Although not necessary for most day-to-day clinical management, precise quantitation of the volume of gas leaked through a bronchopleural fistula during mechanical ventilation is important in research on the natural history or therapy of such leaks. We could find no previous studies evaluating the accuracy of the various devices available for such quantitation in the clinical setting and so constructed an experimental system in which three such devices measured a controlled air leak simultaneously. Using a Tissot 120-liter water-sealed spirometer as a reference standard, we tested a heated pneumotachometer, a Bourns LS-75 ventilation monitor, and a Wright Respirometer, using leaks of 1.07 to 8.66 liters per minute. The pneumotachometer agreed with the Tissot spirometer within 5% except at the lowest flow tested, while the other two devices indicated leak volumes that differed from the standard by at least 10% on 8 of 18 measurements. The Bourns LS-75 indicated only 60.0 and 8.4% of the two lowest leak volumes, respectively. Accuracy of the device used for quantitating air leak should be taken into consideration in studies of bronchopleural fistula during mechanical ventilation, especially when the leaked volumes are small. (Respir Care 1986;31:1065-1068.)
Larson et al. (Sat,) studied this question.