Antenatal Glucocorticoids for Late Preterm Birth?

Late preterm infants (born between 34 weeks and 36 weeks 5 days of gestation) make up the largest proportion of preterm births and are at higher risk for neonatal death and complications including respiratory disease than are infants born at term (at least 37 weeks of gestation).1 Late preterm survivors are also at risk for lifelong health problems, including cerebral palsy and learning difficulties.2 Personal and societal costs of these births are substantial.3 A single course of antenatal glucocorticoids given to mothers at risk for preterm birth is now standard prophylactic treatment against the respiratory distress syndrome and other serious neonatal complications.4 A recent systematic review of 26 randomized trials involving 4469 women and 4853 infants indicated that antenatal glucocorticoids administered between 24 weeks and 34 weeks 6 days of gestation reduced rates of infant death and the respiratory distress syndrome.5 However, the range of gestational ages for which antenatal glucocorticoids are recommended varies among current international clinical practice guidelines, with U.S. guidelines recommending their use up to 33 completed weeks of gestation6 and others, including those of the World Health Organization, recommending their use up to 34 completed weeks of gestation.5,7 There has been limited evidence to guide these recommendations. The three trials that have specifically assessed the use of glucocorticoids for late preterm gestation had low event rates and imprecision in or lack of reporting of neonatal deaths and respiratory distress outcomes.4 In this issue of the Journal, Gyamfi-Bannerman et al.8 report the results of a multicenter, placebocontrolled, randomized trial in which 2831 women between 34 weeks and 36 weeks 5 days of gestation and at high risk for late preterm birth were randomly assigned to receive two injections of betamethasone or placebo 24 hours apart. Women who were expected to give birth in less than 12 hours or who had a multiple gestation or pregestational diabetes were not eligible. Strengths of this trial are the similarity of the study groups at randomization, the use of placebo and masking of study-group assignments, and the nearly complete availability of data for the primary outcome (99.9%). The study showed a significantly lower risk of the composite primary outcome (stillbirth or neonatal death before 72 hours or need for respiratory support by 72 hours of age) in the betamethasone group than in the placebo group. The estimated number of women who would need to be treated to avoid one infant requiring respiratory support was 35, with a lower 95% confidence interval of 19 but an upper level of 259. In keeping with the reduced need for respiratory support, there were significantly lower risks of several secondary outcomes in the betamethasone group than in the placebo group. These risks included severe respiratory complications, transient tachypnea of the newborn, and bronchopulmonary dysplasia but not the respiratory distress syndrome. There were no stillbirths or neonatal deaths before 72 hours. However, there was a higher incidence of neonatal hypoglycemia (relative risk, 1.60; 95% confidence interval, 1.37 to 1.87) among the infants of mothers in