On the night of the 2017 London Bridge attack, police officers found themselves treating seriously injured people while trying to protect each other with little more than their batons. They didn’t know whether the attackers might return, and they couldn’t be sure when medical help would arrive. Elsewhere at the scene, a paramedic faced multiple critically injured patients, some of whom needed immediate treatment and evacuation. He knew the danger hadn’t passed, but he also knew that waiting for the usual ambulance response could have serious consequences.
One of the decisions he made was to put a badly injured woman into a police car and send her directly to hospital. She arrived at King’s College Hospital 37 minutes after being injured, at least 20 minutes before the first casualty left the casualty clearing station. She had sustained major internal haemorrhage and required blood transfusion and emergency surgery.
It’s a remarkable example of decision-making under pressure, although we need to be careful about the lessons we draw from it. The answer cannot simply be that rescuers should take more risks. The 2019 Chief Coroner’s report into the attacks identified problems with inflexible arrangements for designating dangerous areas, but also concluded that those arrangements had not resulted in the loss of an opportunity to save any of those who died. It acknowledged that revised Joint Operating Principles had already addressed the main concerns in 2019, while calling for that flexibility to be reflected in other procedures and training.
The difficult question is how we balance the very real risks of acting against the consequences of waiting. Claire explored this at Tactical Trauma in Sweden in 2024. With the conference returning on 2–4 November 2026, I’ve been revisiting her talk and the interviews Liz Crowe and I recorded there. They cover everything from terrorist attacks to prehospital extracorporeal CPR and ECMO in trauma, but all ask something familiar to anyone making clinical decisions under pressure: how much do we need to know before we act?
How much information do we need before we act?
One of the ideas Claire discusses in our original recording about providing care in the hot zone is the 40/70 rule. The suggestion is that if you have less than 40% of the available information, you probably don’t know enough to make a sensible decision. Once you’ve gathered more than 70%, however, you may already have waited too long.
I wouldn’t suggest for a moment that these are scientifically established thresholds. Nobody is standing in resus calculating whether they’ve reached 63% certainty, and there is no validated clinical rule that tells us precisely when to intervene. It’s a heuristic, intended to encourage us to think about the balance between gathering information and the cost of delaying action.
We recognise the problem in emergency medicine, even when the circumstances are rather less dramatic than a terrorist attack. At the beginning of an assessment, particularly in a seriously ill or injured patient, we often have very little information. Further examination, investigations and observation may reduce our uncertainty, but some decisions simply cannot wait for all of those things to happen.
Simon has written about this previously on St Emlyn’s in his article on time-critical decision-making and finding the right moment to intervene. Waiting for information that could change management is different from waiting because we’d feel more comfortable having it.
Take a patient with life-threatening haemorrhage who needs definitive haemorrhage control. We may be able to gather more physiological measurements, undertake further imaging or involve additional specialists, but each of those steps takes time. Depending on the patient’s condition, some will be entirely appropriate and others may delay the treatment they need. The decision involves judging what we gain from additional information against what the patient stands to lose while we obtain it.
The other side of that judgement deserves just as much attention. Acting prematurely can expose a patient to an unnecessary or dangerous intervention. In the setting Claire describes, it can also place rescuers and other casualties at risk. The challenge is to recognise when information is sufficient for a particular decision, while remaining prepared to change course as circumstances evolve.
Of course, recognising that problem doesn’t necessarily help when you’re standing in front of multiple casualties, frightened, overwhelmed and unsure where to begin. And at a major incident, the person making the initial clinical decisions may not be a clinician at all.
Making the right decision easier
That was the starting point for Sean Brayford-Harris, a London Ambulance Service paramedic who helped develop the Ten Second Triage tool.
Sean had spent years working with the Metropolitan Police, trying to improve the way police officers and clinical services worked together. One of the problems he identified was that the medical responsibilities of police officers were often quite different from the responsibilities for which they had originally joined the service.
If an officer arrives first at a major incident, their initial responsibility may be to confront the threat and protect the public. Yet within minutes they may also find themselves providing lifesaving treatment to several casualties, potentially without immediate ambulance support. We expect them to move between those roles while managing their own safety and making decisions that would challenge experienced emergency clinicians.
Traditional major incident triage systems frequently relied on physiological measurements. Respiratory rates, pulse rates and other assessments make sense when clinical staff have the time, training and circumstances to obtain them. They become considerably less useful when somebody is trying to assess multiple casualties in the dark, in the rain, wearing protective equipment and worrying about whether the scene is safe.
Sean’s team approached the problem by looking at what first responders were already trained to do. Rather than adding more complicated assessments, they tried to develop a structure that would support existing lifesaving skills.
The resulting Ten Second Triage tool, published by NHS England, uses straightforward observations to identify and prioritise casualties. It considers whether someone can walk or talk, whether there is severe bleeding or a penetrating injury, and whether they are breathing. It also prompts immediate actions, including controlling major haemorrhage and opening an airway. It deliberately avoids measuring pulse or respiratory rate and uses a ‘not breathing’ category rather than asking a non-clinical responder to declare someone dead.
The tool is intended to support initial triage and lifesaving interventions by those first on scene. Its ‘not breathing’ category is not an instruction to withhold CPR: the tool says to attempt CPR if resources allow. NHS responders use the NHS Major Incident Triage Tool (MITT) for further triage, with reassessment as the situation changes.
There is an important evidence distinction here. A 2025 evaluation of Ten Second Triage described simulated major incident exercises involving police, fire and ambulance personnel. Participants generally found the system usable and reported confidence in applying it, although the study also identified challenges around communication, training and interagency working. It was an evaluation of usability in exercises, rather than evidence that introducing the tool has improved survival in real incidents.
One of my favourite parts of Sean’s original interview is his description of police officers who had completed the training and subsequently attended a multiple-casualty road traffic collision. They were understandably frightened, but afterwards described how having an established framework helped them know what to do.
There is something useful in that for all of us. When we develop guidelines, cognitive aids and clinical pathways, we have to consider whether they will actually work for the people expected to use them. A beautifully designed algorithm is of limited value if it demands more information, expertise or cognitive capacity than its user has available.
Sean’s work also illustrates why designing these systems with the people who will use them is so valuable. The training included officers describing their own experiences to other officers, rather than simply having healthcare professionals tell them what they should do. That may seem a small difference, but it is a sensible way to make training credible and relevant to the people undertaking it.
When the right decision is to say no
Up to this point, we’ve largely been discussing decisions about doing something. There is another side to clinical decision-making, particularly as our treatments become increasingly sophisticated.
What happens when we have the ability to deliver an extraordinary intervention, but believe it isn’t appropriate for the patient in front of us?
Alice Hutin works with the prehospital extracorporeal CPR service in Paris. Her team can deliver eCPR, using veno-arterial extracorporeal membrane oxygenation (VA-ECMO) to provide circulatory support in selected patients with refractory cardiac arrest, before they reach hospital.
It’s an impressive service, and in our original interview with Alice we explored the considerable organisation required to make it work. Yet one of the aspects that particularly interests me is what happens when the team arrives and decides not to proceed.
The scene may already contain ambulance crews, firefighters and relatives who know that a specialist team has been called. There may be considerable expectation that something exceptional is about to happen. Alice describes the difficulty of explaining that, despite the availability of eCPR, the circumstances of the arrest mean the patient is unlikely to benefit.
Those decisions are not straightforward. The duration of the arrest, whether it was witnessed, the quality of CPR, the patient’s underlying condition and the likelihood of a reversible cause all inform the assessment. Cannulation and extracorporeal support also introduce significant risks and resource demands.
The evidence doesn’t give us an easy answer. Randomised trials of eCPR have produced differing results. Differences in patient selection, treatment systems and the interventions being compared make those results difficult to generalise. The Resuscitation Council UK’s 2025 adult advanced life support guidelines state that eCPR may be considered for selected patients when conventional CPR is failing, in settings where it can be implemented. This is a selective rescue treatment. NICE guidance also stresses specialist services, multidisciplinary patient selection and review of outcomes. Neither recommendation establishes that starting eCPR at the scene is better than starting it in hospital.
We’ve also discussed this on St Emlyn’s in Simon’s recent article about optimising conventional resuscitation before eCPR.
Alice’s account is particularly useful because she describes what her team does when the answer isn’t obvious. They discuss difficult cases together, have colleagues available for a second opinion and routinely debrief afterwards. She explains how simply describing the patient and their resuscitation out loud can make the decision clearer.
I think that’s a practice we could all make more use of. A decision that seems reasonable when it remains in our own head can sound rather different when we have to explain the reasoning to somebody else. We may realise that we’re relying heavily on an assumption, overlooking a relevant finding or allowing the expectations of those around us to influence the decision.
The conversation doesn’t end when the team decides against eCPR. Alice describes the need to explain that judgement to the crews who have been resuscitating the patient and to the relatives who may have hoped the arrival of the specialist team would change the outcome.
Sometimes we need considerable confidence to intervene, but we also need the confidence to explain why an intervention is unlikely to help.
Building a team around the decision
There is a similar problem in Chris Bishop’s work exploring VA-ECMO for patients who develop severe cardiogenic shock following major trauma.
These are exceptionally difficult patients. Chris is interested in a particular group who have suffered major haemorrhage, undergone surgery or interventional treatment to control the bleeding, but subsequently develop significant cardiovascular dysfunction. Despite achieving haemorrhage control, their circulation may remain profoundly compromised.
VA-ECMO offers the possibility of supporting the failing circulation while the patient recovers, but it introduces considerable problems of its own. Extracorporeal circuits are associated with bleeding and thrombotic complications, and anticoagulation may be required to maintain circuit patency. In somebody who has only recently survived major haemorrhage, the consequences of further bleeding could be disastrous.
Reduced or delayed anticoagulation has been reported in selected trauma patients, but this does not remove the risk of bleeding or circuit thrombosis. The 2020 systematic review and a 2023 review and meta-analysis include heterogeneous, largely observational evidence. Much of it concerns venovenous (VV) ECMO for respiratory failure, a different problem from the circulatory shock Chris is investigating. Those outcomes cannot simply be transferred to trauma VA-ECMO.
When we recorded the interview, Chris described four patients who had received VA-ECMO through his team’s work, two of whom had survived. Those cases demonstrate that the treatment can be delivered in selected patients, but without a comparator and with such small numbers, they cannot tell us whether ECMO improved survival.
What I particularly like about Chris’s approach is the way the team has organised itself around these decisions.
Rather than expecting an individual clinician to decide whether to initiate VA-ECMO, they developed a rapidly convened multidisciplinary ‘trauma shock call’. Specialists can review the patient’s injuries, haemorrhage control, evidence of cardiac dysfunction and potential risks before agreeing whether ECMO is appropriate.
It allows people with different expertise to contribute. The surgeon may understand the security of haemorrhage control, the intensive care team the physiological consequences of shock, and the perfusion and ECMO specialists the technical possibilities and limitations of extracorporeal support.
There is still uncertainty, and the conference call cannot manufacture evidence that doesn’t exist. However, it provides an opportunity to identify assumptions, challenge decisions and make the reasoning explicit.
There’s a wider application here beyond ECMO. In emergency medicine we routinely depend on colleagues from other specialties when making difficult treatment decisions. Yet we don’t always create the conditions for a useful shared discussion, particularly when time is short. Simply contacting somebody for permission or agreement isn’t quite the same as asking them to help work through the problem.
Can we get the right people into that discussion quickly enough, especially when the treatment is unfamiliar or depends on several specialties working together?
Decisions under pressure in our own departments
Most of us will never have to make the decisions described during the London Bridge attack, and many will never be involved in prehospital eCPR or trauma VA-ECMO. But the circumstances that make those decisions difficult are very familiar.
We frequently have to act before investigations are complete, decide how much risk is acceptable, determine whether a treatment is likely to help and explain our reasoning to colleagues who may see things differently.
A useful place to start is identifying the information that would actually change our decision. If a patient is deteriorating and we are waiting for another investigation, we should be able to explain what we expect that result to tell us and how it will influence management. If we cannot, it is worth questioning whether the delay is justified.
We can make greater use of simple, rehearsed approaches for circumstances where there is little time to think. That means practising how a decision will be made, rather than just how an intervention will be performed. A team preparing for major haemorrhage, for example, can discuss in advance who will make decisions about definitive haemorrhage control, what information they need and how concerns should be raised.
We can also get into the habit of explaining our reasoning out loud, particularly when the decision is difficult. Asking another clinician what they think, and giving them room to disagree, may expose problems that we would otherwise miss.
Afterwards, our reviews need to consider the information available at the time rather than judging the quality of the decision solely by the eventual outcome. Knowing what happened is a considerable advantage that the original decision-maker didn’t have.
Listening again to Claire, Sean, Alice and Chris, I was struck by how much of their work happens before the difficult decision: rehearsing, agreeing an approach and making sure there is someone to call. None of that guarantees the right answer, but it gives us something to draw on when time is short.
The next time we’re waiting for another result, it may help to ask what it could change and how long the patient can afford to wait. If the answer is still unclear, explain the problem to a colleague and invite them to disagree. That seems a pretty good place to start.
About the guests
Dr Claire Park is a consultant with London’s Air Ambulance and in intensive care and anaesthesia at King’s College Hospital. Her military experience and work with the Metropolitan Police have informed her research into improving casualty care during high-threat incidents.
Sean Brayford-Harris is a London Ambulance Service paramedic who worked closely with the Metropolitan Police on clinical interoperability. He was part of the multidisciplinary team involved in developing Ten Second Triage.
Dr Alice Hutin is an emergency physician involved in the SAMU de Paris prehospital eCPR programme. Her research has included refractory cardiac arrest, extracorporeal resuscitation and post-cardiac-arrest care.
Chris Bishop is a clinical perfusionist with experience in mechanical circulatory support. At the time of the interview, he was undertaking doctoral research at the Centre for Trauma Sciences, Queen Mary University of London, investigating VA-ECMO for cardiogenic shock following major trauma.
References
1. Chief Coroner. London Bridge and Borough Market Terror Attack: Prevention of Future Deaths Report. November 2019. Particularly the discussion of hot and warm zones and the need for flexibility in emergency response.
2. NHS England. Clinical guidelines for major incidents. National guidance on Ten Second Triage and MITT.
3. Davidson L, Vassallo J, Cowburn P, et al. Evaluating Ten Second Triage: A Novel Multi-Agency Prehospital Triage Tool for Major Incidents. Journal of Contingencies and Crisis Management. 2025;33(1):e70025. Evaluation of simulated multi-agency exercises and responder usability.
4. Resuscitation Council UK. Adult Advanced Life Support Guidelines 2025. Includes recommendations on the selective use of eCPR.
5. Extracorporeal Life Support Organization. Extracorporeal Cardiopulmonary Resuscitation in Adults: Interim Guideline Consensus Statement. ASAIO Journal. 2021;67:221–228.
6. Wang C, Zhang L, Qin T, et al. Extracorporeal membrane oxygenation in trauma patients: a systematic review. World Journal of Emergency Surgery. 2020;15:51. doi:10.1186/s13017-020-00331-2.
7. Zhang Y, Zhang L, Huang X, et al. ECMO in adult patients with severe trauma: a systematic review and meta-analysis. European Journal of Medical Research. 2023;28:412. doi:10.1186/s40001-023-01390-2.
8. NICE. VA ECMO for extracorporeal cardiopulmonary resuscitation (ECPR) in adults with refractory cardiac arrest. HealthTech guidance HTG765. 22 October 2025.
About Tactical Trauma
Tactical Trauma is an international conference held in Sundsvall, Sweden, bringing together doctors, paramedics, police officers, military medics and rescue services to share their experiences of delivering care in challenging and sometimes hostile environments. The programme covers prehospital trauma care, resuscitation, major incidents, tactical medicine and human performance, with a particular emphasis on learning from real-world experience. One of the things I particularly enjoy about the conference is the opportunity to hear from people working in very different environments, but often facing remarkably similar challenges. The conversations featured in this podcast were recorded at Tactical Trauma 2024, and the conference returns on 2–4 November 2026, with workshops on the first day and the main conference on 3–4 November.
Learning from podcasts?
If podcasts form part of your CPD, MedPod Learn brings listening, notes, questions and reflection into one place across a range of medical podcasts. I created the app; podcast browsing and listening are free, while the additional learning and CPD features are part of a paid subscription.
