False positive Down syndrome screening — more common than you think

This is shocking

NOTE: This piece was originally sent to paid subscribers only, but I wanted to share it with all of you now. Year end roundup coming this week!

I recently had the opportunity to interview Stephanie (not her real name), about a really bizarre experience she had, not once but three times. This then prompted me to begin investigating what is a very bizarre phenomenon. This post is only the beginning. If you’re interested in testing accuracy across the medical field, this article is for you.

When Stephanie was nineteen and was dating the man who is now her husband, she got pregnant and miscarried at twelve weeks. She told me that because of her age, she was treated quite dismissively by her medical providers. They told her that there was probably something chromosomally abnormal about her fetus that provoked the miscarriage, but no testing was done. (This is not alarming, most of the time testing is not done unless miscarriage is repeated.)

Within six months, she was pregnant again. She was going to an obstetrician at the time, and had an ultrasound late in the second trimester that revealed a hole in her daughter’s heart (which is actually quite common, and almost always heals up.) The doctor recommended an amniocentesis, which resulted in a diagnosis of Down Syndrome. Follow up blood tests confirmed the diagnosis, although as we will learn later in this article, those tests are highly suspicious. (Also, I thought amniocentesis was pretty much perfect. Why bother with a blood test at this point? Is this something I should be investigating further? If anyone knows about false results from amniocentesis I’d love to know.) Later on, another ultrasound was performed, that indicated no markers for Down syndrome, and that furthermore, the hole in her heart had closed up. When the baby was born, she had no markers for Down Syndrome. She is now almost thirteen and developmentally completely normal and perfectly healthy. (I would like to add the caveat that NIPT testing, which this post focuses on, was introduced in 2011 as an improvement on prior screening tests.)

But Stephanie’s story gets stranger. When she was pregnant the third time (her second live birth), she had the blood test for genetic abnormalities. This test is commonly referred to as NIPT, for non-invasive prenatal testing, and it tests for a handful of more common genetic abnormalities, including trisomy 21 (Down syndrome) and trisomy 18 (Edwards syndrome). I think it’s relevant to note that no one at that point had mentioned to her that she could opt out of these tests. The blood test, once again, came back positive for Down Syndrome. By this point, having had the same experience with her first child, she was somewhat skeptical and was also gifted with a very positive personality. They chose not to undergo amniocentesis at this point and her second daughter was born with no indication of Down Syndrome. Again, she is a perfectly healthy and developmentally normal ten year old.

Stephanie’s next pregnancy ended in miscarriage as well, but the pregnancy with her third child was also accompanied by a positive Down Syndrome blood test, at which point her midwife (prior to this she had been seeing MDs) advised her that she might want to refuse the test in the future, because there was clearly something about her that was tripping this test to give positive results every time. Her third child went on to be born with no markers of Down syndrome, and she has since had three completely normal children who were never tested.

This led me to look into the accuracy of these blood tests, and I was not terribly surprised to find out that there is a wide range of alleged accuracy to these tests. An initial google search results in the following (from Nuffield Council on Bioethics:

The websites of most private providers state that NIPT is ‘99% accurate’ or has ‘99% sensitivity’. I think most people would assume that this means their result will tell them pretty much for sure whether their fetus has one of the conditions or not. But as the conditions affect less than 1% of all children born, a dummy test that gave everybody a low chance result would be 99% accurate.

A more helpful statistic is the positive predictive value. Sometimes NIPT gives a high chance result when the fetus does not actually have the condition. If you receive a high chance result for Down’s syndrome, there is a 1 in 5 (20%) chance that the result is wrong and your fetus does not have the condition.

That’s actually a decent explanation of positive predictive value for the layperson, so kudos to them. For a detailed explanation of PPV, please refer to my post from May 2022. But it gets worse. According to the BMJ, “Hindley and Medakkar¹ showed that the clinical diagnosis of Down’s syndrome is inaccurate in one third of cases.” On third is even worse than the twenty percent listed above.

There was even a splashy article in the New York Times, When They Warn of Rare Disorders, These Prenatal Tests Are Usually Wrong. “For every 15 times they correctly find a problem, they are wrong 85 times.” (!!!!!) This article refers to rarer genetic conditions, but should serve as a strong warning.

The tests initially looked for Down syndrome and worked very well. But as manufacturers tried to outsell each other, they began offering additional screenings for increasingly rare conditions.

The grave predictions made by those newer tests are usually wrong, an examination by The New York Times has found.

[…]

Some of the companies offer tests without publishing any data on how well they perform, or point to numbers for their best screenings while leaving out weaker ones. Others base their claims on studies in which only one or two pregnancies actually had the condition in question.

This isn’t the first time Silicon Valley technology has been used to build a business around blood tests. Years before the first prenatal testing company opened, another start-up, Theranos, made claims that it could run more than a thousand tests on a tiny blood sample, before it collapsed amid allegations of fraud.

The whole article is quite bizarre, and this is quite the rabbit hole to go down. You can find false positive rates from between 0.5% to 50%. Clearly there is little oversight involved. One more from the NYT:

“It’s a little like running mammograms on kids,” said Mary Norton, an obstetrician and geneticist at the University of California, San Francisco. “The chance of breast cancer is so low, so why are you doing it? I think it’s purely a marketing thing.”

This reminds me of my last post. Science, and medical science in particular, have become so corrupted. And don’t get me started on mammograms. (Or do. I’m game.)

There is also the phenomenon of inherited chimeric DNA, the phenomenon in which the DNA from a previous pregnancy remains in the mother’s body long term. According to Scientific American, up to 6% of free floating DNA in the mother’s blood coming from prior pregnancies. There have been examples of male DNA remaining in the mother’s body for 27 years. Could it possibly have been the case that that first pregnancy of Stephanie’s that spontaneously miscarried had had Down syndrome and that that DNA remained in her body to trip the subsequent tests? It turns out that this might indeed have been the case. From an article in Pediatric Radiology (referring to the DNA test, not amniocentesis):

Most false positives result from the presence of increased chromosome 21-specific DNA, whose origin does not reflect the chromosome composition of the fetus in the ongoing pregnancy. Possible origins of this increased DNA include the placenta (from confined placental mosaicism), a vanishing twin, maternal mosaicism and other maternal medical conditions.

The implications of this are horrifying. How many wanted pregnancies are being terminated on the basis of a lie, or at best an extremely flawed test? How many parents go through a pregnancy planning and stressed out about a condition their child doesn’t even have? Consider the article from AAMC:

Prenatal screenings can lead to false positives, heightened anxiety

This particular article is alarming because it brings up the ethical issues inherent in diagnosing certain disorders or disabilities, as well as mentioning the possibly alarming fact that: “Since NIPT [non-invasive prenatal testing] became available, the number of patients who choose an invasive diagnostic test has dropped significantly. One study found that at Mount Sinai West Hospital in Manhattan, the percentage of patients who had prenatal diagnostic tests dropped from 38% in 2010 to just 2% in 2015.” Regarding the ethical difficulties (emphasis mine):

While prenatal screening and diagnosis has helped advance pregnancy care, some scientists, bioethicists, and disability rights advocates worry that without proper regulation and education, the practices could reinforce stereotypes and further marginalize people with disabilities.

“Responsible scientists are very concerned about the ethics,” Bianchi says. “Not only for identifying fetal [abnormalities, but] for subsequent care.”

In a 2015 article for the European Society of Human Genetics and the American Society of Human Genetics that Bianchi co-authored with more than a dozen other scientists, the experts raised the concern that the use of selective abortions for fetuses with genetic conditions makes the screening practice “morally sensitive” because it could lead to subtle pressure on the parent to terminate the pregnancy and it could send “a discriminatory message about the worth of the lives of people living with the relevant conditions.”

The National Council on Disability, an independent advisory agency of the U.S. government, published a report in 2019 titled “Genetic Testing and the Rush to Perfection,” which called for the government to further regulate the commercial use of NIPT and its marketing, increase education on disabilities, and ensure neutral genetic counseling is provided to patients along with screening, among other recommendations.

Of course, we have heard for years arguments about such testing, as well as the potential use of in-vitro fertilization to select for certain characteristics, like sex, to create so-called “designer babies.”

There are multiple articles regarding false positive Down syndrome screening tests, and I won’t quote from all of them or even most of them, but I will refer the reader to a few good ones at the end of this post, as well as touching on the wild range of values I’ve found for the NIPT positive predictive value. I do want to draw attention to the following article from The Bump, because it discusses the false negative rate as well, which is dramatically different.

FDA Report Warns Against False Positives in Some Prenatal Genetic Tests

The following quote is long, but it’s important. All emphasis is mine.

NIPTs are blood tests designed to look for genetic (chromosal) abnormalities and disorders in baby. The tests are used by almost a third of American pregnant women, and while the results aren’t a confirmed diagnosis, they do tend to have a good accuracy rate for certain genetic disorders, like Down Syndrome. A negative result is generally greater than 99.9 percent, according to the FDA, while a positive result for Down syndrome is about 90 percent accurate.

However, the American College of Obstetricians and Gynecologists (ACOG) doesn’t recommend using NIPT to detect microdeletions, which occur when a very small part of a chromosome is missing, and the consequential disorders are often rare. A New York Times investigation published in January 2022 looked into high false positive rates for NIPTs that tested for microdeletions and rare conditions. It found that some of the tests were marketed in a misleading way and reported that some women had terminated their pregnancies as a result.

According to the FDA, how accurately a NIPT test can predict baby’s risk for a genetic abnormality is dependent on how common that abnormality is. Due to this, when testing for microdeletions and rare conditions, the likelihood of false positive results is high—according to the FDA, the positive predictive value can range from 2 to 30 percent, depending on the condition.

90% accuracy isn’t terrible, but is this figure true? It seems like no one really knows. A technical article in Molecular Cytogenetics reported a false positive NIPT result (over a group of genetic conditions including Down syndrome) of 47%. Another article from the U.K. Down Syndrome Association also reports a similar positive predictive value of 46% for mothers in their twenties, increasing to 87% at the high end of maternal age (early forties). The FDA seems to like the 90% figure.

To conclude, these figures are somewhat concerning, to say the least, and the implications for prenatal care ought to be seriously considered. This will be a continuing investigation, as this piece only scratches the surface. Stay tuned, because there are more layers to this story, including genetic conditions of the mother that might factor into the phenomenon of false positives. Finally, this should serve as a warning, as well as evidence that terrible tests are not limited to HIV and COVID.