Psychometric validation of the severity of chronic cough diary, leicester cough questionnaire, and a cough severity visual analogue scale in patients with refractory chronic cough.

Abstract

Background: Refractory chronic cough (RCC) is commonly reported in primary care and associated with significant morbidity. Patient-reported outcome (PRO) measures are important for evaluating the efficacy of antitussive medications for RCC in clinical trials from the patient-perspective. Psychometric properties of Severity of Chronic Cough Diary (SCCD) Cough Severity and Cough Frequency, Leicester Cough Questionnaire (LCQ) Total and Physical Domain and Cough Severity Visual Analogue Scale (VAS) scores using data from a 12-week Phase 2b trial evaluating the efficacy of eliapixant in patients with RCC (NCT04562155) are reported.

Results: Quality of completion for the SCCD, LCQ and Cough Severity VAS across the study was high, no ceiling or floor effects were observed at baseline. Internal consistency for LCQ Total and Physical domain scores was also high (Cronbach's alpha = 0.939 and 0.806, respectively). SCCD Cough Frequency and Cough Severity, LCQ Total and Physical domain, and Cough Severity VAS scores demonstrated strong test-retest reliability (Intraclass correlation coefficient ≥ 0.848) among participants defined as stable between Week 3 and Week 4 according to Patient Global Impression of Severity (PGI-S) ratings and Awake Cough Count readings. Construct validity was supported by known-groups comparisons, with large differences (effect sizes 1.99-4.16) observed between groups categorized according to PGI-S ratings and objective Awake Cough Counts. Ability to detect improvement was supported by large effect sizes (≥0.8) observed for mean changes in SCCD, LCQ and Cough Severity VAS scores from baseline to Week 12 among participants classified as 'improved' according to PGI-S/PGI-C ratings and Awake Cough Counts. Triangulated thresholds (score range) for meaningful within-patient improvement based on anchor-based assessments were -0.82 for SCCD Cough Frequency (0-4), -0.69 for SCCD Cough Severity (0-4), 2.36 for the LCQ Total (3-21), 0.77 for the LCQ Physical (1-7) and -17.73 for the Cough Severity VAS (0-100) scores.

Conclusion: Findings support the reliability, validity and responsiveness of the newly developed SCCD Cough Frequency and Severity items as fit-for-purpose PRO measures of cough frequency or severity for use in drug development programs within RCC. The LCQ Total, LCQ Physical Domain and Cough Severity VAS also exhibit acceptable measurement properties for use in this population.