Quantitative Validity of the Sgrq in Patients with Severe Asthma

Rationale: The St George’s Respiratory Questionnaire (SGRQ) was developed for use in asthma as well as COPD, but there are limited data to demonstrate the psychometric properties of the SGRQ in patients with severe uncontrolled asthma.

Methods: Psychometric properties of the SGRQ were tested using treatment-masked data from two clinical trials (MENSA, SIRIUS) of mepolizumab, an anti-IL5 antibody, including: item and scale characteristics, reliability (test-retest and internal consistency), validity (convergent, discriminant and known groups), and responsiveness. Analyses were conducted on treatment masked data guided by an analysis plan developed post-hoc after study unblinding.

Results: Internal consistency reliability was acceptable for the total score and all domains [r=0.71 (symptoms) - 0.94 (total)]. Test-retest reliability was acceptable for the total score and 2 domains; the symptoms domain approached acceptable levels (r=0.60-0.68). Construct validity was demonstrated with moderate to strong correlations with asthma symptoms score and the Asthma Control Questionnaire (ACQ-5) (r=0.40-0.61 & 0.57 to 0.72, respectively); discriminant validity showed low correlations with FEV₁ (% predicted), work absenteeism and nasal symptoms (|r|=0.08-0.33, 0.01-0.34 & 0.02-0.33, respectively). Known groups validity was shown with significant differences in predefined groups of asthma control, baseline exacerbation history, exacerbation counts, and eosinophil counts. Mean SGRQ change score for ratings of mild improvement based on a patient global question were 3.4 and 7.1 compared to the established 4-point minimal clinical important difference.

Conclusions: This analysis suggests that the SGRQ has acceptable measurement properties in a severe asthma population with history of frequent exacerbations. Funding: GSK (NCT01691508, NCT01691521)