9 October 2020

Health apps: what are they good for?

Chronic Disease Telehealth

Chronic disease is a big issue.

This is not news, because the growth in chronic diseases has taken place over the past three decades, and it would be quite hard not to have heard, for example, that diabetes rates have shot up in recent years across Australia.

What is new, and interesting, is the rise of apps. Mobile phone applications have become one of the most talked-about medical interventions, partially because they are flashy and cool, but mostly because they hold such an enormous promise. You can make an app for a fraction of the cost of traditional medical care, and instantly reach millions of users, in theory.

And so, a huge number of apps have sprung up across the world to manage and treat chronic disease, because when we have new and flashy things we want to try them on every problem to see if they work.

But there’s an issue with apps. Anecdotally, people don’t use them for very long, which is a problem when we’re thinking of using them for diseases that last a lifetime. If people only use the app for, say, a month or two, then it’s not going to help very much with diabetes that may not ever go away.

We were interested in investigating this problem, so we conducted a systematic review and meta-analysis into dropout in apps for chronic disease. After combing through about 1,000 studies, we included 34 pieces of research and aggregated their results.

What we found was a bit depressing, but perhaps not unsurprising – people who use apps drop out at a rate of about 43%. On top of this, even the people who keep using apps often do so at levels far below those required to get a benefit – some people only log in once or twice a year!

It’s hard to improve your chronic disease when you’re only using the treatment once every six months.

What this means is that apps have a huge problem, but also a huge opportunity. We found some evidence that linking these apps in with traditional care improved their use, and that there may be other ways to stop people from dropping out. In many ways, this makes apps just like regular medical care, because as evidence shows many people also, for example, stop taking their medications quite soon after they are prescribed.

And in the case of apps, despite the large dropout rates that we found in our study, many of the included pieces of research had positive results. In other words, even though many people didn’t use the app, the people who did had large enough benefits that the whole group looked positive, on average.

So the take-home from our research is twofold: firstly, that people drop out of apps a lot, but secondly that apps are also very useful in the people that they help, and if we work to improve apps they may yet live up to their promise.

Gideon Meyerowitz-Katz is an epidemiologist and PhD student at the University of Wollongong. He blogs at medium.com, hosts the Sensationalist Science podcast and tweets @GidMK.

COVID-19 live update