How Wearable Devices Could Get Doctors' Stamp of Approval

Step 1: actually work. 


Wearables have been hot in Silicon Valley. They refer to little computers and sensors you wear on your body instead of devices you keep on your desk or in your pocket.

The great hope for many wearables is that they can capture data about what you're doing and use it to generate what are invariably called "insights" into your behavior and wellbeing.

Perhaps you've seen the many wristbands that are now for sale: the Jawbone Up, Fitbit Flex, Basis Health Tracker, Withings Pulse, Samsung Gear Fit, and a herd of other tiny devices. In the future, there will be shirts and skin patches and computers you swallow, too.

And yet, as any honest observer will admit: these devices are only good at a few things and only sometimes. For example, they can count steps really well. But for actual medical use, they are just not up to the gold standards, and probably not even the silver or bronze, either. There have been major problems with manufacturing the devices, validating their output, and using the information they generate.

Yesterday, Young Sohn, chief strategy officer at Samsung, the world's largest smartphone maker, revealed the company's vision for digital health. "We are here to outline the single greatest opportunity of our generation," Sohn said, "to better understand our physical well-being, to give a voice to what is happening in our bodies."

And even he admitted on-stage that when it comes to wearable devices, he usually wears one "for two weeks" and then tosses it for not being "good enough for me."

"I become my own guinea pig. I wear anything wearable out there," Sohn explained later in a press conference. "And there are really a lot of issues. We are in a really early stage. They are not very comfortable. I have even worn a patch and a day later, I was very uncomfortable, so I took it off. And there are battery issues, data exchange issues, and also accuracy issues."

"A lot of startup companies are a few people with interesting ideas," he concluded. "If you look at a 30 person startup—like Pebble [a watchmaker]—28 people are trying to make a watch, and two people are working on sensors and algorithms. I want to flip that equation."

So, Samsung is trying to create an open ecosystem that uses common hardware (like the for-developers-only Simband pictured below) and a common data platform to simplify the creation of new devices.

To me, perhaps the most intriguing part of the Samsung plan is the creation of a joint research center with the University of California, San Francisco, the Center for Digital Health Innovation, helmed by Michael Blum, a medical doctor who is the assistant vice chancellor of infomatics at UCSF.

This center would work to validate the sensors and algorithms produced within the Samsung ecosystem. Validation here has a pretty specific meaning that derives from regulatory thinking about medical devices. That is to say, when Blum says "validation," he has a rigorous protocol in mind, not just a rough sense that a device works.

I asked Blum about his plans. How could wearable health devices graduate from "quantified self" science projects to actual clinical use?

"The first projects will be focused on doing the actual technical validation," Blum said. "Doing simple things like looking at the heart rates, the blood pressures, and so on... It's real basic stuff to get started. You can't overstate the importance of getting the validation right and the depth of the validation. Everyone has been joking about the inaccuracy of the prior technologies and that's what we've got to get past to get people really comfortable with this, so you can have the clinical discussions and people will really care about it."

Once they're actually sure that the devices work like they say they do—literally just the basics of input and output—they'll start testing them against The Way Things Have Been Done.

"So we'll really be comparing the heart rates and the blood pressures and the heart rhythms [that come out of the devices] to real gold standards that we use in the medical community," he said. "I don't think that we're expecting out of the gate that these things are going to be perfect. This stuff is going to be an evolution. There is a lot of work to be done to get these to the point where we want them to be."

Then, and only then, can they turn to the fun stuff, where they start trying to use the data in actual clinical trials. "Once we're sure that the data is real, the next step is really starting to do creative clinical trials to say, OK, now that we can collect this data, can we do anything with it?" Blum continued. "Importantly, can we really impact a patient's wellness, impact their health, impact chronic disease management, impact prevention through interventions based on knowing this information?"

And finally we might get to the most interesting part of wearable technology, the kind of potential that people are hyping now, but that is many steps down the road. That's using the continuous monitoring of lots of people to come up with new warning and wellness signs in how their hearts and muscles are working.

"We have the traditional vital signs that we've been looking at for one hundred years—your blood pressure, your heart rate, your temperature, and more recently oxygen saturation," Blum concluded. "But it's quite likely that when we have continuous monitoring of individuals and a lot of new data feeds that when we do the analytics on the combined data there will be new information that we'll get out of it, what we're calling novel vital signs. There will be new markers of health and wellness that come out of these large datasets."

Novel vital signs. Who wouldn't be intrigued by new ways of knowing we are alive and healthy?

For now, though, wearable devices just need to actually work, which is a big enough challenge.