Alphabet’s (Google’s) life sciences division Verily has launched its public pitch for a massive, multi-year health study it’s leading along with Duke University School of Medicine, Stanford Medicine, and Google proper.
Verily is hoping to recruit some 10,000 Americans to volunteer to share their medical records and have blood and other bodily fluids extracted and linked to a Google account in order for the company and its partners to try to “spur the next generation of medical discoveries”, as it couches it. In exchange, participants will get some of their own health data shared back with them (though it cautions not to expect to get “medical care or advice”), plus a small amount of financial compensation for time lost, and a little early intel on what the ad giant might be learning off of their sensitive health data.
The Project Baseline study has been a long time in the works — it was previously slated to launch in 2015 but has evidently taken rather longer to set up; unsurprisingly so, given the scope and size of the longitudinal study. The four-year study will involve volunteers making annual visits to one of the Baseline study sites for a full one to two days of health tests, including giving blood, saliva and other samples; doing specialized tests such as chest X-rays and echocardiogram; and other tests such as assessing physical strength and answering health-related questionnaires.
Some participants will also be asked to visit a study site quarterly for one to two hours to “gather more frequent information about health profiles that we are especially interested in”, says Verily, and some may also be asked in for appointments at other times “following a significant life event, so that we can see if and how your health changes”.
All volunteers will be required to wear an “investigational wristwatch” daily — aka the Verily study watch announced earlier this month; and sleep with a sensor underneath their mattress (so you can add sex-tracking to this study’s scope — a woman on the Project Baseline question phone line said she’d been asked about the movement tracking mattress coil “a lot!”); as well as have a dedicated wi-fi hub device installed in their home to suck up and send the tracking data from the devices back to Verily’s servers.
There’s also trimonthly, half-hour-long online surveys to take, with questions about diet, exercise and well-being; and a mobile app that will push additional questions at participants, perhaps as frequently as daily, such as asking them about their sleep quality or alcohol consumption. Presumably the project researchers want the ability to be able to react to specific tracked activity/events with follow up questions that might shed light on linked factors.
Participants will be compensated $410 per annual site visit; $30 per visit for the shorter quarterly assessments; and $10 a time for the trimonthly questionnaires. So no one except Verily and its various academic and commercial partners stands to get rich from being involved in this lengthy medical research project.
Lastly, but by no means least, study participants will be required to share access to their medical records with Verily — so anyone signing up for this study really will be standing naked in front of Mountain View.
On the website where the company is pitching for volunteers, Verily’s marketing is heavy on trying to stir up stirring historic parallels for this “mission” to — as it puts it — “better understand health and prevent disease”, laying it on thickly that participants will be ‘doing good’ and ‘helping humanity’ by contributing their health data to the research effort.
To the brink, frankly, of emotional blackmail. A glossy marketing video showing a series of people smiling into the camera intones: “what if you could impact the health of millions of people, just by sharing your personal health story?” — with the unspoken implication being: how dare you be so selfish NOT to share your medical records with Google.
The website is rather thinner on detail about what will actually be done with all the sensitive personally identifiable health data that will be obtained from study participants. And there’s also very little about the underlying commercial motives driving the effort to gather health data in “incredible depth and detail”, as Verily’s marketing paints it.
For example an FAQ on the website ostensibly answering what the data will actually be used for is decidedly non-specific — saying only:
We will use it to expand the Baseline database and develop advanced tools for collecting, organizing, and analyzing health information. As well, in partnership with qualified researchers and organizations, we will use the data to uncover new medical insights or develop new health products.
It’s also not clear who else the data might be shared with and for what specific purposes. The FAQ notes that members of the Baseline team will have access to “directly identifying information (your name, street address, phone number, and email)”, but “external researchers and organizations for research” will also be given access to the data — albeit with the directly identifying info removed. Although, on that front, it’s worth pointing out that the re-identification of anonymous individuals attached to health data has been demonstrated by researchers to be disturbingly easy. So the possibility that such granular medical data might be reattached to study participants’ identities in the hands of unknown third parties cannot be ruled out.
On its blog, Verily specifies that it will be working with “partners from academia, medicine, science, patient-advocacy, engineering and design”, adding: “In the future, the intent is to make de-identified data from the Project Baseline study available to qualified researchers to spur new ideas across the broad ecosystem.”
And even if you were to create a new Google ID just to use for Project Baseline it’s rather harder to change your real world identity (i.e. your real name) and thus de-link all of the online information Google might have (or be able to) glean about you based on that rather less mutable identifier. (Something it unintentionally flags up — given it can’t resist linking Baseline’s aim with other Alphabet divisions’ prior efforts; a slogan for the project claims: “We’ve mapped the world. Now let’s map human health.”)
Asked whether Google might share information it has on Internet users who are also participating in the study, the woman on the Project Baseline phone line wasn’t sure. “I don’t believe so,” she said, noting that participants selected to be a part of the study “can ask any questions that I can’t answer at the site visit”. “You can take your head out of the ring at any time,” she added. “You can take your application or your enrollment and withdraw it at any time if anything doesn’t sound good.”
It goes without saying that Verily, a division of Google’s parent company Alphabet, is a for-profit enterprise so is obviously looking for ways to profit from the health data that study participants will be handing it. Yet despite this rather lopsided exchange — your blood, your moods and your medical records in exchange for a vague notion of some possible future health benefits (for someone) — it is not committing to hand over all the data it gathers during the study to individual study participants, saying only that it “may” return information such as laboratory tests and clinical assessments, and that: “We think it’s important to return as much of your information as possible in an ethical, responsible manner and in a format that is interesting and understandable.” So, in other words, volunteers for this four-year “quest” won’t be privy to all the data Alphabet’s divisions will be extracting from their person.
We’ve reached out to Verily with various questions and will update this post with any response. In a blog post about Project Baseline, the company writes that it is focused on “creating new tools to collect and organize information in ways previously not possible so that we can make the information useful” — describing the initiative overall as a “unified effort to map human health”.
“[T]he Project Baseline study dataset will include clinical, molecular, imaging, self-reported, behavioral, environmental, sensor and other health-related measurements. To organize this information, we are creating an infrastructure that can process multi-dimensional health data – much of which have never been combined for an individual. Our vision is that this data platform can serve as a single query source and may be used for more seamless data integration and collaboration,” it adds.
Alphabet/Google has faced controversy in the UK where its AI division, DeepMind unveiled a big health push last year, partnering with a publicly funded National Health Service Trust to get access to patient health data in exchange for building an app. However at the launch of the first partnership it was not made public how many medical records were being shared with DeepMind to power an app for hosting an NHS algorithm designed to spot the early signs of a kidney condition. An FOI request subsequently revealed DeepMind had been given access to some 1.6 million patients’ health data — without their knowledge or consent. The information-sharing arrangement remains under investigation by UK data protection watchdogs.
But while Alphabet has been able to quickly suck up vast quantities of medical data in the UK to power its ambitions for AI-enabled preventative healthcare because a cash-strapped NHS is all too eager to accept the offer of free help from a high profile, high tech outsider, the US healthcare marketplace evidently requires a different approach to outsider corporates gaining access to medical records at scale. Let’s not forget Google’s prior attempt at generating mass adoption for an opt-in, centralized electronic medical records and health data platform of its own failed spectacularly. Seen from that angle, Project Baseline has rather more modest ambitions to only acquire the medical records of ~10,000 US citizens — albeit this project is also couched as only the start. “Eventually we hope to expand internationally to capture health diversity on a global scale,” writes Verily.