Reproducibility of visual analog scale measurements of dyspnea in patients with chronic obstructive pulmonary disease

The purpose of this study was to evaluate the reproducibility of visual analog scale ratings of the effort to breathe (VAS(e)) and the degree of discomfort evoked by breathing (VAS(d)) in patients with chronic obstructive pulmonary disease (COPD) during exercise. Six subjects with moderately severe COPD (FEV₁ = 1.12 ± 0.29 L, FEV₁/FVC = 44 ± 4%) underwent progressive incremental exercise testing to a symptom-limited maximum every week for 8 wk. VAS(e) and VAS(d) were highly correlated in each subject (r = 0.99 ± 0.01). The slope of the VAS(d)/VAS(e) relationship for all trials in all subjects was not significantly different from 1, indicating that our subjects were rating a common sensation with the two scales. VAS(e) at maximal exercise was reproducible in every subject; the within-subject coefficient of variation (CV) was 6% (range, 2 to 10%) and compared favorably with physiologic indices: 7% (range, 3 to 12%) for oxygen consumption and 10% (range, 5 to 16%) for minute ventilation (V̇I). In contrast, submaximal VAS ratings were highly variable. At 66% of the maximal work load, the within- subject CV for VAS(e) was 21% (range, 11 to 28%) compared with 6% (range, 4 to 7%) for V̇O₂ (p < 0.003) and 10% (range, 5 to 16%) for V̇I (p < 0.01). VAS(e) correlated linearly with V̇I and V̇O₂ in all subjects in all trials. However, within an individual subject the slope and position of these relationships varied widely between trials. We conclude that although maximal VAS ratings are reproducible, submaximal VAS ratings and the relationship between VAS ratings and physiologic indices vary considerably when exercise tests are performed at weekly intervals in patients with COPD.