Accuracy and sensitivity of the interrupter technique for measuring the response to bronchial challenge in normal subjects.

The interrupter technique is a non-invasive method for measuring airway calibre. Since the calculation of interrupter resistance (Rint) is critically dependent upon the analysis of the mouth pressure/time (Pmo(t)) curve obtained after flow interruption, we wanted to assess the relative merits of four different analyses of Pmo(t) curves, obtained under basal conditions and following methacholine-induced airway narrowing, in 10 healthy adults. Four methods of analysing the Pmo(t) curves were used to calculate Rint values: RintC-a smooth curve fit with back-extrapolation; RintL-two-point linear fit with back-extrapolation; RintEO-calculated from the pressure change after the post-interruption oscillations had decayed (end-oscillation); and RintEI-calculated from the pressure change at the end of the period of interruption. The airway response measured with the four Rint methods was compared with plethysmographic airway resistance (Raw). The sensitivity of the methods was determined by calculating a sensitivity index (SI), the change in resistance after challenge expressed in multiples of baseline standard deviation. Values of RintC were similar to Raw values under all conditions. Resistance values from the remaining Rint methods significantly exceeded Raw (mean basal difference: 0.13-0.34 kPa.l-1 x s; mean difference after challenge: 0.12-0.42 kPa.l-1 x s. Raw was the most sensitive method for detecting bronchoconstriction (doubling of Raw was equivalent to SI of 10.5). Of the Rint methods, RintEI gave the highest sensitivity index (SI = 3.1), with a 42% mean change; RintC produced the greatest proportionate change after challenge (55%), but with a lower SI (2.2).(ABSTRACT TRUNCATED AT 250 WORDS)