Good estimation of arterial carbon dioxide by end-tidal carbon dioxide monitoring in the neonatal intensive care unit.

End-tidal carbon dioxide pressure (PetCO(2)) was measured in the neonatal intensive care unit (NICU) to assess its reliability and accuracy in predicting arterial partial pressure of carbon dioxide (PaCO(2)). Arterial blood was drawn for gas analysis and compared with exhaled CO(2) measured by mainstream capnography. In total, 130 PetCO(2)/PaCO(2) comparisons were obtained from 61 patients (20 term and 41 preterm infants). PetCO(2) was significantly different from PaCO(2) (PetCO(2) = 42.3 +/- 10.5 mmHg vs. PaCO(2) = 45.8 +/- 12.3 mmHg, P < 0.001, mean +/- SD). The overall PetCO(2) bias (mean +/- SD) was 3.5 +/- 7.1 mmHg. There was a positive correlation between PetCO(2) and PaCO(2) (n = 130, r = 0.818, P < 0.001) in both term (n = 44, r = 0.779, P < 0.001) and preterm infants (n = 86, r = 0.849, P < 0.001). The PetCO(2) biases (95% CI) were 3.5 +/- 9.0 mmHg (0.8-6.2) in the term group and 3.4 +/- 6.0 mmHg (2.2-4.7) in the preterm group. Therefore, PetCO(2) was a valid and reliable method for monitoring PaCO(2) in neonates, especially preterm infants. This method decreases blood loss and prevents complications associated with arterial catheters. In conclusion, we recommend using mainstream capnography to monitor PetCO(2) instead of measuring PaCO(2) in the NICU. Copyright 2003 Wiley-Liss, Inc.