Introduction

Increasing access to care for people with schizophrenia remains a global health priority1. Given their scalable potential, digital mental health and especially smartphone-based solutions have become a topic of great interest2. In the last decade, research on smartphone apps for people with schizophrenia has expanded, with results suggesting uses cases around early diagnosis, real-time monitoring, psychoeducation, relapse prevention, and even therapeutic interventions3,4,5. But to what degree has this research translated into digital tools that patients today can access and utilize? This paper thus aims to review both the research and commercial marketplaces for apps for people with schizophrenia and explore their availability and accessibility.

Smartphone apps represent an important solution to increasing access to care for people with schizophrenia. A preponderance of data suggests that not only do people with schizophrenia own smartphones at high rates but that they also are interested in using them as part of their recovery5,6,7,8,9,10. Despite often raised questions by the public regarding whether smartphone-based monitoring or interventions may make people with schizophrenia more paranoid, there remains scant data to support this unfounded claim11. Rather, people with schizophrenia have themselves been at the forefront of developing and researching new advances in the uses of smartphones for their condition12,13.

The driving force behind the interest in smartphones centers on increased access to care. The simple reality of the lack of a mental health workforce able to deliver care to people with serious mental illness was clear before COVID-19, and now is even more apparent14. Given the enormous mental health gap in services, impacting all countries in the world, digital solutions that are scalable represent a critical target for increasing access to care15. As the unmet need for care for people with schizophrenia is at the higher end of the spectrum, the concomitant interest in smartphone apps for this condition is clear. Apps have been proposed as tools to help with screening and monitoring of schizophrenia as well as tools to offer on-demand as well as just-in-time adaptive interventions16,17; and an active area of research for the last ten years.

Beyond access, interest in these apps is also fueled by the potential of smartphone apps to deliver more holistic and eclectic treatments beyond those readily accessible today. For example, smartphone apps can facilitate cognitive remediation treatments, peer support, cognitive behavioral therapy for psychosis, and other services that may be challenging for patients to find today even in well-resourced countries5.

However, despite this potential, it is also clear from evidence across all of digital health that smartphone apps alone are not a panacea. Concerns around privacy, effectiveness, engagement, and clinical integration are now well-recognized barriers for all health-related apps, including mental health apps18. Even in 2022, mental health app privacy concerns continue to make national news19 and there is rising awareness for high quality studies that assess the impact of these apps against appropriate control conditions20,21. While practical experience suggests that apps for people with schizophrenia have not yet transformed care in 2022 and that there is no well-defined or practical distinction between clinician-prescribed apps and self-prescribed apps related to schizophrenia, little is actually known about the current state of the field or the availability of apps for practical patient use.

To understand the current state of apps for schizophrenia, this review aims to catalog these apps developed in the research space based on platforms and assess their current availability. In parallel, this review also aims to investigate the quality of apps currently offered on the Apple and Android marketplaces and assess any overlap or differences in features between the research apps. Towards understanding factors that may impact availability, this review focuses on apps as platforms rather than the clinical results of any one study.

Methods

Research articles search and a literature review

To analyze the current research on smartphone apps for individuals with psychosis, we searched recent reviews22,23, the gray literature, and standard academic databases including PubMed and Google Scholar. This intentionally surface-level search strategy of the academic literature was intentionally employed as a primary purpose of this dual review (which also searched app stores, see below) was to assess and compare the apps used in the most easily-identified research studies, to those available on publicly-accessible app marketplaces.

A search of PubMed and Google Scholar was performed on August 17, 2022, using the following search algorithm: (“smartphone*“ or “mobile phone*“ or “cell phone”)) AND (“app” or “apps” or “application” or “applications”)) AND (“schizophrenia” or “schizo” or “psychosis” or “psychotic”). However, a primary purpose of the review was to gain an assessment of the extent to which apps in readily-accessed research studies are aligned with those apps widely available on the marketplace, and there is no reason to believe that any missed studies more difficult to identify would significantly impact our data collection.

A total of 309 articles were revealed. After removing duplicates, a total of 264 articles remained.

Two authors (SK and DJ) screened each title/abstract for eligibility using the Covidence systematic review management tool24. Articles were excluded if they (1) delivered an intervention that is not mobile app-based (2) were unrelated to psychosis.

A total of 166 articles were reviewed in full text by 3 authors (SK, DJ, and JT) using the following inclusion/exclusion criteria: articles were excluded if they (1) utilized the same research app as other articles (2) were review articles (3) were non-specific to psychosis (4) were commentary or perspective articles (5) were conference abstracts (6) used the apps not intended for patients’ use, (7) were primarily focused on supporting different condition(s) such as smoking, or (8) were primarily focused on digital antipsychotic medication with the smartphone app only supporting the digital medicine. Consequently, a total of 60 research apps were included (Fig. 1). In this article, the term “research apps” will be used solely when referring to the mobile apps utilized by the academic article.

Fig. 1
figure 1

Flow diagram of research articles screening.

Full size image

Public app marketplace mobile application search and an app audit

To analyze the current marketplace for psychosis apps, a search of the Apple App Store and Google Play App Store was conducted on July 24, 2022. There were no inclusion criteria for the commercial mobile application search since the aim of this study was to determine the most accessible and easily obtainable psychosis apps for a layperson. The terms “Schizophrenia” and “Psychosis” were entered. A total of 675 apps returned, which was reduced to 537 apps after duplicates across Apple and Android were eliminated. In the first phase of screening, returned apps were excluded if they (1) were not available in English (2) developed to include contents unrelated to mental health (eg dating apps) (3) were non-specific to schizophrenia or psychosis (4) claimed to primarily support different condition(s) (5) intended for test preparation (6) intended for non-patient use, and (7) cost more than $10.00 to download; this price limit was set on the premise that an app should be economically accessible and determined based on our team’s clinical experience and patient interactions.

Subsequently, a total of 512 apps were excluded. Two authors (SK and JT) downloaded and used the remaining 25 apps and reviewed each app for approximately 30 min. These apps were then further screened and excluded if they (1) contained outdated information as determined by a patient advisory panel (2) were non-functional (3) required access code, and (4) contained potentially dangerous or stigmatizing information as determined by a patient advisory panel. Outdated information included references and diagnosis excluded from the DSM-5. Potentially dangerous or stigmatizing information included home remedy recommendations without a proper side effects warning and a phrase that perpetuates negative labeling and perception. Subsequently, a total of six marketplace apps were included. The term “marketplace apps” will be utilized exclusively when referring to the psychosis mobile apps returned on the public app marketplace.

Therapeutic, monitoring, and psychoeducation (TMP) classification

There is no established nosology for the categorization of the main features or functionalities of research apps and marketplace apps. Thus, we opted to categorize these features or functionalities broadly into three categories: therapeutic, monitoring, and psychoeducation. For ‘therapeutic’ we counted any features that aim to improve symptoms, behavior or cognitive functioning, such as psychotherapy, skills training, peer support, and tailored daily activities. For “monitoring” we counted any features that help patients to track symptoms, treatment progress, or medications. For ‘psychoeducation’, we counted any features that offered reference or didactic information. Using this nosology, it is possible for the same app to offer therapeutic, monitoring, and psychoeducation functionalities.

Specific platform and non-specific platform

We further classified apps as built or running on specific vs non-specific platforms. A specific platform would be a custom app designed to run only that program. For example, an app built on specific platform would only provide psychoeducation regarding schizophrenia. A non-specific app may be a broad survey platform customized with relevant clinical content or an app platform customized to support specific content. For example, a non-specific app could be a survey-administering app that brings up different question sets based on a set-specific code. A researcher may create a set of questions relevant to schizophrenia patients and administer those questions using a non-specific app.

Data analysis process

A total of 60 research articles and their respective research apps were analyzed, and the data extracted including availability on the public app marketplace, access code requirement, supporting study authors and year, research app example features, TMP framework categorization, and specific vs non-specific platform. All research apps available for download were assessed to evaluate accessibility and categorized as restricted access (requires access code) or full access (does not require an access code).

A total of six marketplace apps were analyzed. Data analyzed included the date of last update, app descriptions, and the TMP framework categorization. The six marketplace apps were also rated based on the M-health Index and Navigation Database (MIND) derived from the American Psychiatric Association’s app evaluation framework25,26. MIND is the largest public database that allows any users to make an informed decision in choosing a mental health app. The rating included 105 objective questions pertaining to app origin, functionality, and accessibility; privacy and security; evidence and clinical foundations; features and engagement; interoperability and data sharing27.

Statistical analysis

The download availabilities of specific and non-specific research apps were compared using the chi-square test. Statistical significance was defined as P < 0.05.

Results

A total of 537 unique marketplace apps were identified on the United States Apple App Store and Google Play App Store when the terms “Schizophrenia” and “Psychosis” were searched. As shown in Fig. 2, 256 of 537 apps (47.7%) were not related to psychosis, and 32 of 537 apps (6.0%) were not developed to be used by individuals with psychosis. There were two apps (0.4%) that specifically stated they supported psychosis in addition to two or more different conditions, including eating disorder, PTSD, generalized anxiety disorder, plus 27 other conditions, and 194 apps (36.1%) that claimed to support mental health wellness or provide information about mental health. Only 25 apps (4.7%) claimed to primarily support psychosis. However, of the 25 apps, one app was non-functional, seven were inaccessible without an access code, three apps contained outdated information such as reference to non-current diagnosis like Disorganized, Catatonic and Undifferentiated schizophrenia which have been removed from the DSM-5 for over five years. Eight apps contained stigmatizing or dangerous information such as telling users to “remember that it’s [schizophrenia] all in your head” or providing herbal supplement advice without discussing potentially dangerous medication interactions. Outdated information included references and diagnoses that had been excluded from the DSM-5, and dangerous information included a list of medications that can cause harmful drug interactions with antipsychotics. The accessibility to obtaining an access code was limited as users had to either (1) contact the app developer (2) contact the research group (3) partake in the research study or (4) be a patient at a specific clinic. Thus, a total of six easily accessible, appropriate, and psychosis-specific marketplace apps (1.1%) were evaluated using the MIND framework (Fig. 2).

Fig. 2
figure 2

Flow diagram of marketplace apps screening.

Full size image

Marketplace App

Of the six easily accessible, appropriate, and psychosis-specific marketplace apps, five apps solely presented psychoeducation (83.3%). Only one app, which was last updated 1587 days ago (4.34 years ago), offered therapeutic and monitoring functionalities without providing psychoeducation. All six apps were not updated frequently with the average time since last updated being 1121 days (3.07 years ago)—well exceeding the 180 days update metric adopted by the American Psychiatric Association, which prompts a concern for an app quality and safety28 (Table 1).

Table 1 Accessible, appropriate, psychosis-specific marketplace apps.
Full size table

All six of these apps were free to download with one offering in-app purchases (16.7%) and five totally free (83.3%). One of the six apps was available on the Apple App Store, and all six apps were available on the Google Play App Store. Of the six apps, four apps had a privacy policy and only one app allowed users to delete data. Moreover, only one app allowed users to opt-out of data collection, and none of the apps claimed to meet Health Insurance Portability and Accountability Act, 1996 (HIPPA) – established to protect an individual’s protected health information29. Only one app included a published supporting study examining the app’s feasibility30.

Research App Feature

Of the 60 unique academic articles on smartphone apps for psychosis, 31 articles utilized apps designed with a psychosis-specific platform, and 29 articles used apps built on a non-specific platform to deliver psychosis-specific features. 50 of the 60 research apps provided symptom monitoring features while 30 research apps exclusively offered monitoring features, 14 research apps offered therapeutic and monitoring features, 3 research aps offered monitoring and psychoeducation features, and 3 research apps offered therapeutic, monitoring, and psychoeducation features, see Tables 2, 3. Although nine research apps included psychoeducation in addition to monitoring or therapeutic features, no research app solely provided psychoeducation (Tables 2, 3).

Table 2 Studies of psychosis research apps with psychosis-specific platform.
Full size table
Table 3 Studies of psychosis research apps with non-specific platform.
Full size table

Research app accessibility

Regarding research app accessibility, 31 of the 60 research apps were not available on the public app marketplace. The remaining 29 research apps were available to be downloaded for free; however, 20 of the 29 research apps required an access code or special credential to access the app features. Thus, only nine research apps were available and easily accessible on the public app marketplace.

Research apps with psychosis-specific platform and non-specific platform differed significantly in their download availability on the public app marketplace (χ2 = 4.241, p = 0.039), with 64.5% of psychosis-specific platform apps and 37.9% of non-specific platform apps unavailable to be downloaded. Subsequently, research apps with non-specific platform had significantly higher download availability on the public app marketplace compared to that of research apps with psychosis-specific platform. However, both research apps with psychosis-specific platform and non-specific platform had limited accessibility with 29.0% and 37.9% of the apps requiring an access code, respectively (Table 4).

Table 4 Feature type and availability of psychosis research apps.
Full size table

Discussion

Study results suggest that while many academic apps have been developed to support people with schizophrenia, very few suitable apps are actually available for use today. The situation is related to the commercial marketplaces where hundreds of apps are returned in searches but there are less than ten that are accessible and clinically relevant. Despite impressive research efforts, people with schizophrenia are today able to access a paucity of apps. A focus on translating research efforts into accessible apps should become the priority for the field.

We identified sixty unique apps used in schizophrenia research. Of these 60 apps, 31 (51.7%) were not available at all meaning that it is challenging to replicate or expand on their results. This result is perhaps not surprising given the challenges and costs associated with maintaining apps. For example, Krzystanek, a developer of MONEO app included in the results31,32, explains “the investor who wanted to commercialize it [the MONEO app] was not up to the task, so the project was suspended.”33 However, our results do suggest one potential solution as we found that apps developed on non-specific platforms were significantly more available today than those created on customized app platforms (χ2 = 4.241, p = 0.039). While a customized app offers clear advantages, using a broader platform may offer a more rapid and sustainable approach, especially for early phase work.

While there are many reasons a research app may not be available for easy public use, our result that 20 (33.3%) were also not accessible as they required an access code is notable given the goal of most apps is to increase access to care. While our results cannot directly support why, it is likely that these apps themselves are necessary to use in concert with a care program and not as standalone self-help tools. Using apps to augment care can certainly help increase access to care, but raises the need for concomitant training, workforce, implementation, and clinical infrastructure to support scalability. In reviewing the relevant literature, we found such documentation was often lacking although it represents an important target for new and ongoing app efforts.

Of the 60 research apps, only nine (15.0%) research apps were available to download and directly use. However, these results must be interpreted with caution as only two of these nine apps was created on a schizophrenia specific platform and the other seven on non-specific app platforms. Examples of non-specific platforms include WeChat, MovisensXS, and mindLAMP (developed by the authors) meaning that a patient or clinician would need to customize or add content to the platform for it to be ready for clinical use. This may often involve a licensing fee depending on the app or at least some degree of technical knowledge, reflecting a further barrier to access.

While there is room to improve accessibility of research apps, our results suggest this work is necessary as the current marketplace offerings are concerning. After an exhaustive search of apps on the Apple App Store and Google Play App Store, we found only 25 psychosis-specific marketplace apps from the 537 our search revealed. Even though there exist other curated third-party health app marketplaces (e.g., Psyberguide34) that provide information regarding app availability, there is no reason to believe that any missed marketplace apps would change our results. Of these 25 apps, only six were deemed clinically appropriate with the other 19 offering inaccurate, outdated, stigmatizing, or inappropriate content. For example, one app tells users to “remember that it’s [schizophrenia] all in your head.”35 Two apps recommend “St John’s Wort [which] works as an antidepressant in patients with schizophrenia disorder”36,37 as a part of the home remedies without clear warnings of dangerous medication interactions38. Three apps offered psychoeducation about “Disorganized, Catatonic and Undifferentiated schizophrenia”39,40,41 which have been removed from the DSM-5 for over five years42.

Considering the six (1.1% of 537 marketplace app) that were deemed clinically appropriate, five apps (only available on Google Play App Store) solely provided information/psychoeducation, and only one app (available on both Apple App Store and Google Play App Store) included therapeutic and monitoring features. The paucity of apps with therapeutic features may be that apps with therapeutic or diagnostic features are subject to Food and Drug Administration’s regulation as those apps do not fall under the “wellness” app category43. However, all six apps are not currently updated with an average time since last updated being 1121 days (3.07 years ago). Of note, while these apps were not updated recently, the psychoeducation they provided was so general that the content was not out of date. Still, considering that these are the apps most readily accessible to people with schizophrenia today it is clear that the potential of apps to increase access to care has yet to be fulfilled. Furthermore, the scarcity of clinically appropriate schizophrenia apps on Apple App Store compared to Google Play App Store raises a concern as this disparity can potentially induce health inequalities.

Our results suggest clear and tangible next steps. For research to accelerate to increase access to care, it is necessary to either build apps on non-specific platforms or to make apps built on specific platforms directly available for people to use. While advanced apps are being studied, there is a clear need to provide simple but up to date apps directly on the Apple and Android marketplaces so that people with schizophrenia can at least benefit from higher quality psychoeducation and app offerings.

Our results also suggest a broader question around when an app is necessary for schizophrenia specifically versus when a more transdiagnostic app will suffice. For example, medication tracking is a common feature across nearly every health condition and the added value of a schizophrenia specific vs general medication tracker remains an open question. Similarly, apps which help support people to engage in regular physical activity, adopt healthy eating patterns, quit smoking or even or manage common physical health conditions such as type-2 diabetes are theoretically just as important (or more so) for use in schizophrenia as the general population44. In our team’s work with the MIND website which evaluates apps, we often find ourselves helping people create toolkits of apps that utilize useful features across a range of apps to create the right set of resources for each patient. Research onto a common or transdiagnostic set of app functions for mental health, including schizophrenia, could help focus schizophrenia research on the unique challenges of the field and avoid unnecessary duplication of work.

Our findings expand on prior research results that support our results. Prior studies from our team have examined the top 10 apps for schizophrenia and found similar results28,45, but those studies did not exhaustively search for every app as we did here. Another review found that the number of commercially available apps with academic evidence from across all health fields is scarce but did not focus on schizophrenia and is based on older 2017 data46. Our results complement past reviews of research apps for people with schizophrenia22,23 that also found few available for patients although here we found nearly twice the number of research apps as compared to prior studies – perhaps reflecting accelerating interest in this space.

Like all reviews, ours has several limitations. We only search for research papers in English and are aware that given the many diverse names used to characterize this research, our search term may have missed some studies. However, a primary purpose of the review was to gain an assessment of the extent to which apps in readily-accessed research studies are aligned with those widely available on the marketplace, and there is no reason to believe that any missed studies more difficult to identify would reveal a preponderance of research apps in a way that would change our results. Our marketplace search was exhaustive, but we also realize searches from different regions (ie outside of the United States) may yield different results such as emerging work from China47. A search with more keywords may have included more apps but given only 1.1% of the returned results were relevant, we felt our search was already comprehensive.