Objectives: To evaluate impact of risk estimation parameters for screening for Down Syndrome during the first and second trimesters.
Methods: We prospectively examined for their performance in the prenatal prediction of trisomy 21, alphafetoprotein (AFP), unconjugated estriol (uE3), total human chorionic gonadotropin (hCG), and its free subunits (free alpha-hCG, free beta-hCG) at both the first and second trimesters, and the impact of three sets of published risk estimation parameters. A total of 14,612 pregnancies were studied. All Down syndrome specimens (12 and 11 cases for first and second trimesters, respectively) and a sample of the unaffected pregnancies were analyzed.
Results: The median multiple of median (MoM) for total hCG was lower in the first trimester (1.83 vs. 2.01 in the second trimester) but no loss in discriminative power was observed if the lower variability of the results in the first trimester is taken into account (interquartile range of 0.251 vs. 0.338). The choice of distribution parameters did not alter significantly the detection rates for the various combinations of markers (p > 0.05). False positive rates were affected significantly however and for the combination AFP-uE3-free beta-hCG they varied from 14.6% to 22.6% (p < 0.001).
Conclusions: Our results suggest that specific distribution parameters would be necessary to account for the lower variability of the markers in the first trimester and the peculiarity of the total hCG assay we used.