Validation of a Dyspnea Visual Analog Scale in Fibrotic Interstitial Lung Disease

Rationale: A visual analog scale (VAS) is a simple and easily administered tool for measuring the impact of disease; however, little is known about the use of a dyspnea VAS in interstitial lung disease (ILD). Objectives: To validate the use of a dyspnea VAS in a large and heterogeneous cohort of patients with fibrotic ILD, including its minimal clinically important difference (MCID), responsiveness to change, and prognostic significance. Methods: Patients with fibrotic ILD were identified from a large prospective registry. The validity of a 100-mm dyspnea VAS was assessed by testing its correlation in change score with other measures of ILD severity, including the University of California San Diego Shortness of Breath Questionnaire, the King's Brief Interstitial Lung Disease quality of life questionnaire Breathlessness and Activities Domain, the European Quality of Life VAS, forced vital capacity, and diffusing capacity of the lung for carbon monoxide. The responsiveness of the dyspnea VAS was qualitatively confirmed on the basis of there being an observable difference in the change in dyspnea VAS across tertiles of change in anchor variables. The MCID in dyspnea VAS was calculated using both anchor (linear regression) and distribution (one-half standard deviation) approaches, with anchors including the above variables that had a correlation with dyspnea VAS (vertical bar r vertical bar> 0.30). The association of dyspnea VAS with time to death or transplant was determined. Results: The cohort included 826 patients with fibrotic ILD, including 127 patients with follow-up measurements at 6 months. The mean baseline dyspnea VAS was 53624mm. Dyspnea VAS change scores were moderately correlated with the University of California San Diego Shortness of Breath Questionnaire (vertical bar r vertical bar = 0.55) and the King's Brief Interstitial Lung Disease quality of life questionnaire Breathlessness and Activities Domain (vertical bar r vertical bar= 0.44) and weakly correlated with the European Quality of Life VAS (vertical bar r vertical bar = 0.19), forced vital capacity percent predicted (vertical bar r vertical bar = 0.21), and diffusing capacity of the lung for carbon monoxide percent predicted (vertical bar r vertical bar= 0.05). The MCID was 2.7 to 4.5 using the more reliable anchorbased methods and 12.0 based on distribution-based methods. Dyspnea VAS was associated with time to death or transplant in unadjusted models and after adjustment for age and sex (hazard ratios, 1.16 and 1.15, respectively; P, 0.05 for both). Conclusions: This study provides support for the use of the dyspnea VAS in patients with fibrotic ILD, with an estimated anchor-based MCID of 5 mm.