[Accuracy of ultrasonography fetal weight estimation in twin pregnancies]

Objectives: Ultrasonography is an essential tool in the management of twin pregnancies. Fetal weight estimation is useful to anticipate neonatal care in case of prematurity, growth restriction or growth discordance. The aim of this study was to evaluate the accuracy of ultrasound fetal weight prediction in twin pregnancies.

Patients and methods: Two hundred and twenty-four patients with an ultrasonography within 7 days before birth were retrospectively studied. Estimated fetal weight (EFW) was calculated with the Hadlock's formula and was compared with birth weight (BW). Growth restriction and growth discordance superior or equal to 25% between the first (T1) and the second twin (T2) were recorded.

Results: Absolute differences between EFW and BW were similar for both twins (127g [56.25-210] for T1 and 110g [50-206.25] for T2). Mean absolute percentage error was 5.41% (2.32-9.65) for T1 and 5.64% (2.46-10.37) for T2 and was superior to 10% for 24% of T1 and 27% of T2. Gestational age inferior to 32 weeks, obesity and oligoamnios were associated with a better accuracy of ultrasonography. Chorionicity as well as fetal presentation did not influence fetal weight estimation. Ultrasonography in the diagnosis of growth restriction had a sensitivity of 82%, a specificity of 76%, a positive predictive value (PPV) of 22% and a negative predictive value (NPV) of 98%. For diagnosis of growth discordance, sensitivity was 72%, specificity 95%, PPV 72% and NPV 95%.

Conclusion: Fetal weight can be accurately predicted in twin pregnancies. The contribution of ultrasonography in the diagnosis of growth restriction and growth discordance is mainly due to a high NPV.