Factors affecting the arterial to end-tidal carbon dioxide gradient in ventilated neonates

Abstract

Objective. To determine factors which influenced the relationship between blood carbon dioxide (pCO2) and end-tidal carbon dioxide (EtCO2) values in ventilated, newborn infants. Furthermore, to assess whether pCO2 levels could be predicted from continuous EtCO2 monitoring. Approach. An observational study of routinely monitored newborn infants requiring mechanical ventilation in the first 28 d after birth was undertaken. Infants received standard clinical care. Daily pCO2 and EtCO2 levels were recorded and the difference (gradient: ∆P-EtCO2) between the pairs were calculated. Ventilatory settings corresponding to the time of each blood gas assessment were noted. End-tidal capnography monitoring was performed using the Microstream sidestream Filterline H set capnograph. Main results. A total of 4697 blood gas results from one hundred and fifty infants were analysed. The infants had a median gestational age of 33.3 (range 22.3–42.0) weeks and birth weight of 1880 (395–5520) grams. Overall, there was moderate correlation between pCO2 and EtCO2 levels (r = 0.65, p < 0.001). The ∆P-EtCO2 for infants born less than 32 weeks of gestation was significantly higher (1.4 kPa) compared to infants born at greater than 32 weeks of gestation (0.8 kPa) (p < 0.001). In infants born at less than 32 completed weeks of gestation, pCO2 levels were independently associated with EtCO2, day after birth, birthweight and fraction of inspired oxygen (FiO2) (model r 2 = 0.52, p < 0.001). Significance. The results of end-tidal capnography monitoring have the potential to predict blood carbon dioxide values within the neonatal population.