Digital Phenotyping: How Our Phones Can Monitor Psychiatric Disorders

Over the past few decades, the number of adolescents with mental illnesses has grown at an alarming rate [1]. What if there was a way to help track and diagnose psychiatric disorders with the very thing that most adolescents could not go a day without: their phones? Digital health monitoring is no new concept in our daily lives; many people use Apple Watches to keep track of their physical and cardiac activity [2]. Digital phenotyping, even more discreetly than other methods of digital health monitoring, can track your behavior, from how many texts you send to what you are commenting on Instagram. Although fraught with potential ethical implications, digital phenotyping is making it possible to learn about and continuously monitor behavior associated with certain mental illnesses, allowing our smartphones to become an invaluable tool in psychiatry.

What is digital phenotyping?

Digital phenotyping collects data from a device, such as a smartphone, to measure an individual’s behavior, cognition, and mood. While the concept has existed for quite some time now, the potential is now being realized with the proliferation of personal computing devices in our daily lives [3]. The data collected for digital phenotyping can be “content-free”, such as tracking reaction times for tapping and scrolling, or “content-rich”, like monitoring geolocation, search history, or social media posts [4]. 

Digital phenotyping falls within the field of telemedicine, in which patients are monitored via continuously collected health measurements (e.g., heart rate, blood pressure). Telemedicine can present as wearable digital technology that tracks movement (e.g., step count) and physiological/biochemical processes (e.g., blood pressure and heart rhythm). Since the COVID-19 pandemic, the integration of remote health monitoring has become increasingly relevant and useful. Telemedicine is now penetrating the field of psychiatry [5]. Currently, psychiatric diagnostic practices involve self-report (e.g., questionnaires and momentary assessments) and biological measures  (genetic testing, blood tests for protein levels)  [6, 7]. However, these practices typically cannot give the full picture of a complex psychiatric disorder like digital phenotyping can.

Application of Digital Phenotyping in Bipolar Disorder

Digital phenotyping of smartphone data can be used to detect and diagnose psychiatric illnesses such as Bipolar Disorder [8]. Bipolar disorder is a common mood disorder in which an individual experiences alternating hypomanic/manic and depressive episodes. Hypomania and mania are characterized by an increase in activity, while depression presents as lower energy and depressed mood [9]. One example of a smartphone phenotyping software that can be used in psychiatry is the MONARCA I system (MONitoring, treAtment, and pRediCtion of bipolAr disorder episodes system). The MONARCA I system automatically collects real-time self-monitoring data, such as the number and length of incoming and outgoing calls per day. Scores on the Hamilton Depression Rating Scale (HDRS-17), commonly used to diagnose in-person patients with Bipolar disorder or depression, are positively correlated with the duration of incoming and outgoing calls per day and negatively correlated with self-reported data on mood [8]. Scores on The Young Mania Rating Scale (YMRS), another method of diagnosing Bipolar disorder, exhibited a positive correlation with the number and duration of incoming calls per day and with self-monitored mood data [8]. In this way, smartphone phenotyping software similar to the MONARCA I system can be used to detect, diagnose, and predict episodes of psychiatric illnesses such as Bipolar Disorder. 

Ethics of Technology in Psychiatry

The intersection of healthcare, commercial, and government domains in digital phenotyping positions this technology beyond existing ethical and regulatory frameworks. For this reason, accountability regarding digital phenotyping must be evaluated before implementation on a large scale [4]. The undefined domain that digital phenotyping lies in raises concerns when considering personal data collection. Despite the collection of digital behavioral data for health purposes, it is challenging to claim that tracking phone calls falls under the protection of existing health standards, like HIPAA, since data collection occurs outside of a clinical setting  [4]. 

It is also important for users to understand that their personal information is being collected through digital phenotyping. Information regarding mental illness has the potential to impact employment, insurance, and personalized marketing, such as targeted ads on Facebook [4]. Therefore, digital phenotyping programs must ensure that user data is sufficiently protected, whether the data be content-free or content-rich. Additionally, patients should be made aware of how, when, and what data is being collected in an accessible way [4]. Transparency from the side of the patient is equally important as transparency with clinicians; software developers should reveal the effectiveness and limitations of their technology to clinicians to ensure digital data is properly being translated into clinical practice [4]. Much of this boils down to the need for improving informed consent practices throughout the process of digital phenotyping.

Going Forward

The novelty of digital phenotyping in the psychiatric setting raises the need to address some ethical concerns regarding the convergence of technology and healthcare. As long as further development within this field takes into consideration accountability, protection of user data, transparency, and informed consent, digital phenotyping has the potential to change how we diagnose and treat psychiatric disorders. The implementation of digital phenotyping software is cost-effective and widely accessible when considering how many people already have smartphones that could support this technology. In the coming years, we may start to see software like MONARCA I being promoted by psychiatrists and downloaded onto our smartphones as we enter a new age of psychiatric care..

References

1. Piao, J., Huang, Y., Han, C., Li, Y., Xu, Y., Liu, Y., & He, X. (2022). Alarming changes in the global burden of mental disorders in children and adolescents from 1990 to 2019: A systematic analysis for the global burden of disease study. European Child & Adolescent Psychiatry,31(11), 1827–1845. https://doi.org/10.1007/s00787-022-02040-4 

2. Lui, G. Y., Loughnane, D., Polley, C., Jayarathna, T., & Breen, P. P. (2022). The Apple Watch for Monitoring Mental Health-Related Physiological Symptoms: Literature Review. Jmir Mental Health, 9(9), 20, Article e37354. https://doi.org/10.2196/37354

3. Baker, J. T., Germine, L. T., Ressler, K. J., Rauch, S. L., & Carlezon, W. A. (2018). Digital Devices and continuous telemetry: Opportunities for Aligning Psychiatry and neuroscience. Neuropsychopharmacology, 43(13), 2499–2503. https://doi.org/10.1038/s41386-018-0172-z

4. Martinez-Martin, N., Insel, T. R., Dagum, P., Greely, H. T., & Cho, M. K. (2018). Data Mining for Health: Staking out the ethical territory of digital phenotyping. Npj Digital Medicine, 1(1). https://doi.org/10.1038/s41746-018-0075-8

5. Sakamaki, T., Furusawa, Y., Hayashi, A., Otsuka, M., & Fernandez, J. (2022). Remote Patient Monitoring for Neuropsychiatric Disorders: A scoping review of current trends and future perspectives from recent publications and upcoming clinical trials. Telemedicine and E-Health, 28(9), 1235–1250. https://doi.org/10.1089/tmj.2021.0489

6. Huckvale, K., Venkatesh, S., & Christensen, H. (2019). Toward clinical digital phenotyping: A timely opportunity to consider purpose, quality, and safety. Npj Digital Medicine, 2(1). https://doi.org/10.1038/s41746-019-0166-1

7. García-Gutiérrez, M. S., Navarrete, F., Sala, F., Gasparyan, A., Austrich-Olivares, A., & Manzanares, J. (2020). Biomarkers in Psychiatry: Concept, Definition, Types and Relevance to the Clinical Reality. Frontiers in Psychiatry, 11, 14, Article 432. https://doi.org/10.3389/fpsyt.2020.00432

8. Faurholt‐Jepsen, M., Vinberg, M., Frost, M., Christensen, E. M., Bardram, J. E., & Kessing, L. V. (2015). Smartphone data as an electronic biomarker of illness activity in bipolar disorder. Bipolar Disorders, 17(7), 715–728. https://doi.org/10.1111/bdi.12332

9. Smith, D. J., Whitham, E. A., Ghaemi, S. N., Aminoff, M. J., Boller, F., & Swaab, D. F. (2012). Chapter 15 - Bipolar disorder. In Handbook of Clinical Neurology (Vol. 106, pp. 251-263). Elsevier. https://doi.org/https://doi.org/10.1016/B978-0-444-52002-9.00015-2

10. Shen, F. X., Silverman, B. C., Monette, P., Kimble, S., Rauch, S. L., & Baker, J. T. (2022). An ethics checklist for Digital Health Research in psychiatry: Viewpoint. Journal of Medical Internet Research, 24(2). https://doi.org/10.2196/31146