Transcutaneous monitoring and capnography are 2 surrogate methods of measuring arterial carbon dioxide levels employed by pediatric sleep laboratories. Both techniques are noninvasive, validated, and quantitative indirect predictors of arterial carbon dioxide level, and both have been widely adopted for use during pediatric and adult polysomnography (PSG). We hypothesized that there would be close agreement between the two techniques when compared in a pediatric population.

Children referred for diagnostic polysomnography to the Pediatric Sleep Laboratory at the Alberta Children's Hospital from June 2000 to October 2003 were included. All subjects underwent an overnight computerized PSG as per American Thoracic Society standards, including both transcutaneous and end-tidal monitoring. A registered PSG technician manually scored studies and eliminated all CO₂ data that was not interpretable. Total “uninterpretable data” time was calculated for both channels. Statistical analysis of the level of agreement between transcutaneous and end-tidal signals was performed using a Bland-Altman analysis.

The PSG studies of 609 children (363 males), mean age 7.9 ± 4.6 years (range 0.1–18.4), were reviewed. On average, interpretable data was available for 61.8% ± 35.1% and 71.5% ±25.2 % of total recording time from the end-tidal and transcutaneous channels respectively. The maximum and mean CO₂ measurements obtained by both devices showed close agreement with a mean difference of 0.1+/−5.4 mm Hg and 0.6+/−3.9 mm Hg respectively.

Transcutaneous and end-tidal carbon dioxide monitoring during polysomnography are well tolerated and provide interpretable and comparable results in the majority of children.