There are two pieces to making apps for mental health that work, according to David Mohr, director of the Center for Behavioral Intervention Technologies at Northwestern University’s Feinberg School of Medicine: a human element and making the app as low-touch as possible.

At the HIMSS Pop Health Forum in Chicago this week, Mohr shared data from a number of studies to demonstrate his point. He said that an analysis of more than 50 randomized control trials shows that standalone tech solutions are not effective, but coach-supported web apps can be as effective as psychotherapy. 

“If you give a person who’s depressed an app or access to a website, it’s not going to help, they’re not going to use it,” he said. “But when you add a coach, when you add some human services to it, it drastically changes things and that doesn’t need to be a lot. When you add a human to support it, the effect sizes go up 20-fold. And those folks don’t have to be mental health professionals. They can be care managers. It’s a way of extending the workforce.”

One reason for that is that app adoption is low in general, but especially low among people with conditions like depression, that are linked to apathy and low engagement with the world in general. Northwestern decided to test the theory that small apps for very particular use cases, meant to be used frequently but for short times, could spur more engagement than apps that required more in-depth engagement.

This project, called IntelliCare, worked well in studies. While the benchmark for app engagement is that 2.6 percent of downloads are used more than 10 times, IntelliCare apps in a small study saw between 5 and 36 percent being used more than 10 times. When used in conjunction with coaching, the apps had 96 percent engagement after 4 weeks and 90 percent engagement after 8 weeks, and were showing impressive results in treating depression and anxiety. 

“The take home on this is that I think these digital tools for mental health, they have to fit in the fabric of people’s lives,” Mohr said. “We have to design them so they’re easy to use. The other thing we have to reconceptualize is that these are not products. We’re not going to put some apps in a clinic and that will solve the problem of depression. What we can do is use technology to create new services that can be provided that are more cost effective.”

To that point, the latest work his Center has been doing has been around passive sensing. Using sensors like the phone’s accelerometer, GPS, activity sensors, and more, passive sensing has impressive predictive power around depression, he said.

“We’ve gotten really really good at treating research patients in research setting, but implementation is hard,” Mohr said. “That’s the next big challenge in this area. I think these digital health tools are not just tools, we have to think of them as services. You can’t just develop a tool, you have to develop a tool that fits into a care setting that has a strategy for how to help a person get better.”