Welcome to The Design Loft!
In an earlier issue, we introduced some principles for designing for complexity. In this issue, we explore what complexity looks like in health services and programs—and how to achieve excellence when “best practices” fall short (and why they often do).
We’ll begin with a look at what complexity means in health and how to know you might be dealing with it.
Complexity in the Context of Health
Complexity is mostly about how health systems actually work: with blurry boundaries, layers of internal rules, and constant adaptation. By system, I refer to interconnected relationships that behave in ways that are germane to the context and might produce different behaviour across those contexts. These interconnected, interdependent parts are best understood as a system, not as piecemeal components that can be reduced to their parts.
What makes complexity difficult is that the causes and consequences tied to those relationships are subject to (often rapid) transformation and adaptation that are both emergent from the context and unpredictable. Think of chronic health issues like diabetes, anxiety, or Parkinson’s disease where there might be commonalities among patients, but the way the conditions manifest themselves, their treatment, and the expected outcomes vary among people. What this means is ‘best practice’ requires changing what’s done for each patient to address their needs within the context of their lives and treatment. It doesn’t mean starting fresh, but it does require a different approach.
It starts with reframing what best practice is (and changing our language). The Cynefin Framework is a sensemaking tool that helps us to place ourselves and our situation in context. The Framework, which has evolved for 20 years through a community of scholars under the stewardship of Dave Snowden, provides a useful way to make sense of complexity and identify it when we see it.
The video below from Dave Snowden, the originator and chief proponent of the Cynefin Framework, provides a good explanation. There are more up-to-date examples and other videos (see references), but I like this earlier one because it speaks to some of the fundamentals of how systems work in practice. Dave’s also a good storyteller. The Framework has been updated, but the foundations are still the same.
For health professionals designing programs or systems, complexity requires replacing traditional top-down problem-solving and best practices with approaches that respect local decision-making and adapt to what’s actually happening on the ground. It requires a lot of learning, feedback, and trust within your health team. In highly complex or emergent situations (think about the early months of the COVID-19 pandemic and its outbreaks), the creative workarounds and practical solutions that people come up with locally aren’t side notes—they’re the main story, and they deserve careful attention.
Better Practices
What the Cynefin Framework and the related systems thinking (see our earlier issue) that underpins it provide us with is an understanding of when we might use best practices and when we need something else. The very concept of best practice implies that we know what the best, most preferred (i.e., effective or efficient) options are among all others, we’ve evaluated them, trust them, and can deploy them consistently and reliably to achieve predictable outcomes.
Cynefin ‘types’ situations based on what level of order and interactions are being understood and the level of trust we have in the way they unfold in practice. These situations are often called Clear (or Simple or Obvious), Complicated, Complex, and Chaotic, based on their behaviour, structure, and outcomes.
For much of what we do in the health system, our best doesn’t follow that. Using the Cynefin Framework, we see roles and situation where we find different types of practice, which is what we can design for. The Framework provides recommendations for how to approach these problems, understanding the role of probes, sensing, and responses in different orders based on the context. Here are the four types of practice we find:
Best Practice
Used in the “Clear” (formerly “Simple” or “Obvious”) domain, best practice applies when cause and effect relationships are well-understood, repeatable, and predictable. There is an established right way to do things and adherence to standardized methods or instructions is appropriate. Think of the expression “ready-aim-fire” and you get an idea of how to follow best practice.
Good Practice
Relevant to the “Complicated” domain, where cause and effect exist but aren’t always obvious. Multiple valid approaches may exist and expert analysis is needed to determine what works in specific contexts. Practices can be standardized, but multiple solutions or “good practices” may be valid depending on the expertise applied.
Emergent Practice
This is associated with the “Complex” domain, where cause and effect can only be determined in retrospect. Here, solutions emerge through experimentation, iterative learning, and feedback (often described as “probe–sense–respond”). Practices cannot be predefined and must evolve as patterns become clear from multiple, safe-to-fail experiments.
Novel Practice
Found in the “Chaotic” domain, where there is no clear cause and effect at all, requiring action to be taken immediately to establish order. Once some stability is achieved, newly created (“novel”) approaches may emerge. There is no pre-existing practice; creativity and rapid response are needed (“act–sense–respond”). These practice conditions are often fleeting, have high-energy, but dissipate quickly.
Eight Ways to Design for Complexity
With Cynefin as an assist, how might we design for different levels of order, particularly complex conditions and situations? Drawing on what we introduced in our earlier issue, this list of principles can help ensure you’re not designing for novel or emergent practice conditions under the guise of best practice.
Embrace adaptability - Design health programs that can evolve based on what’s actually happening with patients and communities rather than trying to control everything in advance. Build a strong core purpose with features that can be adjusted as you learn. Forget “best practices”—focus on learning and adapting as you go.
Build modularity - Create health services with components that can work on their own while connecting through clear touchpoints. Think of a spine based on your core values—the things central to your mission—with flexible parts attached. For example, a patient education program might have a consistent message about health goals but different delivery methods for different communities.
Implement redundancy - Give people multiple ways to reach their health goals. Don’t put all your eggs in one basket. If one pathway doesn’t work for someone—say, in-person appointments—have alternatives ready like telehealth, community health workers, or peer support groups.
Incorporate feedback loops - Build in ways to hear from patients and communities about what’s working and what’s not. Create regular check-ins, collect stories, track outcomes, and use what you learn to adjust your approach. Make learning and course-correction part of the design, not an afterthought.
Design for emergence - Stay open to unexpected benefits and solutions that arise from your work. When you pay attention through feedback loops, you’ll notice patterns you didn’t predict—maybe patients are finding creative ways to support each other, or a side effect of your program is building community trust. Let your values guide you, but don’t lock yourself into rigid expectations about exactly what success looks like.
Focus on relationships - Pay attention to how different parts of your health program connect with each other and with people. How does the intake process affect trust? How do providers and community health workers coordinate? Design these interactions intentionally from the start rather than treating them as secondary concerns.
Accept uncertainty - Recognize that you can’t predict exactly how your health intervention will play out. Different patients respond differently. Communities change. Resources shift. Instead of trying to forecast one “correct” future, prepare for a range of possibilities and build in flexibility to respond.
Use simplicity strategically - Keep individual pieces of your health program as simple and clear as possible—simple messages, simple processes, simple tools. But understand that the overall health system will still be complex because you’re dealing with real people in real communities with many interacting factors. Don’t oversimplify to the point where you ignore important context.
We will continue to cover strategies, tools and approaches for designing for complexity within health contexts in future issues. For our Premium and paid subscribers, these will include additional materials to help you use Cynefin and other tools to ‘type’ systems and match that with specific design strategies.
If this sounds interesting to you, please consider becoming a paid subscriber. And if you need specific support in using this within your organization, please reach out and book a no-fee, no-obligation conversation with me to see how I can help you.
Happy designing and thanks for reading!