Ultrasound Determination of Chorionicity in Twin Pregnancy: Accuracy and Operator Experience

Determining chorionicity in twin gestations: three-dimensional (3D) multiplanar sonographic measurement of intra-amniotic membrane thickness

OBJECTIVES

Measuring the thickness of the membranes to determine chorionicity in twins using two-dimensional (2D) ultrasound may provide uncertain results, given the possibility of measuring in an oblique view. We aimed to determine whether chorionicity can be assessed during the second and third trimesters by measuring the thickness of the membranes with three-dimensional (3D) multiplanar ultrasound, which ensures the measurement is perpendicular to the membranes, and to evaluate the reproducibility of the technique.

METHODS

This was a prospective study of 84 twin pregnancies (30 monochorionic and 54 dichorionic) between 20 and 35 weeks of gestation. The dividing membranes between the fetuses were visualized with 2D and 3D transabdominal sonography and the thickness of the membranes was measured by the multiplanar technique. Intraobserver variability was assessed by comparing the measurements obtained in sagittal and axial views by the same observer. We examined the frequency distributions to determine the optimal cut-off point of the thickness of the membranes to predict chorionicity. Kappa index, intraclass correlation coefficients and 95% limits of agreement were calculated to evaluate the inter- and intraoperator variability. We also compared membrane-thickness measurement, during the second and third trimesters, using 2D and 3D ultrasound.

RESULTS

In monochorionic and dichorionic pregnancies, respectively, the mean thickness was 1.42 (SD, 0.31) mm and 2.48 (SD, 0.47) mm. With 3D ultrasound, the intraclass correlation coefficient was 0.99 in all cases. The 95% limits of agreement were all within +/-0.2 mm. The best cut-off for membrane thickness for discriminating monochorionic from dichorionic twinning was 1.8 mm. Using this cut-off, there were one false-negative (sensitivity, 97%; 95% CI, 83-99%) and three false-positive (specificity, 94%; 95% CI, 85-99%) diagnoses of monochorionic twins. Kappa indices for intra- and interoperator variability were all above 90%, suggesting almost perfect agreement. Measurements with 2D sonography were less accurate (sensitivity and specificity, 83%).

CONCLUSION

Prenatal assessment of dividing membranes using 3D multiplanar ultrasound is reproducible and may be a useful alternative method for determining chorionicity in twin pregnancies during the second and third trimesters.

Scanning for chorionicity: comparison between sonographers and perinatologists

OBJECTIVE

In most prenatal settings, twin pregnancies are initially evaluated by sonographers. Pregnancies diagnosed as monochorionic are subsequently referred to perinatologists or specialists in fetal medicine for the confirmation of chorionicity. In order to assess this screening strategy, we have compared the diagnosis of chorionicity made by the sonographers in the ultrasound department with the diagnosis done in the fetal medicine unit.

METHODS

A cohort of women presenting with twin pregnancy and booked for prenatal care at University College London Hospitals over a 4-year period were investigated prospectively. All women were scanned at their initial visit at 11-14 weeks in the ultrasound department (US), and were subsequently referred to the Fetal Medicine Unit (FMU) for a second ultrasound evaluation. Ultrasound data were compared and diagnosis of chorionicity was confirmed by examination of the inter-twin membranes after delivery.

RESULTS

Chorionicity was determined in 172 twin cases by the two different departments. The overall rate of concordant chorionicity determination between both units was 90.1%. The rate of discordant results in dichorionic pregnancies was extremely small, 1 in 119 pregnancies (0.8%). The rate of discordant results for monochorionic diamniotic pregnancies was 5.5%. Monoamniotic pregnancies were over-diagnosed by the US technicians.

DISCUSSION

These results demonstrate that DC/DA chorionicity is accurately determined by sonographers at less than 14 weeks. In our opinion, it is both efficient and safe to rely on the diagnosis of the sonographers in DC/DA pregnancies in early pregnancy. In such pregnancies, a decision can be made either not to refer these patients for further evaluation of chorionicity by the fetal medicine team or to postpone the referral to after 14 weeks.