23andMe flap at FDA indicates fundamental dilemma in health reform

We must go beyond hype for incentives to provide data to researchers

The FDA order stopping 23andM3 from offering its genetic test kit strikes right into the heart of the major issue in health care reform: the tension between individual care and collective benefit. Health is not an individual matter. As I will show, we need each other. And beyond narrow regulatory questions, the 23andMe issue opens up the whole goal of information sharing and the funding of health care reform.

I’m attempting in this article the opposite of what the FDA, 23andME, and other participants are doing in this case. They’re intent on reducing the issues to narrow regulatory questions such as what constitutes a “device,” what constitutes a “medical claim,” and whether all health claims must be based on randomized clinical trials. If you’re interested in those questions, I invite you to read other articles. I’m seeking big answers to big questions.

The problem is not that 23andMe is a bad product. It seems to work quite well, which is impressive for a service that has reduced the test for selected genes to a $99 kit. The problem is that average folks don’t know how to interpret results. As journalist Christina Farr documents, a few people have gone too far and taken drastic (or at least expensive) actions that have little value. I’ll have more to say about this article in a moment.

Ignorance and recourses

The problem with 23andMe has been known among health professionals for some time. Suppose you find that you have 100 times as much chance as the average person to succumb to a certain form of cancer. If this causes you to take modest precautions that are good for you anyway, such as eating more leafy greens, that’s fine. But it’s not hard to freak out.

One has to establish balance by understanding basic statistics, which few people do. Suppose the chance of getting that form of cancer, on average, is one in a million. If your chance is 100 times as great as that, is worth taking harmful medical actions or limiting your life to cut the risk?

Farr’s suggestion that 23andMe “more readily link consumers with genetic counselors” is reasonable, given consumer ignorance. We could all use more guidance in general. Doctors should spend more time with patients (as should financial advisors and other people who understand the complexity of modern life) or, if they’re too busy, should designate team members to spend time with patients, perhaps in groups.

I’m not surprised that Farr reports, “the majority of geneticists and medical professionals I’ve spoke with have sided with the Food and Drug Administration.” These are the professionals that want to take our information and operate on us. It’s hard for them to trust patients and consumers with their own information.

I’m seeking a middle ground, which is why I haven’t signed the petition trying to let 23andM3 off the hook. With statements on its web site like:

Find out if your children are at risk for inherited conditions

Understand your genetic health risks. Change what you can, manage what you can’t.

one can understand why the FDA is concerned. Is 23andMe promising too much? And if so, can they afford to ratchet back their promises?

Individual and collective health knowledge

So we see that 23andMe doesn’t offer (for most people) information that can lead to major decisions and improvements in health. The value of 23andMe lies in the data it collects and uses on a collective basis. Like another popular health site, PatientsLikeMe, 23andMe can use what it learns about you, and me, and others to find new solutions to health problems. Data crunching can turn up correlations that randomized clinical trials cannot, or that wouldn’t otherwise be checked because those trials are so expensive.

But who will plunk down $99 in order to submit data that may pay off for someone else 15 years from now? 23andMe’s business model depends on promising results right away, to each individual who joins.

They’re not alone in this dilemma: it is faced by every company developing a popular health app for consumers, and is often handled poorly, as I point out in another article.

The data you rack up on your sleep or exercise patterns have little value unless you have a norm to compare it to. So even individual QSers need collective data. And contributing your data to a repository where researchers can crunch it will accomplish more than anything you can do on your own. And we will do this, if we are assured of reasonable anonymization. (Here I’ll pause to plug the book Anonymizing Health Data, which will help researchers do the right things.)

Beside 23andM3 and apps developers, who else is hamstrung by the gap between individual and collective benefit? Basically every heath institution in the US. Most see it in their business interests to hoard patient data, and make it hard even for other doctors treating the patients to get the data. This undermines both good patient care and valuable medical research.

So what the 23andMe dilemma shows us is that we need new incentives for collecting and sharing data. This will be the basis of the “learning health care system” so many reformers talk about.

23andMe will someday be obsolete, because we’ll be able to do full-genome sequencing at a cost that opens it to everyone. We probably won’t do it automatically on every baby brought to its first pediatric visit, for privacy reasons. We should wait until a child develops a serious condition that requires this step, or reaches the age of maturity where she can make an informed decision about whether to uncover her genome’s secrets. But we already have large genetic databases, such as The Cancer Genome Atlas, and they’ll get bigger.

Privacy issues are definitely in play here. One of the reasons for educating people about their health care is to help them make reasonable decisions about sharing. The incentives should also not be hard to find: we can lower health care costs and save lives through our collective data. The question is who will take the first step to make the collective responsibility for data into an individual choice. That’s the “individual mandate” I’d like to see.

Update, December 11: 23andMe is not actually likely to go away if people use other services to sequence their genomes. 23andMe believes their big added value is the accumulating and filtering of research, and presenting relevant results to users.

Strata SC 2014 Image

O’Reilly Strata Conference — Strata brings together the leading minds in data science and big data — decision makers and practitioners driving the future of their businesses and technologies. Get the skills, tools, and strategies you need to make data work.

Strata in Santa Clara: February 11-13 | Santa Clara, CA

tags: , , , , , , , ,