Feasibility and outcomes of anaesthesiology- and intensive care-led ECPR in a hospital without cardiac surgery: a 2.5-year prospective registry

Abstract

Background

Extracorporeal cardiopulmonary resuscitation (ECPR) has been demonstrated to yield positive outcomes in cases of refractory cardiac arrest when implemented within strict time frames. The majority of ECPR programmes are operational within the confines of cardiac surgery centres, with the assistance of surgical teams and perfusionists for the processes of cannulation and circuit management. The present study evaluated the feasibility, safety and outcomes of an anaesthesiology-intensive care-led ECPR programme delivered in a university hospital without on-site cardiac surgery.

Methods

A prospective registry study was conducted of consecutive adults who underwent extracorporeal membrane oxygenation (ECMO) for refractory out-of-hospital cardiac arrest at a single supraregional centre from May 2023 to October 2025. The selection of patients followed predefined criteria adapted from Extracorporeal Life Support Organization (ELSO) guidance. All cannulations were performed percutaneously under ultrasound guidance by anaesthesiology and intensive care specialists. The primary outcomes of the study were survival to hospital discharge and 3-month survival with favourable neurological status, as measured by the Cerebral Performance Category Scale (CPC) score, which ranges from 1 to 2. Secondary outcomes encompassed the success rate of cannulation, the time interval between cannulation and the occurrence of arrest, and periprocedural complications.

Results

A total of 42 patients were included in the study, with a median age of 48 years (IQR 22). The male: female ratio was 83.3:16.7. The most prevalent aetiologies were acute myocardial infarction (38.1%), electrical storm (21.4%), and pulmonary embolism (14.3%). Cannulation was performed in the emergency department (ED) (57.1%; n = 24), the intensive care unit (ICU) (31.0%; n = 13), and the cardiac catheterisation laboratory (Cath-lab) (11.9%; n = 5). The median cannulation time was 15 min (IQR 6), and the median arrest-to-flow time was 59.5 min (IQR 32). It was found that all cases achieved an ECMO flow rate of at least 3.0 L·min⁻¹. At the 3-month mark, the overall survival rate was recorded at 28.6% (12 out of 42 patients), while the favourable neurological outcome, categorised as CPC 1–2, was documented at 23.8% (10 out of 42 patients). Cannulation-related complications occurred in 4.8% (2/42) of cases, and no instances of pneumothorax, cardiac tamponade, aortic dissection, or mechanical ECMO complications were observed. Peripheral cannulation-site bleeding occurred in 31.0% (13/42), distal limb ischaemia in 38.1% (16/42), and these were managed conservatively. Renal replacement therapy was required in 26.2% (11/42) of cases. In 7.1% of cases (3/42), veno-veno-arterial (VVA) ECMO conversion was necessary due to inadequate venous drainage.

Conclusions

The implementation of a protocolised, ultrasound-guided, anaesthesiology-intensive care-led ECPR programme is a feasible prospect for a centre that does not have on-site cardiac surgery or perfusionists. This model achieved universal cannulation success, rapid time to ECMO support, and neurological outcomes within ranges reported in the literature.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12871-026-03678-2.

Introduction

Despite considerable progress in advanced life support and post-resuscitation care, survival after refractory cardiac arrest remains poor, with registry data showing hospital discharge rates typically below 10% [1]. ECPR is defined as the initiation of veno-arterial (VA) ECMO during ongoing resuscitation. It has emerged as a promising option for carefully selected patients who fail to achieve return of spontaneous circulation with conventional cardiopulmonary resuscitation. The extant literature has demonstrated that, when performed within defined time frames, ECMO can significantly improve survival and neurological outcomes. This finding has been replicated in both randomised trials and large registries [2–5].

Conventionally, ECPR programmes have been situated in cardiac surgery centres, where teams benefit from the support of cardiac surgeons and perfusionists during cannulation and circuit management [6, 7]. In such cases, the development of ECMO proficiency is typically initiated in the context of post-cardiotomy ECMO cases, where cannulation is performed in controlled circumstances and time is not a limiting factor. In many centres, this experience facilitates subsequent implementation of ECPR; however, ECPR programmes may also be developed and maintained in non-cardiac surgery settings. In addition, well-functioning ECMO teams led by emergency physicians have also been reported, demonstrating effective ECPR delivery outside cardiac surgery centres [8].

Data from the ELSO registry [9], supported by observational studies including that of Belle et al. from Annecy, France [10], suggest that ECMO initiation for ECPR may be undertaken by appropriately trained non-surgical teams in centres without on-site cardiac surgery. This approach requires clearly defined, protocolised pathways, the use of percutaneous cannulation techniques, and dedicated team training. Such models may be particularly relevant in healthcare systems where the majority of cardiac arrests occur outside specialised centres and where delays associated with inter-hospital transfer can limit timely access to extracorporeal support.

In May 2023, the university hospital launched a programme in anaesthesiology and intensive care led ECPR. The programme was operated without on-site cardiac surgery or perfusionists’ support. The programme was conceived as a systematised, protocol-based initiative that is supported by interdisciplinary collaboration, simulation-based preparation, and the availability of an anaesthesiology-intensive care cannulation team on a 24/7 basis.

The present study aims to evaluate the feasibility, safety, and clinical outcomes of this non-surgical ECPR model, and determine whether such a programme may serve as a scalable framework for implementing extracorporeal resuscitation in hospitals without cardiac surgical facilities.

Methods

Setting

The University Hospital in Kraków is a supraregional academic centre with 1,287 beds, and it functions as the referral hub for refractory cardiac arrest in the region. Notwithstanding its multidisciplinary tertiary profile, the hospital has no on-site cardiothoracic surgery facilities and no perfusionists. In response to this limitation, a protocol-driven, non-surgical ECMO pathway was developed. This pathway is led by anaesthesiology and intensive care specialists. In May 2023, a 24/7 ECPR programme was initiated, incorporating a predefined activation algorithm and a dedicated multidisciplinary cannulation team that has been trained to facilitate expeditious ECMO initiation. All consecutive patients undergoing ECPR from the programme’s inception were prospectively registered. The present analysis encompasses the initial 2.5 years of programme activity (May 2023-October 2025).

Qualification protocol

The patient eligibility criteria were determined by the application of a set of predefined inclusion and exclusion criteria. These criteria were adapted from the guidance provided by ELSO and published protocols from high-volume ECMO centres. The inclusion criteria that were employed in this study were as follows: the occurrence of witnessed cardiac arrest; the administration of cardiopulmonary resuscitation by a bystander; a no-flow time of ≤ 5 min; an age of ≤ 65 years; and an estimated collapse-to-extracorporeal flow time of ≤ 60 min. The exclusion criteria encompassed instances of unwitnessed arrest, the presence of terminal comorbidities, irreversible brain injury, active major bleeding, and a do-not-resuscitate status. The end-tidal carbon dioxide (EtCO₂) value of less than 12 millimetres of mercury (mmHg) during cardiopulmonary resuscitation (CPR) of a high quality was designated as an additional exclusion parameter. An EtCO₂ threshold of 12 mmHg was selected to enhance the sensitivity of identifying patients who may be amenable to rescue therapy. This was undertaken with consideration for variability associated with mechanical chest compression devices and ventilation strategies, consistent with thresholds documented in preceding observational ECPR studies [11]. The on-call ECPR team leader was responsible for making the final eligibility decision. This decision was made after a bedside assessment, handover from emergency medical services, and review of initial arterial blood gas results. Patients with significantly low arterial oxygen partial pressure (PaO₂) or a markedly elevated lactate level (> 18 mmol/L) on initial blood gas analysis were also excluded, as these findings indicated a low likelihood of favourable recovery. No predefined PaO₂ cut-off was utilised for ECPR qualification; arterial oxygenation was interpreted in conjunction with PaCO₂ values within the context of the overall clinical condition by an experienced anaesthesiology-intensive care specialist. In patients ultimately qualified for ECPR, the lowest PaO₂ values were observed in cases of deep hypothermia and were approximately 30 mmHg. By contrast, patients who were disqualified typically presented with profoundly severe hypoxaemia, with PaO₂ values below 10 mmHg.

Team composition and cannulation technique

In Poland, anaesthesiology and intensive care medicine specialists are trained as a single, combined specialty. The ECPR team comprised two anaesthesiology and intensive care specialists responsible for vascular access, a senior specialty trainee overseeing airway and ventilation, a paramedic coordinating advanced life support, and an ICU nurse handling equipment setup. The ECMO circuit was prepared in advance and then transported to the cannulation site, which could be either the ED, the ICU, or the Cath-lab. Furthermore, an ECMO cabinet was assembled and equipped with all necessary apparatus for cannulation, advanced monitoring, and the preparation of VA ECMO patients for transportation.

All ECMO cannulations were performed percutaneously under real-time ultrasound guidance using a modified Seldinger technique, with arterial (15–19 Fr) and venous (19–23 Fr) femoral cannulae employed. Cannulation and circuit initiation were performed exclusively by anaesthesiology-intensive care specialists with extensive experience in ultrasound-guided percutaneous vascular access, including routine placement of permanent dialysis catheters, vascular ports, and management of difficult vascular access. Consequently, percutaneous cannulation techniques were well established and continuously refined within the team. The programme incorporated a range of educational approaches to reinforce the knowledge and skills required for specific ECMO procedures, including cannulation, circuit priming, troubleshooting, and ECMO console management. These approaches included repeated high-fidelity simulation, dedicated ECMO courses, and in-hospital multidisciplinary medical simulations, both prior to and during programme implementation. Moreover, the team’s proficiency in the management and troubleshooting of ECMO consoles has been further refined through the provision of routine clinical care to patients experiencing acute respiratory failure who have been treated with veno-venous (VV) ECMO.

No cardiac surgeons or perfusionists were involved in any part of the procedure. Following connection to the circuit, the pump speed was increased stepwise until ECMO flow rate reached a minimum of 3.0 L·min⁻¹. At this point, chest compressions were terminated. The final ECMO flow was then titrated to the right upper limb arterial pressure, with the objective of achieving a MAP (mean arterial pressure) of 60–80 mmHg.

Post-ECMO management

Following ECMO initiation, the resuscitation management was conducted in the ICU under the supervision of the anaesthesiology and intensive care team, in conjunction with cardiology and vascular surgery specialists. Post-ECMO advanced care was delivered in accordance with standard institutional pathways, which encompassed the following:

• It is imperative to initiate early hemodynamic optimisation with flow titration to ensure the achievement of adequate MAP,

• Coronary angiography is indicated for the diagnosis of coronary artery disease, and subsequent immediate revascularization is recommended for patients presenting with acute myocardial infarction,

• The management of temperature is to be focused on in a targeted manner,

• Continuous invasive monitoring is to be employed,

• It is imperative to implement judicious fluid management and vasopressor adjustment.

During the ECMO run, the attending anaesthesiology-intensive care physician, in conjunction with a dedicated ICU nurse who had received training in ECMO management, performed continuous circuit and console surveillance. This surveillance included monitoring of flows, pressures, alarms, and oxygenator performance, and was conducted using structured daily checklists.

At the time of initial cannulation, distal perfusion cannulation was not a routine procedure; rather, a selective approach was employed, with the decision to proceed based on a combination of serial physical examination, duplex ultrasound assessment of lower limb vessels, and tissue oximetry. This approach was adopted only in the event of clinical signs indicating limb ischaemia. When indicated, distal perfusion cannulation was performed percutaneously under ultrasound guidance by the anaesthesiology-intensive care specialist, with 24/7 vascular surgery support available in case of technical difficulties or complications.

ECMO weaning was performed when native cardiac output was sufficient to maintain perfusion with decreasing ECMO flow, as confirmed by echocardiography and haemodynamic stability. The decannulation procedure was carried out in the operating theatre by a vascular surgeon.

Data collection and outcomes

All ECPR cases were prospectively recorded in a local registry from the programme’s launch. The present analysis was conducted by extracting demographics, arrest circumstances, prehospital care, and procedural variables. These included cannulation time (from skin preparation to established ECMO flow) and the total arrest-to-flow interval (collapse to ECMO flow of 3 Lmin^− 1). Furthermore, EtCO₂ measured during advanced life support was also captured and utilised as part of the qualification process. The primary outcomes were defined as survival to hospital discharge and 3-month survival with favourable neurological status, which was defined as a CPC score of 1–2.

Secondary outcomes included: periprocedural complications (bleeding, pneumothorax, cardiac tamponade or aortic dissection), the cannulation success rate, timing measures (cannulation time and arrest-to-flow interval), in-hospital adverse events and causes of death, classified as multiorgan failure, cardiogenic shock, refractory cardiac arrest, haemorrhagic shock or insufficient ECMO flow.

As a supplementary analysis, an investigation was conducted into the correlation between procedural and outcome data. The data set included cannulation times, arrest-to-flow intervals, short-term survival and 3-month survival with favourable neurological status. These data were then compared with corresponding results from recent landmark randomised trials of ECMO support (ARREST [4], PRAGUE-OHCA [5], and INCEPTION [2]). The purpose of this comparison was to contextualise the performance of a non-surgical programme.

Statistical analysis

Categorical variables are presented as counts and percentages, and continuous variables as mean (standard deviation) or median with interquartile range, as appropriate to their distribution. In consideration of the descriptive and exploratory nature of the study, formal hypothesis testing and comparative statistics were not performed. The analyses were conducted using SPSS^® Version 22 (IBM Corp.^®).

Results

Study population

In the period spanning from May 2023 to October 2025, a total of 42 patients underwent ECMO for refractory out-of-hospital cardiac arrest at our centre. The median age recorded was 48 years (IQR 22), and the proportion of males was 83.3%. The most prevalent underlying causes of arrest were identified as acute myocardial infarction (38.1%), electrical storm (21.4%), and pulmonary embolism (14.3%). The baseline demographics and clinical features of the study population are summarised in Table 1.

Table 1.

Baseline demographics and clinical features

Variable	All patients	Survivors	Non-survivors
Demographics
Age, years-median (Q1-Q3)	48 (38–60)	38 (34–51)	57.5 (41–62)
Sex (male), n(%)	35 (83.3)	8 (66.7)	27 (90)
BMI, kg·m− 2-median (Q1-Q3)	27.7 (24.7–31.1)	27.8 (24.1–31.1)	27.7 (24.7–31.0)
Etiology
Coronary artery disease, n(%)	18 (42.8)	3 (25)	15 (50)
Electrical storm, n(%)	9 (21.4)	2 (16.7)	7 (23.3)
Pulmonary embolism, n(%)	6 (14.3)	3 (25)	3 (10)
Hypokalemia, n(%)	1 (2.4)	1 (8.3)	0
Hyperkalemia, n(%)	1 (2.4)	0	1 (3.3)
Drowning, n(%)	2 (4.8)	0	2 (6.6)
Subarachnoid hemorrhage, n(%)	1 (2.4)	0	1 (3.3)
Abdominal aortic aneurysm rupture, n(%)	1 (2.4)	0	1 (3.3)
Beta blocker overdose, n(%)	1 (2.4)	1	0
Hypothermia, n(%)	1 (2.4)	0	1 (3.3)
Unknown, n(%)	0	0	0
ACLS descriptors
Witnessed, n(%)	42 (100)	12 (100)	30 (100)
Bystander CPR, n(%)	42 (100)	12 (100)	30 (100)
Initial shockable rhythm, n(%)	22 (52.4)	5 (41.7)	17 (56.7)
Mechanical CPR, n(%)	42 (100)	12 (100)	30 (100)
Presence of signs of life, n(%)	2 (4.7)	2 (16.7)	0
Initial lactate, mmol-median (Q1-Q3)	13 (9.9–15.6)	13.75 (8.7–15)	12.35 (9.8–15.8)

Procedural characteristics

All patients underwent percutaneous cannulation, a procedure performed by the ECPR team, which was led by the anaesthesiology and intensive care specialists. The cannulation procedure was performed in three distinct settings: the ED (57.1%; n = 24), the ICU (30.9%; n = 13), and the Cath-lab (11.9%; n = 5). All of the patients included in the study suffered from out-of-hospital cardiac arrest. The location of cannulation was determined by various factors, including the time since cardiac arrest, the clinical circumstances, and logistical considerations, such as the availability of the Cath-lab. Patients presenting with shockable rhythms were preferentially referred to the catheterisation laboratory when available; when it was unavailable, cannulation was performed in the ED procedure room with immediate consultation with an interventional cardiologist during VA ECMO support. Patients with deep hypothermia, as well as patients identified as ECPR candidates during the preliminary telephone assessment and subsequently transported to the hospital with a short collapse-to-arrival time (< 40 min), underwent cannulation in the ICU procedure room. The median cannulation time was found to be 15 min (IQR 6), while the median arrest-to-flow interval was 59.5 min (IQR 32). In all cases, the target ECMO flow rate of 3.0 L·min⁻¹ was successfully achieved.

Survival and neurological outcomes

At the 3-month mark, the overall survival rate was recorded at 28.6% (n = 12). A favourable neurological outcome, defined as a CPC score of 1–2, was achieved in 23.8% (n = 10) of cases.

Secondary outcomes

Cannulation-related complications occurred in 4.8% (n = 2), including one vascular injury requiring immediate arterial compression with successful cannulation of the contralateral groin and one femoral artery perforation treated with endovascular repair. No cases of pneumothorax, cardiac tamponade or aortic dissection were observed. According to the ELSO Complication Codes, no mechanical ECMO circuit failures were reported. No cases of air embolism were observed during ECMO circuit priming or pump initiation. This is notable given that ECMO initiation was performed exclusively by anaesthesiology and intensive care physicians without perfusionist support. Peripheral cannulation-site bleeding was observed in 13 patients (31.0%), constituting one of the haemorrhagic complications. Distal limb ischaemia was observed in 16 patients (38.1%), all of whom were managed conservatively. Eleven patients (26.2%) required renal replacement therapy during their stay in the ICU. In three cases (7.1%), insufficient ECMO flow necessitated the insertion of an additional drainage cannula and conversion to a VVA ECMO configuration.

Comparative analysis of the outcomes: our registry and landmark ECPR studies

The median cannulation time recorded in the present registry was 15 min, which is intermediate between the 7-minute median recorded in the ARREST trial [4] and the 12-minute median recorded in the PRAGUE-OHCA trial [5], and shorter than the 20-minute median recorded in the INCEPTION trial [2]. The median interval between collapse and ECMO flow in the Kraków cohort was 59.5 min, closely aligning with the 59-minute median observed in ARREST [4] and the 62-minute median documented in PRAGUE-OHCA [5]. This interval was found to be significantly shorter than the 74.7-minute median reported in INCEPTION [2]. The survival outcomes observed in Kraków were consistent with those reported in randomised ECMO trials, exhibiting a slight decrease compared to the ARREST trial [4] but demonstrating similarity to PRAGUE-OHCA [5] and superiority to INCEPTION [2]. These findings substantiate the feasibility and efficacy of a non-surgical ECMO model. The ARREST trial [4] reported an overall survival rate of 43% (n = 6) at three months, while PRAGUE-OHCA [5] and INCEPTION [2] did not disclose 3-month overall survival or 3-month neurological outcomes. For the present trials, only 30-day survival with favourable neurological outcome was available, which was 31% (n = 38) and 17% (n = 12), respectively. It is imperative that these comparisons are interpreted with caution due to discrepancies in reporting windows, available endpoints, and sample sizes (Fig. 1).

Fig. 1.

Comparison of clinical outcomes and procedural time intervals across ECPR studies. (A) Survival at 3-month follow-up and survival with favourable neurological outcome (CPC 1–2) comparing the Krakow 2.5-year registry with the ARREST, Prague OHCA, and INCEPTION studies. (B) Cannulation time and time from cardiac arrest to ECMO flow across the same cohorts.Values are presented as reported in the original studies and the current Krakow registry (median [IQR] or mean, depending on source).Notes: For the Prague OHCA and INCEPTION studies, data for survival irrespective of neurological outcome at 3-month follow-up were not available; only survival with favourable neurological outcome (CPC 1–2) at 30-day follow-up was reported. For ARREST, data for neurological outcome at 3-month follow-up were not available. For some studies, interquartile range data were not available.Abbreviations: ECPR, extracorporeal cardiopulmonary resuscitation; ECMO, extracorporeal membrane oxygenation; CPC, Cerebral Performance Category

Discussion

The present study demonstrates the successful implementation and sustainability of an ECPR programme led by an anaesthesiology and intensive care department in a hospital setting, in the absence of on-site cardiac surgery or perfusionists. Over a period of two and a half years, a 100% cannulation success rate was observed, along with encouraging survival outcomes and favourable neurological results. The present findings contribute to the expanding corpus of evidence that suggests ECPR can be organised effectively in non-surgical centres, provided that clear protocols, targeted training, and multidisciplinary governance are in place.

Conventional ECMO pathways characteristically comprise cardiothoracic surgeons and perfusionists within the core team, a model that has been endorsed by expert consensus and has been extensively adopted in high-volume cardiac centres. Perfusionists, as specialists in extracorporeal circulation, are believed to enhance safety and reliability across initiation, troubleshooting, and weaning [6, 12, 13], while surgeons commonly lead vascular access-particularly in complex or central cannulations, and are considered integral to ECMO readiness [12, 14]. Multidisciplinary teams incorporating these roles have been shown to result in favourable haemodynamic control and a reduced incidence of complications in both paediatric and adult settings. Furthermore, surgical availability has been demonstrated to be associated with reduced ICU mortality in post-cardiotomy and perioperative contexts [15, 16]. It has been suggested by economic evaluations that there may be certain advantages to the utilisation of perfusionist-led services in selected settings [17–19]. Nonetheless, reports from hybrid and non-surgical programmes suggest that comparable outcomes can be achieved through the implementation of alternative staffing models. Percutaneous access, guided by ultrasound and performed using a Seldinger technique, has been reported as both efficient and safe in the context of ECMO [20, 21]. In Poland, these disciplines constitute a single, integrated specialty, performed by intensivists or anaesthesiologists. A considerable number of large metropolitan programmes have adopted this approach. For instance, a Milan cohort described approximately 29% survival using exclusively percutaneous cannulation without routine surgical backup [22]. Furthermore, the Paris mobile ECMO service demonstrated successful out-of-hospital initiation by mixed teams, including emergency and intensive care teams [23, 24]. A post hoc analysis combining data from the INCEPTION [4] and PRAGUE-OHCA [5] trials found that percutaneous techniques were associated with shorter procedural times and no increase in major complications when compared with surgical cut-down [2, 10]. Within this framework, our centre initiates ECMO exclusively via percutaneous cannulation performed by anaesthesiology and intensive care specialists. Throughout the duration of the study, cannulation proved to be a successful procedure in all instances. Moreover, the 3-month outcomes aligned with the prevailing ECMO literature. It is important to note that survival with good neurological function has varied across trials. Higher rates have been observed in the ARREST trial [4] and more modest rates in the PRAGUE-OHCA trial [5] and the INCEPTION trial [2]. These variations can be attributed to differences in patient selection, logistics, and time to ECMO support. It has previously been demonstrated that extracorporeal support can be initiated safely in hospitals without on-site cardiac surgery, including by interventional cardiologists, for refractory arrest and cardiogenic shock [23]. The present programme extends the aforementioned concept by demonstrating its feasibility within an anaesthesiology and intensive care-led model. This supports the view that, with appropriate governance, simulation and protocolisation, non-surgical teams can deliver ECMO with outcomes comparable to those reported by experienced centres.

It is imperative to emphasise that the objective of this study is not to advocate for the exclusion of cardiothoracic surgeons or perfusionists from ECPR programmes. Conversely, their proficiency remains indispensable, and close collaboration with surgical services should be maintained whenever feasible. The findings of this study demonstrate that, in well-defined organisational conditions, a team specialising in anaesthesiology and intensive care is capable of safely initiating and managing ECMO in the absence of on-site cardiac surgery. It is important to note that our institution maintains active collaboration with a cardiac surgical centre located within another hospital in the same city. Both hospitals operate under the umbrella of a joint university cardiology institute, ensuring surgical consultation and support whenever required.

Limitations

The present study is subject to several limitations. Initially, the study was conducted as a single-centre observational cohort study. The absence of a contemporaneous control group limits the study’s ability to make causal inferences and consequently, the possibility of excluding residual confounding cannot be determined. Secondly, the sample size was modest, which resulted in wider confidence intervals and an elevated risk of type II error, particularly for subgroup analyses and complication estimates. Thirdly, patient selection followed pragmatic, protocol-based criteria that were applied in real time. This suggests that selection and survivorship biases are possible, and that certain time metrics (e.g. collapse-to-flow interval) relied on estimates or external documentation from emergency medical services. Fourthly, the implementation of process measures and post-resuscitation management (including coronary angiography, targeted temperature management, and anticoagulation protocols) occurred in accordance with local pathways. It should be noted that these measures may not be universally applicable to other systems. Finally, the cross-study comparisons with ARREST [4], PRAGUE-OHCA [5], and INCEPTION [2] should be interpreted with caution due to differences in inclusion criteria, logistics, endpoints, and follow-up durations, as well as incomplete reporting (e.g. missing IQRs or 3-month outcomes in some trials). It is important to note that no formal hypothesis testing was undertaken, which is consistent with the descriptive design. The need for further research is indicated by the requirement for multicentre studies in the future. Such studies would need to be standardised in terms of data collection, and would require larger cohorts and longer follow-up periods. The validation of these findings is paramount.

Conclusions

The implementation of an anaesthesiology- and intensive care-led ECPR programme in a hospital environment is possible without the necessity of on-site cardiac surgery or perfusionists, and can be achieved in a safe and sustainable manner. The findings of this study, based on two and a half years of experience, demonstrate that universal cannulation is successful, and that the complication rate is acceptable. Furthermore, the neurological outcomes are comparable to those reported by established centres. The results of the present study lend support to the hypothesis that protocolised, ultrasound-guided percutaneous ECPR is feasible in non-surgical settings, provided that structured training is in place and collaborative cardiac support pathways are in operation.

Abbreviations

ARREST

Advanced reperfusion strategies for patients with out-of-hospital cardiac arrest and refractory ventricular fibrillation

CPC

Cerebral performance category

CPR

Cardiopulmonary resuscitation

Cath-lab

Cardiac catheterisation laboratory

ECMO

Extracorporeal membrane oxygenation

ECPR

Extracorporeal cardiopulmonary resuscitation

ED

Emergency department

ELSO

Extracorporeal life support organization

EtCO₂

End-tidal carbon dioxide

ICU

Intensive care unit

INCEPTION

Early initiation of extracorporeal life support in refractory out-of-hospital cardiac arrest trial

IQR

Interquartile range

MAP

Mean arterial pressure

mmHg

Millimetre of mercury

OHCA

Out-of-hospital cardiac arrest

PaO₂

Arterial partial pressure of oxygen

PaCO₂

Arterial partial pressure of carbon dioxide

PRAGUE-OHCA

Prague out-of-hospital cardiac arrest study

SPSS

Statistical Package for the Social Sciences

VA ECMO

Veno-arterial extracorporeal membrane oxygenation

VV ECMO

Veno-venous extracorporeal membrane oxygenation

VVA ECMO

Veno-veno-arterial extracorporeal membrane oxygenation

Authors’ contributions

AK, PB, OS, WS, MT contributed to conceptualization, methodology, investigation, and supervision. AK, PB, OS, TL, TS, WS, AŻ, PM, TD, MT were involved in investigation, data curation, literature review, and visualization. AK, PB, MT performed formal analysis, validation, writing – original draft, and visualization. AK, PB, OS, TL, TS, WS, AŻ, PM, TD, MT contributed to project administration, resources, writing – review & editing, and funding acquisition.All authors read and approved the final version of the manuscript.

Funding

No external funding was obtained for the conduct of this study, its authorship, or publication.

Data availability

The underlying data supporting the findings of this study are available from the corresponding author upon reasonable request, subject to appropriate institutional permissions and ethical oversight.

Declarations

Ethics approval and consent to participate

The study protocol was approved by the Jagiellonian University Bioethics Committee (approval no. 118.0043.1.385.2025). The study was conducted in accordance with the principles of the Declaration of Helsinki and reported in compliance with the STROBE Statement. As this study involved only the analysis of existing anonymized data and posed no risk to participants, the Institutional Review Board waived the requirement for obtaining informed consent, in accordance with Polish national regulations and the EU General Data Protection Regulation (GDPR, Regulation (EU) 2016/679).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.