Transcutaneous CO₂ Monitoring and Therapeutic Hypothermia

Abstract

Background: Transcutaneous CO₂ (TCM) monitoring is the accepted standard in assessing carbon dioxide pressure (CO₂) in the tissue and can be helpful in the management of newborns requiring ventilatory support. Infants with hypoxic ischemic encephalopathy (HIE) are routinely treated with therapeutic hypothermia (TH) to 33.5°C. It is well established that fluctuations in the partial pressure of CO₂ in arterial blood (PaCO₂) during TH is associated with poor neurodevelopmental outcomes, thus close monitoring of PaCO₂ is key to good outcomes. TCM correlation is perfusion dependent. TCM heats the skin at the sensor site to arterialize the sample. The non-invasive device produces a PCO₂ value corrected to 37°C (TcPCO₂). The TCM is mostly regarded as a reliable noninvasive manner to continuously monitor tissue CO₂. Our objective was to determine the accuracy and precision of TcPCO₂values in newborns undergoing TH.

Methods: We conducted a retrospective chart review of 10 neonates (3F/7M) undergoing TH for HIE requiring respiratory support. The weight of the neonates ranged from 2,230 to 3,810 g. We compared TcPCO₂to PaCO₂ values (all at 37°C) using a Bland-Altman plot for multiple measurements per patient. A Wilcoxon test for paired samples (non-normally distributed data), and a repeated measures ANOVA (rmANOVA: two types of measurements; five measurements per subject, eight subjects) was also used to aid in the analysis. IRB exempt status was gained through UC Davis.

Results: Ninety paired samples were collected in 10 patients with a range of PaCO₂ 28–94 mm Hg. Median difference was 2 mm Hg higher in TcPCO₂ (Wilcoxon test for paired samples P < 0.01). However, the precision was poor (95% CI was -13 to +18 mm Hg). The rmANOVA showed no main effect on measurement number (P = 0.38), method of comparison (P = 0.78), or the interaction (P = 0.67).

Conclusions: TcPCO₂ showed a strong median correlation to PaCO₂ values at 37°C but lacked precision. TcPCO₂may not be appropriate for estimating PaCO₂ in patients undergoing TH. Care should be taken when utilizing TCM in these patients. TCM may prove to be a reliable trending tool but acute changes in values should be confirmed with PaCO₂. Given our small sample size (N = 10) and the retrospective nature of our study, more investigation into the usefulness of TCM during TH is warranted. Discrepancy in TcPCO₂and PaCO₂ values during TH justifies further exploration in transcutaneous monitoring.