Intended for healthcare professionals


Antenatal screening for Down's syndrome

BMJ 2004; 329 doi: (Published 07 October 2004) Cite this as: BMJ 2004;329:811
  1. Zarko Alfirevic (zarko{at}, professor of fetal and maternal medicine,
  2. James P Neilson (jneilson{at}, professor of obstetrics and gynaecology
  1. School of Reproductive and Developmental Medicine, University of Liverpool, Liverpool L8 7SS
  2. School of Reproductive and Developmental Medicine, University of Liverpool, Liverpool L8 7SS

    Nuchal translucency plus biochemical tests has the lowest false positive rate

    Many pregnant women wish to undergo antenatal testing for Down's syndrome—for reassurance that their unborn child does not have Down's syndrome, to allow the option of termination if it does, or to allow preparation for the birth of a baby with the condition. Unfortunately, invasive tests required to obtain tissue for fetal karyotyping (chorionic villus sampling, amniocentesis) cause loss of the pregnancy in about 1% of cases. The challenge of an antenatal screening programme is, therefore, to identify women in whom a risk of Down's syndrome is sufficiently high to justify such an invasive test and to minimise the risk of miscarrying a healthy baby.

    Initially, invasive testing was offered only to women over 35 years, but this identified only one third of fetuses with Down's syndrome. Universal screening started with the observation that serum concentrations of α fetoprotein, used to screen for neural tube defects, tended to be lower when the fetus had Down's syndrome. Several other biochemical tests were combined with age related risk to calculate an individualised risk for Down's syndrome. Anyone with a risk of 1:250 or greater was offered amniocentesis; others were reassured with a low risk result. Another breakthrough was the advent of screening for increased nuchal translucency, a fluid filled space behind the fetal neck, which tends to be more prominent in fetuses affected by Down's syndrome between 10 and 13 weeks of pregnancy.1 Nuchal translucency screening allowed earlier testing, but required expertise and equipment not readily available.

    The competing claims of advocates of different screening approaches have not made it easy for health planners, clinicians, or pregnant women to reach a balanced decision about what should be offered, or chosen.

    The publication of the Serum, Urine and Ultrasound Screening Study (SURUSS) has advanced our knowledge of the efficacy and safety of antenatal screening for fetal Down's syndrome and placed choices on a firmer platform of evidence.2 Over 47 000 singleton pregnancies were studied prospectively in 25 maternity units. Ultrasound scans of fetal nuchal translucency in the first trimester and assays of several biochemical markers in the first and second trimesters combined were done to test the screening performance of various “packages.”

    For the principal analysis false positive rates (when the test is positive but the fetus does not have Down's syndrome) were compared for each test, and combined test packages, assuming a fixed detection rate for Down's syndrome of 85%. The best performer was the integrated test, comprising ultrasound measurement of fetal nuchal translucency and assay of serum pregnancy associated plasma protein-A (PAPP-A) at 10 weeks, combined with quadruple tests of serum α fetoprotein, unconjugated oestriol, β-human chorionic gonadotrophin (HCG), and inhibin-A during the second trimester (after 14 weeks). This two step package had a false positive rate of only 0.9%. The best first trimester screening package was a combination of nuchal translucency scan, serum free β-HCG, and pregnancy associated plasma protein A, which had a false positive rate of 4.3%. Second trimester quadruple testing alone had a false positive rate of 6.2%.

    Any screening programme package needs to include a risk cut-off point with an expectation that “high risk” women will opt for an invasive prenatal diagnostic test while the rest will be sufficiently reassured not to seek further testing. The UK National Screening Committee has set a screening programme target of a detection rate of at least 75% with a false positive rate of 3% or less by April 2007. Using a currently recommended cut-off of 1:250, the estimated performance of the two step integrated test is clearly superior, with a detection rate of 91% and false positive rate of only 2.6%. But we know that some women ask for the one stop first trimester package, accepting a slightly inferior screening performance (85% detection rate, 4.2% false positive rate). Also we should not forget that some women access maternity services for the first time in the second trimester and others will request definitive diagnosis by chorionic villus sampling or amniocentesis irrespective of their risk. The integrated test may be the most cost effective, but any “one size fits all” policy sits uncomfortably with pregnant women and clinicians.

    The main challenge for pregnant women is to absorb all the relevant information in early pregnancy to allow them to make an informed choice about which, if any, screening option they wish to undergo. The main challenge to health systems will be to ensure that there are enough adequately trained sonographers to deliver an ultrasound based screening programme on a national basis—certainly a major issue.3

    Other remaining questions are behavioural and contextual. How many women will tolerate the delay between the two gestational stages of the integrated tests? And what is the importance of establishing top quality Down's syndrome screening programmes, relative to other priorities in the maternity services—notably tackling inequalities and ensuring that all women in labour have enough midwives to meet their needs?4 5


    • Competing interests None declared.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    View Abstract