Using Passive Sensor Data From Smartphones To Measure Mental Health


Burnout costs companies in the United States $300B annually [1]. 77% of full-time employees report having experienced burnout at least once in their current job [2]. As one of the fastest-growing public health concerns, the World Health Organization, WHO, recently classified burnout as an occupational phenomenon in the 10th release of their International Classification of Diseases, ICD-10 [3]. Studies demonstrate strong correlations [4] [5] between elevated anxiety and depression levels with high Maslach Burnout Inventory, MBI scoring [6].

Studies have shown burnout is a leading cause and correlates strongly with poor mental health. Poor mental health has been proven to increase physical health risks by up to 6.5x [7]. Those with poor mental health can see up to a 10 to 25 year decrease in life expectancy, twice the effect smoking has on longevity [8]. Additionally, poor mental health can increase the risk of musculoskeletal pain by 201% [9] and morbidity risk by 35% in hazardous workplaces [10]. In medicine, it is estimated that 78% of malpractice claims involved a physician who reported burnout symptoms [11] and further research has shown physicians who are burnt out are 2.2x more likely to malpractice [12], costing insurers $3.12B every year in the United States alone.

Products such as Calm [13], Headspace [14] & Happify [15] claim to alleviate mental health and burnout issues. Our previous research and user testing identified a major gap in these products’ long-term efficacy. Products such as Headspace require users to regularly (1) open a mobile app, (2) search through a variety of interventions and, (3) self-select an intervention, or complete a questionnaire for a recommendation. Products such as Calm require surveying of users to identify which intervention should be delivered to the user [Figure 1]. 

Our user studies identified a gap that needed to be filled for the long-term sustainability of positive outcomes to alleviate mental health & burnout issues. For a digital solution to be effective, the user experience must be improved. Passive measurement and detection of mental health scores could fill this gap by removing the onus on the user to seek help, but rather push the right help at the right time.

©2023 Daniel Leung. All rights reserved