MEASURING OXYGEN-UPTAKE AND CARBON-DIOXIDE PRODUCTION IN CRITICALLY ILL PATIENTS USING A STANDARD BLOOD-GAS ANALYZER

Objective: The measurement of oxygen uptake and CO2 production in crit ically ill patients requires invasive monitoring or complex analysis e quipment. This study investigates the hypothesis that oxygen uptake an d CO2 production can be accurately determined by measuring oxygen and CO2 concentrations in samples from inspiratory and expiratory ventilat or circuitry, using a standard blood gas analyzer. Design: Prospective comparison of CO2 production and oxygen uptake measurements determine d by use of a blood gas analyzer vs, a mass spectrometer. Setting: Uni versity teaching hospital medical and surgical intensive care units (I CUs). Patients: Critically ill patients (n = 46) receiving mechanical ventilation in the ICUs. Interventions: Po-2 and Pco(2), were obtained with two new techniques and compared simultaneously with measurements on a mass spectrometer in critically ill, mechanically ventilated pat ients. Two methods were evaluated: a) arterial blood gas analyzer meas urements of Po-2 and Pco(2) from fluid collected in traps on the inspi ratory and expiratory limbs of the ventilator circuitry; b) Po-2 and P co(2) measurements of inspiratory and expiratory gas samples collected in bags and injected directly into an arterial blood gas analyzer. Ox ygen consumption and CO2 production were compared, using both methods of gas measurements. Measurements and Main Results: Direct injection o f gas samples collected in a bag from inspiratory and expiratory limbs of a breathing circuit into the arterial blood gas analyzer correlate d very closely with mass spectrometer measurements for all variables ( n = 32 sample measurements in 25 patients): fractional oxygen (r(2) = .99, Slope = 1.02, bias = 0.37%, precision = 0.54), fractional expired CO2 (r(2) = .90, slope = 0.86, bias = -0.10%, precision = 0.15), oxyg en uptake (r(2) =.87, slope = 0.99, bias = 21.6 mL/min, precision = 38 .0), and CO2 production (r(2) =.98, slope = 0.95, bias = 7.90 mL/min, precision = 15.3). In contrast, although fractional oxygen and CO2 con centrations were approximated by analysis of fluid collected from insp iratory and expiratory traps, the values did not correlate well enough with mass spectrometer values to yield reasonable oxygen uptake or CO 2 production results. Conclusion: We have demonstrated that direct Pic k oxygen uptake and CO2 production can be accurately determined in mec hanically ventilated patients, using direct injection of collected gas samples into standard blood gas analyzers, This simple, inexpensive t echnique can be performed using equipment readily available in any hos pital.