End-Tidal Carbon Dioxide Monitoring in Neonates Receiving Therapeutic Hypothermia for Hypoxic-Ischemic Encephalopathy.

Abstract

Introduction: Primary aim was to assess the agreement between end-tidal carbon dioxide (etCO₂) monitoring and arterial, capillary and venous PCO₂ values in mechanically ventilated patients receiving therapeutic hypothermia (TH) for hypoxic-ischemic encephalopathy (HIE). Secondary, to assess the percentage of time spent in predefined PCO₂ ranges based on continuous etCO₂ monitoring.

Methods: In this prospective observational single center trial, infants with moderate-to-severe HIE receiving conventional ventilation with sidestream capnography were enrolled. Blood gas measurements were performed based on clinical indication. The mean of 12,000 etCO₂ values obtained over 10 min before each corresponding blood gas was used for analysis. The agreement between mean etCO₂ and temperature corrected and uncorrected PCO₂ at 37^◦C were analyzed using Bland-Altman (BA) plots.

Results: A total of 262 paired PCO₂ and etCO₂ values were analyzed from 35 patients. The bias between temperature corrected arterial PCO₂ and etCO₂ (n = 116) was 1.87 mmHg (SD 5.54) with -8.99 and 12.73 limits of agreement; whereas the bias between capillary PCO₂ and etCO₂ (n = 132) was 7.22 mmHg (SD 6.08). EtCO₂ underestimated PCO₂ of any source at 37°C. Excluding patients with lung diseases from BA analysis did not show improvement in the agreement. Infants spent median 23.9% [IQR 8.5; 36.7] of monitoring time in etCO₂ range < 35 mmHg and median 75.0% [IQR 61.1; 87.7] in etCO₂ range of 35-55 mmHg.

Conclusions: EtCO₂ monitoring may be a valuable addition to neurocritical care of infants with HIE as it showed a strong level of agreement with temperature corrected arterial PCO₂.