IEPA 11

The proliferation of smartphones has launched a new era of research providing insight into the dynamics and underlying mechanisms of mental health and mental illness. In this symposium, we will illustrate the potential of different types of digital data and levels of analysis for predicting and intervening in symptom patterns, risk, phenotypes, and trajectories. Sarah Lynch, MSW, will present the clinician’s perspective, illustrating how time-lagged analysis of daily affect and psychosis ratings can inform clinical intervention, from psychoeducation to psychopharmacology. Kristen Woodberry, MSW, Ph.D., will expand on this, examining how symptom dynamics may improve on static predictors in the early stages of mood and psychotic disorders. Preliminary experience sampling data from youth ages 15-25 with psychotic spectrum disorders will illustrate different patterns of affect and psychosis variability. Zuzana Kasanova, Ph.D., will discuss the association of reward-oriented behavior in daily life and striatal reward-related dopamine release in the lab using data collected in healthy volunteers and first degree relatives of patients with psychosis. John Torous, M.D., will illustrate how a combination of smartphone active data (surveys and voice samples) as well as passive data (geolocation, accelerometer, call/text logs) can be combined to create a digital phenotype of early psychosis. Using freely available and open source research apps, the talk will feature examples and discuss the ethics, implementation, and analysis of digital phenotyping data. Finally, Inez Myin-Germeys, Ph.D., will discuss the overall implications of different digital data and levels of analyses for the science of early intervention.