Matthew Katz on 23andMe: “Return to Sender, Genome Unknown: Seven Reasons I Will Return My Personal Genome Kit”

Matthew Katz, MDRadiation oncologist Matthew Katz is a lifetime member of SPM who blogs regularly for the American Society of Clinical Oncology and Mayo Clinic Social Media Health Network. In this guest post he offers his view of 23andMe, the personal genomics service that’s in the news because of the FDA’s recent ruling.

That ruling has generated enormous controversy, on the SPM member listserv as well as in the general media. Matt’s not on the member listserv, so he didn’t know that, but he offers these views … which, as with all guest posts, are his own, not SPM’s. What do you think?

Given recent controversy about the U.S. Food and Drug Administration’s letter to the personal genome services company 23andMe, I just had to learn more about what was truly involved. After buying a kit on Thanksgiving, I’ve gone from being a skeptic to wanting a full refund for seven reasons:

1. Lack of transparency. The company apparently wants to reach out to the health care market for revenue, but seems to claim it is only an “ancestry” company to avoid regulatory scrutiny. Dr. Michael Eisen, on the scientific advisory board, writes: “Genetic tests are simply not … medical devices in any meaningful sense of the word. They are far closer to family history than to an accurate diagnostic.” So why the slogan “Health and ancestry start here”?  Targeting patients through direct-to-consumer marketing might boost sales, but will also guarantee FDA involvement. Why would I buy a personal genome product from a company that refuses to acknowledge it is selling a medical test?

2. Lack of adequate testing and subsequent data. The FDA issued a “cease and desist” letter explaining that because “more than 5 years after [23andMe] began marketing, [23andMe] still had not completed some of the studies and had not even started other studies necessary to support a marketing submission.” In addition, 23and me doesn’t do the testing itself, but outsources it to LabCorp. Unfortunately, that company has made at least one publicized mistake with 96 samples. Why should I pay for a test that I don’t know works? 

3. Privacy risks. 23andMe frames testing as self-empowerment: “Knowledge is power.”  But its privacy policy makes it clear that de-identified information may be shared with third parties. It begins by sending your DNA to a different company, LabCorp, before transferring it to 23andMe. Recent research shows that 45% of healthcare organizations had more than five security breaches in a two year period.  It could happen with 23andMe despite its best efforts. Another privacy concern is that genetic information may be shared with the NSA or legal enforcement. Why would I take a test whose results could be used against me for insurance, employment, or in legal matters?

4. Am I a research subject or not?  23andMe highlights the opportunity to participate in “a new way of doing research.”  As a cancer specialist, I’m all for new research, but what is involved with giving the company my genetic information? To buy the kit, I didn’t have to read the terms of service (TOS) or review the risks and benefits of genomic testing. I didn’t see a consent form to review. Why should I pay for testing without being certain if I was consenting to unknown research on my genome?

5. Possible research on vulnerable populations. Children are considered a vulnerable population in the Belmont Report, which guides ethical clinical research in the US. I previously raised concerns about the ethics of social networks data mining health information, thinking mainly about cognitively intact adults.

Apparently 23andMe permits testing on minors. Buried in the Customer Care section, the company is advising how to swab children. 23andMe isn’t the only company doing this, and there are medical reasons to consider genetic testing, but offering a discount on extras kits for family members seems a bit off. Clearly it’s in the financial interest of the company to get the whole family tested, but is it in mine?  Why would I risk making my children research subjects when they are otherwise healthy?

5. Problematic content organization. The website design is frustrating. Not only did the TOS not show up as required reading before purchase, I had to dig through the site map to find “Before You Buy.” The cool video on the process doesn’t mention that they outsource the testing — it’s “our lab.” That they really use LabCorp is hidden.

6.  Lack of balance in presentation. Technically the company can say it thoroughly lists possible risks. In medicine, it’s an obligation to educate patients with a balanced view for informed decision-making. 23andMe doesn’t do that, as the examples above demonstrate. Why should I give my money to a company that doesn’t give a balanced view from the outset?

7. Eugenics. A company that collects genomic data and then files for a patent for designer babies, raises additional concerns for me. Even if 23andMe doesn’t use it now, I do not want my genome to become part of a eugenics debate. And if the company practices what it preaches, it should be equally as transparent in its practices as it encourages its customers to be. Why would I risk that my genomic information might be used for a different kind of genetic discrimination?

I am not an expert on the FDA’s regulatory process, and it has its much-deserved share of critics, but many of my concerns are reflected in its letter to 23andMe. The purpose isn’t to stop personal genomics, but to ensure testing doesn’t hurt people.. This must begin with a valid test supported by clear and easily understandable marketing materials developed to reassure customers that it’s a trustworthy steward for genetic information.

As a citizen, I expect corporate transparency for any new health product. As a patient, I think the risks of taking the test outweigh the benefits for my health. As a doctor, I have my concerns for people with distress or misinformation from results of an unproven genomic test. As a human being, I worry about misuse and unintended social consequences of our genetic heritage.

With time, the pro-autonomy advocates and more conservative skeptics will find middle ground. But for now, I’ll send my kit back unopened.

Those are Dr. Katz’s concerns, obviously including a lot of participatory / empowered thinking. Many of our members feel strongly similar or opposite. What do you think?


Posted in: policy issues | research issues | shared decision making | trends & principles




10 Responses to “Matthew Katz on 23andMe: “Return to Sender, Genome Unknown: Seven Reasons I Will Return My Personal Genome Kit””

  1. I had “issues” with 23andMe from the get-go and have been spending this past week saying so on every article I’ve read!

    My primary issue was one of ethics after hearing Brin present at a epatient conference years ago. Embedded in that was a concern about how data was being mined and used. Now it turns out the data might not even be all that valid due to possible “chain of custody” issues relative to the test.

    Yes, I believe we have a right to our “damn data,” but when that data is suspect and/or misused? What then?

    • Matthew Katz says:

      Thanks for your feedback Meredith. Done in the right way, the positive potential of crowdsourced science is exciting. But good science has an ethical component and requires rigor. I haven’t used the kit, so I may be missing some good features to the 23andme program. I’m just not willing to risk the small possibility this company will permit my genetic information to be used against me or my family.

      As for getting our health data…first and foremost it should accurate and relevant. Lousy tests are noise, not signal. It’s one thing to want our “damn data” that is already available in the EMR. It’s another to use a test with unclear utility that may compromise my ability to get insurance or lead to unnecessary further anxiety and testing.

  2. As a physician, I value open-access research, understand the implications, and am a participant in the Human Genome Project.
    The points Michael makes are all valid, but I believe miss the real reason 23andMe must raise gigantic red flags of caution. The FDA, in its “cease and desist” order missed it, too. The concern is that one company now has access to a huge and exploding database — one that effectively has created a monopoly. Who can compete with its pricing and aggressive marketing?
    In that data lies information to which the subjects have given up their rights to ever remove. That data is the real thing of value to the company, much as Google has amassed with your on-line presence. 23andMe admits it is willing to sell aggregate data from its “research” to any willing payor including government, employers, insurers and pharmaceutical firms who will pay handsomely for your genetic information.
    And when, not if, security is breached, Pandora’s box will seem as a fairy tale in comparison.

  3. Great post, I really like how the author has distilled out the key questions we should be asking ourselves.

    I wonder if eventually there will be a way for users to get their data back somehow (of course, it will be “out there” no matter what, but might discourage use somewhat)

  4. Matthew Katz says:

    Thank you for your comments, Theodore and Leslie. My understanding is that some of your data are portable. But I won’t get far enough to find out.

    There is now an eighth reason I will return my kit (which just arrived yesterday): 23andme has stopped any health use for the kit effective 11/22 — so I’m out of luck even if I wanted to.

  5. Jeri Burtchell says:

    I only found out about 23andme a month or so ago and decided to take the dive and ordered my kit — on Nov. 18th. Had I read this post, or even just the FDA letter, prior to ordering, I’d have taken a pass and given this company a wide berth.

    Spitting into the wind.

  6. Sounds to me like your mind was made up before you even purchased the kit. Maybe you even purchased it with the intent of returning it to give your complaints more validity.

    1. Sounds like you’re cutting off your nose to spite your face because you don’t like the words they chose. The information provided by your genome is still useful.

    2. Genome sequencing is really not a medical device nor is it a medical test, therefore the FDA shouldn’t be regulating it. I don’t think any agency should be regulating it, but if it has to be regulated then it probably belongs more under the FTC. All labs make mistakes. One time a lab told me that I tested positive for Rubella which prompted a whole host of calls from the lab, the doctor and the CDC, including more follow up phone calls from the CDC 6 months later to make sure I really was still healthy. If you read about that incident you also read that they have put additional fail safes in place to ensure that kind of mistake doesn’t happen again.

    3. Keyword is “de-identified.” And who says you have to use your real name anyway? Pay cash for a visa gift card, set up a fake email address, use a fake name, have it shipped to work or a friends house, and, if you’re really paranoid,and order it from a computer at the library so they can’t track your IP address.

    4. You get the ability to choose whether you want your de-identified data included in studies or not after you register the product. It is my understanding that they won’t even analyze your sample unless you’ve answered those questions during the registration process.

    5. How about because you value the information the test could give you? But guess what, you don’t have to test your kid! You make it sound like 23andme wants children to make their own decision about whether or not to get tested. Call me old fashioned but I think parents should have control over what happens to their kids. If the parents want their kids tested that’s up to them. If the parents are the ones ordering the test, then the child is not “vulnerable.” Kids are subjects in medical studies all the time. At least this one doesn’t involve potentially dangerous drugs.

    5 (part 2). Again, it sounds like you had decided to return the kit before you even purchased it. Who cares where if its their lab or someone else’s? I actually like the fact that they call it “our lab.” Why? Because using the word “our” means 23andme takes responsibility and credit for all the good and bad things that might happen to my sample.

    6. This argument doesn’t make sense. Results from genome sequencing are the most unbiased results you’re going to ever get. Ever. Every doctor who specializes in any branch of medicine is always biased towards their specialty’s way of fixing a patient’s problem. The genome is just data. No opinions. Just facts. That’s as balanced a view as you’re going to get. Combine that data with peer-reviewed double-blind studies published in respected medical journals and you’ve got some very reliable statistics about your likelihood to have or be diagnosed with a disease.

    7. I think your fears in this one are misdirected. The people from the human genome project are who you should really take issue with. Everything needed to make an aesthetically pleasing designer baby was discovered by them. As far as using Pre-implantation Genetic Diagnosis to make a designer healthy baby, I would think as a doctor you would be all for that. Why make a kid suffer with cystic fibrosis or other genetic conditions if he doesn’t have to? There are thousands of couples who have DNA that, when paired, makes having biological children risky for the kid’s health. Considering your concern for this “vulnerable” demographic in #5, I’d think you’d be thrilled at the opportunity to reduce suffering.

    Believe what you want. You seem inclined to expect the worst from 23andme. With every great invention/discovery there will be abuse by someone. Whether its the Internet abused by the porn industry or the discovery of flight abused by terrorists on 9-11 or anything else you can think of, that doesn’t mean we should try to stop progress or boycott it. We can’t live in fear of what someone might do someday.

    Take one for the team. Get your genome sequenced when they get this FDA idiocy worked out.

  7. Ps. If by portable you mean can you get a copy of your entire genome? Yes. Once you’re logged into your account you can download all 900k rows of snps into a simple .txt file.They call it your Raw Data. There are other programs like and and many others that will take your raw data file and give you even more information than 23andme does. For example, will give you your MTHFR mutations.

    Re Theodore’s comment, Although 23andme is willing to share data, it is de-identified and aggregate data, meaning they won’t trace it back to you. I don’t understand the fear associated with this. Then again, I don’t think like a “bad guy” so maybe my imagination just isn’t thinking outside the box enough to imagine the dangers of this. But aggregate data especially seems very innocent to me.

  8. LF Safer, MD says:

    In the kindest sense, involving a member of the non-cognoscenti in strategic decision-making is a reverse paradigm.
    In reality, it is nonsense!
    This is nothing more than pandering.

Leave a Reply