How to apply a behavior design framework to digital health solutions

What is behavior design?

Behavior design facilitates the development of products and interventions that are informed by behavioral and cognitive science principles. When applying behavior design methods, an understanding of how people make decisions is translated into products that are designed to encourage engagement in healthy behaviors as easily as possible.

How to apply a behavior design framework

To create products that are informed by behavioral science, a behavior design framework is applied to ensure a systematic and balanced approach. The following framework integrates behavioral science principles with UX design, research and iterative testing methods. It is made up of three core phases: ‘Foundations’, ‘Behavioral Science’ and ‘Iterative Design’. Together, these represent a series of iterative and inter-connected activities that streamline the design of user-centered digital health applications.  

A framework for behavior design of a digital health application.

Phase 1: Foundations

The first phase of the framework establishes a foundation upon which behavior science methods can be built. The process of applying a behavioral science model relies on a robust understanding of the user experience in the context of the care ecosystem.  

To ensure a comprehensive view of this experience that is informed by evidence, exploratory research and experience mapping should be conducted. In the development of digital health applications, this phase needs to contextualize use within broader care pathways, from diagnosis to treatment. 

During this phase, a variety of insights can be gathered, including:  

1. Standards of care 
2. Touch-points with healthcare professionals 
3. Tools and techniques currently implemented  
4. Biases, stigmas and barriers to access 
5. Pain-points in the care journey  
6. Internal factors influencing engagement in the care pathway (e.g. motivations, knowledge, emotions, cognitive ability)  
7. External factors influencing engagement in the care pathway (e.g. resources, time, physical abilities, accessibility, devices)  

There are a number of methods available for collecting these insights, typically including a combination of semi-structured interviews, observational studies and literature reviews. The research should be designed to systematically understand the circumstances that influence behaviors of different user groups, while respecting individual differences between them.  

Learnings can then be translated into a comprehensive experience map, highlighting clinical touchpoints, tools, pain-points, emotions and decision points, as well as the differences between the journeys of different user personas and demographics. For example, people who have been living with a chronic condition for many years will have different needs and experiences than people who have only recently been diagnosed. People of different ages, sexes and socioeconomic backgrounds may have different experiences in the healthcare system to account for, or different levels of comfort with and access to digital solutions.  

Phase 2: Behavioral science

In the second phase, the foundational research is analyzed with the application of a behavior science model. The COM-B model of behavior from UCL is one of several models that can be applied. In the behavior design framework discussed in this article, this model is recommended as it facilitates a systematic approach to not only the understanding of target behaviors, but also the identification of relevant evidence-based interventions. 

Here are the 3 key steps of applying a behavior science model, drawing on the COM-B model as an example. 

Step one – Identify, diagnose and prioritize behaviors 

First, insights from research are translated into prioritized target behaviors that have the greatest impact on clinical outcomes and are therefore most relevant to address with the digital health solution. These should be specific, measurable behaviors contributing to engagement and adherence and will vary depending on the user and the clinical context. Some examples might include tracking a symptom when it is experienced, or logging questions ahead of a clinical appointment. 

These behaviors are then “diagnosed” based on data in the experience map that was created during the first phase of the behavior design framework. Each behavior should be analyzed in relation to the potential barriers that may prevent a user from achieving it. The COM-B model defines behavior as a culmination of factors relating to capability, opportunity and motivation.* By identifying barriers related to these three factors, we can better understand where to focus digital interventions.  

We can then prioritize behaviors to target during product design based on readiness for change, impact on health outcomes and spillover effect – meaning the extent to which the behavior also influences other important behaviors in the user journey. 

Example of an experience map.

Step two – Identify behavioral interventions 

The COM-B model is part of a broader approach to behavior change intervention design developed at UCL that includes a taxonomy of Behavior Change Techniques (BCTs) and interventions within a practical framework that provides guidance on how and when to implement them. 

This approach can be applied alongside an investigation of analogous scenarios, interventions and efficacies to select techniques that are most relevant to the target behaviors that have been identified. These behavior change techniques should be prioritized based on literature citations, implementation feasibility and applicability to this specific demographic.  

It’s important to note that these BCTs are not prescriptive – but rather provide a starting point that helps de-risk the innovation process and helps to ensure the concept development is informed by evidence. For example, by following this framework a digital solution developer may prioritize the inclusion of BCTs such as ‘information about health consequences,’ ‘goal-setting’ and ‘self-monitoring of behavior.’ These can then be combined and translated into features and design principles for the digital health solution.  

Step three – Conceptualize  

Once a set of behavior change techniques has been selected, these techniques can be translated into tailored concepts. In an innovation process it’s important to provide structured prompts that guide ideation without constraining creativity. BCTs and other insights can be embedded into design challenges to ensure the team does not lose sight of the user and the opportunities for intervention.

Phase 3: Iterative design

Once concepts have been generated, these are iteratively designed and tested in the third and final phase. 

Design core features 

We need to design specific features that bring the concepts to life, to test whether they have the intended impact with the specific user demographic. It can be useful at this stage to outline a blueprint of what the digital product should look like, answering the question: how can we embed the behavioral techniques into the user experience? 

Prototype, iterate and test with users 

Using rapid prototyping techniques, the next step is to design and test different behavior change techniques, iteratively drawing on feedback from experts and representative samples of users. This collection of evidence is combined with insights from the behavioral science application to build confidence in the solution. 

Finalize designs 

With input from users, developers can validate design decisions before formalizing them in the product architecture. The combination of key user insights as well as the theoretical foundations behind the techniques and features selected serve as guides for the rest of the development program. Design iterations, product requirements and technical decisions can all be evaluated back against the original aims so that as the product evolves, the core behavior science principles are sustained. This saves time and resources by ensuring that potential engagement and adherence challenges are addressed up front in the design.  

Our design process for exploring how behavioral interventions can be embodied as a digital rehabilitation app.

Conclusion

The behavior design framework provides a structured approach to developing digital health applications that optimize user experience and outcomes, making the most of what digital tools have to offer. By applying this framework, digital health solution providers can de-risk their development by ensuring it is rooted in insights from both the literature and users.