RESUMO
OBJECTIVE: To determine the optimal sonographic dating of dichorionic gestations. MATERIALS AND METHODS: We reviewed dichorionic pregnancies conceived with assisted reproductive technologies (ART) at 2 institutions between 2006-2016, excluding fetuses with major anomalies. Gestational age was calculated with smaller, larger, and mean of the crown-rump lengths (CRL) and biometry midgestation and compared to the ART age. The mean and mean absolute deviation of the observed gestational age from the ART age was calculated to assess accuracy, precision, and presence of bias. The incidence of small for gestational age using the smaller and larger CRLs was compared to the ART age via McNemar's test. RESULTS: Based on 140 ultrasounds, the CRL from the smaller twin best approximates the true gestational age with least bias compared to the larger twin or average (mean absolute deviation: 2.8, mean deviation: -0.1 [95% CI: -0.4, 0.2] versus 2.7, -0.9 [-1.1, -0.6] and 2.9, -1.5 [-1.8, -1.3], in days, respectively). Based on 165 ultrasounds, biometry from the smaller fetus is least accurate compared to the larger or mean (11.8, 2.5 [1.5, 3.6] versus 11.7, 0.8 [-0.3, 1.8] and 11.9, -1.0 [-2.0, 0.04], respectively). The incidence of small for gestational age neonates did not differ from the true rate using either the CRL from the larger or smaller twin (p > .05). CONCLUSION: In ART dichorionic gestations, ultrasound of the smaller fetus is most accurate in establishing gestational age in the first trimester but least accurate in the second, though all methods performed well with little clinical difference.
