Harnessing Context Sensing to Develop a Mobile Intervention for Depression

• DOI: 10.2196/jmir.1838
• Publication Date: 2011-08-12T15:46:27Z
• Authors: Michelle Nicole B...; Mark Begale; Jennifer Duffecy; Darren Gergle; Chris J Karr; Emily Giangrande; David C Mohr
• Keywords: mHealth; Depression

Mobile phone sensors can be used to develop context-aware systems that automatically detect when patients require assistance. phones also provide ecological momentary interventions deliver tailored assistance during problematic situations. However, such approaches have not yet been treat major depressive disorder.The purpose of this study was investigate the technical feasibility, functional reliability, and patient satisfaction with Mobilyze!, a mobile phone- Internet-based intervention including context sensing.We developed application supporting architecture, in which machine learning models (ie, learners) predicted patients' mood, emotions, cognitive/motivational states, activities, environmental context, social based on at least 38 concurrent sensor values (eg, global positioning system, ambient light, recent calls). The website included feedback graphs illustrating correlations between self-reported as well didactics tools teaching behavioral activation concepts. Brief telephone calls emails clinician were promote adherence. We enrolled 8 adults disorder single-arm pilot receive Mobilyze! complete clinical assessments for weeks.Promising accuracy rates (60% 91%) achieved by learners predicting categorical contextual states location). For rated scales mood), predictive capability poor. Participants satisfied improved significantly symptoms (beta(week) = -.82, P < .001, per-protocol Cohen d 3.43) interview measures -.81, 3.55). became less likely meet criteria diagnosis (b(week) -.65, .03, remission rate 85.71%). Comorbid anxiety decreased -.71, 2.58).Mobilyze! is scalable, feasible preliminary evidence efficacy. To our knowledge, it first unipolar depression, one attempts use sensing identify mental health-related states. Several lessons learned regarding functionality, data mining, software development process are discussed.Clinicaltrials.gov NCT01107041; http://clinicaltrials.gov/ct2/show/NCT01107041 (Archived WebCite http://www.webcitation.org/60CVjPH0n).