Noninvasive monitoring of end-tidal carbon dioxide tension via nasal cannulas in spontaneously breathing children with profound hypocarbia.

OBJECTIVE

To determine the correlation between end-tidal CO2 and PaCO2 measured via nasal cannulas in spontaneously breathing children with profound hypocarbia (PaCO2 < 30 torr [< 4.0 kPa]).

DESIGN

Prospective evaluation.

SETTING

Pediatric intensive care unit (ICU) in a tertiary care referral center.

INTERVENTIONS

None.

PATIENTS

Patients admitted to the ICU with a diagnosis of diabetic ketoacidosis in whom invasive arterial access was deemed necessary for clinical care. The patients were spontaneously breathing, without intubation. The study included nine patients, with an average age of 9.9 yrs (range 4 to 17) and weight of 38.7 kg (range 17 to 68).

MEASUREMENTS AND MAIN RESULTS

End-tidal CO2 was sampled from nasal cannulas by a sidestream aspirator and estimated by infrared spectroscopy. The correlation between arterial and end-tidal CO2 was compared using linear regression analysis. A total of 65 arterial blood gases were obtained from the nine patients. The PaCO2 was < or = 30 torr (< or = 4.0 kPa) in 38 of the samples. The PaCO2 to end-tidal CO2 gradient was < or = 4 torr (< or = 0.5 kPa) in 64 of 65 samples and 4.8 torr (0.6 kPa) in one sample. Linear regression analysis of arterial vs. end-tidal CO2 yielded a slope of 0.99, an r2 value of .97, and a p value of .0001.

CONCLUSIONS

End-tidal CO2 measurement by infrared spectroscopy provides an accurate estimation of PaCO2, even during episodes of severe hypocarbia. Its use may limit the need for invasive monitoring and/or repeated arterial blood gas analyses.