Taking steps to improve activity-tracking results

Fitbit, Samsung Health, the Apple Watch—dozens of devices and programs have emerged in recent years to massage the 10,000-steps-a-day ideal into public consciousness. But what remains elusive is whether these tools actually improve health.

Or, more fundamentally, if they’re being used at all.

“If you search for it, you’ll get ‘20 percent,’ or ‘one in four people,’ have a wearable and carry it with them wherever they go, but if you go ask 100 random people on the street, it’s clearly not true,” says Mitesh Patel, director of the Penn Medicine Nudge Unit, who last fall led a study published in the Annals of Internal Medicine that monitored the rate of use of physical activity trackers among people enrolled in an insurance-based wellness program. “I think this [study] was a nice, objective way to show it’s not 20 percent; it’s probably a couple percent.”

Patel, who has spearheaded several studies about activity trackers since their advent, partnered with Humana between 2014-15 for the largest study of its kind, totaling approximately 4.5 million people who had access to the HumanaVitality wellness program, which incentivized continued use with financial rewards of as much as 25 cents to 40 cents per day. Remarkably, during the two-year monitoring period, only 1.2 percent of the sample—or, 53,245 people—activated a tracking device. 

But the study also offers good news about trackers with its insights into the habits of those who did use them, even after six months had passed. Eighty percent of those who signed up continued to use the trackers after six months; to boot, though only .1 percent of the elderly in the sample population activated their device, 90.4 percent of those people—more than any other demographic—sustained use. Individuals younger than 34 were the least likely to stick with it.

Mitesh Patel, director of the Penn Medicine Nudge Unit, the world’s first behavioral design team embedded within a health system. Photo by Leonard Davis Institute of Health Economics.

Patel says the findings are useful for designing future health tracking programs and informing insurance companies, not to mention medical practitioners, who to focus on in the future.

“Really, the goal of that study was to look at what people are using, in general, and then see what demographics an employer might want to think about,” he says. “Maybe they do want to offer extra promotions to target the elderly.”

Patel, who is a part of the world’s first-ever behavioral design team in a health system, says the study data suggests that the design and framing of incentives are crucial in persuading people to be more active. Access to the technology and financial incentives alone, he concluded, may not be enough.

“One challenge is most of these incentives are designed under the standard economic approach: ‘You lose weight, be active, get your metric screening, then we’ll pay you. We won’t give you cash; it’ll come as a deduction in your health insurance premium, which you’ll never see because it’s tied to your bi-weekly paycheck and shunted to your bank account through direct deposit,’” Patel says. “It’s hidden; it’s delayed. These are all things we know from behavioral economics are not very motivating to people.”

In essence: Build it and, in reality, they may not come.

“These are behavioral issues, not technological ones,” adds David Asch, executive director of the Penn Medicine Center for Health Care Innovation, which collaborates with the Nudge Unit. “Everyone already knows seatbelts save lives and smoking kills, but people still smoke and still people get into a car without buckling up. Behavioral economics recognizes that people don’t always act in their own rational best interest. The key observation of behavioral economics is not just that people sometimes behave irrationally, but that they do so in highly predictable ways. It is the predictability of these decision errors that allow us to design strategies to overcome them.

“That’s what has me optimistic.”

Therefore, Patel has since evolved his research—at present, he has about a dozen ongoing trials he’s juggling in tandem with the Center for Health Incentives and Behavioral Economics, many with at-risk populations like those living with diabetes, to focus on concepts like social incentives and gamification. That is, testing competitive, collaborative, and supportive design structures—all cornerstones of behavioral science—in a game-like interface to encourage sustained use.

An interface might, as in more recent work Patel conducted with Framingham Heart Study, for example, rope family members into groups and track step progress through leaderboards, level-ups, and rank status, drawing on feelings of anticipated regret to encourage use. By design, that intervention might then be married with the financial concept of what Patel calls loss-aversion, or the theory that when presented with incentives, people are more likely to participate if they’re told they’re losing money as opposed to gaining, even if the end-dollar amount they receive is the same.

The ultimate point being that most people need a nudge to stay active. How big a nudge, then, is the variable.

“As we continue, we’re trying to make the intervention better, trying to tailor it by matching it to people’s baselines and characteristics, their personality,” Patel says. “And the goal, over time, with these studies we’ve done and especially the ones that last a year or more, is if we can get people to be active for longer, we can start measuring changes in health outcomes.”

step tracker