Abstract
Background: The minimal clinically important difference (MCID) for fractional exhaled nitric oxide (FeNO) in asthma is uncertain. The American Thoracic Society (ATS) suggests this may be a relative change ≥ 20%; however, this recommendation is not validated against a clinical outcome measure. A secondary analysis of two randomised controlled trials (RCTs) examining associations between changes in FeNO and the Asthma Control Questionnaire-5 (ACQ-5) may help estimate the MCID.
Methods: The PRACTICAL and Novel START were 52-week RCTs that compared as-required short-acting beta-agonists with or without maintenance inhaled corticosteroids (ICS) versus as-required ICS-formoterol in mild-moderate asthma. In this secondary analysis, participants with measurements of FeNO and ACQ-5 were included. FeNO was analysed on the logarithm transformed scale. Logistic regression assessed associations between FeNO and ACQ-5 changes from baseline to end of study, relative to the ACQ-5 MCID (0.5 points). A t-test was used to compare the difference in log FeNO for those whose ACQ-5 fell by the MCID for an improvement in ACQ.
Results: Data from 1553 participants were included. There was a weak association between the change in FeNO and change in ACQ-5; correlation coefficient was 0.08 (p = 0.002). The mean (SD) change in log FeNO for those with a clinically important improvement in ACQ-5 was -0.25 (0.64), N = 530; geometric mean ratio was 0.78, representing a 22% change in the geometric mean of FeNO. The change in log FeNO was -0.14 (0.60) for those without an improvement in ACQ-5, N = 852; geometric mean ratio was 0.87. The sensitivity and specificity for a 20% change in geometric mean FeNO were 47% and 57%, respectively.
Conclusion: Changes in FeNO are a poor surrogate for changes in ACQ-5. The magnitude of FeNO changes in participants with an MCID improvement in ACQ-5 provides weak support for the ATS recommendation that the MCID for FeNO may be about a 20% relative change.