Tutorial: MobileCoach Intervention Platform

MobileCoach: An Open Source Software Platform for Digital Health Interventions

About MobileCoach:
MobileCoach is an open-source software platform for the design of smartphone-based health interventions and ecological momentary assessments. Since its first release in 2015, MobileCoach-based studies have been conducted in the areas of health literacy in asthma [1] and nutrition [2], subclinical depression [3], mindfulness and stress reduction [4], chronic disease management (e.g., hypertension [5], diabetes [6], asthma [7]), physical activity [8], personality development [9], chronic pain [10], childhood obesity [11], and holistic preventive care [12,13,14].

MobileCoach is a client-server system. On the MobileCoach server, which is written in Java, intervention content (e.g., images, video clips, links to external resources such as websites) and logic (e.g., rules that define when to deliver what intervention or data collection request) are defined via the MobileCoach Designer. The MobileCoach Designer is a web-based graphical user interface. It is also used to implement a rule-based conversational agent (aka chatbot) that imitates the conversation with a human being. This conversational agent is then used to deliver the intervention content or ecological momentary assessments via the MobileCoach client applications. Conversational agents allow for building up a working alliance, an important relationship quality between caregiver and caretaker, that is robustly linked to treatment outcomes. The MobileCoach client applications are written in React Native and are available for iOS and Android devices.

MobileCoach can be extended by various other intervention or data collection components. These components must be developed according to the specific needs of the envisioned intervention or ecological momentary assessment study. Examples of such extensions include web-based dashboards for (patient) remote monitoring purposes [1,6], app-based cockpits with gamification features that indicate the study progress [1,2,6,9], audio recordings for cough detection via the smartphone’s microphone [7], use of sensor data streams from medical devices (e.g., blood glucose data) [6], state of receptivity models that trigger interventions or data collection requests at the most opportune moments [15], or integration of Unity-based intervention components (e.g., the gamified slow-paced breathing training Breeze [16]).

Objectives:
1. To provide an overview of MobileCoach through various MobileCoach-based studies and live demonstrations
2. To showcase the design of a digital health intervention with MobileCoach Designer
3. To identify and discuss novel features and open research questions with the workshop participants
4. To motivate participants to join the open source community and to exchange ideas about novel health intervention approaches (e.g., just-in-time adaptive interventions that bring together the digital biomarker and health intervention communities)

The MobileCoach team behind this session:
Prof. Dr Tobias Kowatsch, Professor for Digital Health Interventions, Institute for Implementation Science in Health Care, University of Zurich (UZH), Director, School of Medicine, University of St.Gallen (HSG), and Scientific Director, Centre for Digital Health Interventions, UZH, HSG & ETH Zurich, Switzerland.
Prabhakaran “Prabhu” Santhanam, MSc in Computer Science, MobileCoach Software Engineer, and Community Manager at the Centre for Digital Health Interventions, ETH Zurich, Switzerland.
Fabian Schneider, BSc in Computer Science, MobileCoach Software Engineer and Community Manager at the Centre for Digital Health Interventions, ETH Zurich, Switzerland.