Prenatal screening for Down syndrome: the problem of recurrent false-positives

It has been reported that, in prenatal screening programmes for Down syndrome, women who have false-positive results in one pregnancy have an increased risk of a false-positive result in a subsequent pregnancy. We examined the effect of this in the screening programme conducted from the Wolfson Institute of Preventive Medicine with a view to determining the magnitude of the effect, and to describe a method of avoiding the problem.

Six thousand four hundred and forty-eight women were identified who had had two singleton pregnancies without Down syndrome in the screening programme based at the Wolfson Institute of Preventive Medicine, in which both pregnancies were screened using a Quadruple test (maternal age with alphafetoprotein (AFP), unconjugated oestriol (uE(3)), total or free beta-human chorionic gonadotrophin (hCG) and either free alpha-hCG or inhibin-A as the fourth serum marker).

Among women who had a false-positive result in their initial pregnancy, the false-positive rate in the subsequent pregnancy was high: 20% (46/229), about three times higher than both the overall observed false-positive rate (6.6%), and the expected false-positive rate, in subsequent pregnancies that were false-positive in their initial pregnancy (7.5%) (p < 0.001). This arises because serum marker levels in one pregnancy are associated with the levels in a subsequent pregnancy. Using the slope (the regression coefficient b) of each marker level in a subsequent pregnancy regressed on the value in the first pregnancy, it is possible to adjust all marker values in a subsequent pregnancy to allow for the higher-than-expected false-positive rate. This can be done by dividing the observed MoM value for each marker by the 'expected' MoM, which is the MoM value in a previous pregnancy raised to the power b.

If a woman has had a false-positive result in one pregnancy, she is much more likely to have a false-positive screening result in a subsequent pregnancy than women in general. The problem can be avoided by adjusting the serum markers in all women who have been screened in a previous pregnancy and who have not had a previous pregnancy with Down syndrome.