Pulse oximeter-enhanced accuracy of capnometry in children with cyanotic heart disease

Objectives: To evaluate the relationship between the arterial end-tidal partial pressure of carbon dioxide (PCO2) difference (ΔpCO2) and the degree of desaturation in children with cyanotic heart disease (CHD) and to come to a more reliable estimation of the arterial carbon dioxide partial pressure (PaCO2) from the end-tidal carbon dioxide partial pressure (PET-CO2). Design and setting: In part retrospective, in part prospective observational study at a university children's hospital. Subjects and interventions: We retrospectively assessed the relationship between the arterial oxygen saturation as measured by means of pulse oximetry (SpO2) and the arterial to end-tidal PCO2 differences (ΔPCO2) from the records of medical or surgical interventions in 43 patients with CHD. We derived a PaCO2-PET-CO2 correction formula that was prospectively validated in 34 patients with CHD. Measurements and results: In the retrospective part we found a significant correlation between SpO2 and ΔPCO2 (r2 = 0.84, p < 0.001). The regression equation (corrected PET-CO2 = raw PET-CO2 - 0.36 × SpO2 + 39) was used in the prospective part to calculate for the corrected PET-CO2. The r2s for the correlations between PaCO2 and uncorrected and corrected PET-CO2 were 0.17 (p < 0.05) and 0.94 (p < 0.001), respectively. The uncorrected PET-CO2 bias was 13.0 mmHg, the bias ± 2SDs was -0.1 and 26.2 mmHg. The corrected PET-CO2 bias was -0.6 mmHg, the bias ± 2SD's was -4.0 and 2.9 mmHg. Conclusions: Correcting the PET-CO2 for the degree of hypoxia using the SpO2 in artificially ventilated infants and children with CHD results in a clinically applicable estimation of the PaCO2. As both SpO2 and PET-CO2 can be monitored continuously and non-invasively, this could facilitate artificial ventilation management in children with CHD.