Mobile Health Interventions

Behavioural change is essential in reducing chronic-disease risk in the general population (e.g. obesity prevention), managing chronic diseases (e.g. diabetes) on a day-to-day basis and preventing hospitalisation and long-term complication (e.g. kidney failure).

However, it is often difficult to initiate behavioural change and lifestyle interventions and to maintain them in the long term. While mobile behavioural interventions in the form of smartphone apps that target lifestyle changes are plentiful, low user engagement has limited the effectiveness of many mobile digital health applications.

To overcome this challenge, the researchers from ETH Zurich has developed and employed MobileCoach, an open-source platform that relies on an automated conversational agent (chatbot) to deliver interactive and engaging mobile health interventions.

Building on the experience and success in developing MobileCoach, a systematic development process underpinned by empirical studies and clinical trials will be used to develop and evaluate MobileCoach interventions for the prevention and management of non-communicable chronic diseases in Singapore, together with partners from Singapore.

In addition, the team will leverage an innovative engagement approach that combines storytelling, motivational interviewing, cognitive behavioral therapy, progress feedback visualisation, gamification, and builds upon leading evidence- and theory-based frameworks in the areas of behaviour change to build their intervention, named LvL UP.

LvL UP will target young and middle-aged adults living in Singapore (including vulnerable individuals) to achieve the following outcomes:

  • Increased physical activity and healthy eating
  • Improved mood and reduced stress
  • Reduced risk of developing NCDs and CMDs
  • Improved self-reported quality of life

LvL UP’s digital lifestyle coaches are four characters that portray typical Singaporean lifestyles and who are embedded within an overarching storyline to build a working alliance with intervention users (i.e. a shared understanding about treatment goals and tasks). This attachment bond is robustly linked to treatment success. More specifically, coaching strategies will be personalised according to (1) an individual’s preferences in lifestyle behaviour and (2) the current stage of change. In doing so, the coaching will be rooted in psychoeducation, and cognitive behavioral therapy, covering content in the following three areas:

  1. Move More: focused on physical activity, exercise, and sedentary behaviour
  2. Eat Well: focused on nutrient-rich food, disease-preventing diets, and barriers to healthy eating
  3. Stress Less: focused on emotional regulation strategies for stress, anxiety, and low mood management

LvL UP will be adapted to the characteristics of the population and the healthcare ecosystem in Singapore. In addition to the digital coaching provided, it will be explored whether adding human support can help to build a therapeutic bond with users and improve adherence to and outcomes of LvL UP.

To carry out this research, the team follow the core principles of the multiphase optimisation strategy. In the preparation phase, the team will conduct literature reviews, interviews and focus group discussions with individuals at high and low risk of developing NCDs and CMDs, experts, health and lifestyle coaches, and (potential) future providers of LvL UP. The team will also develop the conceptual model of the intervention and define the optimisation criterion to develop a first MobileCoach-based prototype of LvL UP. A feasibility trial will be conducted with adults living in Singapore, without major exclusion criteria for recruitment. Afterwards, the research plan will involve conducting a Micro-Randomised Trial (MRT) to strengthen the core machine learning technologies behind the engagement mechanisms of LvL UP. Finally, a Sequential, Multiple Assignment, Randomised Trial (SMART) will be conducted to inform the implementation of human support in, and evaluate the effectiveness of, LvL UP.

JavaScript has been disabled in your browser