Abstract
BACKGROUND: Infants born extremely preterm often suffer from respiratory disease and are invasively ventilated. We aimed to test the hypothesis that gas exchange in ventilated extremely preterm infants occurs both at the level of the alveoli and via mixing of fresh deadspace gas in the airways. METHODS: We measured the normalised slopes of phase II and phase III of volumetric capnography and related them with non-invasive measurements of ventilation to perfusion ratio (V(A)/Q) and right-to-left shunt in ventilated extremely preterm infants studied at one week of life. Cardiac right-to-left shunt was excluded by concurrent echocardiography. RESULTS: We studied 25 infants (15 male) with a median (range) gestational age of 26.0 (22.9-27.9) weeks and birth weight of 795 (515-1,165) grams. The median (IQR) V(A)/Q was 0.52 (0.46-0.56) and shunt was 8 (2-13) %. The median (IQR) normalised slope of phase II was 99.6 (82.7-116.1) mmHg and of phase III was 24.6 (16.9-35.0) mmHg. The V(A)/Q was significantly related to the normalised slope of phase III (ρ = -0.573, p = 0.016) but not to the slope of phase II (ρ = 0.045, p = 0.770). The right-to-left shunt was not independently associated with either the slope of phase II or the slope of phase III after adjusting for confounding parameters. CONCLUSIONS: Abnormal gas exchange in ventilated extremely preterm infants was associated with lung disease at the alveolar level. Abnormal gas exchange at the level of the airways was not associated with quantified indices of gas exchange impairment.