Subclinical symptoms of depression or mania, present prior to clinical manic or depressive episodes in individuals with bipolar disease. Mobile applications that measure habitual behavioral patterns such as mobility, physical activity level and social engagement present an opportunity to identify state transitions preceding relapse. This could lead to earlier intervention and improved treatment outcome.

To identify associations between changes in habitual behavioral patterns measured with smartphone usage and psychiatric emergency room visits or admissions for bipolar patients.

In a case-crossover study, around 70 patients at a specialized outpatient clinic for bipolar disorder will be offered to install MoodMapper^®, an application that registers text message and calls (but not communication content or recipients), mobility (but not location), number of steps taken, screen-time, and battery level. A baseline level for each individual will be established. Deviations from baseline habitual behavioral patterns in the time preceding a psychiatric emergency room visits or admissions will be analyzed in order to establish predictor variables. The patients will be followed up to 18months.

Preliminary results will be presented.

Smartphone-collected behavioral data can be used to predict worsening or improvement of a patient's condition. As smartphone ownership becomes more commonplace, mobile apps present a unique opportunity to detect digital real-time signatures of mental illness in a way that is minimally invasive to individual's daily life and privacy.