Cross-Validierung des Screening-Algorithmus der FMF London an 38 700 Ersttrimester-Untersuchungen in Deutschland

Abstract

Validation of the performance of the new algorithm of the FMF London for screening for trisomy 21 using a combination of maternal age, fetal nuchal translucency (NT) and maternal serum free β-hCG and PAPP-A. Between 2002 and 2007, NT was measured prospectively in 39,004 pregnancies in the context of routinely performed first trimester screening in Germany. Individual trisomy 21 risks were calculated by a combination of NT, maternal age, free β-hCG, and PAPP-A using the FMF algorithm in force at the time of investigation. In this study we recalculated the trisomy 21 risks applying the new algorithm of the FMF UK that includes the new mixture model for the NT measurement. 38,751 singleton pregnancies could be included in the study of which 109 (0.3 %) had a trisomy 21. Only 35 % of the NT measurements of euploids were above the median and 25 % of the NT measurements were below the 5th percentile of the FMF UK. For sonographers that were qualified according to level II or III of the German DEGUM system, the median NT of fetuses with trisomy 21 was 0.9 mm above the median of the FMF UK and only 0.5 mm above the median for all other sonographers. Despite the limited performance of the NT measurement, the overall detection rate for a trisomy 21 was 90.8 % when combining the NT with maternal age, PAPP-A and free β-hCG. The overall false-positive rate for a trisomy 21 was 6.5 % at a cut-off value of 1:300. In this study we were able to show that the use of the new risk algorithm of the FMF UK leads to a trisomy 21 detection rate of about 90 % at a 5 % false-positive rate in a German collective despite a significant underestimation of the NT.