Deletions of Concern: Noninvasive Prenatal Screening and Disability, Part One
On January 1st, The New York Times kicked off the new year with a front-page investigative article: “When They Warn of Rare Disorders, These Prenatal Tests Are Usually Wrong.” The early-pregnancy blood tests in question—NIPS1, usually aimed at Down syndrome—can also be used to screen for rare “microdeletions,” for which positive results are inaccurate around 85% of the time. That fact has human consequences:
In interviews, 14 patients who got false positives said the experience was agonizing. They recalled frantically researching conditions they’d never heard of, followed by sleepless nights and days hiding their bulging bellies from friends.
That NIPS does poorly for rare conditions, that its marketing is deceptive, is not news. But by linking these problems to women’s actual experiences, the article’s authors, Sarah Kliff and Aatish Bhatia, threatened a decade of industry effort to shape public perception of NIPS. This may explain the significant backlash, especially from industry professionals.
Because I have a daughter with Down syndrome, the conversation interests me. How do we talk about biotechnologies that bear on what it means to be human? Writing about NIPS for this blog in 2012 (1, 2, 3) eventually became part of a book in which I tried to think through these questions. I won’t rehearse those arguments here, except to answer a couple of questions likely uppermost in some readers' minds: that I approach these issues from a pro-choice standpoint; that I have no interest in debating or judging any woman’s decision to keep or terminate a fetus; and that prenatal tests, including NIPS, can be useful tools. I am interested in talking about prenatal testing as a way of thinking through the climate of assumption in which the tests are offered; in which decisions take place; and in which people with disabilities, and the people who love them, have to live.
What do we think about the conditions the tests attempt to detect? What do we think about the people who have these conditions—not only as potential family members, but as people in their own right? And since the tests are proprietary products, how does the marketing for them influence the conversation?
Because I began writing about the tests at this blog, I’m glad to have the chance to think aloud here about the article’s implications. To see what has changed in ten years and what remains unchanged, to point to a few complexities beyond matters of test performance, and to point to the silences in the conversation.
NIPS is a corporate product. It exists to make money. This doesn’t mean it is without benefits: in the Times article’s comment section, some women praised the test and their experience with doctors or genetic counselors. But to focus only on benefits, or to claim perfect alignment between profits and health, is problematic at best. The picture is more complex.
Unlike screening tests like mammograms or PSA tests, NIPS is proprietary: LabCorp has the MaterniT21Plus test, Natera sells Panorama, Myriad sells Prequel, and so on. And unlike breast or prostate cancer, microdeletions are extraordinarily rare. Not only are most positives false, but a negative result doesn’t tell you much: a pregnant woman can already know with 99.975% certainty that her fetus does not have DiGeorge syndrome, without taking any test at all.2 It’s unsurprising, then, that marketing exists to boost demand. As Kliff and Bhatia write,
Some said the blood screenings that look for the rarest conditions are good for little more than bolstering testing companies’ bottom lines.
“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.”
Given the tests’ mediocre performance for microdeletions, it’s unsurprising that company marketing makes the tests sound more accurate than they are:
. . . on product brochures and test result sheets, companies describe the tests to pregnant women and their doctors as near certain. They advertise their findings as “reliable” and “highly accurate,” offering “total confidence” and “peace of mind” for patients who want to know as much as possible.
In one egregious example, the report given to pregnant women for Roche’s Harmony test (for the 22q11.2 deletion) announces “High probability of a deletion – recommend genetic counseling.” But as Kliff writes, “A footnote defines ‘high probability’ as ‘1% or greater.’” That approach to truth—bold print promises, fine print disclosures—is ubiquitous in advertising, but especially damaging in products that bear on human health. It seemed and seems to me a kind of anti-art, a writing in which truth and implication are opposed. Call it Soft-Orwellian: less War is Peace or Ignorance is Strength, more Uncertainty is Certainty.
Profit drives NIPS marketing appeals, but it also appears to drive problematic billing practices. Kliff and Bhatia note, in passing, that the cost of the test varies widely: “Patients whose insurance didn’t fully cover the tests describe being billed wildly different figures, ranging from a few hundred to thousands of dollars.” Problems with billing, raised in the article’s comment section, have been reported elsewhere (at Fast Company and CBS-DFW for instance), and are a common subject at online forums, including Babycenter, Reddit, and the Better Business Bureau. (The most common claim is a shockingly high bill, followed by phone calls with company representatives and insurers to get the bill reduced.)
On a larger scale, Natera settled with the Justice Department for over $11 million (admitting no fault) when accused of overbilling TRICARE, which provides government-sponsored health care for service members and their families; the complaint specifically cites the use of microdeletion panels for low-risk patients. Progenity, which recently got out of the prenatal testing business, paid $49 million in fines while admitting to fraudulent billing and a kickback scheme.
To make sense of NIPS, we need to think of it not just as a test, but as a product. Doing so helps us understand the technology’s real- world effects, not its performance in the abstract, and it helps us see past the bland deceptions of marketing. What does it mean to say a test is “simple,” or “accurate,” or that a pregnant woman can have “total confidence” in test results? Will it truly provide “peace of mind”? To accept these claims at face value is to risk being deceived. To remember that they’re being used to sell a product is to get a little closer to clarity about the real-world effects of the test.
There were just over 1100 comments on the article. I read them all. Many readers approved of the article’s critique of for-profit marketing. Many shared their own nightmare stories of false positives. But the backlash dominated, and it was immediate, intemperate, and selective. It was also mostly focused on semantics. No one that I saw argued that the article got its numbers wrong, or that the women interviewed were made up, or that they did not suffer anguish at false positives. Instead, the critics claimed since the test was only a screening test—a first step before a test with diagnostic certainty, like amniocentesis—it was inaccurate to say that the results were “wrong.”
In fact, the article consistently identified the tests as screening tests. It used “screen,” and its variations, 38 times. Still, several critics objected that this point should have been emphasized more heavily, and right up front. It was claimed, too, that the article was “clickbait” and “fearmongering.”
To anyone familiar with the history of marketing for NIPS, the irony of these objections was exquisite. For years, the companies selling the tests have blurred the line between screening and diagnostic tests, emphasizing the accuracy of NIPS with bold print claims, putting disclaimers in the fine print. They have also done a fair bit of fearmongering of their own, repeatedly highlighting the dangers of diagnostic testing, emphasizing that NIPS carries “no risk to your baby.” You can see the foundation for that marketing approach in the generic, industry-chosen name for the test: noninvasive. By implying that the tests are an alternative to diagnostic testing, the marketing has created more confusion than any single article ever could.
At the same time, the backlash had a problematic, math-heavy caveat emptor frame, a tone of What do you expect? That’s literally how Bayesian statistics works! The problem with this frame is that it shifts responsibility to the consumer. It focuses on the pregnant woman’s responsibility to be statistically literate and to know the difference between screening and diagnostic tests, and not on the companies’ responsibility to be fully truthful or the government’s responsibility to regulate them.
Some were explicit about this, like the commenter who wrote, “The headline should be ‘bad math skills contribute to great unhappiness in .01% of people getting 99.9% accurate test.’” (Whether the test is “99.9% accurate” depends on which numbers you look at, but what struck me was more the tone: Well, little missy, maybe you should’ve learned yourself some Bayesian statistics before you found yourself in this situation.)
Other comments offered variations on the theme. “The entire premise of this article is faulty,” wrote one commenter. “The one salient point is that parents should be aware of the false positive rate.” “The tests are not inaccurate,” wrote another. “The expectations of those being tested are unrealistic.” One stood out to me, in that it praised the company by name, while expressing disapproval for some potential parents:
I can't speak for every company offering NIPTs, but Natera (which is the most popular) seems to be doing its job well IMO. People who aren't bothering to ask a few questions (or do their own research) about testing while pregnant perhaps aren't responsible enough to have a baby. The information is readily out there.
On the commenter’s good experiences with Natera, I take her at her word. At the same time, I’m troubled by the inflection of the word “responsible”: being a responsible shopper is not like being a responsible parent, and equating them suggests a problematic commodification of parenting and childbirth. Further, the comment is revealing in its idealized image of a prospective parent: someone with the book smarts to understand statistics, the leisure time and wherewithal to research her decisions online, the critical thinking skills to resist deceptive marketing, insurance for prenatal care, and enough money to afford expensive add-on tests for unlikely disorders. At best, it assumes a reality that does not exist for many. At worst, it advances a class-based vision of who is “responsible” enough to reproduce—one which is eerily close to the images of well-off women that have dominated ads for NIPS.
Statistics were invoked repeatedly by commenters, but in the situations portrayed by the article, it’s hard to see how much difference statistics would make. It seems to be widely accepted that many doctors could do a better job of explaining the tests. But don’t the companies have a similar obligation to refrain from deceptive marketing? In any case, statistics are not a magic bullet for worry. The prospect of a disabled child is a fearful result for many, and will, almost by definition, lead to a distress immune to math. You can’t recalculate anguish, and statistics don’t disprove experiences.
Humans aren’t computers: anyone who has received a life-changing diagnosis, or has known someone who got one, knows that the emotional impact of the diagnosis tends to undo your ability to run impersonal calculations.3 I found this to be true for my father, an engineer, when he was diagnosed with lung cancer. I found it for myself when my younger daughter was diagnosed (after birth) with Down syndrome. It was incredibly difficult in the short term for many reasons—there were Laura’s immediate health issues, the shock of upended expectations, and the sudden consciousness of my own internalized ableism, for starters—but at the same time, the news unsettled my understanding of probability itself.
I hadn’t thought about the odds much, but I realized that they had always been, on some level, irrelevant. A hypothetical child has a 1 in 500 chance of Down syndrome, or 1 in 100, or 1 in 1000; an actual child either has it or she doesn’t. Laura’s odds were always 1 in 1. Learning that, though, was the beginning of learning that I did not know what the fact meant, and that the diagnosis was not a sentence; that, as I wrote elsewhere, its meaning was up to Laura, and to us. In time I came to see that the meaning of her diagnosis depended, in part, on what other people thought about Down syndrome and disability. That’s one reason to study the marketing for NIPS, and to listen to the way people talk about it.
In the comments that centered statistics, what interested me most was the reason for invoking statistics in the first place—usually, to excuse the industry while accusing someone else (a patient, a doctor, the journalists, the newspaper, Americans ignorant of statistics). More than that, invoking statistics seemed to really be a proxy for the idea that the issues at hand were simple. NIPS was a good screening test to catch bad conditions. Its performance could be calculated. The failures were matters of information: a doctor had failed to explain the test, or a patient had failed to understand it.
But as the genetic counselor Robert Resta argues in a superb blog post, screening tests raise questions beyond the realm of math. Reflecting on the backlash to the New York Times report (“Clearly the article touched a collective genetic counselor nerve—a lot of the reaction has been more reflexive than reflective”), Resta calls attention to the economics of testing:
. . .the demand for genetic counseling would likely drop off significantly if fewer tests were available. How much does this conflict of interest influence genetic counselors’ attitudes toward, and willingness to adopt, genetic tests, and how much does it subtly and subconsciously influence some of the strong reaction to the Times article?
He also voices a perspective almost completely absent from the Times article’s comment section:
What message does it send to people with disabilities, their families, and their advocates if we continually add, seemingly willy-nilly, more and more genetic conditions to the prenatal testing list, especially, as I noted above, if they obtain no tangible benefit from testing?
Read Deletions of Concern: Noninvasive Prenatal Screening and Disability, Part Two
George Estreich is the award-winning author of The Shape of the Eye: A Memoir, a meditation on the differences that matter, the ones that don’t, and what it means to belong; and of Fables and Futures: Biotechnology, Disability, and the Stories We Tell Ourselves, reflections about the rise of innovations in human biotechnologies, from non-invasive prenatal testing to heritable genome editing.
- In its brief history, the technology has acquired a startling number of names: NIPD (D for “diagnosis”), NIPS (S for “screen”), cfDNA (“cell-free DNA”), cffDNA (“cell-free fetal DNA”). I will use NIPS, to emphasize that the test is a screening test. CfDNA is probably the most neutral and accurate name, but at the same time it does not specify prenatal screening.
- As genetic counselor Katie Stoll points out, microdeletion panels are not currently recommended by the relevant professional societies:
"Generally, with screening tests, the rarer the condition, the more likely that a positive result will be a false positive. For these reasons the American College of Obstetricians and Gynecologists (ACOG) and the American College of Medical Genetics and Genomics (ACMG) recommend against including microdeletion screening on prenatal cfDNA screening."
- In the comment section, Danielle, in California, wrote: “I did genetic testing for my first pregnancy. I was 32 at the time. My fetus tested positive for a condition…which meant for my age, it was a greater than 1/190 chance of having this genetic abnormality. My logical side can do that math and feel very confident. But I’m not a robot and the emotions were reeling. It was such a mind and heart ****.”