Complications of Antenatal Corticosteroids in Infants Born by Early Term Scheduled Cesarean Section

It is unknown whether early term (37 + 0 to 38 + 6 weeks’ gestation) antenatal corticosteroid administration is beneficial among women with gestational diabetes mellitus (GDM), diabetes in pregnancy (DIP), or pre-existing DIP (1–4). In pregnancies uncomplicated by diabetes, antenatal corticosteroid administration prior to early term scheduled cesarean section (ETSCS) reduces admissions to the neonatal intensive care unit (NICU) for respiratory distress syndrome (RDS)/transient tachypnea of the newborn (TTN) by >50% (1,2). Among pregnant women with diabetes, corticosteroid administration results in unpredictable maternal hyperglycemia, which is associated with neonatal hypoglycemia (5,6). We have compared the rates of neonatal RDS/TTN and neonatal hypoglycemia before and after the introduction of a policy ceasing the use of corticosteroids for pregnant women with diabetes undergoing an ETSCS.

This retrospective study was undertaken at Campbelltown Hospital, Sydney, Australia, where, as per national guidelines, the practice was to consider the administration of corticosteroids to women undergoing an ETSCS, although there was no formal policy (6). It was agreed by the obstetric, pediatric, and endocrine departments, following safety concerns, that women with any form of diabetes would be excluded from this policy after 1 May 2017.

Hospital medical records were obtained for all women who underwent a cesarean section (CS) between 1 May 2016 and 30 April 2018. Women who underwent an emergency CS, scheduled CS at <37 + 0 weeks’ or >38 + 6 weeks’ gestation, or scheduled CS for twins and those without diabetes were excluded. Baseline data collected included demographic, obstetric/medical/medication history, BMI at first visit, and factors for both RDS/TTN (gestational age, neonatal male sex, maternal asthma, maternal diabetes, type of anesthesia, and macrosomia) and neonatal hypoglycemia (gestational age, maternal diabetes, macrosomia) (5,7–9). A fasting glucose or 75-g oral glucose tolerance test was used to diagnose GDM (fasting glucose ≥5.1 mmol/L, 1-h glucose ≥10.0 mmol/L, 2-h glucose ≥8.5 mmol/L, following World Health Organization criteria) and DIP (fasting glucose ≥7.0 mmol/L and/or 2-h glucose ≥11.1 mmol/L) (10). The local “neonates at risk of hypoglycemia” policy included monitoring the glucose levels at 1, 4, and 7 h for all neonates with mothers who had diabetes and defined the diagnosis of neonatal hypoglycemia (requiring admission to NICU) as a heel-prick blood glucose of <2.6 mmol/L. The diagnoses of RDS and TTN are not clearly defined (as there is no local policy) other than respiratory distress requiring support (and admission to NICU), including subcostal recession, or a respiratory rate of >60 breaths per minute.

Analyses used SPSS version 25 (IBM) including the independent samples t test and χ² test of independence to compare continuous and categorical variables, respectively. Forward binary logistic regression compared neonatal hypoglycemia and RDS occurrence among women with and without corticosteroid administration after adjustment for neonatal sex, macrosomia, prepregnancy BMI, smoking, asthma, gestational age in days, second or third trimester HbA_1c, and type of diabetes (GDM versus DIP/pre-existing DIP). The study was approved by the South West Sydney Local Health District (SWSLHD) Quality Improvement Committee (reference number CT02_2017) as approved by the SWSLHD Health Research Ethics Committee.

Overall, 102 women with diabetes underwent an ETSCS, of whom 33 (32.4%) received corticosteroids (Supplementary Fig. 1). Of the 49 women with diabetes prior to the new policy, 18 did not receive corticosteroids following endocrinologist advice, obstetric decision, patient choice, or with no recorded reason. Of the 53 women with diabetes after the policy change, 2 women with GDM received corticosteroids, 1 following obstetric decision and the other with no reason recorded. The mean interval between corticosteroid administration and CS was 5 days (median [range], 2 [1–46] days). Those with corticosteroid treatment had a higher BMI and shorter gestation, but they were otherwise similar to untreated women and their neonates (Table 1). Overall, neonates of women administered corticosteroids (versus those without corticosteroids) were significantly more likely to be admitted with neonatal hypoglycemia (24.2% vs. 4.4%, P = 0.003) but were not less likely to be admitted to the NICU with RDS/TTN (15.2% vs. 7.2%, P = 0.209). Neonates of women administered corticosteroids had an increased risk of neonatal hypoglycemia with an adjusted odds ratio (aOR) of 18.96 (95% CI 2.18–165.23) (Supplementary Table 1) and with no significant difference in RDS (aOR 1.47; 95% CI 0.27–7.93) (data not shown).

We have shown that antenatal corticosteroid administration in ETSCS among women with GDM, DIP, or pre-existing DIP is associated with more neonatal hypoglycemia requiring NICU admission with no increase in RDS/TTN. In prior studies, women with pregestational diabetes were excluded (3), remained unidentified (2), or were included in small numbers (10/819 women with diabetes) (1). The impact of the associated neonatal hypoglycemia on the number of NICU admissions has also not previously been considered. This is particularly important with a high prevalence of GDM in parts of Australia (11).

The strength of this study is the completeness of data of both maternal and neonatal characteristics including the known risk factors for both RDS/TTN and neonatal hypoglycemia (5,7–9). There are several limitations. This was a single-center retrospective cohort study. The sample size of 102 women is small compared with studies conducted in the general population (819 and 452 women) (1,2). As a result, CIs are wide, and numbers may have been too small to detect any limited excess risk of RDS/TTN and the effect of several potential confounding factors (e.g., neonatal macrosomia). Both the treated and untreated pregnancies were within the previously reported ranges of neonatal RDS/TTN for the untreated general population (5.1–23%). As expected, the treated group (15.2%) had a higher rate than among women without diabetes (2.4–7.9%) (1,2). Although over 90% of women had GDM, the remaining women had pregestational diabetes, which carries the additional risks of adverse maternal and neonatal outcomes (12). Gestational age was, on average, 2 days longer in women not receiving corticosteroids. While this was statistically significant, we felt that this was unlikely to be clinically significant given the potential for error in calculating gestational days (13). Even though maternal BMI was higher in women receiving corticosteroids, the aOR of neonatal hypoglycemia remained increased after adjusting for maternal BMI (Supplementary Table 1). Maternal asthma is a risk factor for both RDS and TTN and had a higher rate than the national average (11.2%) in this cohort (8,14). The reasons for this high prevalence are unclear, but there was no significant difference between those treated and not treated with corticosteroids. There were no data available for additional potential confounders including maternal blood glucose levels prior to delivery and ethnicity (a known risk factor for RDS) (5).

In conclusion, on the basis of the results of this small study, there is some evidence to suggest that corticosteroids may be safely omitted in women with any form of diabetes in pregnancy undergoing ETSCS. Such management would potentially significantly reduce admissions to the NICU for neonatal hypoglycemia without increasing admissions for neonatal RDS/TTN. A larger, randomized controlled trial would be required to strengthen the validity and generalizability of these results.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. K.G. contributed to the design of the study, collected data, analyzed the data, interpreted the data, and wrote the manuscript. R.R. contributed to data interpretation and manuscript review and editing. F.K. collected data. D.S. designed the study, interpreted the data, and reviewed and edited the manuscript. K.G. and D.S. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Parts of this study were presented as an oral presentation at the Australasian Diabetes in Pregnancy Society 2018 Annual Scientific Meeting, Adelaide, Australia, 24–26 August 2018.