OBJECTIVES: To evaluate psychometric properties of the Saint George’s Respiratory Questionnaire Symptoms Domain (SGRQ-SD) and the Quality of Life-Bronchiectasis Respiratory Symptoms Domain (QOL–B-RSD) using data of a phase 3 randomized, placebo-controlled, multicenter study of Ciprofloxacin DPI administered for 28 days on/28 days off or 14 days on/14 days off compared to placebo for the treatment of non-cystic fibrosis bronchiectasis (NCFB).

METHODS: Confirmatory factor analysis (CFA), item response theory (IRT), reliability, criterion validity, construct validity, ability to detect change (responsiveness), and responder definition (RD) thresholds using anchor- and distribution-based methods were assessed for the SGRQ-SD and QOL-B-RSD.

RESULTS: CFA and Rasch analysis indicated misfit of the SGRQ-SD and item response categories were poorly distinguished. The Cronbach’s alpha for internal consistency of the SGRQ-SD was moderate (0.646), and convergent validity with dyspnea severity was moderate (0.44). For the QOL-B-RSD, analysis indicated misfit and poorly distinguished response categories for three of nine items. Internal consistency and test-retest reliability of the QOL-B-RSD was good (Cronbach’s alpha=0.81, ICC=0.70), and convergent validity with dyspnea severity was moderate (-0.46). The known-groups validity of the SGRQ-SD and QOL-B-RSD was supported with scores significantly distinguished among the levels of dyspnea severity, and responsiveness of both instruments was supported by the dyspnea criterion measure. RD analyses were limited by the instruments’ relationship with the anchors available in this phase 3 trial.

CONCLUSIONS: The classic psychometric analyses performed demonstrated good reliability and validity estimates, known groups validity, and responsiveness when dyspnea severity was used as a criterion variable. However, items of the SGRQ-SD demonstrated poorly-ordered responses, or did not distinguish well among respondents per CFA and IRT analyses. Additional qualitative and quantitative research is needed for further development of NCFB patient-reported outcomes.