Healthcare will not improve dramatically if quality improvement remains the domain of a relatively small group of specialists.
We need many more people across healthcare to understand how to improve the systems around them.
But I’m not sure we always teach beginners the most important things first.
In this piece, I’ll explore five habits I would want someone to learn before worrying too much about the tools:
Get unusually precise about the problem.
Assume you understand less of the system than you think you do.
Treat resistance as information.
Make the problem small enough to learn from.
Measure to learn, not merely to prove that you succeeded.
Healthcare has no shortage of problems worth improving.
Patients wait too long. Important information gets lost. Discharges are delayed. Clinicians spend time on work that adds little value. Processes that seem straightforward on paper become frustratingly difficult in practice.
And yet the number of people who are actually trained to improve these systems remains relatively small.
If we want healthcare to get dramatically better, improvement cannot remain the work of a small group of quality professionals. We need many more physicians, nurses, pharmacists, therapists, administrators, patients, and others to see improving the systems around them as part of their work.
But I’m not sure we always introduce people to quality improvement in the right way.
We often begin with the machinery: PDSA cycles, process maps, driver diagrams, run charts, Lean tools, project charters.
Those things can all be useful.
But before someone learns the tools, there are a few ways of thinking that matter even more.
If I were introducing someone to healthcare improvement for the first time, these are the ideas I would want them to understand.
1. Don’t start with the solution
Healthcare professionals are trained to solve problems.
A patient has a symptom. We gather information, make a diagnosis, and decide what to do. And that instinct is useful clinically but it can also get us into trouble when we try to improve systems.
We notice that discharge takes too long and decide we need a discharge checklist.
We see medication errors and decide staff need more education.
We notice delays getting patients to imaging and decide we need another porter.
The solution can arrive almost simultaneously with the recognition that a problem exists. And while that can sometimes be the right solution, more often it falls short. In complex systems like healthcare, heuristics and knee-jerk solutionizing seldom works.
One of the most useful habits in improvement is learning to separate the problem from your first idea for fixing it. Start by forcing yourself to describe the problem precisely.
Who is experiencing it?
Where does it occur?
How often?
Under what circumstances?
What is happening now that should be happening differently?
Write the problem down. Show it to other people who work in the system. Ask whether they recognize the same problem.
This sounds almost absurdly basic. I promise it isn’t.
A vague problem such as “discharge is inefficient” can support almost any proposed solution. A much more specific problem might be: patients medically ready for discharge frequently wait several hours for a particular step in the process. This framing dramatically changes what you need to understand next.
Problem definition is not administrative housekeeping before the real improvement begins. It is part of the improvement.
2. Assume you understand less of the system than you think you do
When I was a resident learning to place central lines, one lesson was drilled into me repeatedly: prepare the environment before you start.
Position the patient properly. Make sure the ultrasound is working. Have the equipment you need within reach. The procedure starts before the needle ever touches the skin.
Improvement works the same way.
One of the most important things you can do before changing a system is understand who actually knows how that system works.
That sounds obvious. It often isn’t.
Healthcare creates expertise in narrow slices of enormously complicated processes. A physician may understand one part extremely well while knowing very little about what happens before or after it. A hospital leader may understand the intended process without seeing dozens of workarounds that frontline staff use every day.
I once watched a physician pursue what seemed like a straightforward improvement: introducing a better patient gown.
The new gown appeared superior for patients and clinicians. But the project quickly ran into problems because some of the people responsible for purchasing, infection control, laundering, and managing the gowns had not been involved early enough.
They knew things about the system that the project leader did not.
This is one of the easiest mistakes to make in improvement: confusing expertise in your own work with expertise in the whole system.
Before changing something, ask:
Who actually performs this work?
Who depends on it?
Who works immediately upstream and downstream?
Who will have to do something differently if the change succeeds?
Who sees failure modes that I never encounter?
This is often called stakeholder engagement, which makes it sound like a communications exercise. I believe it is more important than that.
You are trying to assemble a more accurate picture of reality.
3. Treat resistance as information
Once people start proposing changes, another predictable phenomenon appears: someone objects.
The new workflow will take too long. The technology won’t work in a particular setting. The proposed process creates extra work. A department says it cannot support the change.
It is tempting to categorize these people as “resistant to change.” Sometimes people really are resistant to change. But that explanation is often too convenient and deserves scrutiny.
The person objecting may understand something about the system that you don’t. Perhaps your proposed workflow works beautifully on weekdays but falls apart overnight. Perhaps it solves a problem for physicians while inadvertently creating more work for nurses. Perhaps an apparently redundant step exists because of a failure that happened years ago.
Perhaps the person who seems difficult has watched three previous improvement projects arrive with enthusiasm and disappear six months later.
An objection is not automatically correct. But it is data.
Instead of asking, “How do we overcome the resistance?” a better first question is:
What might this person know that we don’t?
That question changes the conversation.
It also prevents stakeholder engagement from becoming a euphemism for persuading people to accept something that has already been decided.
Good improvement work does not merely seek buy-in. It seeks understanding.
4. Make the problem smaller
Ambitious people tend to choose ambitious improvement problems. We want to fix patient flow, improve safety, reduce burnout, and redesign disharge.
These are worthy goals. They are also enormous systems.
One of the most valuable things a beginner can learn is that narrowing the scope of a project is not the same as lowering its ambition. It is often what makes learning possible.
Suppose your hospital has a discharge problem. You could attempt to redesign discharge across every medical and surgical unit simultaneously.
Or you could ask a much narrower question:
Can we improve one part of the discharge process, for one patient population, on one unit?
While the second version can feel less impressive, it also gives you something the first version often does not: the ability to see what happens when you change the system.
You can learn quickly, discover unintended consequences, and revise the intervention.
And, if it works, you now have something concrete to build on.
Trying to change too much at once makes it harder to know why things improved, or why they didn’t.
Small scope is not small thinking. It is often disciplined experimentation.
5. Measure to learn, not just to prove that you succeeded
Eventually, improvement requires data.
But beginners are often introduced to measurement as if it were the scoreboard at the end of a project.
Did the intervention work?
Did the metric improve?
Can we demonstrate success?
Those questions matter. But measurement can do something more valuable: it can tell you whether your understanding of the system was correct.
Suppose you believe a new process will shorten a delay.
How much?
For which patients?
How quickly?
What else might change as a result?
Before making the change, make the prediction explicit.
Then look at what actually happens.
If the result is different from what you expected, that is not merely a disappointing outcome.
You have learned something.
Perhaps the bottleneck was somewhere else. Maybe the staff used the new process differently than anticipated. The intervention may have worked only under certain conditions. Or, perhaps you improved one measure while worsening another.
This is why measurement in improvement should not be thought of primarily as proof. It is feedback.
And feedback is what allows a team to move from “we think this will work” to “we are beginning to understand how this system behaves.”
Improvement starts before the tools
None of this means that improvement methods are unimportant. Quite the opposite, they are essential.
People must ultimately learn how to map processes, develop measures, test changes, understand variation, and use structured methods if they want to pursue improvement seriously.
But those tools work much better when they sit on top of a few foundational habits:
Get unusually precise about the problem.
Assume that other people understand parts of the system that you do not.
Treat disagreement as a potential source of information.
Make the problem small enough that you can actually learn from changing it.
Use measurement to test your understanding of the system, not merely to validate your solution.
Healthcare needs more people who know how to improve it. That does not mean everyone needs to become a quality improvement expert. But imagine what would happen if many more people working in healthcare developed these habits.
They might be slower to jump to solutions and more curious about how work actually happens. The might be more willing to test their assumptions and pay closer attention to the ideas of the people around them.
And better equipped to turn the frustrations they encounter every day into opportunities to make care better.
That seems like a very good place to start.
Continue the conversation
If you were teaching quality improvement to someone for the first time, what is the single most important idea you would want them to understand?
I’d especially like to hear from people who teach improvement, lead QI programs, or remember what they found confusing when they first entered the field.
If this was useful
Subscribe to Leading Quality for future essays on healthcare quality, safety, improvement and organizational learning.
Share this article with someone who is learning quality improvement—or someone who teaches it.
Respond to the question above. I read the comments and often use the discussion to shape what I explore next.
Partnerships
Interested in partnering with Leading Quality?
If your company is aligned with our goal to improve healthcare, click here to discuss sponsorship opportunities.