(and what I learned from watching it all unfold)
A St. Emlyn’s retrospective
I’m writing this on a beautiful Somerset evening in September 2038, just back from work, on the eve of my 63rd birthday. Nearly forty years in emergency medicine – nine in South Africa, 28 in the UK, 25 in global health – and I can honestly say that today’s shift would have been unimaginable to the young doctor who started out on the Cape Flats back in 2001.
We managed 194 patients today (our AI system tells me that’s optimal flow for our department size – funny how we now have algorithms for everything). About 60% were seen physically in the department, and the rest through our virtual emergency department (ED) network. Not one patient waited over four hours for discharge, admission or another disposition. More importantly, not one trainee hesitated to escalate when they needed help – something that would have been remarkable in 2025, but has become beautifully normal by 2038.
It got me thinking about how we got here, and what someone who lived through the entire arc of change might want to tell the next generation. Because honestly? The transformation from 2025 to 2035 was more dramatic than anything I witnessed across my diverse career.
The golden years (and why they mattered)
Before I tell you about the crisis and recovery, let me paint a picture of what emergency medicine looked like during its peak years – roughly 2002 to 2010. I went from working on the Cape Flats in one of the busiest Emergency Centres in South Africa, to working at Derriford in Plymouth, and the contrast was stark.
NHS EDs in the mid-2000s were genuinely world-class. The four-hour target was achievable (and being achieved – routinely hitting 95%+ performance). Departments were reasonably staffed. Consultants had time to teach. The specialty was growing, attracting the brightest trainees, and patient satisfaction was consistently high.
Coming from South Africa in 2005, the NHS felt like an aspirational gold standard. When I’d lecture or write about emergency medicine in my global health role, I’d often draw inspiration from UK EDs as examples of what was possible with some grit, proper resources and political support.
[Ed – Context matters here. Having seen healthcare systems at their absolute resource limits in Africa, Stevan brings a unique perspective on what “crisis” actually means in UK healthcare]
The slow-motion crisis (2010-2025)
The decline wasn’t sudden – it was a slow-motion disaster that played out over fifteen years. I moved back to South Africa in 2014 and then back again to the UK, settling in Dorset near Yeovil in 2018. The contrast with the NHS I’d previously admired was jarring.
Four-hour performance dropped from 95% to 71% – and that was prior to the pandemic. By 2021, the whole system started operating in permanent crisis mode. Access block into majors and resus from patients awaiting admission became routine. Winter 2022 saw performance drop to just 51%! But here’s what struck me most: we were failing not because we lacked resources, but in spite of available resources.
In South Africa, emergency medicine (a brand new speciality) developed around optimising every patient interaction as we couldn’t afford waste. It was the main premise of the design of the South African Triage Scale and ultrasound adoption throughout African resource-limited settings. But in the UK, Trusts with ten times the resources per patient struggled with basic flow management. The irony wasn’t lost on me.
By 2025, the statistics were genuinely frightening:
- Only 60.4% of patients attending ED reached a disposition within four hours,
- 30% of emergency medicine trainees reported high burnout risk (GMC 2025),
- 45% of the ED workforce consisted of trainees (RCEM 2023), and
- Consultant-to-patient ratios climbed to over 1:7,000 – nearly double the recommended safe level of 1:4,000 (RCEM 2023)
But the statistic that hit hardest? Twenty-six percent of our emergency medicine trainees felt hesitant to escalate sick patients to their seniors (GMC 2025).
It made no sense: we were supposed to be the happy specialty, the friendly specialty, the accommodating specialty. Escalation hesitancy is bad – in resource-limited settings, everyone has to communicate well or patients die. Seeing it emerge in a well-resourced system felt like watching preventable system failure in slow motion.
The moment everything changed (early 2026)
Working as one of the regional specialty advisors for emergency medicine in April 2026, I knew that although our recent Winter was slightly better than previous years, it was not nearly the renaissance the government had hoped for. Our new RCEM President was trying to build consensus for what would become the transformation agenda. His keynote at the RCEM’s annual conference did not disappoint.
I remember he said: “We have two choices. We can manage decline gracefully, or we can demand the resources to build the emergency medicine system this country deserves. There is no middle ground.”
It was a hell of a statement. But having worked and researched settings where emergency medicine was built from nothing, I knew transformation was absolutely achievable. The question wasn’t whether we could fix emergency medicine – it was whether we had the political will to do what needed to be done.
Three things converged that autumn to create what policy people now call “the perfect storm of change”:
The political moment arrived
On the back of another disappointing Winter, there was renewed commitment for the NHS 10-Year Plan (including the £29 billion to community care promised in 2025). Various recent reports were more closely considered. Such as the Messenger Report that found the type of leadership to drive real change in the NHS severely lacking. The Dash Report that reported on the deaths attributed to poor urgent care system flow and risk management. And the Darzi Report that called to move from “diagnose and treat” to “predict and prevent”.
The workforce crisis became undeniable
The RCEM projections showed a 600-consultant shortfall by 2038 without major intervention. Even Treasury officials couldn’t argue with the numbers. And the 2026 GMC training survey deteriorated further, showing hesitation to escalate rise to 25% and burnout to 34%.
The technology was finally ready
AI triage, virtual consultations, integrated digital records – all had moved from experimental to deployable. We weren’t asking for pie-in-the-sky funding anymore.
The commitment to 120 additional emergency medicine training places annually for six years was announced in August 2026. I’d heard promises before, but this felt different. It came with accountability frameworks, substantial job planning restructure to support supervision and training, and genuine political ownership.
The early wins (2027-2028): Learning from global experiences
The virtual ED pilots launched in January 2027. Having worked in settings where telemedicine was essential rather than optional, I was more optimistic than most UK colleagues about the potential.
The Australian Victorian model we adapted was brilliant – 80% diversion rate for direct consultations. But what excited me was how it connected to my experience in resource-limited settings, where every decision point had to be optimised. Virtual EDs weren’t just about convenience – they were about getting the right care to the right patient in the right setting.
By year-end 2027, our virtual ED was handling 600 cases monthly. But more importantly, patient outcomes improved. Turns out people preferred being treated at home when clinically appropriate (revolutionary insight, I know).
The AI triage pilots began that autumn. I would lie if I say I wasn’t slightly apprehensive. Having developed a major triage system from scratch, I understood the risks and potential better than most. Built on the success of the Netherlands Triage Standard, AI triage seamlessly integrated with our virtual ED.
Here’s what I hadn’t anticipated: AI didn’t replace clinical judgment – it augmented it in ways that reminded me of why I became an emergency physician. Backed by the National AI Triage Accountability Framework, I could focus on time critical or life threatening diagnostic challenges, rather than endless phone calls to get the right patients to the right care without a detour through ED.
The culture transformation (that mattered most)
2027 was also the year we fixed the culture problem
The implementation of the Messenger Report’s Collaborative Leadership Framework became mandatory, and tied to funding. As someone who’d worked in hierarchical systems where escalation hesitancy literally killed patients, I knew this was make-or-break.
The leadership training I underwent that year fundamentally changed how I supervised juniors. We moved from military-style hierarchies to something more collaborative, more human. The old “sink or swim” mentality that had defined emergency medicine gave way to genuine psychological safety.
Our first geriatric emergency medicine consultant also started that year – one of the first cohort to complete the new college accreditation. I was amazed it had taken so long. But the impact was immediate: 18% reduction in length of stay for elderly patients, 23% reduction in readmissions.
System integration (2028-2030): When everything clicked
2028 was the year the technology actually worked
The NHS App became the genuine front door to emergency care. The single patient record system meant seamless information flow across all settings – something I’d dreamed about since my South African days, where patient information regularly disappeared between consultations, let alone different providers.
At this point, a patient with chest pain, reluctant to attend ED, could complete a system assessment through the NHS app, upload smartwatch ECG readings, and be risk-assessed by our AI. If triaged to our virtual ED, a ten minute video consultation could direct them for community point-of-care troponin testing (and discharge) at a diagnostic hub, face to face assessment in ED or Same Day Emergency Care (SDEC), and/or a hot virtual cardiology review.
This was system integration in practice – not grand policy statements, but seamless patient journeys that made clinical sense.
2029 brought the hub-and-spoke networks
Led by stroke teams, specialist virtual consultations increased 300% as GPs, urgent treatment centres, ambulance services and community hospitals accessed expertise through video rather than transfers. Having seen how resource sharing could function on a shoestring in several African emergency care systems, I knew this model could transform UK emergency medicine.
2030 introduced the balanced performance measurement system
The National Speciality Scorecard incorporated outcomes, safety, equity, and system integration into job planning. These were centred around a two-tier framework. Tier 1: specialty-specific productivity measurements that captured how different specialties create value; and Tier 2: intensity coefficients based on objective workforce wellbeing data.
For the first time, productivity was measured based on why emergency medicine became a specialty in the first place: identifying, coordinating and managing time critical, life threatening risk, thus prioritising care whilst preventing unnecessary attendances, diagnostics and cost. Not just processing patients quickly.
Moving beyond the four-hour target obsession felt revolutionary. Our stats that year told the story:
- Emergency attendance growth slowed to 2% annually (from 5-7% historical),
- Virtual ED handled roughly 35% of traditional presentations,
- Patient satisfaction: 6.8/10 to 8.4/10,
- Staff retention: 73% to 89%, and
- Burnout rates below pre-2019 levels for the first time since the pandemic.
Global leadership (2031-2035): Coming full circle
By 2032, we weren’t talking about crisis management anymore – we were just doing emergency medicine well. The first cohort of new trainees started to slot into consultant posts, with five more trainee cohorts at various stages of training filling rotas. Combined with standardised less than full-time job planning, retention rates reached 89%. Sustainable staffing was finally within reach. Nurse staffing followed suit.
International recognition followed
Korean delegations, Australian physicians, European system leaders – they came to learn from what we’d built. As someone who’d spent years advocating for this type of emergency medicine on the global stage, seeing the UK become the destination for international learning felt like coming full circle.
We published the Manchester Emergency Medicine Transformation Study in NEJM in 2033:
- 40% reduction in emergency presentations,
- 60% improvement in patient experience,
- 45% reduction in preventable admissions, and
- 87% of trainees would “definitely recommend” emergency medicine (vs 34% in 2025).
But the statistic I’m proudest of? Trainee hesitation to escalate sick patients reduced to below 5%, and burnout to below 15% – the lowest rates on record. We had fixed the culture that nearly broke us.
What I learned (that might be useful)
1. Culture trumps technology every time. All the AI and virtual systems wouldn’t have worked without psychological safety. The Collaborative Leadership Framework wasn’t management theory – it was the foundation everything else was built on.
2. Global perspectives accelerate innovation. The Netherlands Triage Standard, Australian virtual ED models, and hub-and-spoke networks – we didn’t invent most innovations, we simply adapted and refined proven international approaches. My global health experience of building systems from nothing informed every UK improvement we made.
3. System integration beats departmental optimisation. Moving from reactive to proactive emergency medicine required breaking down silos. Success came from integrating with primary care, community health, and specialist teams – not optimising EDs in isolation.
Looking ahead: The next chapter
As I prepare for retirement in the next five years, I’m optimistic in ways I couldn’t imagine in 2025. But challenges remain. Climate change, aging populations, persistent inequalities – the work continues.
The incoming generation inherits the system we built, but they’ll transform it further. The challenges of 2045 will differ from 2025, requiring different solutions. But having seen transformation happen once, I’m confident it can happen again.
Final thoughts
The decade from 2025 to 2035 proved that healthcare transformation is possible when crisis meets opportunity and leadership embraces change. We didn’t just survive – we built something better than what existed before.
As I finish writing (it’s past midnight now; some habits die hard 😉), our night team is managing Saturday’s influx with the calm competence that would have been impossible in 2025. A consultant is teaching ultrasonography to a trainee. The virtual ED team handles cases across our network. The AI system quietly ensures nothing slips through cracks.
This is what success looks like: not the absence of emergencies, but systems that predict, prevent, and expertly manage whatever comes through our doors.
Working in resource-limited settings taught me that good emergency medicine isn’t just about having resources – it’s about using them wisely. The UK’s transformation proved that wisdom applied to well-resourced systems can create something truly extraordinary.
We went from system failure to global leadership. The specialty that nearly broke us became the one that showed the world what’s possible.
And that’s not a bad legacy for someone who originally wanted to be an orthopaedic surgeon.
