; P, 0.001); GAD-7 specificity was higher than AIR specificity (P, 0.001). Conclusions: Symptoms of anxiety among patients with COPD as identified by screening questionnaires were common and significantly higher than the prevalence of anxiety disorder meeting DSM-V criteria. The GAD-7, the HADS-A and the AIR questionnaires had fair to moderate psychometric properties as screening tools for anxiety in individuals with COPD, indicating the need for improved measures for this patient population.