Triaging and Referring In Adjacent General and Emergency Departments: a six-year follow-up study after a cluster randomised trial

This six-year follow-up study demonstrates that a nurse-led triage tool successfully redirecting low-acuity emergency department patients to a co-located general practitioner cooperative remains a safe, efficient, and sustainable intervention in routine practice without dedicated post-trial support.

The Gist

Imagine the Emergency Department (ED) of a hospital as a busy, high-speed train station. It's designed for people with urgent, life-threatening emergencies—like a heart attack or a broken leg from a car crash. But often, the station gets clogged with people who just need a simple ticket check or a minor repair, like a stomach ache or a small cut. These "minor" passengers slow down the train, making it harder for the "urgent" passengers to get on board quickly.

In Belgium, researchers tried a clever solution: a "Side-Door" Detour.

The Experiment (The Original Trial)

Back in 2019, a team of researchers set up a pilot program at a train station in Antwerp. They placed a specialized traffic controller (a nurse using a smart checklist called the "eMTS") right at the entrance.

• The Rule: If your problem was urgent (like a heart attack), you went straight to the main platform (the ED).
• The Detour: If your problem was minor (like a sore throat or a mild stomach ache), the traffic controller gently guided you through a side door to a General Practitioner Cooperative (GPC). Think of the GPC as a cozy, nearby repair shop right next to the train station, run by local doctors.

The trial worked well. It was safe, and it helped clear the main station. But the big question was: What happens when the experiment ends? Usually, when a research project finishes, the special tools are packed away, and everyone goes back to the old way of doing things.

The Six-Year Follow-Up (The Real-World Test)

This paper is the "six-year later" report. The researchers wanted to see if the side-door detour was still open, or if it had been boarded up.

The Good News:
The side door is still wide open, and it's actually busier than before!

• The Traffic Flow: Over six years, about 6% of all people arriving at the ED were successfully sent to the repair shop (GPC). That's like diverting 6 out of every 100 passengers to the side shop, freeing up space for the critical cases.
• The Surge: Interestingly, after the pandemic (when things were chaotic), the system bounced back and became even more efficient, sending 11% of patients to the GPC in recent years.

The Safety Check (Did anyone get hurt?):
Critics might worry: "What if the traffic controller sends someone to the repair shop who actually needs the hospital?"

• The researchers checked the files of everyone who was sent to the GPC but then had to be sent back to the ED because the doctor there said, "Actually, you need the big hospital."
• The Result: Only 3% of the diverted patients needed to come back. That's actually lower than the rate of people who just walked into the GPC on their own (5%).
• The "Big Mistakes": In the six years, there were only four serious "safety scares" (major triage issues) where a patient might have been delayed in getting critical care. For context, in the original trial, there was one fatal error. The fact that the system has remained safe for six years without a dedicated research team watching over it is a huge win.

The "Appendix" Problem

There was one specific headache: Stomach aches.
The researchers found that the most common reason people got sent back from the GPC to the ED was to rule out appendicitis (a burst appendix).

• The Metaphor: It's like sending a car to a mechanic to check if the engine is seized. The mechanic says, "I can't be 100% sure, take it to the big garage."
• The study found that while the GPC doctors were good at spotting most stomach issues, they still sent about 24% of the "suspected appendix" cases back to the ED for surgery. While this didn't cause harm (the patients were still safe), it meant they had to travel back and forth, wasting time. The authors suggest the "traffic controller's" checklist needs a software update specifically for stomach pain.

Why Did It Stick? (The Secret Sauce)

Usually, when a research grant runs out, the project dies. Why did this one survive?

1. It Was Built into the System: The checklist wasn't a piece of paper; it was built into the computer software the nurses used every day. You couldn't easily ignore it.
2. Everyone Liked It: The nurses and doctors felt less stressed because the "side door" actually worked. The main station wasn't as crowded.
3. No Extra Cost: The hospital didn't lose money by sending people away, and the local doctors didn't lose patients. Everyone was happy.

The Bottom Line

This study proves that you can build a "side door" for minor emergencies, and if you design it right, it can stay open for years.

It's like installing a smart traffic light that automatically directs slow cars to a side street. Even after the engineers who installed it left, the light kept working, the traffic flowed better, and no one crashed. The only thing left to fix is making the light smarter about stomach aches so fewer people have to drive back and forth.

The Lesson for the Future: If you want a new medical idea to last, don't just test it for a year and hope it sticks. Build it into the daily routine, make sure the staff likes it, and plan for the long haul before you even start the experiment.

Technical Summary

1. Problem Statement

Emergency Department (ED) overcrowding is a persistent challenge in European healthcare systems. While General Practitioner Cooperatives (GPCs) exist to handle out-of-hours (OOH) primary care, they have not significantly reduced ED utilization. Previous interventions to redirect low-acuity patients from EDs to primary care have shown mixed results regarding long-term sustainability and safety. Specifically, there is a lack of evidence regarding whether triage protocols designed for clinical trials can be successfully integrated into routine practice over extended periods without dedicated research resources.

2. Methodology

This study is a six-year retrospective observational follow-up of a cluster randomised controlled trial (RCT) conducted in Antwerp, Belgium (the original "TRIAGE trial," 2019).

• Intervention: A nurse-led triage tool based on an extended Manchester Triage System (eMTS). This tool allows ED nurses to refer patients with low urgency (MTS categories 4 and 5) to a co-located GPC, provided they do not require hospital-specific resources (e.g., labs, imaging, suturing).
• Study Period: Data was collected from March 1, 2019, to February 9, 2025 (approx. 6 years).
• Data Sources:
  • Routine Clinical Data: Extracted from the GPC's electronic health record (Mediris Wachtpost) and the ED's system (E.care).
  • Case Review: A detailed retrospective review of medical records for all patients referred from the GPC back to the ED ("back-referred").
• Study Populations:
  • Primary Outcome Group: Patients triaged at the ED and successfully managed at the GPC.
  • Referred Back Group: A subset of the primary group who were sent back to the ED by the GPC physician.
  • Control/Comparison: The overall GPC population and non-diverted ED patients.
• Outcomes Measured:
  • Efficiency: Proportion of ED patients diverted to GPC; proportion of GPC patients originating from ED.
  • Safety: Rates of back-referral to ED; clinical characteristics of back-referred patients; identification of major or minor triage errors (defined by potential for increased morbidity/mortality).
• Ethical Considerations: Ethical clearance was obtained with a waiver for individual informed consent.

3. Key Contributions

• Long-term Sustainability Evidence: Unlike most studies that measure immediate post-intervention effects, this study provides a 6-year longitudinal analysis of a triage intervention after the conclusion of the original RCT and the cessation of research funding.
• Real-World Implementation: It demonstrates that a complex triage protocol can be successfully embedded into routine clinical workflows without a dedicated post-trial implementation plan or continued research resources.
• Safety Validation: Through a rigorous case review of 155 back-referred patients, the study provides robust evidence regarding the safety of diverting low-acuity patients to primary care, identifying specific failure points (e.g., abdominal pain).
• Implementation Science Insights: The paper highlights the importance of stakeholder engagement and integration into routine software as key drivers for sustainability, even in the absence of formal implementation frameworks.

4. Results

Efficiency and Adoption:

• Diversion Rate: Out of 110,941 triaged ED patients, 6,722 (6.1%) were managed at the GPC.
• GPC Contribution: These diverted patients accounted for 11% of all GPC consultations.
• Trends: Diversion rates dipped during the COVID-19 pandemic (dropping to 3-4%) but recovered and increased to 9-11% in 2023–2025.
• Patient Profile: Diverted patients were slightly younger (mean age 30 vs. 31) and had slightly shorter wait times (23.5 min vs. 24.4 min) compared to the general GPC population. Their clinical urgency was comparable to the general GPC population, with 0% classified as "Highly Urgent" or "Life-threatening."

Safety (Back-Referrals):

• Back-Referral Rate: Only 3% (185/6,722) of diverted patients were referred back to the ED, which is significantly lower than the 5% back-referral rate for the general GPC population.
• Clinical Outcomes of Back-Referrals:
  • 83% were treated as outpatients; only 17% required hospitalization.
  • Triage Errors: A detailed case review identified 4 major triage issues (e.g., unstable angina, arterial occlusion, active labor) and 18 minor issues. Crucially, no evidence of increased morbidity or mortality was found due to the delay.
  • Appendicitis: Abdominal pain was the most common reason for back-referral (29 cases). Of these, 7 were confirmed appendicitis (24% positive predictive value), but none were judged as triage errors because the time loss (average 13 minutes) did not compromise outcomes.

Operational Changes:

• The intervention survived the loss of research funding and staffing support.
• Organizational changes occurred, such as expanding GPC hours to weeknights during the pandemic and restricting referrals between 00:00 and 06:00 starting in 2023, yet the system remained functional.

5. Significance and Conclusion

• Clinical Feasibility: The study confirms that nurse-led eMTS triage is a safe and sustainable method for reducing ED overcrowding. The intervention not only persisted but improved in efficiency over six years.
• Policy Implications: The authors argue that successful healthcare interventions should be designed with explicit post-trial implementation strategies from the outset. However, this study proves that even without such a plan, interventions embedded in routine practice with strong stakeholder buy-in can achieve long-term sustainability.
• Future Directions: While the system is safe, the study identifies abdominal pain (specifically ruling out appendicitis) as a persistent challenge for diversion algorithms. Future protocols should refine these specific flowcharts to reduce unnecessary back-referrals.
• Broader Impact: This research fills a critical gap in the literature by moving beyond short-term RCT results to demonstrate the viability of primary care diversion models in real-world, long-term settings.