Missed opportunities for optimal antenatal corticosteroid timing in medically indicated preterm births

Objective: Although delivery timing is physician dictated in indicated preterm births, suboptimal antenatal corticosteroids (ACS) administration occurs in most cases. We aimed to characterize the patterns of use of ACS in indicated preterm births and identify missed opportunities of optimal ACS administration. Methods: We reviewed the records of women who received ACS and were delivered due to maternal or fetal indications at 24?34?weeks of gestation during 2015?2017 at a university hospital. Optimal ACS timing was defined as delivery ?24?h ?7?d from the previous ACS course. Results: Overall, 188 pregnancies were included. The median gestational age at delivery was 32?weeks. Considering only the initial ACS course, the rate of optimal timing was 32.4%. Of 105 (55.8%) women eligible (delivery >7?d since the initial ACS course), only a third (n?=?38) received a rescue ACS course. Among women who did not receive rescue ACS course despite their eligibility (n?=?67), the decision-to-delivery was ?3?h in 36 (53.7%), and ?24?h in 20 (29.9%), representing 19.1 and 10.6% of the entire cohort, respectively. The urgency of the decision to deliver (i.e. in the upcoming 24?h and later) and allowing a trial of labor, were both positively associated with decision-to-delivery interval ?3?h and ?24?h. The rate of delivery within any optimal window (either initial or rescue course) was 40.4%, with gestational hypertensive disorders (OR [95% CI]: 2.40 (1.23, 4.72), p = .01) and decision to deliver made at first hospitalization (OR [95% CI]: 2.27 (1.04, 4.76), p = .04) as independent positive predictors of optimal ACS timing. The rate of composite adverse neonatal outcome was significantly lower in those with optimal ACS administration as compared to those with suboptimal timing (32.9 versus 50.9%, OR [95% CI]: 0.47 (0.26, 0.87), p = .02). Conclusions: Suboptimal ACS administration occurred in most indicated preterm births. Underutilization of rescue ACS course and a substantial rate of missed opportunities for optimal ACS administration were identified as potentially modifiable contributors to improve ACS timing.