Microprocessing

Could Your Therapist Be Replaced with an App?

While the field of internet cognitive behavioral therapy holds a lot of promise, the best mental health care still requires a human touch

In Microprocessing, columnist Angela Lashbrook aims to improve your relationship with technology every week. Microprocessing goes deep on the little things that define your online life today, to give you a better tomorrow.

EEvery evening, I receive a text message from a robot. Its name is Woebot, and sometimes the messages are as simple and straightforward as “Hello there, my friend!” Yesterday, the text was a bit odd: “I flossed my grills today.”

The bot’s earnestness is strange and cute, but it’s not just there to entertain me with bizarre quips. Woebot’s job is to coach me through my stress and anxiety in 10-minute segments each night.

Woebot is a part of a burgeoning field of therapy called ICBT, or internet cognitive behavioral therapy. Cognitive behavioral therapy is a decades-old, highly studied, and remarkably effective form of psychotherapy that focuses on problem-solving and emotional troubleshooting. Yet whether because of geographic distance from a therapist, financial issues, or insurance struggles, many Americans lack access to the mental health care they need. Hence the appeal of ICBT: If much of the costly and time-consuming therapy patients currently receive could be outsourced to a program, considerably more people could access needed mental health treatment.

At least that’s the hope. But there are significant limitations to getting ICBT to the general public. It isn’t yet clear to researchers and app developers how to best implement ICBT outside of clinical trials; what works in a research setting may not necessarily translate to the less-controlled real world. And a lack of regulation or even guidelines means developers have little incentive to use actual scientific findings when designing their apps.

There’s a large body of research showing that ICBT can work. It can relieve symptoms of anxiety, depression, and even harder-to-treat issues like obsessive compulsive disorder or body dysmorphic disorder (BDD), an underdiagnosed and undertreated disorder in which patients experience intrusive and obsessive thoughts about perceived flaws in their physical appearance. A 2018 study out of the Karolinska Institutet’s Rücklab in Stockholm looked at how a 12-week internet-based cognitive behavioral therapy program affected patients with BDD. Participants in the program read traditional cognitive behavior therapy self-help texts and completed worksheets in interactive modules, similar to how one would complete homework in an online class. They also had unlimited access to a therapist via a messaging system within the platform.

The study found that, two years following treatment, 69% of participants had responded positively to the treatment, and 56% no longer had a full diagnosis for BDD. This finding was significant, says Christian Rück, the group leader of the Rücklab, because many therapists shy away from treating BDD since they view patients with the disorder as being unmotivated and difficult to treat. This is because many patients with BDD strongly believe there is something wrong with their bodies, and it can be difficult to compel them to think otherwise.

“Therapists would rather treat people who will be successful cases,” Rück says. For people with BDD, as a result of so few therapists who will work with them, treatment options are limited. “With BDD, we felt we were actually at the frontier of what we thought was possible, but we have had huge successful trials there… the potential there for patients is huge in the way that if they don’t get this treatment, they probably don’t get any treatment.”

Rück estimates there have been about 200 randomized controlled trials finding ICBT to be an excellent treatment option for people with various mental disorders. But things get trickier when the treatments are transferred from the steady hands of psychologists into those of app developers and entrepreneurs. Scientists, says Rück, “are not good at making sexy apps but rather boring-looking things where we claim they are effective and we can reference a scientific paper. But in the app-store world, that’s not too important, and I think we are looking at an avalanche of crap apps in mental health.”

Rück says he thinks there are likely thousands of mental health apps that are rankable by reviews and download numbers but not necessarily scientific rigor. A 2019 study looked at 73 Android and iPhone mental health apps and found that while most of them claimed the ability to diagnose or treat mental disorders, only two of them described evidence of it and just one cited research that backed up those claims.

While there’s no evidence yet that ICBT is harmful in a clinical setting, it isn’t hard to imagine the dangers inherent in having a bunch of unregulated, unscientific mental health apps potentially taking on the role of a traditional therapist or psychiatrist. If someone is seriously mentally ill and decides to use an app rather than a certified therapist, they could be much less likely to get better. They could even, in severe circumstances, be in danger. And there’s the not-insignificant issue of data privacy; our health data is perhaps the most private — and most valuable — data we produce, and the protection of it is crucial.

The American Psychiatric Association is aware of these problems. The group’s proposed solution is a hierarchical rating system that evaluates an app based on a number of factors:

  • Risk, privacy, and security: Is the app safe for a user? Is there a privacy policy, and is data de-identified, and with whom is it shared?
  • Evidence: Does the app actually do what it claims that it can do? What peer-reviewed evidence does it provide?
  • Ease of use: Is it easy and compelling to use? Is it accessible for people with disabilities?
  • Interoperability: Who owns the data? Can patients print it out and share it with their psychologist?

It isn’t clear how this evaluating system will be deployed, however. It certainly isn’t available in the Apple App Store as a way for consumers to judge which mental health app is best for them, and the complexity of the system, which fairly acknowledges the nuances that make an app good for some people and insufficient for others, might require too much for the average user. People interested in using mental health apps currently have little to rely on other than user reviews, which research shows are actually terrible indicators of its efficacy.

Woebot, which wasn’t one of the iPhone apps included in the study, appears to be one of the few ICBT apps that conducts its own peer-reviewed research of its program. Its study, published in 2017 in the peer-reviewed Journal of Medical Internet Research: Mental Health, found that participants using Woebot experienced a greater decrease in depression symptoms than a control group, which was assigned a book about depression published by the National Institute of Mental Health. Among the more surprising findings of the study was that participants found the bot “empathic,” implying that artificially intelligent, nonhuman therapists could play a key role in further development of ICBT apps.

Rück, who is currently conducting a study comparing ICBT without a human therapist to ICBT with a coaching component, says that as of right now, the question of whether or not it’s better to practice ICBT with a therapist is “up in the air.” But, he adds, “it’s not unthinkable that you could, in the future, replace that with some other kind of A.I. bot.”

I personally have found Woebot more engaging than the other apps I tried for this story. Rather than merely completing a worksheet, I chat about whatever is making me feel anxious or depressed that day, and Woebot replies with ways to dismantle my negative thoughts and form newer, more rational versions. It’s a tool I’m familiar with from my face-to-face cognitive behavioral therapy: The goal isn’t to deny the feeling or the thought but rather to transform it into something that’s less harmful and black-and-white.

Woebot is far from perfect. It has typos, and I find myself simplifying my language (and thus the anxiety-inducing situation I’m writing about) to increase the likelihood that it understands me. This is a problem that Athena Robinson, a former associate professor of psychiatry at Stanford University and now Woebot’s chief clinical officer, says the company is working on. Woebot has further research in the pipeline. Right now, the company is partnering with the Lucile Packard Children’s Hospital at Stanford to study how Woebot can help women who have just given birth.

I don’t know if Woebot or another ICBT app would be enough for me on its own. The work I’m able to do with my actual therapist is considerably more nuanced and complex than anything I can do on my phone. But what I like about the app is that it compels me to keep challenging my anxieties and negative thoughts regularly in between therapy appointments. It’s a bit like homework, meant to reinforce what I learned in class.

But so far, the field of mental health apps hasn’t caught up to the research. David Mohr, the director of the Center for Behavioral Intervention Technologies at Northwestern University, says that too few researchers and developers take into account how patients engage with the ICBT programs or what role caregivers, like coaches or therapists, are supposed to fill.

“We are on the verge of a digital revolution in mental health,” he says. “But we won’t get to a revolution without a paradigm shift. And we won’t have a paradigm shift if we keep trying to tie digital mental health to constructs from other domains, such as CBT… I think for digital mental health to become successful, we have to stop thinking about it as ICBT.” He says Mindspot, an Australian company, is a successful example of ICBT practice; it offers anxiety and depression assessments as well as eight weeklong treatment courses for free to any Australian adult not currently in a mental health emergency. Unlike many ICBT programs, the service heavily relies on human coaches.

ICBT holds a lot of promise. Depression is the leading cause of disability worldwide, and mental illness generally costs Americans nearly $200 billion in lost earnings every year. Successful deployment of clinical practice in the real world could change and even save human lives. But for ICBT to have an effect on these statistics, those programs need to be paired with a coach or therapist, and further work needs to be done integrating scientific findings with app development.

In the meantime, there’s probably no harm in downloading an ICBT app and taking it for a spin — just don’t use it as an excuse to skip out on your regular therapist appointment.

I’m a columnist for OneZero, where I write about the intersection of health & tech. Also seen at Elemental, The Atlantic, VICE, and Vox. Brooklyn, NY.

Sign up for Pattern Matching

By OneZero

A newsletter that puts the week's most compelling tech stories in context, by OneZero senior writer Will Oremus. Take a look.

By signing up, you will create a Medium account if you don’t already have one. Review our Privacy Policy for more information about our privacy practices.

Check your inbox
Medium sent you an email at to complete your subscription.

The undercurrents of the future. A publication from Medium about technology and people.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store