Early genetic sonogram for Down syndrome detection.

OBJECTIVE The purpose of this study was to determine the Down syndrome sensitivity of early genetic sonography (14-<16 weeks of gestation) and to compare its diagnostic accuracy with that later in the mid trimester (16-24 weeks of gestation). STUDY DESIGN Nuchal thickness, humerus and femur lengths, hyperechoic bowel, hypoplastic fifth digit (clinodactyly), and any gross anatomic defects were measured or ascertained in singleton pregnancies that were undergoing genetic amniocentesis. Multiple stepwise logistic regression analysis was used to determine the significant sonographic markers for Down syndrome detection in each group. Multivariate gaussian algorithms that included maternal age were used to estimate patient-specific Down syndrome risk. Sensitivity and false- positive rates, receiver-operating characteristic curves, and area under the curves were calculated and compared for both groups. RESULTS There were 1,727 pregnancies with 22 Down syndrome fetuses (1.27%) in the early group versus 3,914 pregnancies with 86 Down syndrome fetuses (2.2%) in the later group. The mean +/- SD ages were 15.5 +/- 0.4 weeks versus 17.6 +/- 1.4 weeks, respectively. Early genetic sonography (14-<16 weeks) had a 100% detection rate, with a 21.2% false-positive rate. The early versus later genetic sonography had an 81.8% versus 61.6% detection rate, respectively, at a fixed 4.8% false-positive rate. Early sonography had significantly higher diagnostic accuracy (area under the curve, 0.962 vs 0.871, respectively; P =.005). In fetuses at 14 to 15 weeks, the genetic sonography was also highly accurate, with 100% detection with a 21.9% false-positive rate. CONCLUSION Early genetic sonography is highly sensitive and statistically superior to later ultrasonography for Down syndrome detection. Early midtrimester sonography achieved a diagnostic accuracy similar to that currently reported for first-trimester nuchal translucency.