2002 OPEN FORUM Abstracts
ACCURACY OF TRANSCUTANEOUS CARBON DIOXIDE MONITORING DURING HIGH FREQUENCY OSCILLATORY VENTILATION.
John W. Berkenbosch1, Joseph D. Tobias1,2. Departments of Child Health1 and Anesthesiology2, The University of Missouri-Columbia, Columbia, MO, USA
Introduction: During mechanical ventilation, continuous, non-invasive monitoring of CO2 may be of benefit by limiting the need for repeated arterial blood gas (ABG) analysis and facilitating proactive rather than reactive ventilator adjustments. Although end-tidal CO2 monitoring is frequently used in the pediatric population, technical considerations prevent its use during high frequency oscillatory ventilation (HFOV). We have previously documented the accuracy of transcutaneous CO2 (TC-CO2) monitoring during pediatric conventional ventilation. The current study evaluates the accuracy of TCCO2 monitoring during pediatric HFOV.
Methods: The study was approved by the University of Missouri Institutional Review Board. TCCO2 was continuously measured using a TCM3 TC-CO2/O2 device (Radiometer, Copenhagen, Denmark). Prior to placement, the TC-CO2 monitor was calibrated according to the manufacturerÕs recommendations. As clinically indicated, ABGÕs were obtained and the simultaneous values for the PaCO2 and TC-CO2 recorded. Statistical evaluation of the arterial-transcutaneous CO2 difference was performed using Bland-Altman and linear regression analyses. To evaluate the accuracy of the TC-CO2 device at higher PaCO2 values, we also compared the arterial-transcutaneous CO2 difference for PaCO2 values ² 50 mmHg and with the difference for PaCO2 values > 50 mmHg using an unpaired StudentÕs t-test. A p value of < 0.05 was considered significant.
Results: 100 sample sets were obtained from 14 patients (8 boys, 6 girls) ranging in age from 1 day to 16 yr (3.5 ± 4.6 yr), and in weight from 3.1 - 85 kg (18.5 ± 22.4 kg). Eleven patients were receiving vasoactive agents including dopamine (2-15 mcg/kg/min), dobutamine (5-10 mcg/kg/min), epinephrine (0.05-0.2 mcg/kg/min) or nicardipine (1 mcg/kg/min). The absolute arterial-transcutaneous CO2 difference was 2.8 ± 1.9 mmHg. The arterial-transcutaneous CO2 difference was 0-2 mmHg in 51/100 values, 3-5 in 41/100 values and >5 in only 8/100 values. Linear regression analysis of arterial vs TC-CO2 revealed a slope of 1.04, an r value of 0.96 and an r2 value of 0.94 (p<0.0001). Bland-Altman analysis of the entire group revealed a bias of 2.1 mmHg with a precision of 2.7 mmHg. When PaCO2 was ² 50 mmHg, Bland-Altman analysis revealed a bias of 1.9 mmHg and a precision of 2.8 mmHg compared to a bias of 2.3 mmHg and a precision of 2.6 mmHg when PaCO2 was > 50 mmHg (p=NS).
Conclusions: TC-CO2 monitoring provides a continuous, clinically acceptable, estimate of arterial CO2 over a wide range of PaCO2 values during HFOV in infants and children. Given that patients receiving HFOV tend to be more critically ill, accurate, continuous monitoring of ventilation in these patients is particularly desirable and may facilitate more proactive ventilator adjustments. Additionally, this may be especially useful during the initiation of HFOV when large changes in PaCO2 may occur.