Psychometric Properties of the Asthma Symptom Index in Patients with Severe Asthma.

Abstract

Evaluation of symptoms is essential in asthma clinical research. Daily symptom diaries require once- or twice-daily recording, making them less suited to real-world use. A modified version of the established Asthma Symptom Utility Index (ASUI) that includes symptom attributes (Asthma Symptom Index [ASI]) was developed as a less-burdensome measure of asthma symptoms. To evaluate the ASI and replicate ASUI psychometric properties, including validity, reliability, responsiveness, and minimal clinically important difference (MCID) estimation in patients ≥12 years of age with severe asthma. Patients from a randomized trial (MUSCA [n = 497]) and a cross-sectional study (IDEAL [n = 721]) were analyzed post hoc. Demographic information, spirometry, ASI and ASUI scores, and other patient-reported outcome measures such as Asthma Control Questionnaire (ACQ-5) and St George's Respiratory Questionnaire (SGRQ) at baseline and during follow-up (MUSCA only) were obtained. Internal consistency reliability and test-retest reliability were considered good (>0.70 [Cronbach's alpha] and 0.87-0.90 [intraclass correlation]). ASI/ASUI scores correlated strongly with ACQ-5 and SGRQ scores (spearman correlation [r_s] magnitude: 0.67-0.85). ASI and ASUI scores differed for asthma control (defined by ACQ-5) and lung function (% predicted forced expiratory volume in 1 second). Changes in ASI and ASUI scores from baseline to week 4 and week 12 had high correlations with changes in ACQ-5 (r_s magnitude: 0.57-0.69). MCIDs ranged from -0.42 to -0.26 (ASI) and 0.07 to 0.11 (ASUI). Findings show the good reliability, validity, and responsiveness of the ASI, indicating potential value for real-world symptom assessment in severe asthma.