Sensors and Sensibilities

Companies and insurers love fitness trackers.
Should you?

A scenario from the closer-than-you-think future: It’s the end of a great, healthy day. You achieved a grueling pace on a 25-mile bike ride through the East Bay hills without your heart rate ever venturing into dangerous territory. You didn’t smoke, drank only one beer, chowed down on a kale salad for dinner. Your blood pressure and blood sugar levels are optimal — satisfyingly better than your average for the last month. You know all this because your wearable fitness tracker is loaded with sensors that collect all your vital medical information. The best news of all? The discount on your health care insurance premium, which is dynamically linked to your real-time medical data, is safe. You will be financially rewarded for taking care of your health and it is awesome.

But the following week, your personal analytics tell a different story. You drank one too many margaritas at the sports bar while watching the World Series. The next day, you blew off your bike ride because you were tired and hungover, and then you binged on chips and soda. You might even have taken a few hits off your vaporizer before collapsing on the couch for a three-hour stretch of DVRed “Sons of Anarchy.” Just before crashing into bed, bloated and stoned, you receive a notification from your always-attentive smart watch: Keep up like this, and your premium discount will be history. You will be financially punished for your drunken sloth. Not so awesome.

Jeff Temple’s Twitter bio reads “Diabetic Triathlete Homebrewer and Cloud Computing Fan.” He wears an insulin pump and normally takes good care of his health, fueled by a competitive streak fierce enough to accomplish a grueling Ironman Triathlon a couple of years ago.

But after achieving that feat, he grew sedentary. He gained twenty pounds, he recalls. He spent too much time on the couch, or in front of a computer.

“I did the real pinnacle endurance event,” says Temple, “and then I fell off.”

Temple turned things around after his employer, Appirio, a San Francisco-based cloud computing consultancy, encouraged him to start wearing a fitness tracker. Appirio had recently joined the burgeoning ranks of companies aiming to improve employee health and cut overall health care costs through the establishment of “corporate wellness programs.” The company set up intramural competitions in which Appirio and his colleagues could compare how many steps they racked up per day, or how many pushups and situps they did per week, or even how well they were sleeping.

The fitness trackers, says Temple, made a real difference.

“Tracking my steps, made me more aware of my activity,” he says. “It got me fired up about getting back out there, getting back into the groove. It’s been a year now, and I’ve lost the 20 pounds. I’m way more active, running every day. It was a very cool experience.”

Cool, not just for Temple’s health, but also for Appirio’s bottom line. In the summer of 2014, citing data gathered by fitness trackers along with other benchmarks detailing employee participation in the wellness program, Appirio convinced its health insurance provider, Anthem, to cut the company’s insurance bill by 5 percent — a tidy $300,000.

For at least a decade a growing number of corporations have been attempting to lower their health care costs by encouraging employees to voluntarily sign up for incentive programs that stress exercise, weight loss, and a healthy diet. Since 2009, the inclusion of a provision in the Affordable Care Act that allows employers to offer discounts of as much 30 percent of their premium costs to employees who participate in such programs has made them even more attractive. Some 50 million Americans are now “affected by workplace wellness programs.”

The use of fitness trackers as part of these wellness programs has become increasingly common. But the interest of health insurance companies willing to draw a direct line connecting the data generated by trackers and cold hard cash is relatively new. Only in the last two years have some of the biggest insurers in the country started deploying programs that offer premium discounts to both companies and individuals who can prove that they are taking care of their health by quantifying their health data with an “activity device.” It’s still too early to know for sure how far this trend will continue, or how the dollars will add up in the end, but the potential ramifications — for our health, our health care costs, and our privacy — are enormous.

Health analysts estimate that around 70 percent of the annual $2.6 trillion bill for health care costs in the U.S. are the consequence of potentially changeable human behavior. The avoidable price tag for complications from obesity and diabetes alone adds up to hundreds of billions of dollars a year. The Centers for Disease Control and Prevention reports that lifestyle changes in diet and behavior resulting in the sustained loss of just 10–15 pounds can reduce the risk of getting diabetes by 58 percent. Imagine if a million, or ten million, or fifty million Jeff Temples got off the couch, and onto their bikes. Our entire economy would be better off. Health care costs wouldn’t be breaking the federal budget. We’d have more money to spend on nice things.

But we’d also be giving more access to our most personal medical data to more people than ever before — a scenario that raises a whole host of troubling questions. Do we want our health insurers and employers to get their hands on real-time data that could conceivably be tracking our alcohol consumption or sexual experiences or mental health? Who will own this data, ultimately — the manufacturers of the devices that collect it, the insurers and employers who act upon it, or us, the creators? Where does the data live? In our devices? In the cloud? How easy will it be for hackers to steal or the NSA to gobble up?

The “full-disclosure future” is upon us. We should probably take a stab at figuring out answers to some of these questions.

“We’re rewarding folks for taking accountability for their health,” says Kristine Mullen, VP of Wellness for Humana.

If their biometrics are “in range,” says Mullen, participants in Humana’s Vitality program are awarded graded status levels — silver, gold, platinum.

“The higher the status level, the more rewards you get,” she says. “It’s akin to an airline miles program. Based on your status, employers are beginning to offer discounts on your premium contributions. If you figure the average health plan member is paying $350 a month in premiums, and you get ten percent off — we’re talking about a good amount of money.”

Humana’s program, along with a similar offering from Cigna, includes multiple metrics for determining participation in a wellness program beyond just the data generated by trackers, but our new wristband accessories solve a core problem that’s been bedeviling insurers and the operators of wellness programs for years. How do you ensure that people are really participating?

Fitness trackers address that dilemma, says Amy McDonough, who runs the wellness programs at Fitbit, manufacturer of one of the most popular fitness tracking devices, says her company has worked with “thousands” of clients integrating Fitbits into wellness programs.

“It’s about participation, and engagement,” says McDonough. “The trackers bring not just ease of use but also a kind of validation of the data. There’s a level of granularity that we haven’t had before.”

“Sustained behavior change is not a trivial thing to accomplish,” notes Christine Lemke, CEO of The Activity Exchange, a San Francisco startup that works with health care companies to manage “digital health interventions.” “But trackers are part of a constellation of things that work.”

As anyone who has ever stepped on a scale knows, being able to measure something makes a difference to your motivation. And now, with Fitbits and Samsung Gears and Jawbone Ups and, soon, Apple Watches, we can measure all kinds of things better and more easily than before. And then, in turn, all that data can be aggregated to give employers and insurers a more accurate view of employee health and risk profiles.

No wonder a report by Rock Health, a San Francisco based seed accelerator for digital health startups (and a funder of The Activity Exchange,) dubbed fitness trackers as “the source of truth for biometric based measuring systems.”

The source of truth! The entire concept of insurance is based on the fact that, historically, we haven’t been able to pinpoint an individual’s exact risk for anything. But if the promise of wearable biosensors integrated into corporate wellness programs bears fruit, the business of insurance is about to change.

When Apple CEO Tim Cook introduced the Apple Watch, he described it as a device that will “greatly improve the way the way we keep track of our activity and change how we look at fitness.” That’s no small thing, when you consider it applied to the finances of health care.

“We are getting out of just paying the claims for your sickness and into the business of improving health,” says Joe Mondy, director of public relations for Cigna.

Mondy is describing to me how Cigna has incorporated fitness trackers into the wellness programs operated by the insurance company’s corporate clients. So far, says Mondy, only “a relative handful” of forward-looking corporations are participating in Cigna’s “digital engagement ecoystems” — wellness programs that come with wearable biosensors. But he expects the number to rise. Because according to Mondy, the program works. Surveys at the end of a Cigna test pilot in 2013 involving 600 subjects indicated 80 percent of the participants were “more motivated to manage their health at the end of the study than at the beginning.” 60 percent, said Mondy, continued exercising afterwards.

“Typically,” says Mondy, “you get 20 percent continuation. So willingness to participate on a longer term tripled.”

The clincher: The “absolute total cost of medical for people that participate in a series of our health improvement plans actually goes down,” says Mondy.

There is, to date, no independent confirmation of the efficacy of wellness programs that incorporate fitness trackers in bringing health care costs down. There is even some doubt as to whether corporate wellness programs, with or without trackers, work all that well. A Rand study that looked at 600,000 employees at seven corporations concluded that wellness programs had “little if any immediate effects on the amount employers spend on health care.” A much-debated study published in Health Affairs in 2013 said the programs may actually “shift costs to workers who can least absorb them” because healthy workers end up getting discounts while sick workers pay the maximum premium.

But that hasn’t slowed interest. Apple has reached out to both Humana and UnitedHealth about bringing the Apple Watch into their plans, and is already testing out its new HealthKit package of medical data apps in medical trials. Cook told BusinessWeek “that employers eager to see their workers become healthier may subsidize the [Apple Watch].” A major case in point: BP distributed thousands of Fitbits to its employees and claims to have lowered its health care costs below the national growth rate. And whether they work or not, it’s clearly good business for the tracker makers: Bloomberg News reported in August on a study by Parks Associates asserting that by 2018, some 66 million fitness tracking devices will be sold annually, “with about a third coming from corporate-wellness programs.”

“That’s where the industry is going,” says The Activity Exchange’s Lemke.

The auto insurance industry has been leading the way on the integration of tracking and insurance for years. Almost every major auto insurance provider now offers discounts for drivers who allow their driving habits to be tracked. It’s called usage-based insurance, and it is, on the surface, quite seductive. Plug a “telematics” device into your computer’s dataport and it will gather data on how fast you drive, what time you drive, where you drive, how quickly you accelerate or brake, and even how hard you make a turn. Stay within the right parameters, and you keep your discount. Let your teenage son go joyriding in the bad part of town at midnight…. and maybe you lose your discount.

News coverage of a conference on usage-based car insurance organized by the Center for Insurance Policy & Research in November 2013, summarized “the expectation of the insurance industry as a whole that all policyholders will eventually be subject to participation in some form of usage-based insurance.” The apparent reason: better risk algorithms will lead to higher profits.

A policy paper published by the National Association of Insurance Commissioners lays out the logic:

UBI has the advantage of utilizing individual and current driving behaviors, rather than relying on aggregated statistics and driving records that are based on past trends and events, making premium pricing more individualized and precise.

UBI is poised for rapid growth in the U.S., with many experts predicting that up to 20 percent of all vehicle insurance in the U.S. will incorporate some form of UBI within five years. This growth is being propelled by technology advances, which continue to substantially improve the cost, convenience, and effectiveness of using telematics devices. It is through the use of telematics that insurers are able to collect driving data that better enable them to more closely link a driver’s individual risk with premium. Through UBI programs, insurers are able to differentiate products, gain competitive advantage, and attract low-risk policyholders. Recognition of the societal benefits and growing consumer acceptance of personal data collection will only serve to further increase demand for UBI products in the future.

But one attendee at the CIPR conference who is frequently at odds with the insurance industry, Birny Birnbaum, the executive director of the Center for Economic Justice, warns that the “societal benefits” may not be as cut-and-dried as the insurance industry would like us to think.

“There’s no guarantee that they are only using that data for purposes of determining your premium based on what your driving profile is,” says Birnbaum.

The data gathered on your driving habits, suggested Birnbaum, will be detailed enough to suggest whether you are the kind of person who wouldn’t blink if your premium was raised.

“Predictive analytics,” says Birnbaum, crunching the “big data” generated by telematics, will be able infer quite a bit about you from where and when and how you drive — whether, for example, you might be a good candidate for other products, like life insurance. Or whether you might be a bad risk, not because of how you drive, but because of where you live or what your income level is. And then, of course, there’s the inescapable flip side to getting a discount for participation. If you don’t participate, you are effectively paying a higher rate.

Translating those warnings from auto insurance into the world of health insurance raises all kinds of gnarly questions, especially when you consider the far greater intimacy of information that sophisticated devices like the Apple Watch will be able to gather about us. Few people have been paying closer attention to this emerging new world than Scott Peppet, a law professor at the University of Colorado who has been tracking the evolution of the sensor-everywhere “Internet of Things” in the context of insurance systems for several years.

You name it: Blood glucose levels, arrhythmia, epilepsy, stress, body temperature, heart rate, respiration, sleep patterns, weight, physical movement, and even “medication compliance” (whether or not you take your drugs as directed) — Peppet is convinced it’s all going to be tracked.

“The sensors will get smaller and cheaper and easier and less invasive,” says Peppet, “and you are going to be able to build more and more of them into a device. At this point it’s easiest to just assume we’re going to be able to track most health variables. And then the question is: What does that mean?”

Echoing Birnbaum’s concern about predictive analytics, Peppet observes that the spread of wearable biosensors entails a major shift in the quality of information that can be known about us.

“The sensors that are in these kinds of devices generate such rich information that they blow away anything that we’ve seen before. We’ve spent the last ten years perfecting how to draw inferences about a person from how they click around their Web browsers,” he says. “That’s baby steps compared to this.”

(Just a few days before I interviewed Peppet, researchers at Dartmouth announced that they were able to monitor students for signs of developing mental illness over the course of a semester through their iPhones, while another study published by the University of Pennsylvania concluded that data from a wristbound accelerometer could tell whether a person had fired a gun. The source of truth, indeed.)

“Where is that data, who is holding it, and what are the terms under which it is being shared with my employer?” asks Peppet, noting that these are not always easy questions to get clear answers to. Privacy policies can change drastically over time, as anyone with a passing acquaintance with Facebook knows. Your biosensor data might also turn out be very interesting to law enforcement if you’re accused of a crime. And so on.

Peppet also has concerns about whether bottom line considerations will result in pressure from insurers and employers for ever greater levels of participation in these programs. What happens to privacy when “wellness” becomes a condition of your employment? After the drug test, here’s your tracker. Don’t take it off.

Every single person I talked to who was involved in setting up wellness programs involving activity trackers was quick to stress that participation was always voluntary, and that employees should and would have control over whether to share their information. “We only share data with the active consent of the creator,” swears Fitbit’s McDonough. Humana’s Mullen and Cigna’s Monday made similar declarations.

The Activity Exchange’s Lemke acknowledges that “there will be a very strong incentive for folks to report their data.” But she is convinced that “it will always be consumer-controlled.”

In the post-Snowden era, “people want to control what’s exposed, and they are going to reject surveillance,” says Lemke. “Employers moving into a world where talent is more important than physical labor will not want to make their employees angry about this, no matter what it ends up costing.”

Peppet disagrees.

“Do I really have a choice?” asks Peppet. “If I am the only person in my office who is not wearing my fitness tracker — doesn’t that start to make me look bad? You end up with stigma if you are not participating and the stigma is you must be hiding something: You’re the guy who is not exercising, you’re the guy who is driving poorly. If you start to see the rise of corporate cultures in which everyone has a comprehensive health tracker, and the boss is regularly saying how much money the company is saving because we have got such great participation, but we have a corporate participation goal of 95 percent and we’re not quite there…. There’s going to be a negative consequence for not participating.”

So do we have real autonomy in the age of wearable biosensors? In Peppet’s view these are the kinds of questions we should be thinking about before we plunk down $349 for an Apple Watch. “The reality is that the legal system is way behind preparing for this. It is kind of wild how quickly these devices are proliferating and how slowly we are moving.”

Go for a long bike ride — your health insurance premium shrinks. Drive under the speed limit — your auto insurance premium is discounted.

What, asks Peppet, does that mean for how we have traditionally thought about “insurance”?

As implemented over the last couple of hundred years, Peppet explains, the business of insurance has been built on the concept of a community pooling risk together, precisely because we don’t know exactly what our individual risk is.

“It’s a form of social sharing, of shared social obligation,” says Pennet. “I’m going to carry a little bit of your risk and you are going to carry a little bit of mine, because the truth is neither of us knows who is less safe; who is going to have the car accident or whose house is going to burn down. But now we know exactly how people drive and exactly what their pre-diabetic risk is. We can price everybody perfectly, because we are tracking everyone’s individual behavior. That seems to me like a different social experience. It’s not the same sense of shared social obligation.”

For many people — particularly the safely-driving, actively-exercising people who don’t play stupid games with fire — that future is going to look awfully attractive. But it will also be defined by what Peppet calls “a different kind of discrimination” from the racial or gender discrimination that we’re used to.

“This basic sorting of people into exercisers and non-exercisers,” says Peppet, “and responsible drivers and irresponsible… I think we’re going to start to see people get very uncomfortable with that level of sorting. We may decide this seems like a level of discrimination that we don’t want.”

The discrimination isn’t just between responsible and irresponsible. Peppet believes there could be a form of economic discrimination as well. The rich will always be available to afford their insurance, no matter how fast they drive or many french fries they binge. But the poor won’t have a choice — stay within the right biometric range, or you won’t be able to make the rent.

The age of “pervasive lifestyle incentive management” is upon us. It is, in some ways, a logical, foreseeable conclusion of the digital revolution. We always knew that our data would define the truth of who we are. Now we should prepare to pay accordingly.

But in any given trade-off we need to be clear about what we are willing to give up for a potential benefit — whether that be cheaper insurance or better health or both. If the merger of fitness tracking with health insurance helps an overweight nation get its diet under control, that’s great news. But if a future where health risks can be priced exactly means that the sick and lazy are penalized, and the poor are monitored every second of their lives, we may regret grabbing for that discount so eagerly.

All illustration by Clay Rodery
Follow Backchannel: Twitter | Facebook