Is prenatal identification of small-for-gestational-age fetuses useful?

Abstract

Introduction Small-for-gestational-age (SGA) neonates have an increased risk of neonatal mortality. The utility of prenatal detection of these fetuses ( Methods This historical cohort study included women giving birth at 247 French maternity units. Women with a singleton delivery (≥ 32 weeks), in cephalic presentation were eligible for the study. Medically-indicated terminations of pregnancy, pregnancies with intrauterine fetal deaths, and women with missing delivery data were excluded. Among the SGA infants, we compared those who had been identified as such in utero (i.e., exposed group) (n = 4525) with those who were not (n = 18,516). The term “SGA” at birth describes a neonate whose birth weight is at least 1.64 standard deviations (SD) lower than the mean for the infant's gestational age and sex, that is, the 5th percentile for gestational age, based on data derived from the reference population included in our national database. The principal outcome was measured by a composite variable, defined as resuscitation in the delivery room, or death in the delivery room or the immediate postpartum period, or transfer to a neonatal intensive care unit. The secondary outcomes were the 5-minute Apgar score, and the rates of neonatal traumatic lesions, operative vaginal deliveries, and cesareans. Study ethics approval was obtained on June 30, 2017 (CECIC Rhone-Alpes-Auvergne, Grenoble, IRB 5921). A log-binomial model was used to adjust for covariates previously reported in the literature as either a risk factor or a confounding factor for each outcome we studied. Results The mean birthweight in the cohort was 2472 ± 343 g. Risk of resuscitation in the delivery room or of death in the delivery room or the immediate postpartum period or of neonatal transfer to the NICU was higher for the babies in the prenatally suspected group: RR = 2.31 (95% CI: 2.20–2.42). The adjusted RR was 1.22 (95% CI: 1.09–1.36) (39.7% in the group identified prenatally as SGA vs. 17.2% in the other group). The 5-min. Apgar scores, and the rates of neonatal traumatic lesions, operative vaginal deliveries, and global and elective cesarean deliveries did not differ significantly between the 2 groups. The a posteriori study power with a = 0.05 was 99%. Conclusions Contrary to what we expected, among children born SGA, prenatal identification did not improve neonatal or maternal outcomes. Therefore, we cannot currently recommend the systematic screening of SGA fetuses to improve their neonatal prognosis. More particularly, this strategy improved the diagnosis of SGA fetuses, but paradoxically this diagnosis was not followed by improved fetal outcome. Further studies on this topic should be based on customized in utero weight estimates.