Using Multi-modal Assessments to Capture Personalized Contexts of College Student Well-being in 2020: A Case Study.

BACKGROUND 2020 has been a challenging year for many, particularly for young adults who have been adversely affected by the COVID-19 pandemic. Emerging adulthood is a developmental phase with significant changes in the patterns of daily living; it is a risky phase for the onset of major mental illness. College students during the pandemic face significant risk, potentially losing several protective factors (e.g., housing, routine, social support, job and financial security) that are stabilizing for mental health and physical well-being. Individualized multiple assessments of mental health, defined as multimodal personal chronicles, present an opportunity to examine indicators of health in an ongoing and personalized way using mobile sensing devices and wearable internet-of-things. OBJECTIVE To assess the feasibility and provide an in-depth examination of the impact of the COVID-19 pandemic on college students through the utility of multimodal personal chronicles, we present a case study of an individual monitored using a longitudinal subjective and objective assessment approach over a nine-month period throughout 2020, spanning the pre-pandemic period of January through September. METHODS The individual completed psychological assessments measuring depression, anxiety, and loneliness across the 4 time points of January, April, June, and September. We used the data emerging from the multimodal personal chronicles (i.e., heart rate, sleep, physical activity, affect, behaviors) in relation to psychological assessments to understand patterns that help to explicate changes in the individual's psychological well-being across the pandemic. RESULTS Over the course of the pandemic, the individual's depression severity was highest in April, shortly after shelter-in-place orders were mandated. His depression severity remained mildly severe throughout the rest of the months. Associations in their positive and negative affect, physiology, sleep, and physical activity patterns varied across time periods. Lee's positive and negative affect were positively correlated in April (r = .53, P = .04) whereas they were negatively correlated in September (r = -.57, P = .03). Only in the month of January was sleep negatively associated with negative affect (r =-.58, P = .03) and day BPM (r = -.54, P = .04), and then positively associated with resting RMSSD (r = .54, P = .04). When looking at his available contextual data, the individual noted certain situations as supportive coping factors and other situations as potential stressors. CONCLUSIONS We observe more pandemic concerns in April, and notice other contextual events relating to this individual's well-being, reflecting how college students continue to experience life events during the pandemic. The rich monitoring data alongside contextual data may be beneficial for clinicians to understand client experiences and offer personalized treatment plans. We discuss benefits as well as future directions of this system, and the conclusions we can draw regarding the links between the COVID-19 pandemic and college student mental health. CLINICALTRIAL