Improving decision-making for prehospital Resuscitative Thoracotomy in traumatic cardiac arrest: a data-driven approach

Abstract

Prehospital Resuscitative Thoracotomy (RT) can be life-saving in traumatic cardiac arrest (TCA), particularly in patients with cardiac tamponade. Yet selecting who may benefit is challenging, as survival depends on two often uncertain factors: the underlying aetiology and the duration of arrest. Drawing on extensive prehospital RT experience, we propose a pragmatic framework based on two simple clinical surrogates. Injury location provides a useful surrogate for the likely cause of arrest, helping direct the initial resuscitation strategy, while presenting ECG rhythm offers a rapid marker of physiological viability when timelines are unclear. Combining these surrogates offers a practical decision aid: likely tamponade should prompt immediate RT when an organised rhythm is present, or when arrest duration is short even in asystole, while cases where tamponade is unlikely should prioritise transfusion and haemorrhage control, with RT reserved for those in whom tamponade is subsequently confirmed. This perspective highlights how simple bedside information, supported by governance, can guide timely, pathology-specific care and maximise the benefit of this life-saving intervention.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s13054-025-05705-z.

Background

The challenge of prehospital RT decision-making

Prehospital Resuscitative Thoracotomy (RT) can be a life-saving intervention in traumatic cardiac arrest (TCA), particularly when caused by cardiac tamponade [1–4]. However, identifying which patients may benefit from RT in the prehospital setting is challenging. Survival depends largely on two factors: the aetiology of arrest (most commonly tamponade or exsanguination), and the duration of arrest, which determines physiological viability [1]. Both are usually uncertain during prehospital resuscitation.

This diagnostic uncertainty has important consequences. In our analysis of over 600 prehospital RTs, survival was high in patients with tamponade who underwent immediate RT, but markedly lower when RT was prioritised in patients with exsanguination. RT is the definitive treatment for tamponade and must be delivered rapidly to be effective. In contrast, patients with exsanguination require haemorrhage control and blood transfusion as primary interventions. Prioritising RT in these cases risks delaying the treatments that are most likely to restore circulation and may worsen outcome. RT may, in select cases, still have a role as an adjunct—for example, enabling thoracic haemorrhage control or temporary aortic occlusion—but only in patients who do not respond to aggressive volume resuscitation and standard prehospital haemorrhage control methods [1, 4].

As prehospital transfusion becomes more widely available, pathology-specific resuscitation strategies are increasingly feasible before hospital arrival. Yet the lack of reliable diagnostic tools remains a major challenge. While point-of-care ultrasound (POCUS) may support aetiology assessment [5], its role in TCA is limited by time pressure, operator skill, and evolving evidence [6]. Importantly, no validated decision aid currently exists to guide prehospital clinicians in determining which patients should undergo RT, and when [2, 4, 7, 8].

This perspective draws on a large, prospectively maintained cohort of prehospital RTs to explore how simple clinical surrogates of aetiology and arrest duration may help guide decision-making. Our aim is to outline a structured, evidence-informed approach that may help clinicians rapidly determine the most appropriate resuscitation strategy based on the likely cause of TCA and the patient’s physiological viability.

Diagnosing TCA and the initial priorities

London’s Air Ambulance (LAA) approach to managing penetrating trauma emphasises minimising on-scene interventions and prioritising rapid transfer to the nearest Major Trauma Centre (MTC). Resuscitative Thoracotomy (RT) is only undertaken in patients with confirmed or imminent traumatic cardiac arrest (TCA), where immediate intervention is required. The technical aspects of the procedure have been described previously [9].

Because prehospital RT is only appropriate in TCA, it is important to first clarify how TCA is defined. In prehospital practice, TCA is defined pragmatically as circulatory arrest following trauma [2, 10]. It is a clinical diagnosis and refers to a patient who is no longer able to generate effective cardiac output, resulting in unconsciousness, agonal or absent breathing, and no palpable central pulse [11]. This includes patients with organised cardiac activity but insufficient vital organ perfusion to sustain critical functions such as consciousness, respiration, and cardiac contractions — a low-flow, peri-arrest state that rapidly progresses to death if not treated immediately. Early recognition enables early intervention, which is essential for survival [1]. Even brief delays in initiating effective resuscitation are associated with a rapid decline in the probability of survival and waiting for complete asystole risks missing the narrow window in which resuscitation is possible.

Successful TCA resuscitation depends on the rapid identification and correction of reversible causes. The HOTT framework (Haemorrhage control, Oxygenation, relief of Tension pneumothorax, and relief of Tamponade) [12] provides a useful structure for this, but because TCA resuscitation is so time-critical, effective treatment ultimately requires prioritising interventions according to the most likely underlying pathology. In practice, this means immediate chest decompression to relieve suspected tension pneumothorax, immediate RT when tamponade is likely, or prioritising blood transfusion and haemorrhage control when exsanguination is suspected. Tailoring the sequence of interventions to the pathology offers the best chance of rapidly restoring circulation and improving survival.

Aetiology: using injury location to distinguish tamponade from haemorrhage

When suspected, tension pneumothorax can be rapidly treated by thoracostomy [13]. Accordingly, bilateral thoracostomy is the first step in our protocol before RT. The greater challenge is distinguishing tamponade from exsanguination. Because the underlying cause of TCA determines the most appropriate treatment—RT for tamponade versus transfusion and haemorrhage control for exsanguination—it is vital that clinicians can estimate the likely cause as early as possible.

We examined whether the location of penetrating injuries could predict arrest aetiology. Each wound was mapped to one of nine predefined torso regions and compared to the pathophysiological cause of TCA (eSupplement 1). This analysis showed that the surface location of wounds is a strong and clinically useful surrogate for arrest aetiology, consistent with seminal trauma literature defining the ‘cardiac box’ as a high-risk region for cardiac injury (Fig. 1; eSupplement 2). However, its predictive value may vary with mechanism.

Fig. 1.

Association between injury location and underlying cause of traumatic cardiac arrest (TCA). Body surface regions were mapped from operative and post-mortem data. Shading indicates the proportion of patients in each region with tamponade T and exsanguination E. Injury location of injuries was strongly related to TCA etiology

In penetrating trauma, wound location was strongly associated with the underlying cause of arrest. Injuries within the anterior ‘cardiac box’ and epigastrium carried the highest probability of tamponade, often as the sole pathology. By contrast, penetrating injuries to other thoracic regions were only rarely associated with tamponade, and those below the diaphragm (outside the epigastrium) were invariably associated with exsanguination. In the neck, tamponade was not observed in zone 2 or 3 injuries and was rare in zone 1, where exsanguination predominated.

These findings suggest that when wounds are located outside the cardiac box or epigastrium, isolated tamponade is unlikely, and initial management should prioritise transfusion and haemorrhage control. If tamponade is subsequently identified, RT can then be performed. Conversely, injuries within these high-risk regions should prompt early consideration of tamponade and immediate RT.

Ballistic trauma presents a different challenge. While such injuries may cause cardiac injury, isolated tamponade is less common than in knife-related trauma. In our subgroup analysis, wound location was a weaker but still informative predictor: nearly half of cardiac box or epigastric stab wounds produced isolated tamponade, compared with only around a quarter of ballistic injuries, where exsanguination predominated (eSupplement 2). This likely reflects both the higher energy transfer and the greater likelihood of the projectile injuring multiple cavities. For this reason, even in central chest wounds, exsanguination should be regarded as the more likely cause of TCA and treated first, with RT reserved for cases where tamponade is subsequently identified.

In blunt trauma, exsanguination was the most common cause of TCA, and outcomes were poor when RT was prioritised. Although uncommon, isolated tamponade was observed in around 5% of patients, consistent with previous reports [14, 15] and these patients had good outcomes following RT, comparable to tamponade after penetrating injury [1]. Unlike penetrating trauma, however, body surface injury patterns do not reliably predict tamponade following blunt mechanisms. Initial treatment should therefore focus on exsanguination, with clinical assessment and point-of-care ultrasound used to identify tamponade before RT is considered.

Taken together, these insights suggest that injury location can be a valuable surrogate for estimating the likely cause of TCA, particularly in knife-related penetrating trauma. Its predictive value is lower in ballistic and blunt mechanisms, where tamponade is less common and the role of surface injury is limited. Nonetheless, recognising when tamponade is unlikely can help clinicians prioritise transfusion and haemorrhage control, while reserving RT for those most likely to benefit.

Viability: ECG rhythm as a surrogate for time in arrest

Estimating the duration of TCA is challenging in the prehospital environment, yet it is central to determining whether RT may still be effective. Difficulties arise because low-flow states may mimic TCA, timelines are hard to determine rapidly during resuscitation, and events are prone to recall and cognitive bias [9, 16, 17]. However, survival is inextricably linked to the duration of arrest, with previous analyses showing a potential window of up to 15 min for tamponade and under 5 min for exsanguination [1].

In this context, presenting ECG rhythm can serve as a practical surrogate for arrest duration and physiological viability (eSupplement 3). Across our cohort, cardiac activity deteriorated in a predictable sequence over time: organised sinus rhythm in the first minute after arrest, sinus bradycardia between 2 and 5 min, agonal rhythms with wide QRS complexes predominated by 6–10 min, and asystole was the most frequent rhythm beyond 10 min of arrest (Fig. 2, eSupplement 4). These patterns were consistent across tamponade and exsanguination, and align with contemporary clinical data showing predictable deterioration of ECG activity during cardiac arrest [18].

Fig. 2.

Predictable deterioration of ECG rhythm with increasing duration of Traumatic Cardiac Arrest (TCA). Forest plot representing mean (95% CI). A Cardiac tamponade (n = 69) and B exsanguination (n = 273). Presenting rhythm is a practical surrogate for arrest duration when timelines are uncertain

Presenting rhythm was also strongly associated with survival (Fig. 3). Organised rhythms were associated with higher survival, particularly when tamponade was the cause. In contrast, agonal rhythms and asystole typically reflected longer arrest duration and poor survival. Notably, around one in ten tamponade patients presenting in asystole survived, whereas no patient with exsanguination survived once the rhythm had deteriorated to asystole.

Fig. 3.

Survival and neurological outcome in traumatic cardiac arrest (TCA) by presenting rhythm. A Cardiac tamponade (n = 69) and B exsanguination (n = 273). Survival was strongly associated with presenting rhythm

These observations suggest that ECG rhythm can provide a rapid, clinically useful surrogate for physiological viability when timelines are uncertain. In practice, the presence of an organised rhythm should always support an aggressive resuscitative approach, while agonal or asystolic rhythms in exsanguinated patients are strong markers of futility in the prehospital setting [1, 19, 20].

Towards an integrated decision framework

By combining these two readily available clinical surrogates—injury location to estimate the likely cause of TCA, and presenting rhythm to estimate arrest duration—it is possible to create a pragmatic and evidence-based framework to guide prehospital RT decision-making. This approach balances the potential benefits of early RT against the risks of delaying treatment of exsanguination.

When tamponade is the likely cause of TCA (based on injury location) and the patient still has an organised ECG rhythm or has been in arrest for less than 15 min, immediate RT should be considered. In contrast, when isolated tamponade is unlikely (as with penetrating injuries outside the cardiac box or epigastrium, ballistic injuries or blunt trauma), initial management should prioritise exsanguination: rapid blood transfusion, haemorrhage control, and expedited transfer to definitive surgical care. Performing RT first in these patients is harmful, as it delays the interventions needed to restore circulation. By contrast, prioritising volume resuscitation— even if tamponade is present—may transiently improve preload and physiology, providing a window to assess for tamponade. If tamponade is subsequently identified by clinical assessment or POCUS, RT can then be performed. [10, 21, 22]

For exsanguinated patients who present in asystole or with an agonal rhythm, survival in the prehospital setting is highly improbable, and RT is unlikely to change the outcome. In practice, RT for exsanguination is only likely to offer benefit if arrest occurs in the presence of the team. Framing decision-making around these simple surrogates therefore supports prehospital teams to deliver the right intervention for the right pathology at the right time (Fig. 4).

Fig. 4 .

Integrated decision support for prehospital resuscitative thoracotomy (RT) in traumatic cardiac arrest (TCA). Framework combines suspected aetiology (estimated from injury location) with presenting rhythm (surrogate for time in arrest)

Generalisability, limitations, and ethical considerations

This framework is derived from the experience of a high-volume, urban prehospital service with over three decades of delivering resuscitative thoracotomy, and a case mix dominated by penetrating trauma, particularly knife-related injuries [1, 9, 23]. While this provides a uniquely detailed dataset, it inevitably limits external applicability. In systems where blunt trauma is more common, or where ballistic injuries predominate, the decision framework may be less reliable [11, 24−26]. Our framework focused on the difficult distinction between tamponade and exsanguination as underlying aetiologies. Other reversible causes of TCA, such as tension pneumothorax, critical hypoxia from airway compromise, or traumatic asphyxia, also require prompt recognition and treatment. Our analysis was limited to patients within the London trauma system attended by LAA, but the potential for selection bias is low, with a dispatch miss rate of only 0.6% and an operational miss rate of less than 10%.

Generalisability is also constrained by system-level differences. LAA operates with physician-led teams, established governance and training processes [27], and immediate access to prehospital blood transfusion — conditions that are not universal. As a surgical procedure, RT can only be performed by physicians, and even then, should only be undertaken within a well-governed, highly trained system. As such, the decision support described here may not be applicable to prehospital systems without physician involvement or equivalent governance structures.

Alongside these practical considerations, ethical dimensions must also be acknowledged. RT is a rare, invasive, resource-intensive intervention that carries risks for patients, clinicians, and bystanders. Performing it in cases without realistic potential for benefit may expose teams to both physical and moral injury [28], while also delaying other resuscitation priorities. Clear governance and well-defined decision-support frameworks are therefore essential—not only to maximise patient benefit, but also to safeguard teams and maintain public trust.

Conclusion

Prehospital resuscitative thoracotomy remains one of the most challenging decisions in trauma care. Survival depends on rapid recognition of both the underlying pathology and the patient’s physiological viability. Injury location and presenting ECG rhythm provide simple, readily available surrogates to guide these judgements, helping clinicians prioritise the right intervention for the right pathology at the right time.

This perspective outlines a pragmatic framework to support those decisions, derived from substantial urban prehospital experience. Its value lies in sharpening indications for RT while ensuring that alternative priorities, such as transfusion and haemorrhage control, are not delayed. Importantly, this framework, and the data on which it is based, has already informed changes to our service’s practice, shifting from broad mechanism-based triggers to pathophysiology-based decision-making, embedded within formal governance and training processes (see eSupplement 5 for RT SOP). Ultimately, the safe use of prehospital RT requires not only timely decision-making but also robust governance, training, and system-level support to maximise benefit while minimising risk.

Abbreviations

ILCOR

International liaison committee on resuscitation

LAA

London’s air ambulance

MTC

Major trauma centre

POCUS

Point of care ultrasound

RT

Resuscitative thoracotomy

TCA

Traumatic cardiac arrest

Author contributions

EtA, LK, ZBP: Conceptualization, Data curation, Formal Analysis, Writing – original draft; DJL, MDC, TH: Conceptualization, Writing – original draft; DB, CA: Data curation, Writing – review & editing.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The data collected and analysed during the course of this observational cohort study are available from the corresponding author on reasonable request. All requests for data access will be considered by the corresponding author and the research team, pending approval from the institutional review board. The data will be shared in a de-identified format.

Declarations

Ethics approval and consent to participate

The project met local criteria for, and was registered as a service evaluation, waiving the need for full ethics committee oversight in alignment with UK Health Research Authority guidance. The study protocol was prospectively registered (UIN: researchregistry6529) [11] and received approval from the institutional Clinical Effectiveness Unit (Registration number: 10445, Bart’s Health NHS Trust).

Competing interests

The authors declare no competing interests.