METHODS
This is an observational cohort study of singleton term (≥ 37 weeks) neonates born in a single tertiary centre between October 2016 and September 2019. STROBE guidelines were followed. Exclusion criteria were diagnosis of neonatal sepsis 17, multiple pregnancies, those with uncertain dates, and those complicated by major structural fetal abnormalities, aneuploidy, or genetic syndromes.
Outcome measures were 1) severe neonatal hypoglycaemia, defined as capillary blood glucose < 2.0 mmol/l in the first 24 hours of life and requiring admission to NNU and 2) the need for mechanical ventilation or CPAP, (selected due to inconsistency in diagnosis of respiratory distress syndrome (RDS)). Where measured, the lowest blood glucose in the first 24 hours was also recorded.
Blood glucose level was measured in neonates considered at risk (BAPM) using an Abbott ‘i-STAT handheld’ glucometer 2-4 hours after birth, (usually before the second feed) and immediately if there were clinical signs suggestive of hypoglycaemia (lethargy, abnormal feeding behaviour, high pitched cry, altered level of consciousness, hypotonia, seizures, hypothermia (<36.5°C), apnoea). Criteria for admission to the Neonatal Unit (NNU) due to neonatal hypoglycaemia were single glucose measurement < 1 mmol/l; two or more consecutive measurements between 1.0-1.9 mmol/l despite ongoing feeding support; or clinical signs consistent with hypoglycaemia.
One complete corticosteroid course of two doses Betamethasone 12 mg 24 hours apart was administered where the risk of preterm birth (< 34+0 weeks) in the following 7 days was perceived to be high. Betamethasone was also used at the discretion of the clinician from 34+0 weeks, prior to planned elective late preterm or early term birth. If a course of betamethasone was previously administered for any indication, a second course was not given. Breastfeeding was supported within the first hour and for those declining, formula feed 10-15ml/kg was offered, with 3-hrly subsequent feeds.
Pregnancies were dated using crown-rump length (CRL) between 9+0 and 13+6 weeks. Birthweight (BW) was converted into centiles according to Intergrowth-21st standards18: small for gestational age (SGA) was defined as < 10th centile. Gestational diabetes and hypertensive disorders were defined according to NICE criteria19, 20.
Prospectively collected data were merged from electronic maternity record (Cerner Millennium) and neonatal records (Badgernet, Clevermed, Edinburgh, UK). Statistical analysis was performed using Statistical Package for Social Sciences (SPSS) v26.0 (IBM Inc., Chicago, IL, USA).
The study cohort was divided into three groups according to corticosteroid exposure: the non-exposed (comparison) group; the exposed group 1 where ACS were given before 34+0 weeks; and the exposed group 2 where ACS were given at or after 34+0 weeks. A subset of the latter group, where ACS were administered within 7 days before planned caesarean birth was further analysed (exposed group 2a).
As neonates born from diabetic mothers represent the group with the highest risk of developing neonatal hypoglycaemia, a secondary subgroup analysis was carried out to assess relationship between ACS exposure and neonatal outcomes in pregnancies affected by pre-existing or gestational diabetes.
Continuous variables were presented as median and interquartile range (IQR), while categorical variables were presented as absolute numbers and percentages. Demographic and pregnancy variables were calculated for each group of neonates. Demographic characteristics, pregnancy characteristics and corticosteroid exposure, including timing, were then compared between hypoglycaemic and normoglycaemic neonates. Group comparisons of variables were performed with Mann–Whitney U test, χ2 test or Fisher’s exact tests where appropriate. Differences were considered significant when p value was <0.05.
The association between ACS exposure and neonatal outcomes was assessed for each group. Demographic and pregnancy characteristics significantly associated with neonatal outcomes on univariate analysis were then included into a multivariate backward-stepwise regression model to adjust for potential confounders. Adjusted odds ratios (aOR) and 95% confidential intervals (CI) were calculated.
Finally, among neonates exposed to ACS who had a glucose measured within 24 hours of birth, the correlation between time interval between corticosteroid exposure and birth, and minimum neonatal glucose value, was assessed using scatter plots. Pearson’s correlation was performed to assess the relationship between variables.
NHS Health Research Authority ethical approval for this analysis was granted on 27/07/2017 (IRAS project ID 222260; REC reference 17/SC/0374).