Can computer power in apps augment clinical treatment for psychiatric disorders?

Mobile apps can potentially tap into the computer processing power of smartphones to crunch data in real time during therapy sessions. But do any apps currently do that?
I and my colleague published a literature review to survey how many apps are designed to augment one or more of four common elements of clinical treatment for psychiatric disorders: 1) assessment/diagnosis, 2) treatment planning, 3) treatment fidelity monitoring, and 4) outcome tracking (Pacheco and Scheeringa, 2022).
There is considerable interest in this question because the assessment and treatment of psychological conditions can be complex, and the competency of clinicians to perform these with reliability has been questioned for a variety of reasons. To be fair, no two patients are alike, and therapists who, by nature, are skilled in empathic listening and complex relational dynamics are not always skilled or enthusiastic about working with numbers. Perhaps computers can help.
In our review, we analyzed all commercially available mobile and computer applications (apps) focused on treating psychiatric disorders. Searches of the Google Play Store, the Apple App Store, and the One Mind PsyberGuide found 676 apps in English labelled for mental health use (as of November 2020), of which 513 apps were mostly wellness apps that solely focused on single techniques such as meditation, journaling, or trackers for thoughts or emotions. The remaining 163 apps were judged by us to address one or more of the four common clinical elements, of which 59 had a computer-driven function for at least one of the four main elements of clinical treatment.
Our definition of computer-driven processing was that data were processed in a way that would be difficult or time-consuming for clinicians. For example, apps that simply summed short (less than ten items) clinical measures did not meet this definition and were excluded. Apps that summed longer measures and multiple types of measures and provided interpretations of the scores were included.
Telehealth was excluded because it is simply a different interface between therapists and clinicians. Telehealth, by itself, does not use computer processing power to perform one of the four elements of treatment that were our focus.
 
Our findings:

1. The most common element that employed computer-driven processing was assessment/diagnosis (55/59 apps)
   
2. This was followed by outcome tracking (34/59 apps)
   
3. Six apps addressed treatment planning by updating treatment plans using user input.
   
4. Only one app tracked treatment fidelity.
   
5. None of the apps contained computer-driven functions for all four elements. All 59 apps are described in a table in our paper (link below).

 
We also found that twelve apps were supported in randomized clinical trials to show greater efficacy compared to either wait list or other active treatments. That is encouraging, but it is not, however, our opinion that randomized trials are very enlightening  for judging apps because nearly all apps are simply digitizing methods that have already been proven effective.
In summary, results showed that these four clinical elements can be implemented through apps, but the full potential of computer processing appears unreached in mental health-related apps.
 
Michael Scheeringa, MD, founder
 
Postscript: The Scheeringa Mind Company’s Treatment Fidelity app was not available at the time of this literature review. It is the first known app to use computer-driven processing for all four elements. It is also the only app that can monitor treatment fidelity for CBT protocols, and can be modified for nearly any therapeutic technique.
 
REFERENCES
Pacheco CR, Scheeringa MS (online 8/19/22) Clinical wisdom in the age of computer apps: A systematic review of mental health apps. the Cognitive Behaviour Therapist 15:e40  doi:10.1017/S1754470X22000368.