In Las Vegas: A Tale of Two Devices at CES

The FDA's job may need clarification as technology changes. Fitness devices are becoming more intelligent and will soon produce data that approaches medical advice.

Two years ago, at the end of my on-site roundup of the building digital health excitement at the International CES show, I wrote about Scanadu, a company developing a noninvasive vital sign reader called the Scout. At the time, Scanadu's founders showed their prototype at a table at one of the off-floor CES press events. It was to my knowledge one of the only announced products at the show waiting for FDA clearance before going on sale.

Today, two years later, Scanadu still doesn't have FDA clearance, and you can't buy the Scout on the company's website. It's a shame, really. The product is more elegantly designed than many consumer electronics products, appears to be simple to use, and provides a lot of valuable biometric information to someone possessing one.

But here is where we get into what's wrong with CES, or what's wrong with the FDA, or perhaps both.

The digital health section of CES is substantially larger than it was two years ago, but few if any devices on display are FDA-approved, and thus have everything to do with fitness and virtually nothing to do with health monitoring. This is by design, a grand bargain FDA reached with industry many years ago, which I studied at length in 2013 when I earned a certificate in clinical trial design and management at the University of California San Diego Extension.

Devices that purport to treat disease must get the FDA's blessing. Devices that simply provide information but do not claim to treat disease may proceed without FDA approval, but must warn users they are not designed to treat disease.

Scanadu's journey to approval as a disease-treating device is an instructional story.

"We were one of the pioneers in this, but the amount of work we had to do to actually qualify for FDA is enormous with consumer electronics," says Walter De Brouwer, co-founder and CEO of Scanadu. Belgian-born De Brouwer has been something of a media darling in Silicon Valley due to some previous entrepreneurial activity.

"It's not a problem with FDA," De Brouwer told me in a phone interview last month. "It was a problem with our device actually. We do vitals and we fuse data, but if one of the vital signs which is measured by a sensor doesn't work, we have to start all the data and analysis again, so we have to see thousands of patients again.

"A big part of my company are mathematicians. FDA is very mathematically oriented. They tell you, 'This is the standard deviation. These are two tests. Here we want 285 people. Here we want 80 people. The repeatability must be three. The standard deviation must be six. Our mean absolute error must be that.' And that means that you have to have teams that do hardware, who do manufacturing, who do firmware, software, mathematics, and if one of them makes one mistake, because it's fusion, it pervades through all the others, and you have to start all over again."

This month, Scanadu will ship 8,600 Scouts, but without FDA approval. According to a company spokesperson, this batch consists of "investigational devices" being provided to participants in the company's Indiegogo crowdfunding campaign (which exceeded its goal of $100,000 by more than $1.5 million). Feedback from the Indiegogo participants is a valuable part of Scanadu's usability study for the company's road to FDA clearance, the spokesperson says.

De Brouwer now hopes to receive FDA approval in the third quarter of 2015.

Despite all this, last year a study described in JAMA reported that 42 of 50 selected medical devices lacked publicly available scientific evidence to verify the devices' safety and efficacy. In response to these findings, the FDA said it "reviews a significant amount of data—far more than what is publicly available" and that the agency "has served the American public well by balancing the need for robust evidence to assure safety, while expeditiously bringing new technologies" to the market.

Meanwhile, back on the CES show floor, it's anything goes as long as you carry that catch-all disclaimer. Fitness trackers and trainers abound, usually dispensing nothing but raw data (heart rate, distance, cadence) and often now incorporating a social media aspect that urges weekend warriors to go faster and further. Sharing results is typically "gamified," which means friendly competition abounds among fitness gadget wearers. These devices often carry disclaimers advising the user to consult with a doctor before starting any exercise program, disclaimers that may be ignored.

While FDA has its own work scrutinized, it has to be aware that the more intelligent the average fitness device gets, the more likely it will be that those devices will be aggregating data, as Scanadu is doing, and produce something that approaches medical advice—the exact kind of innovation CES attendees are all talking about.

In 2013, a variety of House Republicans did their table-pounding best to take shots at the FDA on the impact the agency's heavy-handedness was having on innovation in the U.S., where only the stalwart folks like Scanadu or AliveCor have the fortitude to make health-oriented consumer devices safe, effective, and easy to use.

Now, as if to illustrate some Congressman's point about stifling U.S. innovation, a Taiwanese company today will exhibit on the CES show floor with what it calls the "world's first stamina sensor and app."

Runners or cyclists wear the sensor with a chest strap, and the sensor's two-lead EKG starts to generate data. The system alerts wearers through different vibrations when wearers are at 60% of their stamina, 30%, or reaching a limit where they should relax to recover, potentially avoiding injury.

Two ironies about this device. First, its name is GoMore—a name more appropriate for devices that do far less and simply egg on athletes—and the company developing it has the name bOMDIC, which in my book looks a lot like the word biomedical, certainly suggestive of the kind of device that the FDA would have to approve.

And yet, while the U.S.-created Scanadu Scout waits its turn, GoMore goes to market. The GoMore manual doesn't really explain this, but there is first a calibration phase, when the wearer is supposed to exercise all-out, to allow the GoMore algorithm to know the wearer's capability. With this data, GoMore learns the capability of the wearer, and knows where to set the thresholds for its later stamina alarms and further learning about the wearer's performance. But it's really the wearer who controls the gadget's upper boundaries of knowledge.

While the product is not classified as a medical device, bOMDIC also throws in the standard disclaimer that the product is "not to be used to diagnose, treat, or prevent any disease or medical condition."

"Our algorithm is a self-learning algorithm, which means you have to disclose your limitation, so the next time it will feed back more precise results," says HsinFu Kuo, CEO of bOMDIC.

GoMore's algorithm infers the wearer's remaining energy and buildup of lactic acid, Kuo says. "If you ask the people in the FDA, they do not give you a very clear definition of stamina, so we try to sell the product as consumer electronics."

Nevertheless, products like GoMore are starting to make the same kind of advice your doctors might make if they were running alongside you. I predict that if the product is wildly successful, the FDA will have to revisit its policy, because so many people might be coming to depend on GoMores and the like for what amounts to a kind of medical advice.

Or, as De Brouwer puts it, "The FDA might not be their worst enemy. Their worst enemy may be the ambulance-chasing lawyers. America is a litigious society."