[Airway challenge testing - accuracy of the interrupter technique]

According to national and international recommendations the bronchial sensitivity should be determined based on the decrease of the FEV1 by 20 % (FEV1 - 20) or the increase of the airway resistance by means of body plethysmography by 100 % (Raw + 100). Measurement of airway resistance by interrupter technique (Rint) is a simple method and needs no active cooperation of the patient, but is not recommended in airway challenge testing. We investigated the role of the increase of Rint by 100 % (Rint + 100) compared to Raw + 100 and FEV1 - 20 during carbachol airway challenge testing by means of dosimetry. We examined 123 patients with following symptoms: 85 x coughing, 31 x coughing and dyspnea, 7 x medical opinion. Significant correlations between Rint and Raw were found before and after the challenge tests (Rint before/after 0,3 +/- 0,13/0,36 +/- 0,25 kPa*s/l; Raw before/after 0,24 +/- 0,09/0,50 +/- 0,41 kPa*s/l; r = 0,504/0,672; p < 0,001 [Pearson]). The median values of Rint and Raw were significantly different (p < 0,001 [Wilcoxon]). Moreover Rint systematically overestimated airway resistance in the normal range and underestimated the increase of airway resistance during challenge testing (r = 0,783; p < 0,001 [Pearson]). In 58 patients an increased airway responsiveness was found. In 21 oft these patients there was no increase of Rint above the initial value. Sensitivity/specificity/positive predictive value/negative predictive value in % to the detection of airway hyperresponsiveness were in Rint + 100 9/95/63/54, in FEV1 - 20 61/100/100/66 and in Raw + 100 98/100/100/98. In conclusion we found significant correlations between Rint and Raw, but the median values were systematically and significantly different. Rint + 100 had a low sensitivity to detect airway hyperresponsiveness and is not comparable with FEV1 - 20 or Raw + 100.