ObjectiveTo determine the effect of heliox, nitric oxide (NO), and perfluorocarbon on differential pressure pneumotachometer characteristics and to determine the effect of heliox on volumes delivered by the Siemens S900C (S900C), and Servo Ventilator 300 (SV300) ventilators.DesignProspective, laboratory study.SettingPulmonary laboratory of a tertiary care, nonprofit children's hospital.ApparatusSV300, S900C ventilator, differential pressure pneumotachometer.InterventionsDual pneumotachometers were connected in series to a 0.5-L calibration syringe and a 1-L anesthesia bag creating a closed system. Calibration of the pneumotachometers was done in room air at ambient temperature with 100 strokes. Accepted accuracy of measured volumes is within 0.5%. Flow-conductance curves were constructed using 100 strokes each for heliox (70:30 mixture), NO, and perfluorocarbon. Expired gases of room air and a 70:30 mixture of heliox from the above ventilators were collected into a nondiffusing gas collection bag, and the volume was measured in a chain-compensated gasometer. Ten sets of 500-mL breaths (20 breaths each set) and 100-mL breaths (40 breaths each set) were collected. The paired Student's t-test was used to detect significant differences in measured volumes, with significance defined as p < .01.Measurements and Main ResultsVolumes measured with the pneumotachometer using 25 ppm of NO, 50 ppm of NO, and perfluorocarbon were within +0.25%, -0.7%, and +0.4%, respectively (p = .155, p = .001, p = .06). Heliox decreased the conductance of the pneumotachometer, thereby increasing the measured volume by 15% (p < .001). However, heliox did not affect its linearity. Heliox had no affect on volumes delivered by the S900C. However, the SV300 delivered 7.9% less volume of heliox at a set tidal volume of 500 mL and 10.8% less at a set tidal volume of 100 mL.ConclusionsA 70:30 mixture of heliox caused a significantly overestimated gas volume measured and, therefore, an underestimated gas volume delivered by SV300. NO at 25 ppm and perfluorocarbon did not interfere with the accuracy of a differential pressure pneumotachometer. However, at 50 ppm, NO caused a difference in measured gas volume that was statistically, but not clinically, significant. Application of pneumotachometers in critically ill children receiving heliox requires recalibration. Heliox did not affect volumes delivered with the S900C ventilator. Although volumes delivered with the SV300 were significantly reduced by heliox, the difference can be corrected easily by increasing minute ventilation until expired tidal volume equals desired tidal volume. (Crit Care Med 1999; 27:1603-1607)