Résumé
Plethysmographic airway resistance (Raw) and specific airway resistance (sRaw) are widely used to assess baseline lung function and bronchodilator reversibility in children unable to perform spirometry. However, no consensus exists on cut-offs for significant reversibility. We conducted a retrospective study on a large cohort of asthmatic children tested during panting without electronic thermal compensation, to determine optimal Raw and sRaw cut-offs for detecting significant forced expiratory volume in 1 s (FEV 1 ) reversibility. A smaller cohort undergoing tidal breathing plethysmography with electronic thermal compensation was also analysed. From 2436 tests without thermal compensation, ROC curve analysis identified cut-offs of −34.4% (Raw tot ) and −36.0% (sRaw tot ) from baseline to detect significant FEV 1 reversibility, with sensitivities of 70% [95%CI: 65%;74%] and 72% [95%CI: 67%;76%] and specificities of 70% [95%CI: 68%;72%] and 73% [95%CI: 71%;75%], respectively. Expressing reversibility as a percentage of predicted did not improve accuracy but required larger cut-offs (−42.6% for Raw tot , −56.1% for sRaw tot ) and increased the influence of baseline value. In the 106 tests performed using electronic thermal compensation, Raw tot/eff and sRaw tot/eff cut-offs were lower to that without electronic thermal compensation (−25% of baseline) with sensitivities and specificities ranging from 63% to 70%. In conclusion, Raw tot and sRaw tot reversibility effectively identify FEV 1 reversibility with cut-offs around −35% of the baseline value, showing no added benefit when using larger cut-offs expressed as a percentage of the predicted value. When electronic thermal compensation is applied, it may necessitate lower cut-offs of −25% of the baseline, compared to those determined without it.