Screening for small for gestational age infants in early vs late third-trimester ultrasonography: a randomized trial

Abstract

Recent studies have demonstrated that a routine third-trimester ultrasound scan may improve the detection of small for gestational age infants when compared with clinically indicated ultrasound scans but with no reported reduction in severe perinatal morbidity. Establishing the optimal gestational age for the third-trimester examination necessitates evaluation of the ability to detect small for gestational age infants and to predict maternal and perinatal outcomes. Intrauterine growth restriction most often corresponds with small for gestational age infants associated with pathologic growth patterns. This study aimed to assess the performance of routine early ultrasound scans vs late ultrasound scans during the third trimester of pregnancy to identify small for gestational age infants and fetuses with intrauterine growth restriction. This was an open-label, randomized, parallel trial conducted in Upper Normandy, France, from 2012 to 2015. The study eligibility criteria were heathy, nulliparous women older than 18 years with gestational age determined using the crown-rump length at the first trimester routine scan and with no fetal malformation or suspected small for gestational age fetus at the routine second trimester scan. Pregnant women were randomly assigned to a third-trimester scan group at 31 weeks gestational age ±6 days (early ultrasound scan) or at 35 weeks gestational age ±6 days (late ultrasound scan). The primary outcome of this trial was the ability of a third trimester scan to predict small for gestational age infants (customized birth weight <10th percentile) and intrauterine growth restriction (customized birth weight <third percentile) using birth weight as the gold standard. The purpose of these adjustments was to optimize the detection of fetal weight associated with pathologic growth patterns. It was calculated that a sample size of 3720 women would be required to obtain 80% power at a 2-sided level of 0.05 with a 15% difference in sensitivity between the 2 intervention group to detect small for gestational age fetuses in favor of the late ultrasound scan group and considering that small for gestational age infants would represent 10% of all live births. Secondary outcomes were maternal and perinatal morbidities with interventions reported. The analysis was based on the intention-to-treat principle. Results from 1853 women assigned to the early ultrasound scan group and 1848 women assigned to the late ultrasound scan group were analyzed. The sensitivity was found to be higher in the late ultrasound scan group than in the early ultrasound scan group, both for identifying small for gestational age infants (27%; 22%-32% vs 17%; 13%-22%; P=.004) and intrauterine growth restriction (44%; 35%-54% vs 18%; 11%-27%; P<.001). There was little difference in the specificity between the late ultrasound scan and early ultrasound scan groups in identifying cases of small for gestational age (97%; 96%-98% and 98%; 97%-99%, respectively; P=.04) and intrauterine growth restriction (96%; 95%-97% and 97%; 96%;-97%, respectively; P=.24). Overall, the maternal and neonatal outcomes were comparable between the early ultrasound scan and late ultrasound scan groups with the exception of additional (at least 1) ultrasound scans performed (25% in the early ultrasound scan group vs 19% in the late ultrasound scan group; P<.001). Rates of perinatal death (0.4% vs 0.8%; P=.12) and adverse perinatal outcomes (1.8% vs 2.7%; P=.08) were comparable between the early ultrasound scan and late ultrasound scan assigned groups, and the overall sensitivity to detect small for gestational age infants and intrauterine growth restriction, including in the last ultrasound scan performed before delivery, were also similar (30%; 25%-36% vs 26%; 21%-31%; P=.23; and 50%; 40%-60% vs 38%; 28%-48%; P=.07). A late ultrasound scan performed in the third trimester increases the probability of detecting small for gestational age infants and intrauterine growth restriction with fewer additional scans reported than for the early ultrasound scan group. The overall perinatal outcome risk was comparable between the 2 groups. However, the overall sensitivity for detecting small for gestational age fetuses and intrauterine growth restriction, including in the last ultrasound scan performed before delivery, remains comparable between the late ultrasound scan and early ultrasound scan groups.