Simplifying Systems Thinking - Part 1
You don't have to swim faster than the shark, just faster than the slowest diver. While I originally found this a kind of inelegant advice about avoiding burnout by being just good enough, this analogy has grown on me as a way to explain optimizing systems.
I like to think of myself as someone who practices quality improvement every day, but especially as the COVID-19 pandemic has dragged on; at times I have felt that the number of opportunities to improve is too great, and the need to address them is too urgent. It's important in our quest for progress, focus, efficiency, nimbleness, and speed to not to lose sight of the fact that we need to keep swimming. Sometimes the sustainability of our improvements relies on steady progress that addresses our biggest threats to the whole system. Micro-level innovation for the sake of innovation, or optimization of a piece of a whole can make matters worse. When dealing with complex systems, sometimes the key is doing enough of the right things instead of doing more, or trying to be the best at everything.
You're probably familiar with the concept of systems, and aware we work within them each and every day. However, it starts to get trickier when we actually go about improving systems instead of just putting out fires, or worse, pretending everything is fine as we see the flames growing. Luckily, QI methodologies and concepts give us some practical ways to address systems change.
The ideas in this post owe a huge debt to Eliyahu Goldratt's seminal books, The Goal, and Beyond the Goal, as well as Peter Senge's, The Fifth Discipline, and the work of Gene Kim, Alex Grant, Gerald Weinberg, and Nassim Taleb. All of these are great reads and resources, and I can't do justice to their contributions to systems thinking in this short-form. Instead, I just want to highlight the most relevant application of systems thinking applied to QI in healthcare.
In this first part, we're going to talk about systems thinking in plain terms, discuss the argument for systems change, and determine what needs to change first.
What are systems in healthcare?
Systems thinking, Lean, and Agile approaches are some of the more advanced QI concepts that teams shy away from because of their perceived complexity and steep learning curve. They do require a shift in mindset and metrics, but I really like to encourage them because they help get so much more out of the investment in doing quality improvement.
At the most basic level, a system is an interconnected set of steps and inputs that are structured in a manner to achieve an outcome. The WHO definition encompasses everything from direct caregivers to public and population health administrations, and private multi-conglomerates. For improvement purposes, I like to scale it back some and think about it as a process or processes, plus all the inputs and stakeholders that are part of them.
I often hear systems talked about derisively, even among some of my QI colleagues. There is a temptation to think about systems as monolithic bureaucracies, that are in opposition to people-powered changes - but I actually think of people as the most valuable part of any system. This is how I make the argument for a systems approach -- no system is more important than the people in it, or that it was designed to serve.
The difficulty, I think, is that to harness the power of systems thinking, improvement teams have to be focused on very small interactions between processes, steps, people, and information…but have to understand the ripple effect of these changes throughout the system to avoid local optimization. This is my argument for QI making systems more centered on patients and staff. Things like huddles vs. meetings, training vs. a policy -- these are very much interactive, people-centered changes that impact systems.
What are the biggest opportunities to improve our systems?
Systems thinking is a framework to help us find the biggest limitations, weaknesses, and threats to the capacity and productivity of our processes.
Throwing resources anywhere else is a waste because they are throttled by the constraint – in fact, you often make it worse by putting more into a system that is already maxed out! A process will never be faster than its slowest step, so that is a logical place to start making improvements. How can we find constraints? Here are some simple suggestions:
Use a visual task-board to show bottlenecks and backlogs. Move away from calendars that are useless after the first missed deadline, and prioritize progress and focus on the immediate work to be done.
Collect data on value to the patient, and the flow of materials and information on a process map instead of a static flow chart alone to see the entire system.
Reflect instead of reacting: Firefighting and working overtime can be an action bias that is throwing more work into a system that it is not designed to handle. If your team is on the verge of burnout and losing faith in the project, it's okay to press pause. I'm reminded of another old aphorism, "A stitch in time saves nine." This is definitely true of improvement work.
Use QI tools such as Failure Mode and Effects Analysis (FMEA) to look at steps in the process: I'll probably highlight this in another tool blog post, but FMEA is a great way to find the problems that need addressing.
What should the system be?
In order to change our systems and break constraints, we need a vision of what the system should be, and a rough road map to get there. In QI, we always start by getting a baseline of what is going on at the frontlines of service delivery. Start by walking the actual processes and systems you need to improve.
Current State: The way things actually are in a process at the beginning of an improvement activity.
Ideal State: This is the vision. In a reasonably perfect world, what could this process look like?
Now that we've covered the background, rationale, and identification of systems constraints, we can actually shift to improvements. The important thing with systems thinking is we don't jump right into making improvements without understanding how a small change can impact the big picture. In part two I'll cover:
Systems improvements
Strategies to implement changes
Overcoming the barriers that allowed the constraint to form
Sustaining and spreading improvements for optimal flow throughout the system.
