Adverse neonatal outcomes associated with antenatal dexamethasone versus antenatal betamethasone

For at least 25 years, the delivery of a corticosteroid-usually dexamethasone or betamethasone-has been the most important part of the management of preterm labor at 24 to 34 weeks gestation. Although antenatal corticosteroid treatment has been associated with a significant reduction in respiratory distress syndrome, intraventricular hemorrhage (IVH), and neonatal mortality, both animal and human studies suggest that one of these steroids, dexamethasone, may actually increase the risk of adverse neurologic outcomes. This historical cohort study was an attempt to compare the incidence of periventricular leukomalacia (PVL), retinopathy of prematurity (ROP), IVH, and neonatal mortality in 3600 very-low-birthweight (401-1500 g) infants born after antenatal treatment with betamethasone compared with dexamethasone. A complete course of dexamethasone consisted of 4 6-mg intramuscular doses given at 12-hour intervals and of betamethasone, 2 12-mg intramuscular doses given at a 24-hour interval. Nearly half of the infants (48%) were exposed to betamethasone and 34% to dexamethasone, whereas 18% received no antenatal steroid. The overall rate of PVL was 3.5% and was similar in all 3 groups (4.4% in the no-steroid group, 3.5% in the betamethasone group, and 2.9% in the dexamethasone group; not significant). Infants exposed to either steroid were significantly less likely to have IVH (30.7% in the no-steroid group, 24.4% in the betamethasone group, and 24.8% in the dexamethasone group; P <.05) or severe IVH (12.8% in the no-steroid group, 10.3% in the betamethasone group, and 9.4% in the dexamethasone group; P <.05) than those not exposed to either steroid. Neither steroid changed the risk of severe ROP (8.8% in the no-steroid group, 9.9% in the betamethasone group, and 10.7% in the dexamethasone group), but the overall incidence of ROP was increased in betamethasone-exposed infants (42.9% in the no-steroid group, 46.4% in the betamethasone group, and 42.2% in the dexamethasone group; P <.05). Infants given either steroid were at lower risk of death than those who did not receive antenatal steroids, but betamethasone reduced the risk significantly more than dexamethasone (11.0% in the no-steroid group, 7.4% in the betamethasone group, and 7.9% in the dexamethasone group; P <.05). Odds ratios for all these comparisons were computed by multivariate logistic regression analysis and confirmed the risk reductions. However, when multivariate logistic regression analysis was used to compare the 2 steroid-treated groups with each other, there was only one significant difference; the risk of neonatal death was lower with betamethasone than with dexamethasone (odds ratio, 0.44 for betamethasone vs 0.73 with dexamethasone, P <.05). These and other findings suggest that a randomized clinical trial should be considered to compare the fetal and neonatal effects of dexamethasone and betamethasone. At present, the best management might be to expose very-low-birthweight infants to betamethasone rather than dexamethasone when there is a choice.