Kudos to Sophia Smith, associate professor, and her entire team for the submission of her NIH R21 Subcontract application to University of North Carolina, Chapel Hill, entitled “Digital Phenotyping to Identify Psychological Distress in Adolescent and Young Adult Cancer Survivors." This application requests funding for a two-year period with a start date of July 1, 2021.

While 95% of adolescent and young adult cancer survivors (AYA) are alive five years after their cancer diagnosis, up to 70% experience psychological distress from depression, anxiety, and post-traumatic stress symptoms. Distress is highest among groups with lower access to healthcare—those who are young, single, low-income, or unemployed. Left untreated, distress contributes to social, educational, and occupational difficulties, lower quality of life, negative health outcomes impacting survival, and higher healthcare costs. Simultaneously, the high AYA survival rate translates to a compounding of negative distress-related outcomes spanning decades. An accessible solution could have widespread benefit.

Research has focused on self-guided mobile health (mHealth) apps to overcome barriers. A mHealth solution is particularly attractive for reaching AYAs since over 92% of those aged 18 to 39 in the US own a smartphone, and many younger, lower-income, and non-white Americans depend on it as their only means of internet access. However, most mHealth apps fail to sustain engagement among users as only 4% continue to use them beyond the first 15 days, due in part to high user burden. A convenient, unobtrusive method for identifying distress outside the clinic could be a pathway to improving quality of life. To promote symptom identification and minimize burden, our unique approach to mHealth incorporates digital phenotyping. This approach uses information passively collected from smartphone logs and sensors (e.g., typing speed, words, usage patterns) as ‘digital biomarkers’ to identify clinically meaningful behavioral anomalies and trends.

This project will use digital phenotyping to develop an effective distress-monitoring app named DistressID. We will identify digital biomarkers correlating with psychological distress scores and use those to define the DistressID model to trigger active screening. We will then validate those biomarkers in a randomized controlled trial. This project addresses a critical gap in successfully mitigating psychological distress by leveraging widespread smartphone use for remote monitoring.

Successful completion of this study will advance the science of digital phenotyping and provide the foundation for future studies to use digital phenotyping to guide interventions for mental health. Without this innovative research, the gap in successfully using mHealth tools to address psychological distress is likely to continue.