What prenatal cell-free DNA screening does, and does not, tell you

Prenatal cell-free DNA screening — sold as NIPT, "the blood test for Down syndrome" — is a very good screen and not a diagnosis, and the distinction is the whole point. In the largest trial to test it in a routine pregnancy population, the test flagged all 38 cases of trisomy 21 it was given, but a positive result corresponded to an actual affected fetus only about 81 percent of the time, and the figure is lower still for rarer conditions [s1]. A screen sorts pregnancies into higher and lower risk; it does not establish what is true of a particular fetus, which is why a positive result is followed by diagnostic testing, not by a decision [s2].

Because this piece bears on how a screening result should be read, its risk tier is held: the numbers below describe a test's behaviour across populations, not the probability for any individual pregnancy, which depends on factors an obstetric clinician weighs case by case. It is not medical advice.

What the test measures

Cell-free DNA screening reads fragments of DNA circulating in the pregnant person's blood, a mixture that includes DNA of placental origin, and estimates whether the fetus is likely to have an extra copy of chromosome 21, 18 or 13 [s1]. It can be done from about 10 weeks. In the NEXT trial, researchers enrolled 18,955 women presenting for aneuploidy screening at 10 to 14 weeks across 35 centres, and had analysable results for 15,841; the mean maternal age was 30.7 years and the mean gestational age at testing was 12.5 weeks [s1]. Crucially, this was a general population, not the high-risk women in whom the test was first validated — the setting that matters for how the test performs when offered to everyone.

The trial's numbers

Against standard first-trimester screening — nuchal translucency ultrasound plus blood analytes — cell-free DNA did better on every measure the trial reported for trisomy 21. The area under the ROC curve was 0.999 for the DNA test versus 0.958 for standard screening [s1]. It detected trisomy 21 in 38 of 38 affected pregnancies (100 percent; 95% CI, 90.7 to 100), against 30 of 38 (78.9 percent; 95% CI, 62.7 to 90.4) for standard screening [s1]. And it produced far fewer false alarms: a false-positive rate of 0.06 percent (95% CI, 0.03 to 0.11) versus 5.4 percent (95% CI, 5.1 to 5.8) for standard screening [s1].

Those figures are why the test spread so fast. But detection rate and false-positive rate are not what a person who gets a positive result actually needs to know. That is the positive predictive value — the chance a positive is real.

Why "positive" is not "yes"

In the NEXT population, the positive predictive value for trisomy 21 was 80.9 percent (95% CI, 66.7 to 90.9) for cell-free DNA, compared with 3.4 percent (95% CI, 2.3 to 4.8) for standard screening [s1]. Read the second number carefully: under the older screen, the great majority of "positive" results were false. The DNA test is vastly better, but even at 80.9 percent a meaningful share of positives are still not true trisomy 21 [s1].

Predictive value is not a fixed property of a test; it falls as the condition being screened for gets rarer. Trisomy 21 is the most common of the three, so its predictive value is the most favourable. For the rarer trisomies 18 and 13, and still more for the microdeletion syndromes some laboratories add, the same test throws a much higher proportion of false positives, because a highly sensitive test applied to a very rare condition inevitably flags more people who do not have it than people who do. The general principle — that predictive value tracks prevalence — is the same one that governs multi-cancer blood tests, and it is the single most misunderstood feature of screening.

What it does not cover

The American College of Obstetricians and Gynecologists frames prenatal testing as a menu in which no single test is superior in all circumstances, each offering different information and performance, and it stresses that screening provides a risk assessment rather than a diagnosis [s2]. Cell-free DNA screening looks for specific chromosomal conditions; it is not a scan for the many structural and genetic conditions it was never designed to detect, and a reassuring result does not rule those out [s2]. A definitive answer comes only from diagnostic testing — chorionic villus sampling or amniocentesis — which is why ACOG casts the decision as one requiring counselling rather than a number handed over without context [s2].

One provenance note belongs on the record: the NEXT trial was funded by Ariosa Diagnostics, which sold a cell-free DNA test, and by the Perinatal Quality Foundation [s1]. The result was independently published and blinded in design, but a commercial interest in the outcome is part of how any reader should weigh it.

The takeaway is narrow and worth holding onto: this is among the best screening tests in routine medicine, and it is still a screen. A negative lowers the odds; a positive raises them enough to warrant a diagnostic test, not a conclusion.