Transcutaneous carbon dioxide monitoring during therapeutic hypothermia for neonatal encephalopathy.

ABSTRACT

BACKGROUND: In neonates with post-asphyxial neonatal encephalopathy, further neuronal damage is prevented with therapeutic hypothermia (TH). In addition, fluctuations in carbon dioxide levels have been associated with poor neurodevelopmental outcome, demanding close monitoring. This study investigated the accuracy and clinical value of transcutaneous carbon dioxide (tcPCO2) monitoring during TH. METHODS: In this retrospective cohort study in neonates, agreement between arterial carbon dioxide (PaCO2) values and tcPCO2 measurements during TH was determined. TcPCO2 levels during the first 24 h of hypothermia were tested for an association with ischemic brain injury on magnetic resonance imaging (MRI). RESULTS: Thirty-four neonates were included. Agreement (bias (95% limits of agreement)) between tcPCO2 and PaCO2 levels was 3.9 (-12.4-20.2) mm Hg. No relation was found between the body temperature and tcPCO2 levels. TcPCO2 levels differed significantly between patients with considerable and minimal damage on MRI; after 6 h (P = 0.02) and 9 h (P = 0.04). CONCLUSIONS: Although tcPCO2 provided a limited estimation of PaCO2, it can be used for trend monitoring during TH. TcPCO2 levels after birth could provide an early indicator of ischemic brain injury. This relation should be investigated in large prospective studies, in which adjustments for confounders can be made. IMPACT Transcutaneous carbon dioxide measurements during therapeutic hypothermia in neonates show limited accuracy similar to measurements reported in normothermic neonates and can be used for trend monitoring. Low transcutaneous carbon dioxide levels during the first 24 h were associated with considerable ischemic brain injury on MRI. The value of transcutaneous carbon dioxide measurements during the first 24 h as an indicator of considerable ischemic brain injury on MRI should be investigated in future studies, adjusting for confounders. Transcutaneous oxygen measurements during therapeutic hypothermia showed an inaccuracy that could not be related to a low body temperature.