Abstract
Introduction: Patient-reported outcomes (PROs) are important in COPD for understanding patients’ symptoms and experience of the disease. Responder rates at specific timepoints are frequently used to analyse clinically important improvements in PROs. Alternatively, responders can be defined using average PRO improvement vs. baseline over the whole study period or the last two timepoints, as recently discussed by the FDA.
Aims and objectives: To compare three techniques for determining responder rates of St. George’s Respiratory Questionnaire (SGRQ) and transition dyspnoea index (TDI).
Methods: Two double-blind, replicate, Phase III, 52-week TONADO trials compared tiotropium + olodaterol (T+O) with the mono-components. SGRQ and TDI responder rates were analysed using three different methods: area under the curve (AUC) for the whole duration, and a mean of all timepoints or the last two timepoints.
Results: Response rates were consistent between analyses. Odds ratios of response were similar for each comparison whichever technique was used, and there was an improvement with T+O vs. its mono-components (Figure).
Conclusion: Analysing SGRQ and TDI responders using AUC and mean values allows investigation over a longer period of time and may be a better way to define responders than using only a single timepoint. The results consistently show an improved response of T+O compared to its mono-components.
- Copyright ©the authors 2017
