Pre Hospital Thoracotomy: Data from 21 years at London's Air Ambulance

Background

In recent years, we have learnt a great deal about the management of traumatic cardiac [or circulatory?] arrest (TCA). We now know that resuscitation for it is not as futile as once thought, we have information to suggest that closed chest compressions may not be as beneficial as once thought, and we know that aggressive attention to the potentially reversible causes (usually using the HOTT prompt) is important. Traditionally we have grouped patients in TCA depending on their mechanism of injury – blunt trauma vs penetrating trauma. This is an easy and quick distinction, but is it as clinically valuable as we think?

With huge institutional experience of pre hospital resuscitation thoracotomy (RT) for TCA, London’s Air Ambulance have published their findings, having analysed the data of over 600 patients from 21 years.

Abstract

Objective

To evaluate the association of prehospital resuscitative thoracotomy with survival outcomes for TCA

Design, Setting and Participants

This retrospective cohort study examined all cases of prehospital resuscitative thoracotomy for TCA in London from January 1999 to December 2019. Data were analyzed from July 2022 to July 2023.

Main Outcomes and Measures

The primary outcome was survival to hospital discharge. Secondary outcomes included survival to hospital admission and neurological status at discharge.

Results

Prehospital resuscitative thoracotomy was undertaken in 601 patients with out-of-hospital TCA. The median (IQR) age was 25 (20-37) years; 538 (89.5%) were male and 63 (10.5%) female. A total of 529 patients (88.0%) had a penetrating mechanism of injury. TCA occurred at a median (IQR) of 12 (6-22) minutes after the emergency call, with 491 arrests (81.7%) before the advanced trauma team’s arrival. TCA was the result of cardiac tamponade (105 patients, 17.5%), exsanguination (418 patients, 69.6%), and exsanguination combined with cardiac tamponade (72 patients, 12.0%). Thirty patients (5.0%) survived to hospital discharge, with a favorable neurological outcome observed in 23 survivors (76.6%). Survival varied significantly with the cause of TCA: 22 of 105 patients (21%) with cardiac tamponade, 8 of 418 patients (1.9%) with exsanguination, and none of the 72 patients with combined or other pathologies survived. There were no survivors beyond 15 minutes of TCA for cardiac tamponade and 5 minutes after exsanguination. Multivariable analysis revealed that the cause of TCA (adjusted odds ratio [aOR], 21.1; 95% CI, 8.1-54.7; P < .001), duration of TCA (aOR, 20.9; 95% CI, 4.4-100.6, P < .001), and absence of the need for internal cardiac massage (AOR, 0.2; 95% CI, 0.06-0.5; P = .001) were independently associated with survival.

Conclusions and Relevance

TCA occurs soon after injury, with only a brief window available for effective intervention. This study found that resuscitative thoracotomy is feasible in a mature, physician-led, urban prehospital system and is associated with improved survival for patients with out-of-hospital TCA, particularly when caused by cardiac tamponade, in situations where other treatment options are limited.
Perkins ZB, Greenhalgh R, ter Avest E, et al. Prehospital Resuscitative Thoracotomy for Traumatic Cardiac Arrest. JAMA Surg. Published online February 26, 2025. doi:10.1001/jamasurg.2024.7245.

What Kind of Study is This?

This is a retrospective cohort study, classified as a service evaluation. It included all of the patients who underwent pre hospital RT by the LAA team during the study period. Data were extracted for the patients from the electronic patient record system, patient monitor recordings, and an additional narrative report completed by the clinical team for thoracotomy cases.

Tell Me About the Patients

601 patients underwent pre hospital RT in the 21 year study period. This represents 1.3% of the total number of patients attended by the service, and a high proportion compared to other pre hospital critical care services in the UK. Full demographics are included in the paper, but the headlines are: median age of 25 years (IQR 20 – 37), 89.5% male, 88% penetrating trauma (predominantly stabbing, very few gunshots, which we know have higher mortality).

What Were the Measured Outcomes in This Study?

Primary: Survival to hospital discharge

Secondary: Survival to hospital arrival, and neurological outcome at hospital discharge

What Are the Main Results?

There are so many pieces of fascinating operational information and physiology here, that we’ll tackle them in sections:

Timing: Most patients arrest prior to the HEMS team’s arrival. Median of 12 minutes from 999 call to arrest, and 20 minutes from 999 call to HEMS team on scene. This becomes relevant when you see the survival data…
Pathophysiology: 70% exsanguination, 17.6% cardiac tamponade, 12.1% both (2 cases = other causes)
Timeline: Duration of TCA was very significant for survival. 15.7% < 1 minute, 9.3% 1-5 minutes, 2.6% >5-10 minutes, 0.8% > 10 minutes
Survival: 1.6% exsanguination/other, 21% cardiac tamponade. 14.3% witnessed (by HEMS team), 2.8% unwitnessed. 16.1% PEA, 1.8% asystole/agonal. 2.4% with pre RT CPR, 13% no CPR.

In summary on survival, the optimum set is witnessed arrest from tamponade, in PEA, no CPR given (although the incidence of CPR is probably related to the TCA being unwitnessed by the HEMS team)

Vertical lines within the violin plots indicate the median (IQR). The shaded background indicates the median (IQR) time from the emergency call to arrival of the HEMS team

What happened to the patients?

Overall there were 30 patients (5%) who survival to hospital discharge, the majority of whom (76.6%) had favourable neurological outcome (CPC 1-2). What this 5% doesn’t reflect though, is the substantial differences between the groups, which needs much closer attention.

What are the differences? Is it penetrating trauma vs blunt trauma?

Short answer – no it isn’t. The majority of patients had penetrating trauma, and exsanguination pathophysiology, with its dismal 1.6% survival. This would suggest that simply having penetrating trauma within the accepted timeframes should not necessarily imply candidacy for RT. Another interesting point is that the survival of patients with cardiac tamponade (21%) was the same, whether this tamponade was caused by penetrating or blunt trauma (20.8% vs 25%, p > 0.99). That said, the majority of patients with cardiac tamponade developed it due to penetrating trauma (101 vs 4 patients with cardiac tamponade from blunt trauma). So, most penetrating trauma patients had exsanguinated, but most cardiac tamponade patients developed it after penetrating trauma. 91.4% had penetrating thoracic injuries, 9.5% penetrating epigastric, and none anywhere else. Thus – injury anywhere other than the chest or epigastrium only ever resulted in exsanguination. There is lots more data in the supplemental online content about this, including specific locations of cardiac injuries. I strongly recommend that you review it as it is very generous of the authors to include so much detail.

For tamponade patients, survival decreased with each minute of TCA, and there were no survivors beyond 15 minutes of TCA.
For exsanguination patients, survival deteriorated very rapidly, and there were no survivors beyond 5 minutes of TCA.

Electrical activity seems to be helpful. For tamponade patients, 48.3% of survivors had PEA and 12.3% had agonal/asystole. For exsanguination patients, all survivors had PEA.

Blood transfusion: it would follow that this would be very helpful for exsanguination. It did not significantly improve survival from exsanguination TCA: 2.2% (with blood) vs 1.6% (without) p = 0.73. I remember one of the authors, Zane Perkins, describing exsanguination TCA as ‘the heart has beat itself to death.’ These numbers certainly support that, especially given that transfusion was not associated with survival.

Should We Change Practice Based on This Study?

The usual caveats of reading it for yourself, and of the limitations of retrospective studies (selection bias, data loss etc) all apply, but despite these it would be very hard to say we should not change practice based on this excellent publication. It’s a huge number of patients who will have received very consistent care, and the differences identified are stark. I would suggest that the question is not should we change practice, but how should we, and what should we change?

Firstly: we need ways to get this level of intervention (pre hospital critical care team) to the patients sooner. A complex problem, made all the more complex for the majority of services which, unlike London’s Air Ambulance, do this fairly infrequently.

Next, we probably need to evolve our decision making from the easy penetrating vs blunt, to a more challenging tamponade vs exsanguination. We also need to do this reliably and in a way which does not add substantial time (or any time) which we know is precious. Ultrasound seems like the obvious solution, but it’s time expensive, and risks image acquisition and interpretation errors. What is clear is that penetrating injuries which are not thoracic or epigastric have not caused tamponade in this dataset, and we should consider these patients to have exsanguinated. In contrast, there are rare cases of tamponade from blunt trauma, who need to be identified and treated with RT, rather than risk excluding them based on mechanism of injury – perhaps this is where there is a role for ultrasound. It is probably reasonable to assume that a patient in TCA with penetrating thoracic or epigastric injury has tamponade. The data does show that actually the majority of these patients have exsanguinated, but the risk of missing tamponade or taking too much time to identify it is too great to ignore.

Thinking about external vs internal cardiac massage: survival was substantially lower for patients who received external CPR prior to RT – this makes sense as it implies a delay between cardiac arrest and RT. More interestingly, not requiring internal cardiac massage was independently associated with survival: 15.8% no massage vs 3.5% massage. This implies that we should be intervening with RT before the heart has stopped, and relieving tamponade early, whilst the heart is still beating. This does then beg the question, what is the indication for RT if not traumatic ‘cardiac’ arrest? Cardiac tamponade with severe cardiogenic shock? This is one of the reasons I contextualise this condition as traumatic circulatory arrest – probably the optimum time for intervention is when there is a lack of meaningful circulation, but the heart is still beating.

For patients who have exsanguinated, we are probably using RT too frequently and too late. We should think carefully about identifying these patients in two ways: an alive group who are yet to lose cardiac output and need extremely prompt management, and a group who have lost meaningful circulation for which RT and blood transfusion is probably futile (exsanguination, beyond 5 minutes, without organised electrical activity). LAA are evolving the management of the former group with a ‘Damage Control Prehospital Care’ (DCPC) approach aiming to maximise pre hospital intervention whilst decreasing pre hospital duration. The few potential survivors of exsanguination TCA probably have organised electrical activity and a very recent cessation of circulation. These patients probably need large volume blood product transfusion, and aortic occlusion to prevent the heart from arresting. See the St Emlyn’s blog or paper on P-PRO for more information on this physiology. Whether the method for aortic occlusion is REBOA, lateral thoracotomy, or clamshell thoracotomy, needs more evaluation.

Speed is really of the essence for these patients – the graphs are striking. Look again at the decline in survival for tamponade patients between 0 minutes (~60%) and 5 minutes (~30%). If this patient arrested in a tight spot, and you chose to extricate them prior to RT, what is the implication of that extrication time on their survival? The same goes for patients on ambulances – be honest, how many minutes does it take to offload them, whilst attached to the monitor and oxygen, in order to do the RT outside? Not to mention that outside may be cold, dark, and in the public gaze; the back of an ambulance is none of these things, albeit a little cramped. This is an intervention which needs to be done immediately, and we need to learn to adapt to the situation, rather than adapt the situation to us.

Summary

Pre hospital RT for TCA is feasible. The mechanism of injury is probably overstated, and what is really important is the pathophysiology: cardiac tamponade vs exsanguination. Identifying this early will identify extra potential survivors (blunt tamponade was up to 15% of blunt trauma patients in TCA in another pre hospital paper) and, also importantly, identify futility. There has been mission creep with RT for TCA, and now this data shows that it does not appear to be an effective treatment for exsanguination TCA, we should acknowledge that.

Suspect tamponade with penetrating trauma to the thorax or epigastrium. Intervene as early as possible, within 15 minutes of arrest, and don’t delay to adapt the situation

Suspect exsanguination with penetrating trauma elsewhere, or with blunt trauma. Rule out tamponade where possible – this may require ultrasound or even ‘diagnostic’ RT. Intervention needs to be immediate, else it is futile.

Learn to recognise futility and do not feel compelled to intervene. Injuries (physical & psychological) are risks to ourselves, colleagues, and ‘bystanders’ from this procedure.

The headlines:

Patients are 30 times more likely to survive TCA from tamponade than exsanguination
Patients are 20 times more likely to survive TCA if RT is delivered in less than 5 minutes
Non-chest penetrating injury has to be exsanguination
RT does not appear to be an effective treatment for exsanguination TCA
Chest penetrating injury has 40% chance of being tamponade. RT indicated, but be prepared to stop if there is no tamponade
Unloading a patient in TCA from an ambulance to undertake RT can be the time difference between good neurological survival and death

Hutch