Reliability of cerebral oximeter in non-invasive diagnosis and follow-up of hypercapnia

Abstract

In this study, aimed to evaluate the sensitivity and specificity of cerebral oximetry and EtCO2 values in non-invasive diagnosis and monitoring of hypercapnia. This study enrolled pediatric patients admitted to and mechanically ventilated at the Pediatric Intensive Care Unit of Gaziantep University Faculty of Medicine Hospital between January 2014 and January 2015. Patients’ age, gender, diagnosis, ventilatory parameters, a measured of the mean end-tidal carbon dioxide value stream method, and the simultaneously monitored arterial blood gas PaCO2 level and near infrared spectroscopy device (NIRS) measurements were recorded. The mean age of patients was 61 months (min 4-max 193), and there were 8 (53.4%) female and 7 (46.6%) male subjects. A significant correlation was found between PCO2 and NIRS, PCO2 and EtCO2 (r = 0.571, p <0.001). There was a significant positive correlation between EtCO2 and NIRS (r = 0.479, p <0.001). NIRS levels were significantly higher (p <0.001) in the group with pCO2 >45; EtCO2> 40 and pH <7.35, compared to the group with PCO2<45 EtCO2<40 and ph≥7.35​. The best cut-off point for NIRS to distinguish the groups with PCO2> 45 with PCO2<45 was 80.5, with a sensitivity of 65.3%, specificity of 84.3%, and positive and negative predictive values of 81.9% and 69.1%, respectively. The best cut-off point for NIRS to distinguish the groups with EtCO2> 40 and EtCO2<40 was 81, with a sensitivity of 62.2%, specificity 77.9%, and positive and negative predictive values of 61.2% and 72.3%, respectively. As for pH <7.35 and PaCO2> 45, while sensitivity of EtCO2 (at a cut-off point of 40) was 64.9%; the sensitivity of NIRS (at a cut-off point of 80.5) was 28.1%, with EtCO being significantly more sensitive than NIRS (p <0.001). The results of the2 present study suggest that NIRS values of above 80 should alert clinicians for hypercapnia associated with increased cerebral blood flow.