Level 1: Likely reliable evidence

The efficacy of weekly prenatal corticosteroids was evaluated in a randomized trial of 982 women at risk of preterm birth before 32 weeks’ gestation (Lancet. 2006;367:1913-1919).

Women were treated with one corticosteroid course and randomized to betamethasone sodium phosphate 7.8 mg plus betamethasone acetate 6 mg (Celestone Chronodose) or to saline placebo by IM injection weekly until delivery or 32 weeks’ gestation. The statistical analyses were performed using intention to treat but were adjusted for gestational age, antepartum hemorrhage, and preterm prelabor rupture of membranes. Infant analyses were adjusted to account for multiple gestations. All 982 women were included in the analysis. A total of 1,146 infants were enrolled, including two who were stillborn; 1,090 (95%) survived to hospital discharge.


Continue Reading

Comparing repeat use of corticosteroids with placebo in 1,144 liveborn infants, statistically significant outcomes favored corticosteroids:
• 33% vs. 41% had respiratory distress syndrome (P = .01, NNT 14)
• 60% vs. 67% had any lung disease (P = .01, NNT 14)
• 12% vs. 20% had severe lung disease (P = .0003, NNT 14)
• 56% vs. 63% required oxygen therapy (P = .03, NNT 15)
• 30% vs. 35% required mechanical ventilation (P = .08)
• 7% vs. 12% had patent ductus arteriosus (P = .01, NNT 20)

Outcomes with no significant differences included rate of infant death before hospital discharge (5% vs. 5%), median length of hospital stay (29 vs. 32 days), and rates of intraventricular hemorrhage (6% vs. 7%), periventricular leukomalacia (1% vs. 2%), proven necrotizing enterocolitis (1% vs. 2%), retinopathy of prematurity (7% vs. 8%), any suspected systemic infection within 48 hours after birth (31% vs. 33%), and any proven systemic infection (12% vs. 13%).Maternal outcomes that were more common with repeat corticosteroids than with placebo included 67% vs. 58% rate of cesarean section (P = .01, NNH 11) and 10% vs. 5% rate of any side effect (P = .003, NNH 20), although none of the individual side effects reached statistical significance. There were no significant differences in rates of chorioamnionitis requiring antibiotics (9% vs. 8%), postpartum fever at least 38°C (7% vs. 8%), or postnatal BP ≥140/90 mm Hg (13% vs. 14%).