Most healthcare technologies are not designed in a patient-centered fashion, nor utilize tailored, persuasive approaches, resulting in poor adherence - and ultimately poorer outcomes.
We are fast approaching a connected, digital health world. The great hope for connected healthcare technologies is that they can greatly reduce costs by increasing the reach of existing medical providers; that they will enable more effective forms of self-care and management; and that they will enable “smarter” interventions with patients over time, thus reducing the risk of more costly health interventions.
This digitally connected future brings a lot of opportunity, but there are also associated pitfalls if our digital future is poorly designed. Adherence is already a massive issue; in some disease conditions, more than 40% of patients sustain significant risks by misunderstanding, forgetting, or ignoring healthcare advice. Additional findings from the behavioral therapy literature also suggest most individuals have difficulty maintaining healthy behavior changes, with reports of premature drop-out ranging from 30–60%. Creating more technology, without considering how we will get ourselves and our patients to use it may very well lead to more of the status quo. As designers, engineers, entrepreneurs, and health care change makers, I believe we need to reflect, together, on a more thoughtful and nuanced approach to utility and how utility evolves over time.
I am not the first to talk up the potential of design to help solve the issue of patient adherence, but I do believe most often it is done at face value without instructions for implementation. In the rest of this post, my goal is to unpack how human-centered design techniques can be combined and sequenced to generate engaging connected health applications and services, and why this sequence is often at odds with healthcare practice. In future posts, I will relate additional details relevant for each technique.
In pursuit of longer term adherence through design.
Starting with design research.
Design research is first and foremost concerned with gaining insights and empathy with a target population in order to drive product and service creation. Typically, design research is guided by quantitative research (these are generally the types of people we should target), but is purposefully qualitative in that it goes deep into the why people act the way they do, which may or may not be explicitly stated or even known — information that can’t be gathered necessarily through surveys. Design research is often conducted in the context of the person’s home, where habits and real behavior are more likely to shine through.
The designer applies a lens to sensemaking that introduces bias and risk that will ultimately be remedied through prototyping of potential solutions to the problem or set of problems at hand. In this way design is often a novel approach to health, which is grounded in scientific studies which are more concerned with immediate accuracy, proof, and truth. For wicked problems such as behavior change in the realm of health and wellness, we may be better off disregarding the search for variables that can be tweaked and instead spread out our search for efficacy across a human centered design process.
Identifying personas.
Research studies in health try to limit variables, while design seeks to understand what happens when life gets in the way. As such, design research often finds new “groupings” of behaviors among individuals that stretch across commonly assumed demographics. For example, I once conducted a study about healthy decision making for Whole Foods Market. While we initially thought of groups of people such as “single” “millennial”, and “moms”, there were stronger defining characteristics across a spectrum such as “planners” “self-chastisers” and “speed demons” in the context of where and when the behavior actually occurred. These newly discovered traits had implications for solutions that supported the planning process, encouraged shoppers at moments of discouragement, and thoughtfully considered “sub-mapping” to place healthy products in the same spaces across each aisle section.
Lifestyle influencers within health and wellness application use include living arrangements, social patterns, and even more broadly city culture. Understanding and grouping distinctions for the target population have immediate application for timeline mapping activities that help clarify ideal points to support positive behavior change, while mitigating risks unique to the sub-group.
Mapping behaviors.
Service design has a rich tradition of journey mapping activities that are designed to a specific journey or service experience — but what happens when the service occurs across the minutia of an unspecified timeframe?
We look for meaning and inspiration in periods of time that relate to the cycle of treatment and to the patient’s life more generally. For example, treatment milestones as indicated in assessments and plans, and standard cycles of behavior such as weeks or months across which we can comprehend and map more easily. Similar to engaging in qualitative research, maps take on risk and the potential to not be accurate, as they include both bias from the synthesis of the designer AND a small study group. But they do help tell a story that can form the hypothesis of the right types of intervention, either digital or human led.
Standard customer journey maps consider questions such as what is the user thinking in this moment? What tools or artifacts is the user using to accomplish a task or reach her goals? These types of questions may be correct for the persona group at hand, or they may need to be modified slightly to help tell a coherent story in the context of a lifestyle. Two additional questions that are key when considering health and behavior change for any temporal period are: “where and what are the risks for this persona group to fail?” and “where and what are the opportunities to encourage or inspire adherence behaviors?”
Once these are mapped, the best areas and times for intervention — either subtle, strong, or in between — can be identified. I use intervention instead of technology because there is likely an appropriate mix of human and technology touchpoints that vary for each persona group.
Generating solutions and testing them rapidly, at low cost.
One of the most important pieces of instruction I heard in interaction design school was to “be the interface” — to literally role play whatever combination of technology and service we expected to ultimately build in order to root out issues before spending money on implementation. For a caregiving interface, that meant manually sending texts and creating calendar invites for tasks to be completed to support care using google calendars. In terms of health-related interventions, I would encourage individuals to go through micro-rounds of service that can identify the larger issues in an intervention — the type of “duh” issues that are often missed that could ultimately drive behavior.
Once noticed, these are often low-hanging fruit that can be addressed with the change of a color, word, or addition of a button or icon. In terms of a service dialogue, it could be the quick change of a script or tone that clarifies a direction that is unclear. This type of testing should be done in concert with interviews with the target population of patients to hear their perspectives on what’s working, what’s not, and how they’re feeling about it.
Determining and Measuring Key Performance Indicators (KPIs).
Once it’s clear that there are no more glaring issues in a workflow or micro service round, the service and technology (still cobbled together!) are working in tandem, and the patient is satisfied or even delighted, then it’s time to actually test whether or not impact can be had across a full period of time as influenced by your KPIs. In terms of a health intervention, it may be change in weight within the month, drop of cholesterol by x, or other important subjective quality of life indicators. This is the appropriate time to conduct more standard quantitative studies as required by clinical need.
Commonly missed additions.
As designers and problem solvers, we must not forget that what we’ve designed will be used, repeatedly — and with digital technology, it’s on us to identify ways to keep that experience meaningful to the user of the product and service over the course of treatment. Digital technologies are also relatively cheap to implement, so we don’t get a pass to avoid this level of consideration. We must ask, how does a concert of interactions play out over time, in a way that results in deeper forms of engagement that are driven by sentiments or beliefs of the user?
Choice architecture and service design frameworks are helpful when considering how to persuade an individual to make a decision and to derive satisfaction from an experience, yet sometimes other layers are required to drive behavior change particularly in the consumer setting.
Layering Identity and Authenticity.
If your product or service were a person, who would it be? And how do your target population and persona groups relate?
Understanding this is key to learning how to build trust and a connection with a user that can be leveraged. Without connection, the use of any tool will ultimately fail. The nuance of a healthcare application is that it likely (but not always) will require someone to do something that they are not fond of doing — things that either take additional time in a day, or bring up unpleasant thoughts. This can be counteracted to an extent by thinking about the types of personalities that may be respected and trusted by your persona group.
Unfortunately, with software being as mechanical and structured as it is, there’s often little thought that goes into how it can be LESS mechanical and structured. This also takes time and effort on the part of the team that’s building the software — to figure out how a combination of product and service can be seamless. Often, there’s no immediate reward for the company for this kind of effort, which is why it seems to be an afterthought, or worse, killed in its tracks as soon as its proposed.
“Easter egg” is the congenial term if it’s strictly a software view and is irreverent. But irreverence is what people remember. Surprise is what people remember, and are thankful for, in a computerized world of yes/no decision trees.
Layering Extrinsic Motivators.
Extrinsic motivation is typically defined as behavior that is driven by external rewards such as money, fame, grades, and praise. This type of motivation arises from outside the individual, as opposed to intrinsic motivation, which originates inside of the individual.
Depending on the particular persona group, there are likely to be certain motivators or drivers that can work in concert with the product and service to help drive engagement and adherence. Such a motivation profile can be used to drive slight changes in UI and experience. For example, certain individuals will respond better to social proof than others; for those delighted by data, quantified metrics and tracking updates would be more beneficial.
We should ideate outside of the core intervention when considering engagement. For another example, some individuals may be driven by the potential to access a specific sale from one of their favorite stores once a certain milestone is reached, and driven to greater savings in tandem with their progress. On some level, this speaks to gamification, although providing rewards that are desired by the persona group at hand is more important than generically considering the use of badges.
Final thought.
At its root, the challenge for digital health technology is how to design for persistence, which requires persuasion towards good decision making in the absence of human advisors. Persuasion takes creativity. In the absence of a doctor telling a patient what to do, we need to harness the technology humanizing force of designers to build products and services that can foster engagement with health while simultaneously addressing issues of patient adherence.