AbstractRecording of end-tidal carbon dioxide (EtCO2) noninvasively reflects a real-time estimation of arterial carbon dioxide (PaCO2 [partial pressure of CO2]). However, as the EtCO2 is dependent on metabolism, perfusion, and ventilation, predicting PaCO2 from EtCO2 is not linear. The objective of the study was to find out the predictability of PaCO2 from EtCO2 in PICU and to evaluate the factors affecting the correlation of EtCO2 and PaCO2 in critically ill ventilated children. The design involved was prospective observational study. The setting discussed over here is that of pediatric intensive care unit (PICU) of tertiary care hospital. A total of 160 children between 1 month and 14 years received mechanical ventilation. EtCO2, PaCO2, PaO2/FiO2 (PF) ratio, oxygenation index (OI), and ventilation index (VI) are the factors involved in main outcome measures. A total of 535 pairs of EtCO2 and PaCO2 were recorded in 160 ventilated children during the stable hemodynamic state. Mean age and weight (Z-score) of patients were 31.15 ± 40.46 months and −2.10 ± 1.58, respectively. EtCO2 and PaCO2 differences were normal (2–5 mm of Hg) in 393 (73.5%) pairs. High gradient (>5 mm of Hg) was mostly found with children with pneumonia, prolonged ventilation, and pressure mode of ventilation (p < 0.05). EtCO2 had a strong positive correlation with PaCO2 (r = 0.723, 95% confidence interval [CI] = 0.68 and 0.76) and not significantly affected by PF ratio or OI. However, presence of pneumonia and high ventilation index (VI > 20) adversely affected the relationship with poor correlation coefficient (r = 0.449, 95% CI = 0.30, 0.58 and r = 0.227, 95% CI = 0.03, 0.41, respectively). EtCO2 reading showed good validity to predict PaCO2 and not affected by oxygenation parameters. The correlation was affected by the presence of pneumonia and high ventilation index; hence it is recommended to monitor PaCO2 invasively in these patients till a good correlation is established.
Read full abstract