TY - JOUR T1 - Measuring improvement in dyspnoea: Absolute or relative? JF - European Respiratory Journal JO - Eur Respir J VL - 44 IS - Suppl 58 SP - P4744 AU - Miriam Johnson AU - Martin Bland AU - Stephen Oxberry AU - Amy Abernethy AU - David Currow Y1 - 2014/09/01 UR - http://erj.ersjournals.com/content/44/Suppl_58/P4744.abstract N2 - Background: Whether minimal clinically important differences (MCID) in chronic dyspnoea intensity should be based on absolute or relative measures is debatable. We have shown a moderate effect size of 11.2mm (0-100mm visual analogue scale[VAS]) and patient anchor MCID of 9mm VAS (Johnson et al 2013) but an MCID for relative change is unknown. We addressed both issues.Method: Pooled data analysis from 4 clinical trials of opioids for dyspnoea (213 datasets).1) The variability of difference of dyspnoea intensity from baseline (Dd) against baseline measure (Db) for absolute and ratio values was estimated and displayed graphically.2) Effect size cannot be calculated for relative values, so MCID estimation used a patient anchor (blinded study arm preference).Results: 1) Dd and Db were negatively correlated. Variability of absolute measures was uniform across baseline intensities. Variability of ratio measures was very large for low baseline dyspnoea.2) 113 preference responses were evaluable for 93 participants. Comparison of preferred and un-preferred arm ratios showed a difference of −14.4%, and −20.5% using log ratios.Conclusions: Uniform variation in response for absolute measures and ability to calculate effect size gives preference to absolute measures in sample size calculations. Reported outcomes should be presented as absolute and relative measures. The MCID in chronic dyspnoea expressed as a relative reduction is 15-20% of baseline measures. ER -