Nonsmokers# | Smokers# | |||||||||
Total | D ≤500¶ | D >500¶ | ACQ <2+ | ΔD 0§ | Total | D ≤500¶ | D >500¶ | ACQ <2+ | ΔD 0§ | |
Total events n | 643 | 306 | 337 | 432 | 301 | 92 | 35 | 57 | 53 | 47 |
Positive cases n | 161 | 64 | 97 | 130 | 99 | 26 | 11 | 15 | 17 | 14 |
Prevalence | 25 | 21 | 29 | 30 | 33 | 28 | 31 | 26 | 32 | 30 |
Sensitivity | 51 | 67 | 42 | 54 | 48 | 67 | 70 | 64 | 71 | 57 |
Specificity | 76 | 76 | 78 | 73 | 70 | 77 | 84 | 91 | 86 | 73 |
PPV | 37 | 43 | 44 | 47 | 34 | 52 | 78 | 43 | 71 | 47 |
NPV | 84 | 90 | 77 | 79 | 80 | 86 | 87 | 85 | 86 | 80 |
Accuracy | 70 | 74 | 68 | 67 | 65 | 74 | 86 | 70 | 81 | 68 |
p-valueƒ | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | 0.037 | <0.001 | 0.025 |
Data are presented as %, unless otherwise indicated. PPV: positive predictive value; NPV: negative predictive value. #: cut-off value of +30%; ¶: inhaled corticosteroid (ICS) dose (D) in μg equivalents of beclomethasone diproprionate·day−1; +: ACQ <2 tested the subgroup with an initial ACQ score <2; §: ΔD 0 tested the subgroup without treatment modification between consecutive visits; ƒ: the p-value is the statistical significance of rejecting an area under the curve of 0.5. A positive event is defined as a worsening of asthma control. A true-positive case was defined as an exhaled nitric oxide fraction (FeNO) change of more than or the same as the cut-off value associated with a worsening of asthma control between consecutive visits. When smoking patients were treated with a high ICS dose, FeNO ability to detect a worsening of control was somewhat reduced (p = 0.037).