Abstract
BACKGROUND: The recall periods and response scales of existing surveys of asthma control are poorly suited for studying acute exacerbations. OBJECTIVE: To develop an instrument able to predict exacerbations after the onset of acute symptoms and with a recall window sufficiently short to study recovery. METHODS: We developed the six-item Acute Asthma Exacerbation Survey (AAES). Data were collected at baseline, acute, and recovery visits within an established longitudinal protocol for participants with severe asthma. Participants scheduled acute study visits at the first sign of a cold. Nasal lavage samples and lung function measurements were also collected. The AAES data were analyzed using Cronbach α, Spearman correlations, and Kruskal-Wallace methods. We used logistic regression for predictors of bursts of oral corticosteroids (OCS). RESULTS: Of 130 participants studied at baseline, 52 returned for an acute visit. The AAES scores were elevated at the acute visit and returned to baseline after recovery independently of respiratory virus detection. Cronbach α for the AAES was 0.853, 0.822, and 0.889 at the three respective visits. Compared with participants not needing burst OCS, those with exacerbations had higher acute AAES scores (16 [13.5-18] vs 11.5 [8.2-14], median [interquartile range]; P = .017) and a larger reduction from baseline in lung function. For each 3-point increase in AAES scores, the odds ratio for burst OCS use was 1.64 (95% CI, 1.04-2.57; P = .030). CONCLUSIONS: The AAES is internally consistent and dynamically responsive during acute asthma exacerbations. Additional validation studies are warranted to support future trials and aid in clinical decision-making.