Comparison of different methods for dead space measurements in ventilated newborns using CO 2 -volume plot

Objective: The aim of the study was to test the applicability of Ventrak 1550/Capnogard 1265 (V-C) for respiratory dead space (Vd) measurement and to determine anatomic (VDana), physiologic (VDphys), and alveolar dead spaces (VDalv) in ventilated neonates. Design: Prospective study. Setting: Neonatal intensive care unit. Patients: 33 investigations in 22 ventilated neonates; median gestational age 34.5 weeks (range 27–41), median birthweight 2658 g (range 790–3940). Method: The single-breath CO2 test (SBT-CO2) and transcutaneous partial pressure of carbon dioxide (PCO2) were recorded simultaneously and Vd was determined (1) automatically (V-C software), (2) by interactive analysis of the PCO2 volume plot, and (3) manually by Bohr/Enghoff equations using data obtained by V-C. Results: Vd measurements were possible in all cases by method 3 but not possible by methods 1 and 2 in 22 of 33 investigations (67 %), especially in preterm neonates, because of disturbed signals. VDana/kg (1.6 ± 0.6 ml/kg, mean ± SD), VDana/tidal volume (Vt) (0.36 ± 0.09) were lower compared to published data in spontaneously breathing infants, whereas VDphys/kg (2.3 ± 0.9 ml/kg) and VDphys/Vt (0.50 ± 0.12) are comparable to data obtained from the literature. Five minutes after insertion of the sensor (dead space 2.6 ml) into the ventilatory circuit, the transcutaneous PCO2 rose above baseline for 3.2 % (patients > 2500 g) and 5.7 % (patients < 2500 g). The time necessary for one analysis was 50–60 min. Conclusion: In ventilated newborns, dead space measurements were possible only in one-third by SBT-CO2, but in all cases by Bohr/Enghoff equations. Improved software could further reduce the time needed for one analysis.