Background: Measurements of the subarachnoid space during routine cranial sonography may provide an indirect method of monitoring brain growth in preterm infants. Methods: The width of the subarachnoid space was measured on coronal views during head sonography. Initial scans (within five days of birth) were compared with follow up scans. Results: A total of 361 scans were performed on 201 preterm infants. The mean width of the subarachnoid space was < 3.5 mm for 95% of initial scans. It was slightly larger in neonates born closer to term, the equivalent of an increase of 0.02 mm/gestational week (95% confidence interval 0 to 0.10 mm) for initial scans. When the scans of all infants, born at 24-36 gestational weeks who were 36 weeks corrected gestational age were compared, the mean (SD) subarachnoid space was 60% larger for follow up scans than for intial scans: 3.2 (1.38) v 1.95 (1.35) mm (p = 0.002) or the equivalent of a mean increase of 0.20 mm/week (95% confidence interval 0.15 to 0.30 mm) for follow up scans. At 36 weeks corrected gestational age, mean head circumference was not different between those having initial or follow up scans (33.0 (2.0) v 32.2 (1.9) cm; p = 0.31). Conclusions:The mean subarachnoid space is normally < 3.5 mm in preterm infants. The difference between initial and follow up scans suggests reduced brain growth in extrauterine preterm babies.

Objectives: To report 18 month outcome of a randomised trial of two courses of dexamethasone to prevent chronic lung disease of prematurity. Study design: Babies of birth weight 1250 g or less ventilated at 7 days of age were randomised to a 42 day reducing course (long) or a 3 day pulsed (pulse) course of dexamethasone. Growth, cardiovascular status, and respiratory and neurodevelopmental outcomes were assessed at 18 months. Results: Seventy six babies were enrolled. Nine died and three were lost to follow up. Babies receiving the long course were weaned off oxygen more quickly than those receiving the pulse course (47% v 69% on oxygen at 28 days; p = 0.01), but there were no differences in 18 month outcomes. However, children averaged −1 SD for growth parameters, half had moderate or severe disability, and 35% and 19% respectively required oxygen at 36 weeks and discharge. Conclusions: The dexamethasone course used did not influence long term outcome. However, entry criteria for this study selected a group of babies at high risk of poor long term outcome.

The individualized course of dexamethasone used in this study reduced the total dose of dexamethasone administered but did not significantly reduce side-effects of treatment or alter outcome in infants at risk of chronic lung disease.