Maternal serum screening for Down syndrom is an established practise in man
y countries. In the second trimester human chorionic gonadotrophin (hCG) or
free beta-hCG is the marker of first choice, with alpha-fetoprotein(AFP) a
s the second marker and unconjugated oestriol (uE(3)) the third. Statistica
l models with parameters derived by meta-analysis predict that a three mark
er combination will yield a 67% detection rate for a 5% false-positive rate
. The model prediction have been confirmed in 21 large prospective interven
tion studies. A fourth marker, inhibin A, increases the detection rate by 7
% for the same false-positive rate. In the first trimester, similar models
predict that a combination of pregnancy associated plasma protein A, free b
eta-hCG, AFP and uE(3) will yield a 70% detection rate. This is increased t
o 88% if ultrasound nuchal translucency is used as an additional marker. Sc
reening can also be extended to Edwards' syndrome, yielding high detection
rates with little increase in the false-positive rate. Abnormal marker leve
ls are also associated with a variety of adverse outcomes of pregnancy. Hig
h quality information and decision aids are needed to minimise anxiety amon
g screenees. (C) 2000 Elsevier Science Ireland Ltd. All rights reserved.