What’s the problem?
Having undertaken a quality improvement project for the past year, it’s fair to say I have learned a thing or two about creating sustained change within a busy workplace environment, or, more precisely, how hard it is to do so!
It’s 3:12pm. There are twelve children in cubicles, three in the waiting room, and one worried registrar facing the impossible triage decision of who gets seen next. In this moment, a CXR request feels like a shortcut to clarity.
Within this is a roughly 65-strong team of hardworking and dedicated clinicians, who strive to provide the best care to every child and family member presenting to the department. Last year, one of the paediatric emergency consultants and I undertook a quality improvement project to reduce unnecessary chest X-rays (CXRs) in children presenting with acute respiratory illness.
Chest x-rays are a familiar tool in paediatrics. Faced with a wheezy child, a persistent cough, or a fever with no clear source, it’s tempting to reach for imaging as reassurance—for the clinician, the patient, or both.
But the evidence is clear: for many children with acute respiratory illness, chest x-rays rarely change management and may lead to overdiagnosis, unnecessary antibiotics, and avoidable radiation exposure. In our department, anecdotal concerns about overuse were growing louder.
When we reviewed local data, we found a surprising level of variation in CXR requests—especially among junior clinicians rotating through the department. Some were unaware of existing guidelines. Others defaulted to the “better safe than sorry” approach, particularly during high-pressure shifts.
The result? A pattern of imaging that often added little value, and sometimes introduced harm.
And so our project was born—not just to reduce numbers for the sake of it, but to embed a culture of thoughtful, evidence-based imaging that would outlast any one cohort of clinicians.
Our aim was straightforward: reduce unnecessary chest X-rays in children presenting with acute respiratory illness. But translating that aim into action meant navigating the unpredictable terrain of a tertiary paediatric emergency department—Bristol’s CED. This trauma centre sees around 50,000 children a year and is staffed by a rotating cast of over 60 clinicians.
We used the Plan-Do-Study-Act (PDSA) framework to guide our work—iterative cycles of intervention and reflection, each designed to move us closer to sustained change. Interventions ranged from educational teaching sessions to the development of a new departmental guideline and later, a bespoke induction package for incoming junior doctors.
At first, things looked promising. Through the first three PDSA cycles, we saw steady improvements in the proportion of appropriate CXR requests, each intervention nudging us in the right direction.
But shortly after our fourth cycle, things took a turn.
A new cohort of resident doctors rotated into the department, and, without warning, our carefully tracked improvement curve dropped below the median. Despite months of progress, the new team’s imaging habits hadn’t shifted.
That was our turning point. It wasn’t that the intervention had failed—it was that the system hadn’t been designed to carry it forward. Sustained change, we realised, meant embedding the intervention into the fabric of the department—not relying on individuals to remember or replicate it.
Why is it difficult to implement sustained change?
Emergency departments are built for reactivity, not reflection. They thrive on immediacy—on rapid decisions, shifting priorities, and the unpredictable rhythm of a 24/7 service. It’s what makes them extraordinary places to work. It’s also what makes them deeply resistant to change.
During the course of our project, we uncovered three systemic barriers that quietly sabotaged our early interventions—none malicious, all entirely predictable.
Rotating doctors create a knowledge sieve
In a department like ours, junior doctors rotate every few months. Just as one group starts to embed new habits, another arrives—often from entirely different training backgrounds, with different expectations, and a different understanding of what ‘normal’ looks like.
With every rotation, institutional knowledge thins. Unless change is embedded at the system level, it disappears.
Speciality variation makes consistency a challenge
Our junior doctors don’t just come from one pathway. Some are training in paediatrics, others in emergency medicine, general practice, or anaesthetics. Each arrives with a unique lens—and a different set of priorities.
What feels vital to one cohort may be seen as peripheral to another. Ensuring a shared standard of practice across such diversity is no small task.
Shift patterns shatter educational continuity
In theory, we’d love to deliver teaching to every clinician. In practice, the rota doesn’t care about our plans. Nights, weekends, late finishes, and sick leave mean that many staff simply miss key sessions.
One well-timed intervention isn’t enough. In a department like ours, change has to be omnipresent—because people are not.
So what did we do?
After multiple failed interventions and one spectacular backslide, we stopped asking, “What should we teach?” and started asking, “How do we make this unmissable, repeatable, and sustainable?”
Here’s what worked.
Utilise staff inductions
We built a targeted education package directly into the junior doctor induction process—ensuring everyone started their rotation with the same understanding of our goals and expectations.
This wasn’t just about knowledge. It was about shaping behaviour from day one—before old habits took hold.
Ensure interventions are accessible
Our shift patterns were chaotic; our education couldn’t be.
We developed flexible interventions that clinicians could access on their own time—e-learning modules, recorded teaching, and visual summaries pinned in shared spaces. This meant that education wasn’t just for those who happened to be free on Tuesday at 3 pm.
Create a guideline
Guidelines provide scaffolding. We created and disseminated a departmental SOP (standard operating procedure) outlining when CXRs were appropriate—and, just as crucially, when they weren’t.
By having something to refer to on shift, we reduced reliance on memory or hearsay.
Foster a shared-decision-making culture
We encouraged team-based decisions and open conversations. Educational games, peer discussions, and short in-situ teaching sessions helped normalise evidence-based questioning, even in the middle of a busy shift.
This reduced the burden on individuals to ‘know the right thing’—and made it easier for juniors to pause and ask.
Engage all members of the team
Junior doctors rotate. Senior staff don’t. We made sure to involve long-standing team members—consultants, senior nurses, registrars—in our teaching, data reviews, and project planning.
Their buy-in created consistency. Their voices carried weight. And their presence helped bridge the gap between each rotating group of juniors.
What does sustained change look like?
Sustained change doesn’t shout. It doesn’t come with balloons or press releases. It shows up quietly—in better decisions, fewer unnecessary actions, and a culture that nudges people toward the right choice, even when no one’s watching.
In our case, it looked like this:
- A rise in appropriate chest x-ray (CXR) requests from 1 in 2 (August 2024) to 4 in 5 by August 2025
- 823 fewer CXR requests for acute respiratory illness between January and July 2025 compared to the same period in 2024
- A departmental cost saving of £25,513
- A 0.625-tonne reduction in CO₂ emissions
- And most importantly: fewer children exposed to unnecessary radiation
But numbers only tell part of the story.
Behind each of these figures was a series of decisions made differently—a doctor who paused before ticking the request box, a team that talked instead of defaulting to protocol, a junior who felt confident enough to say, “Let’s think twice.”
That’s what sustained change really looks like.
Not just a temporary dip in the data, but a shift in culture—a new normal that holds, even when the people change.
This project began as an attempt to change a single behaviour—reduce unnecessary chest x-rays in children with respiratory illness. But what we learned along the way was far broader: that lasting change doesn’t depend on the brilliance of the intervention, but on how deeply it’s embedded in a department’s people, processes, and culture.
The most effective strategies weren’t complex. They were consistent. Simple interventions, repeated with care, designed with the messy realities of clinical life in mind. Nothing flashy. Just thoughtful, well-placed nudges that allowed good practice to become routine.
And crucially, we didn’t do it alone. This project worked because it was shared. Because senior staff championed it, junior doctors questioned it, and the department slowly made it its own.
We hope that these lessons can travel. Not because every department has a CXR problem—but because every department grapples with the challenge of change.
So if you’re in the middle of a QI project that feels stuck, slipping, or just short of its potential, know this: change is possible. But it has to be designed to last.
