A systematic method for validation of gas exchange measurements.

The measurement of gas exchange is useful, but thus far, has not been practical during the mechanical ventilation of critically ill patients. To validate two new commercial instruments, (Siemens-Elema Servo Ventilator 900B, Beckman Metabolic Cart), the authors constructed a lung model into which they delivered CO2 and N2 at precise rates to simulate Co2 production (Vco2) and O2 consumption (Vos). The model consists of 13.5-1 gas jar with an attached one liter anesthesia bag. The lung model was ventilated at present tidal volumes and frequencies. The authors also compared the measured respiratory quotient (RQ) with the known RQ of burning methanol (RQ = 0.67) in the jar. When the model was ventilated with levels of tidal volume and gas exchange applicable to adults, both instruments measured V02 within 5 to 13% of predicted values. Varying the FI02 did not significantly affect this accuracy. At tidal volumes below 350 ml, the difference increased between predicted VCO2 and measured VCO2. The difference between measured vs. the actual RQ of methanol was 5 and 1.5% in the Siemens-Elema and Beckman Systems, respectively.