#MyDataBelongsToMe, not @23andMe.

In praise of integrating digital health data to build my “Datanome”

My Datanome, My API

Starting this month, 23andMe, the famous personal genomics company, will stop allowing me to share my genetic data, or rather, its digital summary, with third party app developers through an open API (short for Application Program Interfaces). A longtime advocate of open health data and patient empowerment, I relished the possibility of importing my genome, or any digital health data, from one platform to another. It gave me a sense of ownership. It gave me a sense of playing an active role in helping Precision Medicine come about. To move away from average population treatment, we need better data and we need it to flow freely. To personalize treatments, we need to compile it all: genetic data, medical records and wearable sensors. But 23andMe has just made my “datanome” costlier. It became too valuable for me to own… In many ways, the evolution of the company epitomizes what needs to be fixed today in digital health.

My Data, My Rights

In this example, it feels like my basic rights have been denied. Lawmakers would probably agree. The General Data Protection Regulation (GDPR) introduced earlier in 2018 by the European Union creates new rights associated to our digital data. To avoid abusive data collection, companies must obtain clear and explicit consent from users. In this case, I never gave mine. It introduces a rule of data minimization, forcing platforms to limit collection to what is strictly necessary to the service, nothing more. It entails a “right to be forgotten”, meaning I can ask for my data to be deleted. It is making the “Blue Button” compulsory. This refers to the possibility to download and transfer data across platforms, something the health systems for Veterans (VA) first introduced in the US and which Medicaid and Medicare are extending. This is essential for patients to gain control over their data. It is ironic that 23andMe would close its platform just when a whole continent starts pushing for openness. Even Electronic Health Records in the US are now promoting interoperability…

GDPR Fundamental Rights

23andMe best take notice. Violation can cost a company up to 4% of its annual turnover. And US companies are concerned, because the rules apply as soon as they analyze data from European citizens…

Spitting in the Tube

And thanks for all your data !

My disappointment has been slowly building up. An early believer in the promise of personal genomics, I remember receiving my $199 spit-tube DNA test some 5 years ago, sent from California to Paris, just before 23andMe was banned by the FDA from publishing any information on genetic predispositions. Luckily, I was so early an adopter that I received all my data before the ban. No scoop there: I learned I had 2% Neandertal ancestry, like everyone else really. Against all evidence, I was told I had blue eyes… Like 60% of Westerners, I was told I was likely to become obese or diabetic. I learned an odd is no certainty. The best part of 23andme was meeting a distant relative, who turned out to be a prominent Tech investor, and now we are talking business… I had faith in 23andMe’s pioneering role.

The Dog with Neanderthal Origins

It turns out the FDA had good reasons for banning disease data late 2013. For one, people could not tell the difference between a real threat and the likeliness of a threat. Everyone remembers Angelina Jolie, getting a double mastectomy, when she discovered she had the morbid BRCA2 mutation. She was a symbol for patients taking their health in their own hands. But many physicians consider her a terrible example, because all mutations don’t lead to cancer. The standard protocol is to step up surveillance and go through the surgery only if the odds play out wrong. True or not, I recently heard from an insider that the FDA had other reasons to ban the tests. One agent had sent dog spit to the Californian lab to see what results would bark out. Just like me, the dog had Neanderthal origins…

My Data, My Asset

Like many, I started feeling betrayed when 23andMe doubled down selling my data to life sciences. Don’t get me wrong. I find helping research and life sciences fantastic because it can speed up discovery of new treatments. But learning this from the press, when I should have been asked to give my consent, that’s an ethical problem. If I am the product, should I really have paid $199? In one pharma deal on Parkinson disease, the value of data reached $20,000 per patient… What’s wrong is not letting patients take part in the sale of their own data. What’s wrong is not getting consent when genetic information can hardly be anonymized.

My Data, My Blockchain

What’s to be done? The GDPR won’t save me. Once my data is sold, it’s too late. Citizens can’t fully trust regulators because tech companies are too fast, often innovating in areas where rules are still to be written. We must turn to better technologies, which foster patient control by design. It’s no accident that Blockchain is emerging at a moment where digital mistrust is peaking. Its very properties can give power back to patients.

What if a different kind network could give power to the patients?

What if you could set the terms when you share your data? What you could be compensated for your contribution to research?

When a ledger is distributed across all participants of a network in real-time, each patient can set the terms for data sharing, cutting out middle men, like central genomic banks, Facebook etc. Smart contracts make it possible to link data sharing to tokens, creating fair trade in place of previously opaque and costlier transactions.

Time for Next Generation Real World Evidence

Why does any of this matter? Because securing patient buy-in is key to accelerating the release of new treatments. The 21st Century Cures Act (2016) gave the FDA a mandate to accelerate approvals for new treatments that incorporate the perspectives of the patients. This has led pharma to integrate “real world evidence” (RWE) into clinical studies. Like 23andMe illustrates, the problem is that current methods of collecting RWE use costly intermediaries that capture all the value. Patients are not aware and not compensated. It’s time for them to set the terms of the sale of their data. Active consent will lead to higher resolution individual patient level data, better studies, more personalized drugs. For life sciences, this is a chance to create a more equitable relationship with patients, giving them stronger guarantees as they buy-in the research they ultimately benefit from. This will ultimately speed up the delivery of new treatments.

Beyond the technicalities of the 23andMe API, I stumbled on a much bigger, deeper debate. Would you give the keys to your house to a stranger that wants to sell your furniture? I did worse. I gave my genetic code to a company that does not think it is accountable. We need to change that. We need to own the keys to our health vault.

Alexis Normand

Head of Embleema Health Consortium

Join the movement! #MyDataBelongsToMe