The use of digitally assessed phenotyping parameters to model change in psychological therapy - Development of an EMA feedback tool

Background: In clinical psychological research, Ecological Momentary Assessments (EMA) have been used recently to track a variety of patients’ experiences over time in real-time and real-life situations. However, the measurement of treatment processes and outcome are predominantly based on self-reports. Given new technological developments, other potential sources can be considered to improve measurements. Since physical activity, heart rate variability, and sleep are often associated with health outcomes, recent studies have focused on using digital phenotyping to examine their significance in psychotherapy. Objectives: To detect individual changes in digital phenotyping parameters before treatment onset, to investigate their predictive value for psychotherapy outcome. Research question: The present project aims to examine whether intensive longitudinal assessments using digital phenotyping (stress level, activity, and sleep duration) can be a valuable tool to investigate change during psychological therapy. Furthermore, the project targets to develop a feedback tool displaying the relation of digital phenotyping parameters and self-reported affect with treatment outcome for the individual patient. Methods: Patients filled out self-reports regarding positive and negative affect four times a day over a two-week time period while treated with CBT. Measurements of continuous stress levels (heart rate variability), activity, and sleep duration were assessed via fitness trackers (Garmin) during the same two-week period. Time-varying change point autoregressive (TVCP-AR) models were employed to detect both gradual and abrupt changes in stress levels. Results: First results for 25 patients indicate differential patterns of change processes in stress. Inertia of stress level changed gradually over time in a few participants, whereas most participants showed gradual as well as abrupt change. The information was used to generate a feedback tool using heat plots to show the relationship between stress and affect parameters with outcome. Conclusion and implication: This feasibility study demonstrates that intensive longitudinal assessments enriched by digitally assessed parameters have the potential to investigate intra- and interindividual differences in psychological change processes and outcome. The feedback tool can be used to give personalized feedback to therapists e.g. in the form of warning signals. If replicated such research could be used to support clinical decision making for example in the context of the Trier Treatment Navigator.