Abstract
BACKGROUND: Timely administration of antenatal corticosteroids improves preterm outcomes; however, in low- and middle-income countries (LMICs), multiple barriers often delay treatment. Whether exposure within 24 h before delivery confers benefits remains uncertain. We aimed to evaluate the association between antenatal dexamethasone-to-delivery intervals and preterm outcomes in an LMIC setting. METHODS: This single-center prospective cohort study was conducted at a tertiary neonatal intensive care unit in India from June 2023 to December 2024. The participants were preterm infants born at 24 0/7-33 6/7 week gestation. Exclusion criteria were major congenital anomalies, multiple courses of dexamethasone, administration-to-birth interval > 7 d, and interfacility transfer. The exposure variable was maternal first dexamethasone dose-to-delivery interval, categorized as ≤ 24 h (n = 167), > 24 h-7 d (n = 64), or no exposure (n = 101). The primary outcome was in-hospital mortality. Secondary outcomes included severe necrotizing enterocolitis (NEC), intraventricular hemorrhage, respiratory distress syndrome, and a composite of death or severe morbidity. Propensity scores from 10 covariates generated inverse-probability-of-treatment-weights (IPTW). Adjusted risk ratios (aRRs) and 95% confidence intervals (CIs) were estimated using a doubly-robust, trimmed IPTW-modified Poisson model. Sensitivity analyses used doubly-robust overlap weighting and an entropy-balanced treatment-only model. RESULTS: Of the 371 infants born during the study period, 332 met eligibility criteria (mean gestational age, 30.6 ± 2.2 week; 60.5% male). Overall mortality was 7.8% (26/332): 11.9% (12/101) without dexamethasone, 6.6% (11/167) with ≤ 24 h exposure, and 4.7% (3/64) with > 24 h-7 d exposure. Mortality did not differ by dexamethasone timing intervalgroup: ≤ 24 h (aRR, 0.58; 95%CI 0.26-1.27) and > 24 h-7 d (aRR, 0.33; 95%CI, 0.10-1.12) vs. no dexamethasone. Wide CIs were observed, reflecting imprecision due to lower-than-expected mortality in the unexposed group. Severe NEC was lower with ≤ 24 h exposure (1.8% [3/167] vs. 8.9% [9/101]; aRR, 0.15; 95%CI 0.04-0.54), persisting in sensitivity analyses (aRR, 0.11; 95%CI 0.03-0.38) and after multiple-comparison adjustment. No other secondary outcomes differed significantly. CONCLUSION: Dexamethasone administration within 24 h before delivery did not reduce mortality but significantly decreased the risk of severe NEC. Imprecise estimates due to lower-than-expected mortality, reduced statistical power, and potential residual confounding limit definitive conclusions.