Reproducibility of multiple breath washout indices in the unsedated preterm neonate.

Multiple breath inert gas washout (MBW) is gaining popularity for measurements of resting lung volume and ventilation inhomogeneity. Test reproducibility is an important determinant of the clinical applicability of diagnostic tests. The between-test reproducibility of variables derived from MBW tests in newborn infants is unknown. We aimed to determine the within-test repeatability and short-term between-test reproducibility of MBW variables in unsedated preterm infants. We hypothesized that measurements obtained within a 3-day interval in clinically stable preterm infants would be reproducible and suitable for use as an objective clinical outcome measurement. In this cross-sectional observational study, clinically stable hospitalized preterm infants whose parents had given informed consent for MBW studies were tested twice within 72 hr during quiet, unsedated sleep. Functional residual capacity (FRC), lung clearance index (LCI), and the first and second to zeroeth moment ratios (M(1):M(0); M(2):M(0)) were computed from MBW traces obtained using a mainstream ultrasonic flowmeter and 4% sulphur hexafluoride (MBW(SF6)). Within-test repeatability and between-test reproducibility were determined. Within-test repeatability (expressed as a coefficient of variability (C(v))) for differences between two and four replicate measurements on the same test occasion, were 9.3% (FRC), 9.0% (LCI), 7.6% (M(1):M(0)), and 15.6% (M(2):M(0)), respectively. The within-test C(v)'s were not statistically different to the between-tests C(v)'s, which were 7.7% (FRC), 10.3% (LCI), 6.1% (M(1):M(0)), and 13.0% (M(2):M(0)), respectively. Among unsedated preterm infants, between-test reproducibility over a 3-day interval was similar to within-test repeatability. The wide limits of agreement may limit the application of these measures to detect a clinically significant change in condition in small preterm infants.