Survival effect of prehospital emergency anaesthesia with intubation in risk-stratified major trauma

Author: Kaleb Lachenicht, Chief Clinical Officer of ROCKET and Director Epic EM

Nelson et al., Lancet Respiratory Medicine, 2026

Why Prehospital Emergency Anaesthesia With Intubation Matters to Us in HEMS and Critical Care Retrieval

For years, prehospital RSI has sat in the uncomfortable space between dogma and doubt, and although this paper does a lot to work through some of this... I'm not completely sure we are out of the dogma or doubt just yet.

• RCT evidence? Almost nothing

• Observational data? Confounded and inconsistent.

• Policy decisions? Often ideological or economic.

This study is a little different

It uses high-dimensional machine learning and doubly robust causal modelling on 6,467 UK major trauma patients to estimate the survival effect of prehospital intubation, and it provides the strongest evidence to date that in the right patients, prehospital intubation improves survival.

It still provides this evidence in a system that is VASTLY different to the one we practice in in the South African context, and it still is not using actual data, rather modelling systems to try and predict the possible benefit.

Study Design: What They Actually Did

Dataset

• UK TARN registry

• 6,467 patients

• 30-day mortality outcome

Key Concept

Rather than asking:

“Do intubated patients do better?”

They asked:

“In patients who are predicted to require early intubation, what is the causal survival effect of prehospital intubation?”

That distinction matters.

The Modelling Approach

They built:

1. Early intubation prediction model

2. 30-day mortality prediction model

3. Used doubly robust causal estimation

The Key Result

In High-Risk Patients (n=229):

Conditional Average Treatment Effect (CATE):

• Equivalent to a 10.3% reduction in 30-day mortality

That translates to:

• 28 additional survivors per 229 high-risk patients

Scaled nationally:

• 170 lives saved per year in the UK

• Economic value ≈ £101 million annually (this is the interesting part for me - as we don't always think of the economic benefit or drain on the system)

There are some important nuances

If you intubate everyone:

• Average treatment effect = –1.56%

• Much smaller benefit

The benefit of pre-hospital emergency anaesthesia with intubation is not universal!! We can't make blanket statements like "all trauma patients should be intubated (duh...) but this paper does make the argument that for the correct patients we could be making a difference.

That aligns with what experienced HEMS clinicians intuitively understand:

The question is not “RSI or no RSI.”

The question is “Which patient benefits from RSI now?”

Who Were the High-Risk Patients?

Shapley feature analysis showed strongest predictors of needing early intubation:

• Low GCS

• TXA administration (suspected major haemorrhage) when administered for the right patients

• Age of patient

• Fall mechanisms

• Physiological instability

For mortality:

• Age (older patients tended to fall into the higher risk groups)

• GCS (lower GCS as a surrogate for cerebral function)

• Hypoxia

• No airway support

This is clinically coherent. The model is not magic, it reflects recognisable physiology and is almost the expected outcome... it reflects the experiences of the critical care retrieval environment.

Why This Is Different From Previous Evidence

Previous meta-analyses suggested ED intubation might be safer, but those studies:

• Did not adequately control for injury severity

• Were heavily confounded

• Often mixed paramedic RSI systems with physician-led RSI

This study attempts to address confounding using propensity modelling with a more robust estimation model, and propensity matching to find the right answers for the right patients.

Limitations (And We Must Be Honest because there ALWAYS are)

• Single UK trauma centre (does this relate to our system?)

• Physician-paramedic critical care model (this is vastly different to the model we use in SA and within the system I work)

• 30-day mortality only (not neurological outcome or morbidity status)

• Does not include prehospital deaths not admitted

• Assumes time-fixed treatment and effect

This does not prove RSI works everywhere, it starts to answer the question of where RSI and with whom RSI might work.

What This Means for System Leaders

This paper starts and move the conversation

It suggests:

• Prehospital RSI is not just philosophically justified, it may be physically and clinically justified for the right patient groups in the right situations

• It is economically justified.

• It may save lives at scale comparable to trauma centre implementation.

Systems will need to think about:

• Funding

• Staffing

• Scope of practice

• Dispatch criteria

  
  

The Intub-8 Model

In this paper they also developed a simplified 8-feature model suitable for point-of-care use.

Imagine:

• Dispatch tools identifying high-risk airway patients

• Targeted activation of critical care teams

• Data-guided prehospital airway decision support

This is something that we have been discussing heavily in our system - trying to figure out which patients will benefit the most from a HEMS activation for the purpose of skills and equipment to the patient on the road-side. When we send such an expensive resource we want to be sure that it will be the most useful and efficient use of the recourse possible for the best possible patient outcome.

What I’m Thinking About in ROCKET's Context

From a HEMS leadership perspective:

1. Are we identifying the right airway patients early enough?

2. Is our dispatch aligned with physiology, not geography?

3. Could we use structured risk stratification?

4. Are we measuring survival effect in our own system?

5. Do we have enough governance to safely scale RSI capability?

Because this paper implies something important:

Under-provision of advanced prehospital airway capability may be a preventable mortality driver. That’s uncomfortable (especially within the paradigm of the current airway thinking), but it is powerful.

In Conclusion

• Risk-stratified prehospital intubation appears to reduce 30-day mortality by ~10% in high-risk trauma patients.

• Blanket RSI for everyone is not beneficial! (like any blanket treatment recommendation there is nothing that is likely to be helpful for everyone).

• Targeted RSI, in mature systems, may save lives at population scale.

• This is compelling evidence, supporting prehospital emergency anaesthesia in major trauma, but there is still a LOT of work to be done to bring a model based paper like this to the real world.

If you run a retrieval service or trauma system:

How are you currently deciding who gets prehospital RSI?

Would be very interested to hear how different systems are approaching this?

Contact Kaleb on kaleb@rockethems.co.za