Some Doubts About Unregulated Human Genome Sequencing

This post might put me at odds with much of the science bloggysphere, but I think a lot of the concern over congressional and FDA hearings over whether ‘over-the-counter’ genetic screening and genome sequencing should be regulated is overblown. Maybe my feelings are influenced by my concerns about the misuse of antibiotics–in the case of cefquinome, the FDA did the right thing (and I only wish they had acted sooner).

Consider this article which was recommended by a ScienceBloging, the crux of which is:

The controversy here rests on a single assumption: That the typical Walgreen’s customer, and by proxy the typical American, isn’t prepared to learn about their DNA. Unless a physician guides us, the argument goes, we can’t handle the truth. We’ll quit our jobs, we’ll divorce our spouses, we’ll panic. Without a doctor, we’re told, information can be more than mystifying — it can be downright dangerous.

Now let’s examine the source of this article:

The Huffington Post.

Have you ever checked out its “Wellness section?” This is the same Huffington Post that routinely publishes altie woo in every form imaginable (my favorite is the claim that cancer is a fungus, although the enema fetish is nice too). My point is not to discredit by association those opposed to regulation–I like the Huffington Post‘s political reporting. But let’s return to this:

The controversy here rests on a single assumption: That the typical Walgreen’s customer, and by proxy the typical American, isn’t prepared to learn about their DNA.

I won’t make claims about the typical Walgreen’s customer*, but the typical Huffington Post Wellness section reader is a fucking moron. Having worked in public health, I see serious potential for bad outcomes; public health workers, though hard-earned experience, have learned that even seemingly simple things can go very wrong.

First, many well-intentioned people are ignorant about health (and biology in general). They’re not stupid, but they are misinformed. Remember all the parents who didn’t want their kids to get a flu shot? Talk to any MD, you’ll hear some real howlers (provided you have a dark sense of humor). Second, people have a tendency to selectively hear things. Some patients have a tendency to interpret advice in a way that reinforces bad behavior. Keep in mind that much of the genetic testing–and most of the genome-wide scans–have, to date, disproportionately targeted well-educated, often scientifically literate people. As we move to the broader population, well….

Enemas. Nuff said.

Third, doctors need to learn about this stuff. I don’t mean to disparage MDs, but, in my job, I collaborate with a lot of MDs, and these are MDs who are researchers interested in genomics. If they’re the leading edge (and they are smart people), then most doctors are going to need a lot of training and supplemental materials to be able to explain this to their patients and to use this information appropriately (again, this isn’t a slam against MDs–how they reliably know so much different stuff amazes me, especially when getting it wrong isn’t an academic exercise). But most people aren’t population geneticists/genomicists. Fourth, while I think blaming some things on irrationality is stupid, people do have a hard time assessing risk when it’s expressed in mathematical terms; sadly, not everyone is a statistician in the rough.

Finally, we regulate all sorts of diagnostics (and we should)–why should this be any different?
The point is that I can see some downsides without regulation, and it doesn’t hurt to examine if and what regulation would improve this technology for patients and doctors.

*I regularly shop at CVS, and I doubt, on the whole, we’re any smarter than Walgreen’s customers.

This entry was posted in Genomics, Public Health. Bookmark the permalink.

6 Responses to Some Doubts About Unregulated Human Genome Sequencing

  1. Mary says:

    Hey Mike, I’m with you on this one. I saw that tweet and read the article over the weekend, and had exactly the same reservations.
    And then I also thought to myself: well, warnings on cigarette packs don’t deter smokers…so…meh. Will it really help? But I feel that at least it is worth some effort.
    I think there are good points about control/access to the information. But it doesn’t preclude a little stewardship by the authorities.
    Some of the personal genomics cheerleading worries me. I’m all for it–I don’t want us to stop, but it does carry some risk of misleading people. Just like the HGP discussions did.

  2. Elf Eye says:

    I would want regulation even if only in terms of quality control. How careful are these genome sequencers going to be? Wouldn’t FDA scrutiny be necessary if only to keep DNA sequencing scammers from springing up willy nilly? Can we rely on for-profit DNA sequencers to police themselves? Other industries haven’t succeeded in doing so; why would we expect DNA sequencers to do so?

  3. qetzal says:

    Finally, we regulate all sorts of diagnostics (and we should)–why should this be any different?

    Exactly. Medical tests are required by federal law to be FDA-approved. 23andME were apparently advertising their tests as being useful for determining how you’ll react to drugs like warfarin, or what is your risk of certain diseases. As I understand it, those are clearly medical tests according to the legal definition. (Disclaimer: IANAL.)
    The point of regulating medical tests is to ensure that the test really does what it claims to do, that the results are clinically valid and useful, that it’s reliable, etc. I understand that some people will disagree with the overall premise that medical tests should be regulated. But I’ve yet to see any rational argument why 23andME et al. shouldn’t be held to the same standards as Myriad or Genomic Health.

  4. hibob says:

    But, really, how much better are the MDs? As a group, studies have shown they don’t have a grasp of conditional probability as applied to screening tests, and end up wildly misinterpreting them. For tests that generate large numbers of false positives/negatives in the population groups tested, conditional probability is absolutely necessary when it comes to calculating the actual probability of a patient having a condition. (see below. In short, when given a set of statistical assumptions, doctors in the US interpreted a positive mammogram as indicating a 95% probability of breast cancer when the true number was only 9%).
    It seems to me either MDs should be required to be certified in the interpretation of screening tests, as they are with other techniques, and forced to subscribe to a regularly updated spreadsheet that allows them to do actual risk analysis on different subpopulations when necessary (as opposed to just a SWAG), or the relevant patient history should be given to the testing company along with the sample and the company should supply an interpretation along with the raw results (much more sensible). Otherwise the stupidity is just being moved from Huffpo to the doctor’s office.
    I still think labs should be able to sell raw DNA SNP/sequence data to private customers. Scoring and interpreting actual medical tests is what should classify them as needing certification and regulation.

    In one study, Gigerenzer and his colleagues asked doctors in Germany and the United States to estimate the probability that a woman with a positive mammogram actually has breast cancer, even though she’s in a low-risk group: 40 to 50 years old, with no symptoms or family history of breast cancer. To make the question specific, the doctors were told to assume the following statistics …:
    The probability that one of these women has breast cancer is 0.8 percent. If a woman has breast cancer, the probability is 90 percent that she will have a positive mammogram. If a woman does not have breast cancer, the probability is 7 percent that she will still have a positive mammogram. Imagine a woman who has a positive mammogram. What is the probability that she actually has breast cancer?
    … As for the American doctors, 95 out of 100 estimated the woman’s probability of having breast cancer to be somewhere around 75 percent.
    The right answer is 9 percent.

  5. qetzal says:

    I still think labs should be able to sell raw DNA SNP/sequence data to private customers. Scoring and interpreting actual medical tests is what should classify them as needing certification and regulation.

    I agree. As best I can tell, FDA isn’t trying to stop the personal genomics companies (PGCs) from doing the former. Nor is it concerned if PGCs provide non-medical interpretations of the data. (E.g. your ancestors are from the Mediterranean; you’re heterozygous for this eye color allele so your kids could have blue eyes.) I think FDA only got involved because the PGCs chose to make medical interpretations. (E.g. you may have this reaction to warfarin; you have an increased risk of that disease.)
    You also make a great point that accurate and reliable medical tests aren’t very useful if your MD doesn’t know how to interpret them properly. In Gigerenzer’s study, it’s not clear that he specifically asked docs who normally have to make such interpretations. I’d like to hope that oncologists would do much better than those unspecified “American doctors” but I fear I might be disappointed.

  6. Dan Vorhaus says:

    @qetzal (and others): It is not in fact the case that the FDA regulates all medical/diagnostic tests and that DTC tests have been slipping in the back door. In fact, the majority of genetic tests – both clinical and DTC – are what are known as laboratory developed tests (LDTs) and are not currently regulated by the FDA (although they are regulated by CMS under CLIA to ensure accuracy).
    That may all be changing next month (see:, with the FDA broadening its regulatory gaze beyond DTC genetic tests to capture all LDTs. The concern is the extent to which new regulations will burden genetic test providers – again, of both the clinical and DTC variety – and whether and how effectively those regulations will adjust that burden based on the actual use of the tests (e.g., informational vs. clinical).
    Finally, I appreciate the point that information, particularly genetic information, can be easy to misinterpret. That seems to be the primary focus of this article, and there is a broad consensus that this is a problem right now for both recreational and clinical genetics. But that is a problem that FDA regulation does not solve, at least not directly. And, in fact, there is a strong case to be made that many of the more reputable DTC companies, by encouraging consumers to engage with their own DNA, are doing more to further genetics education than our medical schools or government agencies.
    There’s no doubt that in an industry as new and complicated as DTC genetic testing we need to be paying close attention. The question is whether adopting traditional pathways of FDA regulation is the proper approach.
    Dan Vorhaus
    Editor, Genomics Law Report

Comments are closed.