Administration of antenatal corticosteroids for singleton preterm birth in China, 2017 to 2018

The administration of antenatal corticosteroids (ACS) to women who are at risk of preterm birth has been proven to reduce not only the mortality, but also the major morbidities of the preterm infants. The rate of ACS and the risk factors associated with ACS use in Chinese population is unclear. This study aimed to investigate the rate of ACS use and the associated perinatal factors in the tertiary maternal centers of China. Methods Data for this retrospective observational study came from a clinical database of preterm infants established by REIN-EPIQ trial. All infants born at <34 weeks of gestation and admitted to 18 tertiary maternal centers in China from 2017 to 2018 were enrolled. Any dose of dexamethasone was given prior to preterm delivery was recorded and the associated perinatal factors were analyzed.


Results
The rate of ACS exposure in this population was 71.2% (range 20.2% -92%) and the ACS use in these 18 maternal centers varied from 20.2-92.0% in this period. ACS exposure was higher among women with preeclampsia, caesarean section delivery, antibiotic treatment and who delivered infants with lower gestational age and small for gestational age. ACS use was highest in the 28-31 weeks gestational age group, and lowest in the under 26 weeks of gestational age group (x 2 =65.478, P < 0.001). ACS exposure was associated with lower odds of bronchopulmonary dysplasia or death (OR, 0.778; 95% CI 0.661 to 0.916) and invasive respiration requirement (OR, 0.668; 95% CI 0.585 to 0.762) in this population.

Conclusion
The ACS exposure was variable among maternity hospitals and quality improvement of ACS administration is warranted.

Backgrounds
Preterm birth has been increasing in China in recent decades, and accounted for 6.9% of live birth or 1.1 million preterm infants annually in 2019 1 . Preterm birth is a leading cause of neonatal mortality in China, second only to perinatal asphyxia 2 . Consequently, management of preterm birth and improvement of preterm birth outcomes is a priority for China.
Administration of antenatal corticosteroids (ACS) to women who are at risk of preterm birth has been proven to decrease the mortality of preterm infants and reduce not only major morbidities like neonatal respiratory distress syndrome (NRDS), necrotizing enterocolitis (NEC) and intraventricular hemorrhage (IVH), but also improve long term developmental outcomes 3 . ACS has been widely accepted as standard of care for anticipated preterm deliveries between 24 to 34 weeks of gestational age 4,5 . The best timing of the ACS is within 7 days of and prior to premature delivery 6 . One repeat course is recommended for pregnant women below 34 weeks of gestational age who have received one prior course of ACS for risk of preterm delivery and more than 2 repeated courses of ACS are not recommended 7 . ACS is safe for pregnant women 8 . Nevertheless, the prevalence of ACS administration varies in different gestational ages and different maternal centers and is reported to be between 70-90% among pregnant women less than 34 weeks of gestational age in high income countries 9,10 and 50-53% in China 11,12 . This gap merits investigation and needs to be reduced to improve the care of preterm infants. In this study, we aim to analyze the use of ACS among tertiary level maternity and infant health centers in China, to gain insights that may facilitate development of a strategy of quality improvement to increase ACS use.

Methods
Overview Data for this retrospective observational study came from a clinical database of preterm infants established by REIN-EPIQ (REduction of Infection in Neonatal intensive care units using the Evidencebased Practice for Improving Quality) trial (REIN-EPIQ study, clinicaltrials.gov #NCT02600195) 13

Population
The subjects were preterm infants whose gestational age was less than 34 weeks. The inclusion criteria for the study were: (1) gestational age < 34 weeks; (2) birth weight < 1500g; (3) admission to the NICU of member hospitals of REIN-EPIQ within 7 days of birth; (4) discharge time from May 1, 2015 to April 30, 2018. Exclusion criteria were: children with congenital malformations, including severe organ structural malformations and chromosomal abnormalities.
A total of 27,534 children were included in the REIN-EPIQ database during this period. Only inborn preterm infants were included in this study because there was a high incidence of missing perinatal data among outborn infants. Consequently all 7 children's hospitals were excluded. We excluded data prior to 2017 because twins were not identi ed prior to that time.
During 2017 and 2018, there were 10,598 singleton preterm infants below 34 weeks of GA admitted into the 18 participating maternity hospital NICUs. We excluded 1529 out-born infants as well as 180 in-born infants with missing information on ACS use. The remaining 8,889 infants were included in the analysis.
Of these, 636 infants were discharged against medical advice. (Figure 1).

Data collection and data quality control
A standardized database was used for data collection, including maternal and infant baseline data, and information on clinical care and outcomes of infants. A trained and dedicated data abstractor collected data at each site using standardized data de nitions established prior to study commencement. Data were uploaded monthly to the research center of Children's Hospital of Fudan University, and data integrity and quality were checked by the research center.

Measures and de nitions:
ACS administration was de ned as any dose of dexamethasone administration prior to preterm delivery. Whether the ACS course was complete or incomplete was not identi ed in this database. Hypertensive disorders of pregnancy (HDP) was defined as either chronic hypertension (persistent elevation of blood pressure before 20 weeks of gestation or prior to pregnancy) or pregnancy-induced hypertension if blood pressure >145/95 was first recorded after 20 weeks of gestation. Prolonged premature rupture of membrane (PROM) was defined as membrane rupture more than 6 hours before the onset of regular spontaneous uterine contractions. Prenatal antibiotics was de ned as administration of antibiotics during second and third trimester of pregnancy. The Transport Risk Index of Physiologic Stability (TRIPS) score was used as an illness severity score on NICU admission 14 . Bronchopulmonary dysplasia (BPD) was de ned as mechanical ventilation or oxygen dependency at 36 weeks of postmenstrual age or discharge 15 . NEC was de ned as ≥stage 2 according to Bell criteria 16 . IVH was de ned as ≥ grade 3 according to Papile criteria. Periventricular leukomalacia (PVL) was de ned as the presence of periventricular cysts on cranial ultrasound or cranial magnetic resonance imaging scans 17 . Early-onset sepsis (EOS) was de ned as the presence of clinical symptoms and a positive culture from blood or cerebrospinal uid samples drawn within 72 hours after birth 18 . Retinopathy of prematurity (ROP) was de ned as ≥ stage 3 according to the International Classi cation of ROP 19 .

Statistical analysis
Stata / SE 15.0 software was used for statistical analysis. For normally distributed data, Mean ± SD, and t-test were used for comparison between groups; for non-normally distributed data, Median (Q1, Q3), and rank sum test were used instead. Count data were expressed as frequency and rate, and the χ 2 test or Fisher exact probability method were used for comparison between groups. Logistic multiple regression analysis was used to analyze for risk factors associated with ACS. The P < 0.05 level of signi cance was used.

Results
The prevalence of ACS use (at least one dose) was 71.2% (6325/8889). On univariate analysis, women who received ACS prior to delivery were more likely to have regular antenatal care, HDP, PROM, prenatal antibiotics, and delivery by cesarean section (CS) compared to those with no ACS exposure (Table 1). Infants exposed to ACS during pregnancy had smaller birth weight and gestational age, and were more likely to be SGA and have Apgar score <4 at 1 and 5 minutes of life. Logistic regression analysis of perinatal factors showed that factors independently associated with ACS exposure were small for gestational age (SGA), HDP, CS, PROM and prenatal antibiotics. There was no correlation between the use of ACS and infant gender, GDM and primipara (table 2). Multivariate logistic regression also showed that infants exposed to ACS during pregnancy had smaller birth weight and gestational age, and were more likely to be SGA and have Apgar score <4 at 1   The proportion of ACS use varied from 20.2% to 92.0% in these 18 maternal centers. There was signi cant inter-institutional variation in ACS use for different gestational age groups (Figure 2). The proportion of ACS use was positively correlated with the number of the infants (Pearson coe cient 0.487, p=0.04), and particularly so among very low birth weight infants (Pearson's coe cient 0.524, p=0.03).

Discussion
Antenatal corticosteroids administration has become an important obstetrical practice for improving the outcomes of preterm infants less than 34 weeks of gestational age since 1972 8 . It reduces the risk of neonatal mortality and morbidities including IVH, NEC, and ROP in every gestational age group 10 . ACS use in North America and Europe were reported to be between 70-91.4% 20,21 , which is signi cantly higher than the 50-56% reported previously in China 13 . Although the 71.2% incidence reported in our study is a signi cant improvement over previous reports, there is still room for improvement in China.
Administration of ACS to pregnant women at risk of preterm delivery is standard of care for obstetricians in China. Usually a course of intra-muscular dexamethasone (6mg at 12 hours interval for two days) is used in China instead of the betamethasone (12 mg at 24 hours of interval for two days) used in North America and Europe 22 . Brownfoot et al reported that dexamethasone may be associated with lower incidence of IVH and shorter duration of hospitalization but the data is inconclusive 22 . A more recent study reported no signi cant difference in outcomes at 2 years of age 23 . Dexamethasone and betamethasone are both safe for pregnant women 23 . Although infants previously exposed to these ACS have an increased risk of long-term adverse neurodevelopmental and neurosensory outcomes when delivered at term 24 , there was no evidence that a single course of ACS increased the risk of metabolic disease long term 25 . However, there may be risks in repeated courses of ACS 26,27 . It is very challenging for obstetricians to accurately predict whether preterm delivery will occur within one week and when ACS should be optimally administered 28 . Existing tests for predicting preterm birth are inaccurate and can result in missed opportunities for using ACS 29,30 . In a Japanese report, there was a high chance of missing the ACS for pregnant women who received tocolysis due to the risk of preterm delivery while only 23% were given ACS 31 . In our study, the women who had more preterm related complications and who were 28-31 weeks of GA at delivery were more likely to receive ACS. It is possible that increasing the awareness of ACS and developing a standardized protocol may improve the rate of ACS administration to the women at risk of preterm delivery.

Variation of ACS among maternity hospitals in China
Understanding the reasons for missing ACS in pregnant women less than 34 weeks GA is very important for quality improvement 32 . Regional variations in incidence of ACS administration present an opportunity for improvement. For example, inter-institutional ACS use varied from 23% to 76% with an average of 58% in Canada in 1996-1997 33 . Following a national quality improvement effort, this improved to 91.4% and inter-institutional variation was signi cantly reduced 34 . Outcomes of these infants were also signi cantly improved 21 . Many perinatal collaboratives have worked on quality improvement of ACS administration by focusing on reducing missed opportunities and optimizing the appropriate time of use 35 . By establishing a reliable practice culture, Kaplan et al reported that ACS use increased from 76% at baseline to 86% 36 .
Similarly, in a report from California from 2005 to 2011, ACS use was increased from 82% to 87.9% with a quality improvement strategy. They also found that a lower level of care was associated with lower incidence of ACS use 37 . Of signi cance, the ACS use is lower in low and middle income countries, where the majority of preterm death occur 38 . According to the Every Newborn Action Plan report, the use of ACS varies from 4% to 74% among low and middle income countries 39 . Therefore, reducing regional differences is a viable strategy for improving ACS use and outcomes of preterm infants.

Limitations
This is a retrospective study and only singleton births were included. Information on complete versus incomplete course, or multiple courses of ACS was not available. The knowledge level of obstetricians about ACS was not investigated and may present an opportunity for improvement.

In Conclusion
The overall incidence of ACS use among Chinese level III maternal hospitals in our cohort was 71.2%. The incidence of prenatal ACS use was highest among preterm infants who were 28-31 weeks GA and in pregnancies with medical complications. There are opportunities for improving ACS use in Chinese hospitals.

Availability of data and materials
The data used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Ethical approval: This study was approved by the Ethics Committee of the Children's Hospital of Fudan University.
Consent for publication: Not applicable.
Competing: No nancial or non nancial bene ts have been received or will be received from any party related directly or indirectly to the subject of this article.

Figure 1
Flow diagram of research cohort. ACS: antenatal corticosteroids.