Prenatal diagnosis of esophageal atresia.

The prenatal sonographic detection of esophageal atresia (EA) has been possible for more than a decade and relies on the finding of a small or absent fetal stomach bubble associated with maternal polyhydramnios. The aims of this study were to assess the accuracy of this technique and to determine whether the outcome of prenatally diagnosed EA differs from its postnatal counterpart. All fetal sonograms performed between January 1989 and October 1993 demonstrating a small or absent fetal gastric bubble were reviewed together with all neonates with EA treated during the same period. Eighty-seven fetuses with a small (n = 53) or absent stomach bubble (n = 34) were identified, representing 1.4% of all fetal sonographic surveys. Esophageal atresia was present in 15; in 13 of these, the maternal amniotic fluid volume was increased. The positive predictive value of an absent stomach bubble and polyhydramnios was 56%, and the sensitivity of prenatal sonography in the diagnosis of EA was 42%. One neonate with EA had the prenatal diagnosis established at another institution, yielding a total of 16 cases of prenatally diagnosed EA for analysis. Seven (44%) of these had trisomy 18. Of the remaining nine, two had isolated EA, two had laryngeal atresia and EA, and there were two late-gestational unexplained fetal deaths. Only four (25%) survived through the neonatal period. The prognosis of the fetus with EA is radically different from that of the neonate with EA.