In vitro and in vivo assessment of the Ventrak 1550/Capnogard 1265 for single breath carbon dioxide analysis in neonates

The Ventrak 1550/Capnogard 1265 (V&C) enables deadspace (VD) measurements t o be made in neonates. The aim of our studies was to validate the V&C devic e for VD measurement in vitro (lung model) and in vivo (adult rabbits). Met hods of measurement of VD using the V&C (automatic computation, interactive carbon dioxide-volume plot analysis, Bohr equation) were tested by compari ng known added deadspace volumes (VDadd) with calculated VDadd. After produ cing a change in alveolar (VDalv) and physiological (VDphys) deadspace by i n vivo bronchoalveolar lavage, VDalv and VDphys computed automatically were compared with values calculated by the Bohr-Enghoff equations. VDadd was s lightly underestimated (absolute error in mean: automatically -0.61 ml inte ractively -0.55 ml; Bohr -0.54 ml). The higher the VDadd, the lower the abs olute errors and coefficients of variation (cv). The highest cv occurred fo r automatic analysis (approximately 11%) compared with <6% for interactive analysis or the Bohr equation. Average differences between results calculat ed automatically and by the Bohr-Enghoff equation were -0.79 ml for VDalv ( 95% confidence interval -2.02 to 0.44 ml) and -0.23 ml for VDphys (-0.6 to 0.14 ml). We conclude that the V&C can be used in newborn infants undergoin g mechanical ventilation. If changes in VD are <5 ml, interactive analysis or the Bohr equation should be used.