Digital tools must be sophisticated and reactive guides, capable of offering personalized feedback, encouragement, and tips, according to Brennan Spiegel, a director of health services at Cedars-Sinai. Photo credit: Getty Images

You’re inside the heart. More precisely, you’re floating through an artery. You pass a buildup of waxy plaque and watch as it thickens before your eyes. The 360-degree organ is designed to be manipulated. Arrhythmias, clogged arteries, and valve problems can all be conjured, rendering hard-to-visualize conditions.

Using the virtual reality (VR) platform, doctors can show patients, in an immersive, highly detailed way, what will happen to their bodies if they take their medication, stop smoking, and eat better — and what will happen if they don’t.

“Think of it as a time machine,” says Jeff Hazelton, the CTO of Sharecare’s VR lab, which built the simulator.

See also: Sharecare aims to drive behavior change with VR acquisition

In the future, the company hopes to create virtual avatars whose unique biology responds to changes in a person’s vitals, diet, and exercise routine, among other metrics. There will be rewards for maintaining healthy habits, but the more powerful incentive will be “visualizing how your actions impact your health,” Hazelton predicts. “Once you see it, it becomes more real.”

He isn’t alone in his belief that digital tools can improve healthcare, in large part by motivating patients to develop better habits. Through immersive visualizations or constant feedback loops, the thinking goes, we’ll be alternatively spooked and nudged into leading healthier lives.

It might not be that simple, as a new study shows by looking at wearables — a proxy, perhaps, for assessing the impact of other promising digital health tools. In the study, researchers evaluated 27 randomized trials to assess the effects of using wearables on patient outcomes (the review was published today in Eric Topol’s new journal, Digital Medicine).

See also: Drugmakers turn to tech to better demonstrate empathy

After analyzing the results, which included information on BMI, weight, body fat percentage, and blood pressure, the researchers found using wearables had “no statistical benefit on human health,” reports Brennan Spiegel, one of the study’s authors.

“Digital health is not a computer science or an engineering science. It is a social and behavioral science,” explains Spiegel, who is director of health services at Cedars-Sinai. It’s a phrase he repeats often, an acknowledgment tech’s ability to meaningfully shape behavior must always be viewed through the messy, often glitchy lens of human decision-making.

This isn’t the first time digital health’s ability to improve behavior has been shown to be lacking. A 2016 Journal of the American Medical Association study found dieting adults who wore activity trackers for 18 months lost significantly less weight than those who didn’t.

The behavioral economist Dan Ariely says he doesn’t find the results all that surprising. Instead, they’re an illustration of one of the myriad ways we’re wired to make irrational decisions, such as rewarding ourselves with a high-calorie treat when we hit our daily step count.

See also: Cigna partners with Microsoft HoloLens on biometric game

Data alone might not be enough to fix these tendencies, as Spiegel and colleagues found. “Interventions aimed at increasing physical activity and weight loss using various activity trackers showed mixed results,” the authors wrote, while those “based on health behavior models and personalized coaching were most successful.”

It’s not that wearables, VR, or other emerging tech can’t improve patient outcomes. It’s just that human behavior is complex, multifaceted, and hard to crack. “Data is confusing,” Ariely says.

To truly move the needle, he continues, digital tools will need to do more than present detailed information or immersive experiences. Instead, they must be sophisticated and reactive guides, capable of offering personalized feedback, encouragement, and tips (but only when we’re ready to take them in).

“Some patients aren’t ready to hear advice,” Spiegel concludes. “Some will contemplate it, but won’t act on it. Some are ready to act, but don’t have the tools.”

In other words, good luck meaningfully changing behavior without first identifying where a patient falls on this spectrum. With the “quantified self” hitting a roadblock and the jury still out on VR and other promising digital tools, we’re not quite there yet.

Correction: An earlier version of this article misidentified Jeff Hazelton’s role at Sharecare. He is the CTO of the virtual reality division, not CTO of the company.