[The reliability of proximal and distal gas samples in the capnometry of infants]

To compare the accuracy of end-tidal pCO2 measurement derived from the connector (petCO2 proximal) and from the tip of the endotracheal tube (petCO2 distal), we measured paCO2, petCO2-proximal and distal in 42 infants (weight: 2500-9500 g) manually ventilated with an Ayre's t-piece breathing circuit (Kuhn modification). For distal gas sampling we used a commercially available "uncuffed tracheal tube with monitoring lumen" Mallinckrodt, Hennef, with a separate monitoring lumen leading to the distal tip of the endotracheal tube. Compared with paCO2 (26.2 +/- 5.3 mmHg), proximal gas sampling resulted in a significantly lower petCO2 (17.3 +/- 6.0; P less than 0.001) and a large arterial-end-tidal pCO2 difference (pa-etCO2 9.0 +/- 6.8 mmHg). The correlation between paCO2 and petCO2 was poor (r = 0.29; n.s.). In infants weighing less than 5000 g the arterial-end-tidal pCO2 difference was found to be especially high (12.1 +/- 5.1 mmHg); it was slightly smaller in infants with more than 5000 g (5.6 +/- 7.0 mmHg). Distal petCO2 values approximated arterial pCO2 much better (petCO2 26.0 +/- 5.8 mmHg; diff pa-etCO2 0.4 +/- 3.0 mmHg, r = 0.86) even in very small children. In 12 children (weight: 2500-8000 g) ventilated with a circuit (Dräger-Kreissystem) accurate petCO2 measurements were obtained from both the proximal and the distal sampling site.

When an Ayre's t-piece is used for ventilation of children weighing less than 10 kg, distal gas sampling for petCO2 measurement should be preferred. Proximal gas sampling does not allow correct capnometry except in a circuit with one-way valves.