Strengths and Limitations
Neonatal hypoglycaemia is variously defined, is variably tolerated by neonates and universal screening is unusual 31-34. In our cohort neonatal glucose measurement was not universal but neonatal testing was performed according to BAPM Guidelines; maternal corticosteroids administration was not a criterion 34. For this reason we also chose the stricter outcome of ‘severe hypoglycaemia’ and defined it to include a requirement for NNU admission. This is reflected in our 1.3% overall incidence of severe hypoglycaemia compared to the 4.1% incidence of ‘treated hypoglycaemia’ in a recent analysis 22. It is also possible that the administration of ACS, particularly after 34 weeks, could reflect other risk factors for hypoglycaemia. This is addressed in the multivariate logistic regression, but residual confounding cannot be ruled out. Indeed, this could account for our finding of no reduction in severe respiratory morbidity. We were unable to adjust for confounders when assessing the relationship between time and interval and neonatal glucose level. A further limitation is the largely white ethnicity. Despite the large numbers, the number of term births exposed to ACS was relatively small.