Aim: End-tidal CO 2 (Et CO2 ) is the standard in operative care along with pulse oximetry for ventilation assessment. It is known to be less accurate in the infant population than in adults. Many neonatal intensive care units (NICU) have converted to utilizing transcutaneous CO 2 (tcP CO2 ) monitoring. This study aimed to compare perioperative Et CO2 to tcP CO2 in the pediatric perioperative population specifically below 10 kg, which encompasses neonates and some infants. Methods: After IRB approval and parental written informed consent, we enrolled neonates and infants weighing less than 10 kg, who were scheduled for elective surgery with endotracheal tube under general anesthesia. P CO2 was monitored with Et CO2 and with tcP CO2 . Venous blood gas (Pv CO2 ) samples were drawn at the end of the anesthetic. We calculated a mean difference of Et CO2 minus Pv CO2 (Delta Et CO2 ), and tcP CO2 minus Pv CO2 (Delta tcP CO2 ) from end-of-case measurements. The mean differences in the NICU and non-NICU patients were compared by t-tests and Bland–Altman analysis. Results: Median age was 10.9 weeks, and median weight was 4.4 kg. NICU (n=6) and non-NICU (n=14) patients did not differ in Pv CO2 . Relative to the Pv CO2 , the Delta Et CO2 was much greater in the NICU compared to the non-NICU patients (−28.1 versus −9.8, t=3.912, 18 df, P =0.001). Delta tcP CO2 was close to zero in both groups. Although both measures obtained simultaneously in the same patients agreed moderately with each other (r =0.444, 18 df, P =0.05), Bland–Altman plots indicated that the mean difference (bias) in Et CO2 measurements differed significantly from zero ( P <0.05). Conclusions: Et CO2 underestimates Pv CO2 values in neonates and infants under general anesthesia. TcP CO2 closely approximates venous blood gas values, in both the NICU and non-NICU samples. We, therefore, conclude that tcP CO2 is a more accurate measure of operative Pv CO2 in infants, especially in NICU patients.

BACKGROUND The concept of anaesthesia-related neonatal neurotoxicity originated in neonatal rodent models, yet prospective clinical studies have largely not supported this concern. OBJECTIVES To determine the frequency and magnitude of hypercarbia, hypoxia and death in rodent models of neonatal anaesthetic toxicity and neurodevelopmental delay. DESIGN Systematic review of published rodent studies of neonatal anaesthesia neurotoxicity. We documented anaesthetic, route, dose, frequency and duration of exposures. We further report ventilation method, documentation of adequacy of ventilation [arterial blood gas (ABG), other], mortality and the reporting of mortality. DATA SOURCES A PubMed literature search from 2003 to 2017 was conducted to identify studies on neurotoxicity in neonatal rodent models. ELIGIBILITY Studies were included when at least one group of animals fell within the postnatal age range of 3 to 15 days. Only English language original studies published as full-length articles in peer reviewed journals were included in the final analysis. RESULTS One hundred and three manuscripts were included. Ninety-eight percent of studies were conducted using spontaneous ventilation (101/103), with ABG monitoring used in only 33% of studies and visual monitoring alone for respiratory distress or cyanosis was employed in 60%. Of the 33% who reported ABG results, there were widely divergent values, with most reporting modest-to-severe hypercarbia. Mortality (median 11%, range of 0 to 40%), which infers severe hypoxia, was documented in only 36/103 (35%) reports. CONCLUSION Hypoxia and hypercarbia have known apoptotic effects on developing brains. Hence, the inadequate control of hypercarbia and hypoxia in neonatal rodent models of anaesthetic exposure during spontaneous ventilation suggests that the evidence for developmental delay and neurotoxicity attributed to anaesthesia may not be valid in humans.