Answers are grounded in this paper and cite their sources.
BACKGROUND: Exacerbations are a common outcome measure in chronic obstructive pulmonary disease (COPD)-related studies, despite limitations. COPDCompEx, a composite endpoint that includes moderate and severe exacerbations, acute worsening events and early treatment discontinuations, was developed to address these limitations. This post hoc analysis evaluated COPDCompEx as a rate endpoint. METHODS: Data from seven randomised controlled trials (three development, four test) were used to refine the COPDCompEx rate endpoint. Acute worsening events, moderate exacerbations, severe exacerbations and early treatment discontinuations were considered COPDCompEx events. Acute worsening events were algorithmic combinations of deteriorations in peak expiratory flow, reliever use and symptoms. Aspects of COPDCompEx, used as a rate endpoint, were tested with combinations of model (negative binomial, Andersen-Gill, Cox proportional hazard frailty survival) and inter-event gaps (7, 10 or 14 days). Estimated minimum sample size, treatment effect ratios and annualised event rate were compared between COPDCompEx and exacerbations. RESULTS: COPDCompEx, with a negative binomial model and 14-day inter-event gap, proved optimal based on treatment effect ratios (exacerbations) and estimated minimum sample size. COPDCompEx had an annualised event rate 3.0-times higher in development trials, and 4.5-times higher in test trials, compared with exacerbations. Season- or geography-related treatment effect variations were generally smaller than or comparable to those seen with exacerbations. CONCLUSIONS: COPDCompEx is a viable rate endpoint for clinical trials where time or recruitment are limiting factors. This endpoint generates significantly more events than exacerbations alone and may, therefore, reduce variation in treatment efficacy resulting from seasonal/geographical variation. TRIAL REGISTRATION: NCT00206154, NCT00206167, NCT00419744, NCT00363896, NCT00358436, NCT01854645, NCT01854658.
No takes yet. Share an insight, caveat, or question.
Spata et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: