By Sarah Duranske

In 1978, for the first time, an American woman could take a pregnancy test in the comfort of her own home. Critics claimed that home pregnancy tests only would be used by promiscuous or immoral women who were too ashamed to visit their doctors, but the appeal to women was undeniable. [1] Today, eight out of ten women learn they are pregnant from a home pregnancy test.[2]

In the 1980s, as the AIDS crisis swept the nation, the FDA banned the use of at-home AIDS tests over fears of poor test reliability and insufficient counseling.[3] Advocacy groups bolstered these fears by highlighting the suicide of a man who jumped off the Golden Gate Bridge after testing positive for HIV.[4]  But with improvements in testing technology and treatment options – and stymied by a stubbornly high infection rate – the FDA reversed its stance.  It encouraged home test kit applications in 1990 and approved two home collection kits in 1996.[5]  Within a year, more than 175,000 people purchased kits, and the expanded screening was not associated with any increase in the suicide rate.[6]

In 2013, the FDA shut down 23andMe’s health-related genetic tests due to concerns that users would act, or fail to act, to their detriment based on incorrect test results or unsupported clinical interpretations. In spite of experts’ concerns, 23andMe had genotyped around 400,000 individuals between its 2007 market entrance and the FDA’s 2013 action.[7] Studies consistently demonstrate (for better or worse) that people don’t change their behavior based on genetic tests that report on disease risk.[8] In April 2017, after receiving evidence of accuracy and reliability from 23andMe, the FDA approved ten tests that provide information on users’ genetic factors associated with increased risk for diseases or conditions.

The adoption of these and other home tests for health-related information indicates a robust desire by many consumers to directly access health information outside the confines of the doctor-patient relationship. Consider our use of the internet. Thirty-five percent of U.S. adults have accessed the internet specifically to figure out what medical condition they or someone else might have.[9] In 2015, WebMD reported an average of 210 million unique users a month.[10] Mobile health apps that consumers install on their smartphones or tablets also tap into this demand.  More than 97,000 mobile health apps were available as of March 2013, and the market is growing.[11] Although many consumer abandon health apps after downloading them, the robust market suggests that there is a desire to access health information, even if we don’t actually use it.[12]

Given the public’s thirst for direct access to personal health information, I celebrate the FDA’s recent approval of 23andMe’s genetic health risk assessment system. These tests should be regulated, but lightly, and the FDA struck the right balance.

As a gating issue, regulating direct-to-consumer genetic tests for accuracy is sound policy. Requiring a probabilistic genetic test to demonstrate accuracy based on scientific studies is important because an individual user cannot judge its accuracy based on her own experience. This is because predictive genetic tests report on increased disease risk, but the overall risk may still be so small that she remains unlikely to develop the disease. Also, because other factors affect the likelihood of developing a disease, the user cannot know if the genetic factors identified by the test were significant in her developing (or not developing) a disease. Thus, unlike a user’s ability to judge whether a window cleaner “removes streaks” based on her own experience, the user of a genetic test cannot judge whether it accurately identifies relevant genetic variants and whether the identified variants have the claimed impact on disease risk. Situations like this, where consumers cannot tell whether a product does what it claims to do, can give sellers an incentive to “cheat” by producing inferior quality goods.  Regulations that eliminate this information deficit help consumers by ensuring that the products in the market actually do what they claim to do.

To approve 23andMe’s genetic health risk assessment system, the FDA required evidence of its analytical validity (i.e., that the tests correctly and consistently identified the genetic variants associated with the ten diseases/conditions) and clinically validity (i.e., that the link between the variants and the ten diseases/conditions is supported by peer-reviewed scientific data). These requirements allow users to rely on the accuracy of the tests, and remove any incentive for 23andMe (and future sellers who will be subject to the same requirements) to offer ineffective products.

Second, the FDA’s requirement that 23andMe disclose certain information is an appropriate, and relatively light, regulatory response.  A condition of the FDA’s approval is that 23andMe must provide information to consumers on the uses and limits of the tests, as well as the scientific data that support the accuracy and validity of the tests. In addition, 23andMe must demonstrate that consumers understand the uses and limitations of the test results.  By mandating the disclosure of this information in an easily comprehended form, the FDA addresses the concern that justifies regulating these tests: the user’s inability to judge the test based on her own experience. The information disclosure requirement also decreases the risk that users will rely on test outcomes to act, or fail to act, in ways that damage their health and wellbeing.

Finally, the FDA wisely avoided imposing its (or other experts’) judgment of the value of genetic health risk tests.  Critics often charge that genetic health risk tests like these are a waste of money because they lead to no clinical benefits.  And, at least in some circumstances, they may be right.  As noted above, these tests merely provide a sliver of information relevant to any individual’s likelihood of developing a certain disease or condition.  But there is a demand for direct access to personal health information, and our society tends to rebel against purely paternalistic regulations that substitute an expert’s evaluation of our well-being for our own.

I am a 23andMe customer from pre-2013. I welcome the FDA’s approval of the genetic health risk tests because I am now more confident with the information I receive from the company. Although the test results were not clinically useful to me and did not change my medical decisionmaking, I did enjoy the dinner party conversations they inspired and my increased curiosity about genetics.  As for lack of personal utility, genetic health risk tests fall far behind other unnecessary purchases I’ve made over the years (did I really need that raincoat for my Chihuahua?). I’m glad that 23andMe’s tests are back on the market, and I’m glad that they are regulated.  Now, I’m off to get my microbiome analyzed.

Sarah Duranske is a Grey Fellow at Stanford Law School (and an honorary CLB  fellow, too!)

 

[1] Sarah Abigail Leavitt, A Private Little Revolution: The Home Pregnancy Test in American Culture, 80 Bulletin of the History of Medicine 317, 325 (Summer 2006) (citing Test Yourself for Pregnancy?, Consumer Rep. 644, 645 (November 1978)).

[2] Pagan Kennedy, Could Women Be Trusted With Their Own Pregnancy Tests?, N.Y. Times, , July 31, 2016,.

[3] 54 Fed. Reg. 7171, 7279 (Feb. 17, 1989); Malcolm Gladwell, FDA May Allow AIDS Home Tests; Medical Advances Spur Policy Shift, The Washington Post, Apr. 25, 1990.

[4] Alexi A. Wright, and Ingrid T. Katz, Home Testing for HIV, 345 N. Engl. J. Med. 437 (Feb. 2, 2006).

[5] Malcolm Gladwell, FDA May Allow AIDS Home Tests; Medical Advances Spur Policy Shift, The Washington Post, Apr. 25, 1990; Alexi A. Wright, and Ingrid T. Katz, Home Testing for HIV, 345 N. Engl. J. Med. 437 (Feb. 2, 2006).

[6] Alexi A. Wright, and Ingrid T. Katz, Home Testing for HIV, 345 N. Engl. J. Med. 437 (Feb. 2, 2006)

[7] Diane Brady, 23andMe Wants to Take Its DNA Tests Mass-Market, Bloomberg (September 30, 2013), available at https://www.bloomberg.com/news/articles/2013-09-30/23andme-wants-to-take-its-dna-tests-mass-market.

[8] Gareth J Hollands, et al., The Impact of Communicating Genetic Risks of Disease on Risk-Reducing Health Behaviour: Systematic Review with Meta-Analysis, 352 BMJ (March 15, 2016).

[9] Health Online 2013, Pew Research Ctr. (Jan. 14, 2013), http://www.pewinternet.org/files/old-media//Files/Reports/PIP_HealthOnline.pdf

[10] Maria Armental, WebMD Health Profit Jumps on Ad Revenue, Traffic Growth, Wall St. J., May 6, 2015, http://www.wsj.com/articles/webmd-health-profit-jumps-on-ad-revenue-traffic-growth-1430944191.

[11] Nathan G. Cortez, I. Glenn Cohen, & Aaron S. Kesselheim, FDA Regulation of Mobile Health Technologies, 4 New Eng. J. Med. 371, 372 (July 24, 2014) (citing research2guidance, Mobile Health Market Report 2013-2017, March 4, 2013).

[12] According to a survey of 1604 mobile phone users, 45.7% of those who had downloaded a health app reported that they no longer used it.  Paul Krebs, Health App Use Among US Mobile Phone Owners: A National Survey, 3 JMIR Mhealth and UHealth 2015.