Development of a Novel Infant Volumetric Capnography Simulator: Making the Invisible Visible Improves Understanding and Safety.

Abstract

An infant lung simulator was fed with CO2 supplied by a mass flow controller (VCO2-IN) and ventilated using standard settings. A volumetric capnograph was placed between the endotracheal tube and the ventilatory circuit. We simulated ventilated babies of different body weights (2, 2.5, 3, and 5 kg) with a VCO2 ranging from 12 to 30 mL/min. The correlation coefficient (r2), bias, coefficient of variation (CV = SD/x × 100), and precision (2 × CV) between the VCO2-IN and the elimination of CO2 recorded by the capnograph (VCO2-OUT) were calculated. The quality of the capnogram's waveforms was compared with real ones belonging to anesthetized infants using an 8-point scoring system, where 6 points or greater meant that the simulated capnogram showed good, 5 to 3 points acceptable, and less than 3 points an unacceptable shape. The correlation between VCO2-IN and VCO2-OUT was r2 = 0.9953 (P < 0.001), with a bias of 0.16 (95% confidence intervals from 0.12 to 0.20) mL/min. The CV was 5% or less and the precision was 10% or less. All simulated capnograms showed similar shapes compared with real babies, scoring 6 points for 3 kg and 6.5 points for 2-, 2.5-, and 5-kg babies. The simulator of volumetric capnograms was reliable, accurate, and precise for simulating the CO2 kinetics of ventilated infants.