The Development and Validation of a Novel Anxiety Scale to Measure and Screen Anxiety in Patients with COPD

There is a high prevalence of co-morbid anxiety disorders in patients with chronic obstructive pulmonary disease (COPD). Although co-morbid anxiety impacts negatively upon health-related quality of life, physical functioning and healthcare utilisation, anxiety disorders remain significantly under recognised and undermanaged. One reason for this may be the lack of a validated diseasespecific patient-reported screening tool and outcome measure. Existing scales may be limited by their inclusion of somatic items, which may overlap with symptoms of COPD or the side-effects of medications. This thesis aimed to develop a novel non-somatic self-report anxiety scale that can be used to screen for anxiety disorders and assess the severity of anxiety in patients with COPD. The Anxiety Inventory for Respiratory disease (AIR) was developed using a mixed methods approach to item development that incorporated both emic (interviews with 14 COPD patients) and etic (review of extant anxiety scales) perspectives to generate 16 novel items scored using a Likert-type response set. Patients and clinicians were involved in the development of the AIR to ensure that the scale is user-friendly and clinically relevant. Qualitative findings from the interviews also provide a unique insight into the experience of anxiety from the patients’ perspective and support the non-somatic format of the AIR. The draft 16-item AIR was completed by 88 patients with COPD and refined through rigorous item and factor analysis. Six items were removed to create the final 10-item AIR (score range 0-30). The reliability and validity of the AIR were examined in a sample of 56 COPD outpatients. The AIR proved to have excellent internal consistency in all phases (Cronbach’s α = 0.92-0.95) and test-retest reliability (Intraclass correlation coefficient = 0.81). The AIR also demonstrated high convergent validity with the Hospital Anxiety and Depression Scale (Spearman’s rho correlation = 0.91) and was able to discriminate between patients with and without anxiety disorders (p