Improving people’s health and supporting improved clinical outcomes for patients is predominantly about behavioural change. Certainly digital health interventions are predicated on the idea that we can modify behaviour, whether this is adherence to medication or changes in lifestyle. Unfortunately this is not easy to achieve, human beings are complex and when it comes to health this complexity is magnified.
Application of behavioural change principles within the digital health sphere has been mostly limited in both its understanding and execution. Often the focus has been on gamification, which can be simply defined as the utilization of the games systems to non-game applications (such as setting goals, defining rules, designing user centered feedback systems and motivating social interactions).
Unfortunately, although these systems can be powerful, it takes little expertise to design feedback bars, leaderboards and reward badges. Unless carefully executed they can fail completely to align with the underlying psychology of the user, especially in complex areas such as healthcare. What is more, people often mischaracterise games themselves for gamification, leading to greater confusion.
I always encourage people to start a little further back and instead focus onmotivational design, creating experiences that support the psychological needs of the user. This process seeks to understand the motivational dimensions of a user’s personality and utilises techniques based on procedures, activities and emotions.When we apply these techniques to digital health interventions it quickly becomes obvious we need to really understand how people think and make decisions.
In an ideal world our decisions would be the result of a careful weighing of positive and negatives and informed by our accumulated existing knowledge. We would always make optimal decisions. This is termed Rational Choice and became fashionable in the 1970’s with economists such as Gary S. Becker outlining ideas known as the pillars of ‘rational choice’ theory.
Many digital health interventions rely on this ‘rational choice’ model, assuming people will make the right decisions regarding their health if given the right information and opportunity to do so, unfortunately it has been shown time and again that this is not the case.
If digital interventions, such as mobile health, are going to significantly improve outcomes for patients I feel we need to apply behavioural science in a more sophisticated way to our design process. I have been studying how behavioural economics could be applied to the design of digital healthcare solutions and at The EarthWorks we have been applying it to our health technology design.
Behavioural Economics And Digital Health
Behavioural economics (BE) is a field of study that investigates why people make the decisions they do and then attempts to devise ways in which those decisions might be influenced. The fundamental theory at the heart of the discipline is that decision are predominantly not made rationally but instead are irrational. The easiest example in health is that we continually make irrational decisions that can damage us: such as smoking, drinking, eating unhealthy foods, not exercising and failing to take medications that have been prescribed.
BE blends psychology and sociology with economics to try to better understand the irrational thought systems we all use. I have distilled what I think are the four concepts that can be considered when designing and executing digital health interventions with behavioural change as a core element:
1. Prospect Theory
While this rational view of human decision-making outlined in the 1970’s influenced many people, others began to undermine ideas about human nature held by mainstream economics. Most notably among these people where Amos Tversky and Daniel Kahneman, who are best known for the development of prospect theory. In essence this demonstrates that people value gains and losses differently and will base decisions on perceived gains rather than perceived losses. So if a person were given two equal choices, one expressed in terms of possible gains and the other in possible losses, people would choose the former (for example a 25% chance to win £1000 or a 75% chance to win nothing). While the idea of human limits to rationality was not a radically new thought, Tversky and Kahneman’s‘heuristics and biases’ research program made important methodological contributions, Another takeaway message derived from the study of prospect theoryis that giving something up is more painful than the pleasure we derive from receiving it.
Implications for digital health design
This helps us to focus on how we can best frame motivational feedback systems to support someone living with a chronic disease. Many ‘wellness’ applications set simplistic targets (steps taken, calories consumed) and develop ‘all or nothing’ loss or gain targets. Chronic disease is often multi faceted and improved outcomes rely on small changes being made across a number of daily decisions. For example in diabetes, very small changes to four or more lifestyle decisions can have a significant impact on clinical outcomes, such as substituting one piece of chocolate a day for a piece of fruit, or drinking one less glass of wine. Many interventions in diabetes however have gone down the ‘all or nothing’ approach to lifestyle change: expecting someone to go vegan and take up roller-blading. Our feedback systems focus on small incremental gain and then incentivize behavioural change by the fear of losing this gain. This is especially powerful in the field of mobile health applications.
2. Bounded Rationality
Long before prospect theory there was an understanding that human beings do not always make optimal decisions due to restrictions to human information processing and limitations in knowledge, combined with environmental factors, such as the time someone has to make this decision. This is termed bounded rationality and is one of the topics discussed in the 2008 book Nudge. In the book, Thaler and Sunstein describe how experience, good information, and prompt feedback are the key factors that influence people to make good decisions. Health is a particularly challenging problem as feedback in this area is often poor, and we are more likely to get feedback on previously chosen options rather than rejected ones.
Implications for digital health design
We focus strongly on the principles of experience, good information, and prompt feedback, especially with regard to mobile health solutions. In order to make better decisions people need feedback and support in context, exactly when it can be acted upon. We have utilized this philosophy in a recent design of an application in asthma that looks to be a truly contextual support tool for patients, with the capacity to predict and modify the course of someone’s disease by combining patient data with environmental information.
We also bear this in mind when designing feedback mechanisms for patients with chronic disease where the impact of decisions are at best noticeable over the course of years, while the impact of small decisions on the body is often not evident to the individual. An example of this is smoking cessation. Traditionally, generic feedback aimed at inducing behavioral change has been limited to information ranging from the economic costs of the unhealthy behavior to its potential health consequences (Diclemente et al., 2001). More recent mobile health solutions supporting people to stop smoking tend to provide positive and personalised behavioural feedback, which may include the number of cigarettes not smoked and money saved, along with information about health improvement and disease avoidance in the context of behaviour and not merely as a means of trying to change it.
3. Asymmetric Dominance
Boundedly rational choices made due to limits in our thinking processes are expanded upon in Dan Ariely’s book Predictably Irrational. A good portion of the research is focused on choice and value perception. He describes how humans rarely make choices about things in absolute terms; we don’t have an internal value meter that tells us how much things are worth. Instead we focus on the relative advantage of one decision over another and estimate the value accordingly. People also have a tendency to compare things that are easily comparable. For example, if given the options for a romantic weekend away:
£259 Paris (with free breakfast)
£259 Rome (no breakfast included)
£259 Rome (with free breakfast)
Most people would choose Rome with the free breakfast. It is easier to compare the two options for Rome than the more complex decision between Paris and Rome. This is called the decoy effect (or asymmetric dominance).
Implications for digital health design.
The theory of asymmetric dominance is most often applied to the influence of consumer decisions, especially those involving complex cost and value. Within digital health design it helps us to frame the presentation of complex decisions. This has been applied to incremental decisions regarding planning exercise programs and framing information supporting improved adherence to medication. We have also used prospect theory to map the decision-making process related to BMI, better understanding peoples response to risk in the context of body shape preoccupations and aversion to weight gain. We have used these techniques in the registration process for mobile health and personlised web support programs, helping us to better present information, support and services to patients in psoriasis and diabetes.
4. Social and Market Norms.
In the same book Ariely outlines the differences between social norms and market norms. We live simultaneously in two different worlds: one where social norms prevail, and the other where market norms make the rules. The social norms include the friendly requests that people make of one another (Could you help me move this chest of drawers? Could you help me change this tire?) And are wrapped up in our social nature and our need for community. These interactions do not need instant payback and it provides pleasure for both parties.
The second world is one governed by market norms, is very different. The exchanges are sharp- edged: wages, prices, rents, interest, and costs and benefits. When you are in the domain of market norms, you get what you pay for and that’s just the way it is.
Through a number of experiments Ariely arrives at the conclusion that money can be the most expensive way to motivate people. Social norms are not only cheaper, but most importantly, often more effective as well.
Implications for digital health design.
This has influenced and reinforced two key facets of my design of digital health interventions.
Firstly is the application of social forces, especially for mobile health. There is growing evidence of strong associations between a patient’s social environment and health behaviour. Utilising Social forces as a strategy to influence health behaviours can be highly effective because patients usually engage with their healthcare professionals occasionally but they interact with their social networks much more frequently. Furthermore, social forces may be particularly effective at building enduring habits for healthy behaviour, often the primary motivation for improved behaviours in chronic disease management are due to a desire to reassure close associations, such as family and friends. We utilise this by helping patients build ‘care networks’ that support them to make better decisions.
This leads to the second facet, that of incentives. There has been much discussion and some research into financial incentives as a means to increase health-enhancing behaviours. For example The Centre for the Study of Incentives in Health (a joint initiative between King’s College, Queen Mary, and the London School of Economics; www.kcl.ac.uk/schools/biohealth/research/csincentiveshealth/ ) and the Center for Health Incentives at the University of Pennsylvania (www.med.upenn.edu/ldichi/ ). The evidence at this stage suggests that they might do, if targeted appropriately. However the impact seems to be short term, there are concerns that they may increase health inequality and there is no strong link between financially incentivized choices and improved clinical outcomes.
Financial incentives could be better implemented with mobile and connected health solutions, however they are not necessarily appropriate for pharmaceutical company funded interventions. Building on the premise that money is the most expensive way to motivate people, we are working on mobile health application design in pulmonary arterial hypertension, multiple sclerosis and opioid addiction where we hope to utilise social forces to incentivize improved behaviours, eventually linking the intervention to improved clinical outcomes.
I firmly believe that behavioural design theory can help those of us designing what we hope to be clinically impactful digital health solutions to better understand how people make decisions, therefore providing a better-targeted intervention. However, behavioral science research cannot tell us for certain what will or will not translate into the real world. What it can do is frame a way of thinking about the problems we are looking to solve and provide a road map for experimentation.
I believe that the personal nature of technology, the capacity to integrate with someone’s everyday life combined with the capacity to measure health data, gives us a significant opportunity to influence choice and behaviour. Behavioural economic theory can help us with the choice architecture we design in digital health interventions, we are at the early stages, but exciting opportunities lie ahead.