VALIDATION OF A NITROGEN WASHOUT SYSTEM TO MEASURE FUNCTIONAL RESIDUAL CAPACITY IN PREMATURE-INFANTS WITH HYALINE-MEMBRANE DISEASE

A multiple-breath nitrogen washout system designed to measure lung vol ume in mechanically ventilated infants was validated by assessing thre e performance criteria: 1) accuracy of lung volume measurements in the presence of an endotracheal tube leak was assessed by comparing the m easurements of functional residual capacity (FRC) in a mechanical lung model with and without airway leak; 2) in vivo accuracy was assessed in rabbits by comparing FRC measurements obtained by this system with measurements obtained by helium dilution; and 3) in vivo precision was assessed by analyzing measurements of FRC obtained in replicate measu rements at different times in ventilator-dependent premature infants w ith hyaline membrane disease. The average difference between the measu rements of FRC in a mechanical lung model with airway leak and without leak was 3.0 +/- 9.4% (mean +/- SD, P > 0.2), and no difference was g reater than 20%. There was a significant correlation between the measu rements of FRC in rabbits by nitrogen washout and by helium dilution ( r = 0.93, P < 0.0001), and 65.4% of the paired measurements were withi n 20% of their average. The 95% limits of agreement within pairs of me asurements by the two techniques ranged from -4.0 to +6.5 mL/kg. FRC m easured by helium dilution was slightly higher(1.3 +/- 2.7 mL/kg, P < 0.01) than FRC measured by nitrogen washout, and positive end-expirato ry pressure was a significant predictor of this difference (P < 0.0001 ). The regression between the individual FRC measurements obtained in premature infants and the average of the other replicates was signific ant (r(2) > 0.98, P < 0.0001). The coefficient of variation was 12.3%. These findings provide further validation of this multiple-breath nit rogen washout system for measuring FRC in premature infants during mec hanical ventilation. (C) 1995 Wiley-Liss, Inc.