The "genetic sonogram": comparison of the index scoring system with the age-adjusted US risk assessment.

Abstract

To compare two ultrasonographic (US) methods for prenatal detection of fetal Down syndrome. Genetic amniocentesis was successfully performed in 3,303 consecutive women with high-risk pregnancies (mean gestational age, 17.1 weeks). All patients underwent a complete "genetic US" examination prospectively. Risk was assessed by using (a) various modifications of the index scoring system (ISS) and (b) the age-adjusted US risk assessment (AAURA). The prevalence of Down syndrome in this population was 1.6% (53 of 3,303). By using a threshold of at least 2 points to detect trisomy 21, the best ISS had a sensitivity of 45.3%, false-positive rate of 4.9%, likelihood ratio of 9.3, and positive predictive value in the high-risk population in this study of 13.3%. Lowering the threshold to 1 point increased the sensitivity to 60.4% but increased the false-positive rate to 15.8%. Adding points for age increased the sensitivity to 67.9% but increased the false-positive rate to 24.3%. Results of using AAURA to detect trisomy 21 were nearly identical, with a sensitivity of 43.4% and false-positive rate of 4.9% at a 1 in 36 risk threshold and a sensitivity of 69.8% and false-positive rate of 26.1% at a 1 in 200 threshold. Trisomies 18 and 13 were detected with sensitivities of 80.0% and 100.0%, respectively, with either system. The modified ISS and AAURA are equivalent in screening for Down syndrome, with detection of approximately half of all trisomy 21 fetuses at a 5% false-positive rate.