Objective: Continuous monitoring of ventilation during mechanical ventilation may improve patient management by facilitating proactive rather than reactive ventilator adjustments and may decrease the need for repeated arterial blood gas analysis. Because of their more critical pulmonary status, patients requiring high-frequency oscillatory ventilation may especially benefit from continuous monitoring.
Design: Prospective evaluation of the correlation between transcutaneous CO2 (TC(CO2)) and PaCO2 values.
Setting: Tertiary care pediatric intensive care unit.
Patients: Fourteen pediatric patients receiving high-frequency oscillatory ventilation for severe respiratory failure.
Interventions: TC(CO2) was monitored continuously and compared with PaCO2 values when arterial blood gas analysis was performed.
Measurements and main results: One hundred sample sets were obtained from 14 patients age 1 day to 16 yrs (3.5 +/- 4.6 yrs) and weighing 3.1-85 kg (18.5 +/- 22.4 kg). The mean absolute difference between PaCO2 and TC(CO2) was 2.8 +/- 1.9 mm Hg. Regression analysis of TC(CO2) and PaCO2 values revealed a slope of 1.04, an r value of.96, and an r value of.94 (p <.0001). Bland-Altman analysis revealed a bias of 2.1 mm Hg with a precision of 2.7 mm Hg when TC(CO2) was compared with PaCO2 for the entire group. In the subgroup where PaCO2 was <or=50 mm Hg, the bias was 1.9 with a precision of 2.8, whereas when PaCO2 was >50 mm Hg, the bias was 2.3 with a precision of 2.6 (p = not significant).
Conclusions: TC(CO2) monitoring provides an accurate and clinically acceptable estimate of PaCO2 over a wide range of CO2 values in pediatric patients during high-frequency oscillatory ventilation.