Every caregiver of someone with a serious mental illness knows the moment: You’re searching an app store at 2 a.m., looking for something, anything, that might help your loved one track symptoms, stay on a medication schedule, or just feel less alone between crises. You download it. You show them how to use it, gently, because trust is already fragile. And if you’re lucky, it becomes part of the routine that keeps things stable.
A new study should make every clinician and caregiver think twice before that recommendation. John Torous and colleagues, led by Julian Herpertz, tracked 865 mental health apps listed on MindApps.org over roughly six years and just published the first large-scale survival analysis of this market in the American Journal of Psychiatry. More than half (54 percent) were removed. The median app lasted about three years. And the apps that disappeared fastest were the ones built for schizophrenia, mood disorders, and smoking cessation: precisely the conditions where losing a familiar tool isn’t an inconvenience, it’s a rupture in continuity of care.
Here’s the part that should reframe how we all think about digital mental health tools: The features that get the funding, the press releases, and the pitch-deck slides had nothing to do with whether an app survived. In-app chat, video, peer support, real-time responses, gamification, AI capabilities, even having a randomized controlled trial behind it: None of it predicted longevity. What did predict survival were the unglamorous things: journaling, goal-setting, letting users export or own their own data, being available in Spanish, and working across both iOS and Android instead of locking users into one platform. Those features cut the risk of an app disappearing by 30 to 50 percent. Being platform-exclusive more than doubled it. And in one of the study’s more uncomfortable findings, simply having a privacy policy was associated with higher attrition, while actually giving users ownership of their data was protective.
The authors’ interpretation is worth sitting with: Engagement features cost money to host, moderate, and staff, and venture-backed developers build them anyway because that’s what attracts the next funding round, even though high failure rates come with the territory. Journaling and goal-setting cost almost nothing to maintain. Survival, in other words, tracks a company’s business model and operational discipline far more than it tracks whether the app actually works or feels good to use. As Torous put it, “An engagement layer or an AI layer is really not a sustainability plan.”
Real physician voices, twice a week
Free, and one click to unsubscribe.
I’d push that one step further, from where I sit. A companion analysis the same research team published earlier this year in World Psychiatry (drawing on the same MIND database) found that among the apps currently listed, none (zero) specifically target schizophrenia. Only six target bipolar disorder. And every schizophrenia-specific app that had ever been indexed in that database was, by the time of the analysis, eventually archived. Fewer than 14 percent of all listed apps report any supporting research at all. Non-commercial developers (academic, government, nonprofit) were significantly more likely to handle data responsibly and use evidence-based practices than the for-profit companies that make up nearly 90 percent of the market. And even as more apps now post privacy policies than a few years ago, more of them are also sharing personal health data with third parties than before.
Put plainly: The population least equipped to absorb a tool’s disappearance (people managing psychosis, often distrustful of anything new, for whom rebuilding routine and rapport takes real effort) is the population the market is least built to serve for the long haul. Schizophrenia isn’t an underserved niche in this data. It’s functionally unserved, on a rolling basis, by design.
This matters for how clinicians counsel families, and it matters for how caregivers like me evaluate what we hand our loved ones. A feature list was never the right question. The right questions are: Who built this, how do they make money, do I own my data if the app vanishes tomorrow, and does this company have a track record of still existing in three years? A nonprofit or academically backed tool with a plain interface and no chatbot may be a far safer bet for a family managing schizophrenia than the sleek, AI-forward app with the venture funding and the six-month runway.
It also matters for anyone funding or regulating this space. If longevity is a prerequisite for clinical benefit (and it is, because a tool that vanishes mid-crisis helps no one), then grant programs, health systems, and app-evaluation frameworks need to start weighing a developer’s sustainability and data practices as seriously as they weigh features and evidence. Right now, the apps most likely to survive and the apps most likely to get funded are not the same apps. Families with the most at stake are the ones absorbing that gap.
I’ll keep recommending technology to the caregivers I work with. But I’ll be asking a different question before I do.
Nicole Drapeau Gillen is a mother, advocate, and author who translates the fast-moving landscape of technology in serious mental illness (SMI) care into guidance families and clinicians can use. Thrust into caregiving for a loved one with SMI, with no direction on how to help, she turned that experience into a mission, writing two books and building an ongoing effort to bring families and clinicians into the conversation.
Her first book, Schizophrenia and Related Disorders: A Handbook for Caregivers, is a reference for every stage of caregiving, endorsed by Dr. E. Fuller Torrey as a must-read for SMI caregivers. Her second, Connected Care: A Practical Guide to Technology for Serious Mental Illness, maps apps, artificial intelligence tools, telepsychiatry, and brain-based treatments for a field moving faster than anyone can track. Dr. Akira Sawa, director of the Johns Hopkins Schizophrenia Center, has said the book “directly addresses” significant gaps.
