2003 OPEN FORUM Abstracts
COMPARISON OF MICROSTREAM AND MAINSTREAM CAPNOGRAPHY WITH ARTERIAL CO2 IN NEONATES
Kathleen Deakins RRT-NPS, John Dickson RRT-NPS, Ricardo Rodriguez MD, Michael
Anderson MD, Robert L. Chatburn RRT-NPS, FAARC, University Hospitals of Cleveland, OH
Introduction: End-tidal carbon dioxide (PETCO2) is commonly monitored as an indicator of
ventilation. Some published studies have shown good correlation between PETCO2 and PaCO2 in
low birthweight infants with Respiratory Distress Syndrome [An Esp Pediatr 1997 (2): 171-180].
Microstream sampling technology has been marketed as requiring smaller, lower dead space
adaptors and smaller sampling sizes when compared to conventional mainstream sampling. The
purpose of this study was to test the hypothesis that microstream sampling of PET CO2 is a more
accurate estimate of PaCO2 than mainstream sampling in low birthweight infants and that the end
tidal-arterial PCO2 difference is related to respiratory rate and/or weight.
Methods: A convenience
sample of infants in the NICU or PICU weighing 559-3000 grams with artificial airways and
arterial lines participated in this study. A microstream sampling adapter (Microcap, Oridion
Capnography) was connected to the endotracheal tube for a minimum period of five minutes prior
to sampling an arterial blood gas. A PETCO2 reading was obtained and the respiratory rate was
counted while an arterial blood gas was drawn. The procedure was repeated on the same patient
using a mainstream sampling adapter (Spacelabs) when a subsequent blood gas was indicated. The
order of adapter type usage was randomized. Data collected included age, weight, diagnosis,
respiratory rate and PETCO2 and PaCO2 values using both technologies. The Mann-Whitney Rank
Sum Test was used to compare median values for PETCO2 - PaCO2 difference. The correlation
between PETCO2 - PaCO2 difference and either weight or respiratory rate was evaluated with
Pearson's Product Moment Correlation Coefficient. Significance was set at p = 0.05.
Results: 55
infants entered the study. Both end tidal sampling techniques underestimated arterial PCO2. The
median PETCO2 - PaCO2 difference for the Spacelabs mainstream sampling adapter was
significantly less than the difference for the Microcap microstream sampling adapter (-8 torr
versus -13 torr, p = 0.001). There was no significant correlation between PETCO2 - PaCO2
difference and either birth weight or respiratory rate for either sampling technology. This finding
held true even when the data were restricted to infants weighing less than 1,100 grams.
Conclusions: Mainstream PETCO2 sampling was a more accurate estimator of PaCO2 than microstream sampling in this population of neonates. Errors in estimating PaCO2 do not appear to be related to either birth weight or respiratory rate.