Detection of Carbon Monoxide During Routine Anesthetics in Infants and Children

Background: Carbon monoxide (CO) can be produced in the anesthesia circuit when inhaled anesthetics are degraded by dried carbon dioxide absorbent and exhaled CO can potentially be rebreathed during low-flow anesthesia. Exposure to low concentrations of CO (12.5 ppm) can cause neurotoxicity in the developing brain and may lead to neurodevelopmental impairment. In this study, we aimed to quantify the amount of CO present within a circle system breathing circuit during general endotracheal anesthesia in infants and children with fresh strong metal alkali carbon dioxide absorbent and define the variables associated with the levels detected.

Methods: Fifteen infants and children (aged 4 months to 8 years) undergoing mask induction followed by general endotracheal anesthesia were evaluated in this observational study. CO was measured in real time from the inspiratory limb of the anesthesia circuit every 5 minutes for 1 hour during general anesthesia. Carboxyhemoglobin (COHb) levels were measured at the 1-hour time point and compared with baseline.

Results: CO was detected in all patients older than 2 years (0-18 ppm, mean 3.7 +/- 4.8 ppm) and rarely detected in patients younger than 2 years (0-2 ppm, mean 0.2 +/- 0.6 ppm). Only the relationship between CO concentration and fresh gas flow to minute ventilation ratio (FGF:(.)VE) remained significant after adjustment in longitudinal regression analysis (P < 0.001). Although not powered to determine such a relationship, CO levels were weakly associated with the use of desflurane and female sex. There was no significant association between CO concentration and anesthetic concentration. Baseline COHb levels were higher in children younger than 2 years and decreased significantly at the 1-hour time point compared with baseline and children older than 2 years. However, COHb levels increased significantly from baseline in a predictable manner consistent with CO exposure in children older than 2 years. FGF:(.)VE correlated significantly with change in COHb using simple linear regression (r = 0.62; P < 0.02).

Conclusions: CO was detected routinely during general anesthesia in infants and children when FGF:(.)VE was <1. Peak CO levels measured in the anesthesia breathing circuit were in the range thought to impair the developing brain. Further study is required to identify the source of CO detected (CO produced by degradation of volatile anesthetic versus rebreathing CO from endogenous sources or both). However, these findings suggest that avoidance of low-flow anesthesia will prevent rebreathing of exhaled CO, and use of carbon dioxide absorbents that lack strong metal hydroxide could limit inspired CO if detection was attributable to degradation of volatile anesthetic.