You have been sitting for a long time, and your device of choice nudges you to get up and move around, so you get some activity in – does this sound familiar? In the past, body aches may have been prompted us to move, but these days we have sensors we willingly wear tell us all kinds of things about ourselves, how much we move, our heart rate, our daily steps, how well we sleep; the list goes on. My take on this is we are data rich but insight poor. The solutions that provide actionable insights will be a vast improvement on the current state.
A new paper in Frontiers in Computer Science by Steven Schwartz and colleagues from IndividuALLytics discusses the use of a digital twin in optimally supporting behavior change. The concept of a digital twin comes from NASA and the world of engineering. Full-scale mock-ups of space capsules were developed to mirror real-life simulations under space conditions to pressure test and plan for contingencies. What if all the data collection in digital health, software, and hardware can replicate as a digital twin so interventions could be delivered at optimal times to allow the end-user to experience the intervention and perhaps get new data on how the intervention is impacting their behavior. These self-experiments are data-driven and may lead to changes in behavior becoming more consistent over time.
Data derived from N = 1 or one’s own personal data set has gained prominence in the last decade. As scientists, we are used to looking at randomized control trial data on how interventions work compared to a control group on a population level. The ubiquitous nature of device derived data has moved from quantified selfers, who used data to enhance performance to the general population. The table below, tracking left to right, shows the kinds of data used to support behavior change to prevent or better manage diseases.
How might we take the complex interplay of data and make it relevant for the end-user? What principles of design can guide that user experience?
Creating the digital twin experience is a vital first step; what data visualization is sufficient to inform and motivate a person to change? Being transparent about data use agreements and the process of opt-in or out of data collection is also essential. Knowing the level of data needed to create a compelling digital twin is also crucial. This data can be informative for the consumer and the health care team to support data-based decision making.
Having a straightforward method to prioritize which interventions are offered based on the data can be ordered by level of importance; for example, taking a medication that reduces the risk of a heart attack may be a call to action. Other interventions like increasing levels of physical activity can be added on over time. Ensuring the interventions and data are presented in plain language will ensure better engagement; users can develop mastery of the content as they go deeper into the experience. Choice in data displays gives users agency on what data is essential to them.
One opportunity for improvement in current products on the market is viewing data in context. Often the consumer gets canned messages that aren’t personalized to their experience; this can negatively impact engagement. Giving the user ability to add context is a development opportunity; for example, on tracking sleep, the user can add things like room temperature our outside noise to further explain a poor night’s sleep.
Context on how data changes over time is also essential for the consumer; behavior change is a long term endeavor and may have seasonal aspects. Providing interventions that support the long term nature of change will keep user engagement and allow them to see patterns in their lives with greater clarity.
The possibilities of leveraging digital twins to support behavior change offer an exciting evolution in designing and supporting behavior change interventions. It may also help make the necessary leap to being insight-rich based on the myriad and data streams derived from apps and sensors. This data feed is also continuous, so it offers users and health systems new ways to support care.
Thanks for reading – Trina
(Opinions are my own)
References
Digital Twin
https://www.networkworld.com/article/3280225/what-is-digital-twin-technology-and-why-it-matters.html
Digital Twin and Emerging Science of Self
https://www.frontiersin.org/articles/10.3389/fcomp.2020.00031/full
