The Science Journal of the American Association for Respiratory Care

2010 OPEN FORUM Abstracts


Joseph Orr, Lara Brewer; Anesthesiology, University of Utah, Salt Lake City, UT

Introduction: CO2 excretion (VCO2) monitoring is useful as an indicator of effective ventilation, metabolic activity and pulmonary perfusion. Interpretation of a change in VCO2 subsequent to a change in alveolar ventilation can be difficult because measured VCO2 changes in a transient manner following ventilation adjustments. When ventilation changes occur, the metabolism remains stable while large amounts of stored CO2 are transferred into or out of the body over an extended time. The transferred CO2 is reflected in the transiently changing VCO2 and PaCO2. To facilitate understanding of the VCO2 signal, we developed a computer algorithm that separates the measured VCO2 into two portions: 1) metabolically produced and 2) transferred to and from the tissue stores. Methods: Our system is based on a multi-compartment computer model of CO2 production, storage, distribution and elimination within the body. This computer model simulates the change in stored CO2 volume and partial pressure as ventilation is changed. This computer model simulates the change in stored CO2 volume and partial pressure as ventilation is changed. The computer model inputs are ventilation and VCO2 data collected during the 30 minutes immediately prior to the time of analysis. The model parameters are continuously optimized to minimize the difference between the actual and modeled VCO2. Once optimized, the model is used to predict the future value of end-tidal CO2 (etCO2) and VCO2 following ventilation changes. A volumetric capnometer (NICO2, Philips-Respironics, Wallingford, CT) was used to monitor ventilation, etCO2 and VCO2 in 5 mechanically ventilated piglets. Step changes in respiratory rate or minute ventilation were made periodically. Model predictions of etCO2 and VCO2 were compared to actual data recorded 20 minutes after the prediction time. Results: The difference between actual and predicted etCO2 was 0.0 ± 1.47 mm Hg (n=22). The difference between the predicted and actual VCO2 was 1.55 ± 4.72 mL/min. The squared correlation coefficient for etCO2 was r2 = 0.94 and for VCO2 was r2 = 0.89. Conclusions: Isolating the metabolic portion of measured VCO2 from the total VCO2 monitored at the bedside allows prediction of future values of PaCO2 following a ventilation change. Calculations such as percent of over- or under-ventilation and the ventilation rate needed to achieve a specific PaCO2 value are made possible by estimation of metabolic VCO2. Sponsored Research - Philips