Efficacy of screening for fetal Down syndrome in the United States from 1974 to 1997.

Abstract

To estimate the 16-week prevalence of Down syndrome in the United States from 1974 to 1997 and to determine the efficacy of maternal age cutoffs and triple screens for detecting it antenatally. Using natality statistics for the United States from 1974 to 1997 of maternal-age-specific live births to women 13-49 years old, we evaluated advanced maternal age (35-49 years at delivery) and the triple serum test (maternal serum alpha-fetoprotein, hCG, and unconjugated estriol) as screening tests for Down syndrome. Efficacy was evaluated using sensitivity, false-positive rate, positive predictive value, and likelihood ratio (likelihood ratio = sensitivity/false-positive rate). In 1974, the estimated second-trimester prevalence of Down syndrome was one in 740, but by 1997 that had increased to one in 504. The proportion of Down syndrome fetuses at 16 weeks' gestation in women 35-49 years old increased from 28.5% in 1974 to 47.3% in 1997. However, live births to women 35-49 years old increased more rapidly from 4.7% in 1974 to 12.6% in 1997. The likelihood ratio for maternal age to identify an affected pregnancy decreased during the study period and was substantially lower than that using the serum test. A maternal age cutoff of 35 years in the 1990s resulted in high false-positive rates and was less efficacious based on likelihood ratio and positive predictive value. Serum testing of all pregnant women would reduce the number of amniocenteses and decrease procedure-related losses.