STUDY DESIGN: On the basis of a literature review, we calculated cumulative rates of delivery, neonatal survival, and cerebral palsy for progressively longer gestational age intervals beginning at 24 weeks' gestation. Sample size estimates with two-tailed chi(2) tests without Yates' correction (alpha = 0.05, beta = 0.2, 50% effect size) for a clinical trial were calculated for these rates.

RESULTS: In the 4-week interval from 24 weeks 0 days to 27 weeks 6 days the minimum number of neonates required to conduct a randomized trial would be 1189. Loss to follow-up of neonates, maternal ineligibility, and failure of enrollment would substantially increase the number of necessary gravid women delivered in the 4-week interval to complete such a trial, such that 1 million pregnancies would be required to generate the necessary number of eligible gravid women. Shorter gestational age intervals (e.g., 24 weeks 0 days to 25 weeks 6 days) required fewer neonates (n = 729) but more pregnancies (n = 1.5 million). Because of the decreasing prevalence of cerebral palsy at higher gestational ages, extending the gestational age interval (e.g., to 29 weeks 6 days) markedly increased the minimum number of neonates to 1982 but only slightly decreased the necessary number of pregnancies, to 900,000.

CONCLUSIONS: A randomized trial of maternally administered intrapartum magnesium sulfate to reduce the rate of cerebral palsy in surviving preterm neonates would be a formidable undertaking, requiring a concerted multicentered effort.