OBJECTIVE: To assess the effect of antenatal magnesium sulfate exposure on
neonatal demise.
STUDY DESIGN: A retrospective analysis of prospectively captured data from
100 tertiary centers between May 1997 and January 2000 was performed. Inclu
ded were nonanomalous newborns who were admitted to the neonatal intensive
care unit between 23 and 34 completed weeks' gestation. Predictors of neona
tal demise were determined from a pool of 24 candidate variables in a univa
riate analysis. A multivariate predictive model for mortality was construct
ed by using the variables that had significant interactions with the rate o
f demise (P less than or equal to .1).
RESULTS: A total of 12,876 cases were available for analysis. When these ca
ses were stratified according to gestational age, magnesium was associated
with a significant reduction in neonatal demise (OR, 0.67; 95% Cl, 0.54 to
0.84; P =.0005). The effect remained when controlling for both gestational
age and indication for therapy (adjusted OR, 0.70; 95% Cl, 0.56 to 0.89; P
=.003). The effect was similar in direction and magnitude in the final mode
l after controlling for additional antenatal factors (OR, 0.82; 95% Cl, 0.6
5 to 1.04; P =.108).
CONCLUSION: Antenatal magnesium exposure is not associated with neonatal de
ath, regardless of indication for therapy.