Publication

• Title: Prehospital resuscitative endovascular balloon occlusion of the aorta in non-traumatic out-of-hospital cardiac arrest (REBOARREST): an international, multicentre, open label, pragmatic, randomised, controlled trial
• Acronym: REBOARREST
• Year: 2026
• Journal published in: Critical Care
• Citation: Brede JR, Farbu BH, Gamberini L, Thorsen K, Rehn M, Rognås L, et al. Prehospital resuscitative endovascular balloon occlusion of the aorta in non-traumatic out-of-hospital cardiac arrest (REBOARREST): an international, multicentre, open label, pragmatic, randomised, controlled trial. Crit Care. 2026;30:324.

Context & Rationale

• Background

  Non-traumatic out-of-hospital cardiac arrest (OHCA) remains dominated by failure to achieve sustained return of spontaneous circulation (ROSC), despite high-quality CPR, defibrillation, airway management, adrenaline, and treatment of reversible causes.
  The physiological rationale for thoracic aortic occlusion is compelling: occluding the descending thoracic aorta should reduce the vascular volume being perfused during CPR and preferentially direct limited forward flow to the heart and brain.
  Coronary perfusion pressure is strongly associated with ROSC during human CPR, making augmentation of proximal aortic diastolic pressure a biologically plausible target.¹
  Prior human evidence consisted mainly of feasibility work, uncontrolled case series, and physiological studies showing that REBOA during CPR could be performed and could increase peripheral arterial pressure, end-tidal CO₂, or coronary perfusion pressure, but without randomised evidence of patient benefit.²³⁴
• Research Question/Hypothesis

  REBOARREST tested whether a pragmatic prehospital strategy of adding zone 1 REBOA to standard advanced life support (ALS) would increase sustained ROSC in adults with non-traumatic OHCA.
  The published protocol framed this as a phase II randomised trial of adjunctive REBOA, not as a device-comparison study.⁵
• Why This Matters

  A Norwegian needs assessment estimated that patients potentially eligible for REBOA represented a selected OHCA subgroup with a substantially lower ROSC rate than the overall cardiac arrest population, supporting the trialists’ decision to target a refractory, high-risk population.⁶
  If a prehospital endovascular intervention could meaningfully increase ROSC without degrading CPR quality, it would represent a major escalation in the resuscitation armamentarium.
  Conversely, REBOA is invasive, resource-intensive, training-dependent, and potentially harmful; it therefore required randomised evidence before routine deployment.

Design & Methods

• Research Question: In adults with non-traumatic OHCA receiving ALS, does prehospital zone 1 REBOA as an adjunct to ALS increase sustained ROSC, defined as ROSC lasting at least 20 minutes, compared with ALS alone?
• Study Type: Pragmatic, parallel-group, multicentre, international, open-label, randomised controlled trial conducted at 12 physician-staffed emergency medical service sites in Norway, Denmark, and Italy.
• Population:
  • Adults aged 18–80 years with non-traumatic OHCA.
  • Required less than 10 minutes from arrest onset to initiation of basic life support or ALS.
  • ALS had to be established and capable of continuing after randomisation.
  • Eligibility was determined by the physician on scene during active resuscitation.
  • Exclusions included traumatic arrest, accidental hypothermia below 32 °C, suspected cerebral haemorrhage, suspected non-traumatic haemorrhage as arrest aetiology, obvious or suspected pregnancy, prior inclusion in a site extracorporeal CPR pathway, or other safety/environmental contraindications.
  • Of 3,958 screened patients, 200 were randomised and 179 were included in the final full analysis set after deferred consent was not obtained for 21 patients.
• Intervention:
  • ALS according to national guidelines plus prehospital REBOA.
  • Femoral arterial access was obtained using ultrasound guidance and a 7 Fr introducer.
  • The balloon catheter was advanced 50 cm to achieve zone 1 thoracic aortic occlusion.
  • Left radial pulse was checked during compressions before and after inflation; loss of a previously palpable radial pulse triggered deflation, withdrawal by 5 cm, and reinflation.
  • The balloon was inflated with sterile 0.9% sodium chloride; if ROSC occurred, the balloon was deflated over 30 seconds and left in situ.
  • The ER-REBOA catheter was used in Italy and the REBOA Medical 20 mm balloon at the other sites.
• Comparison:
  • ALS alone according to national and European Resuscitation Council-based protocols.
  • Decisions about total resuscitation time, transport, termination, and post-ROSC care followed site-specific standard practice.
  • Post-ROSC treatment was intended not to differ between groups except for the presence and later management of REBOA equipment in the intervention group.
• Blinding: Treating clinicians and investigators were necessarily unblinded. The patient was unconscious during the primary endpoint window. The statistician performing the primary analysis was blinded to group allocation.
• Statistics: A total of 200 patients were planned to detect a doubling in sustained ROSC from 18% to 36%, with 80% power at a two-sided 5% significance level, allowing for repeated interim testing and loss after deferred consent; 94 patients per group were required, rounded to 100 per group.⁵ The primary analysis followed the intention-to-treat principle within the full analysis set, excluding patients without consent, using logistic regression adjusted for study site; the adaptive group-sequential design used O’Brien-Fleming boundaries, with a final nominal significance level of 0.043.
• Follow-Up Period: Sustained ROSC was assessed during the index resuscitation; 30-day survival, 30-day modified Rankin Scale, adverse events, and organ dysfunction were assessed after hospital admission where applicable; one-year all-cause mortality was prespecified as exploratory but was not reported in the primary results.

Key Results

This trial was not stopped early. It completed planned randomisation of 200 patients after three prespecified DMC interim analyses; final outcome analyses included 179 patients because deferred consent was not obtained for 21 randomised patients.

• The central clinical finding is simple: adding a prehospital REBOA strategy to ALS did not significantly increase sustained ROSC, 30-day survival, or 30-day survival with mRS 0–3.
• There was a physiological signal after actual balloon inflation, with EtCO₂ rising by 0.41 kPa at 120 seconds, but this did not translate into a demonstrable patient-centred benefit.
• Actual delivery was a major determinant of interpretation: only 51/88 patients (58%) allocated to REBOA underwent aortic occlusion, and median arrest-to-occlusion time was 47 minutes [41–54].

Internal Validity

• Randomisation and Allocation: Randomisation used permuted blocks of variable size, stratified by site, with sealed sequentially numbered lightproof envelopes generated by an independent clinical research unit; this is a reasonable emergency-trial allocation method, although envelope systems are intrinsically less robust than central web randomisation.
• Dropout and Post-Randomisation Exclusions: Of 200 randomised patients, 21 (10.5%) were excluded because deferred consent was not obtained: 8 from the REBOA group and 13 from the control group. Two next-of-kin refused; 19 could not be contacted. This was ethically required but weakens the purity of intention-to-treat inference and exceeded the protocol’s expected low dropout rate.
• Performance and Detection Bias: Clinician blinding was impossible. The primary endpoint was objective in concept but assessed by treating clinicians during resuscitation, so unblinded clinical judgement and EtCO₂ changes could theoretically influence ROSC assessment. Blinded statistical analysis partly mitigates analytical bias but not bedside ascertainment bias.
• Protocol Adherence: Aortic occlusion was achieved in 51/88 intervention patients (58%). In the remaining 37/88, occlusion was not performed because of ROSC before inflation in 17 (19%), unsuccessful procedure in 14 (16%), or procedure abortion in 6 (7%).
• Procedure Fidelity: Among cases where the procedure was initiated, cannulation/occlusion success was 51/65 (78%). Median cannulation attempts were 1 [1–2], catheter repositioning after initial inflation occurred in 6/51 (12%), and randomisation-to-occlusion time was 14 minutes [11–16].
• Baseline Characteristics: Groups were broadly comparable, but the REBOA group had a longer median arrest-to-randomisation time (33 minutes [23–39] vs 29 minutes [23–38]) and fewer initial shockable rhythms (39/88, 44% vs 47/91, 52%). These imbalances plausibly favoured control, although sensitivity analysis adjusted for time to randomisation and rhythm still showed no significant difference: adjusted RD 4%; 95% CI -9 to 16; P=0.56.
• Severity of Illness: Patients were already in prolonged cardiac arrest at randomisation, and actual occlusion occurred at 47 minutes [41–54] after arrest onset. This may have selected a group too late in the low-flow trajectory to benefit from proximal flow redistribution.
• Heterogeneity: Heterogeneity was clinically important but pragmatically appropriate. The trial spanned 12 sites, three countries, rural and urban settings, helicopters and rapid response cars, and both shockable and non-shockable rhythms. Subgroup analyses did not show convincing effect modification, but interaction testing was underpowered.
• Timing: The timing problem is central. REBOA may be a time-dependent physiological intervention, and the observed median arrest-to-occlusion time of 47 minutes is late relative to the window in which improved coronary and cerebral perfusion would be expected to change outcome.
• Dose: The intended “dose” was complete zone 1 thoracic aortic occlusion. The achieved biological separation was incomplete because 42% of intervention patients did not receive occlusion and intra-aortic pressure data were unavailable, leaving EtCO₂ as the main physiological marker.
• Separation of the Variable of Interest: Actual aortic occlusion occurred in 51/88 intervention patients versus 1/91 control patients. The intervention group also had a measurable EtCO₂ rise at 120 seconds after occlusion (+0.41 kPa; 95% CI 0.15 to 0.67; P=0.003), but this was modest and not accompanied by direct aortic pressure measurement.
• Key Delivery Aspects: The procedure was delivered by highly trained physician-staffed teams; 168 physicians and 106 paramedics or nurses completed mandatory structured training, including theory, skills training, real procedure performance, and high-fidelity simulation assessment.⁷
• Crossover: One control patient received balloon inflation, and one intervention patient had venous REBOA placement. These events were rare but demonstrate the complexity of performing endovascular procedures during CPR.
• Adjunctive Therapy Use: Prehospital co-interventions were similar overall: mechanical chest compression was used in 58/88 (66%) vs 63/91 (69%), endotracheal intubation in 71/88 (81%) vs 71/91 (78%), median total adrenaline dose was 7 mg [5–10] vs 6 mg [4–8], and median defibrillations were 3 [0–6] vs 1 [0–6]. Among ICU-admitted patients, coronary angiography occurred in 12/16 (75%) vs 15/18 (83%), mechanical circulatory support in 3/16 (18%) vs 2/18 (11%), and renal replacement therapy in 1/16 (6%) vs 3/18 (16%).
• Outcome Assessment: Sustained ROSC is a standardised and clinically relevant early endpoint, but it is not a definitive patient-centred outcome. mRS 0–3 at 30 days was reported, but health-related quality of life was not obtained, despite its inclusion in the cardiac arrest core outcome set.⁸
• Statistical Rigor: The statistical approach was prespecified, used site-adjusted logistic regression, preserved the group-sequential alpha structure, and distinguished the pragmatic ITT analysis from per-protocol and sensitivity analyses. Secondary outcomes were not adjusted for multiplicity and should be interpreted supportively.
• As-Treated Analyses: The apparently favourable as-treated sensitivity analysis 1 (adjusted RD 16%; 95% CI 2 to 29; P=0.02) conditions on post-randomisation events and cannot overturn the primary randomised comparison.

Conclusion on Internal Validity: Internal validity is moderate-to-strong for the pragmatic question of assigning prehospital REBOA during ALS, because randomisation, monitoring, training, and prespecified analysis were robust. It is more limited for the mechanistic question of whether early successful aortic occlusion itself can improve outcomes, because many intervention patients did not receive occlusion, the intervention was late, the trial was open-label, and direct intra-aortic pressure data were unavailable.

External Validity

• Population Representativeness: Only 200/3,958 screened patients (5.1%) were enrolled, reflecting a narrow population: adult, non-traumatic OHCA, early BLS/ALS, continued resuscitation possible, and not already selected for ECPR.
• Patient Profile: Participants were mostly male (136/179, 76%), older (median age 68–69 years), commonly witnessed arrests (82/88, 93% vs 86/91, 95%), and frequently received bystander CPR (79/88, 90% vs 83/91, 91%). Presumed cardiac aetiology dominated: 84/88 (95%) vs 83/91 (91%).
• System Representativeness: Applicability is strongest for advanced European-style physician-staffed EMS systems with ultrasound, Seldinger expertise, mechanical CPR availability, and established hospital pathways for sheath/catheter management.
• Resource-Limited Settings: Generalisability to paramedic-only systems, low-resource systems, or environments without ultrasound-guided vascular access expertise is limited.
• Urban and In-Hospital Settings: The trial does not settle whether earlier REBOA in dense metropolitan systems, emergency departments, catheterisation laboratories, or in-hospital cardiac arrest could be beneficial.
• ECPR Interaction: Exclusion of patients already included in site ECPR protocols may have removed a subgroup with more favourable prognosis and limits generalisability to systems where ECPR and endovascular resuscitation are integrated.
• Trauma and Haemorrhage: The findings should not be extrapolated to traumatic haemorrhage, ruptured aneurysm, postpartum haemorrhage, or other haemorrhagic shock states, where the indication and physiology differ.

Conclusion on External Validity: External validity is good for late prehospital REBOA deployment in expert physician-staffed European EMS systems, but limited for earlier, in-hospital, paramedic-led, ECPR-integrated, or resource-limited settings. The trial evaluates a real-world deployment strategy rather than the best-case physiology of immediate aortic occlusion.

Strengths & Limitations

• Strengths:
  • First randomised trial evidence for prehospital REBOA in non-traumatic OHCA.
  • International, multicentre, pragmatic design across 12 sites and three countries.
  • Independent sequence generation and concealed envelope allocation.
  • Prespecified protocol, SAP, interim-monitoring structure, and blinded statistician.
  • Mandatory structured training programme and external clinical review panel.
  • Directly relevant implementation data: feasibility, procedure time, occlusion success, crossovers, and adverse events.
• Limitations:
  • Late intervention: median arrest-to-occlusion time was 47 minutes [41–54].
  • Incomplete intervention delivery: only 51/88 (58%) assigned to REBOA underwent aortic occlusion.
  • Post-randomisation exclusion after deferred consent in 21/200 (10.5%).
  • Open-label bedside care and clinician-assessed primary endpoint.
  • No intra-aortic blood pressure data, despite blood pressure being central to the physiological hypothesis.
  • Low event counts for patient-centred outcomes: 30-day mRS 0–3 occurred in only 8/179 patients overall.
  • Designed to detect a large effect, so smaller but clinically meaningful effects cannot be excluded.
  • Adverse events were registered only among patients admitted to hospital, limiting safety inference in patients who died on scene.
  • No autopsies were performed, limiting detection of occult vascular or visceral complications.
  • No health-related quality-of-life outcome was obtained.

Interpretation & Why It Matters

• Clinical meaning

  REBOARREST does not support routine addition of prehospital REBOA to ALS for unselected eligible adults with non-traumatic OHCA in physician-staffed EMS systems.
• Physiology versus outcome

  The trial separates a real physiological effect from clinical efficacy: EtCO₂ rose after occlusion, but sustained ROSC, survival, and neurological outcomes did not improve significantly.
• Strategy versus procedure

  The result applies most directly to a deployment strategy: dispatch, arrival, randomisation, cannulation, and occlusion during ongoing OHCA. It is less definitive about immediate successful occlusion earlier in the arrest trajectory.
• Opportunity cost

  REBOA requires expert personnel, equipment, training, ultrasound, and post-ROSC vascular management. A strategy that helps only a small screened fraction and does not improve patient-centred outcomes should not displace proven resuscitation fundamentals.
• Future direction

  Future studies should focus on earlier deployment, better physiological phenotyping, direct coronary/aortic pressure measurement, integration with ECPR or selective aortic arch perfusion strategies, and rigorous capture of visceral and vascular harms.

Controversies & Other Evidence

• The physiological premise remains attractive but not sufficient. Earlier observational studies showed feasibility and pressure/EtCO₂ signals, but REBOARREST demonstrates that a plausible haemodynamic intervention can fail to produce measurable improvement when inserted into real prehospital cardiac arrest systems.²³⁴
• Timing is the dominant unresolved issue. Median arrest-to-randomisation was 33 minutes in the intervention arm and median arrest-to-occlusion was 47 minutes. This is plausibly beyond the window in which improved proximal pressure can reverse myocardial and cerebral injury in many patients.
• The pragmatic design is both a strength and a constraint. Randomising before cannulation gives the clinically relevant answer for a deployed prehospital strategy, but it dilutes the mechanistic effect because 37/88 intervention patients never received occlusion.
• The as-treated signal should not be overinterpreted. Sensitivity analysis 1 suggested higher sustained ROSC with actual occlusion (adjusted RD 16%; 95% CI 2 to 29; P=0.02), but it excluded control patients according to post-randomisation events and therefore cannot provide the causal reliability of the randomised comparison.
• ROSC is a necessary but incomplete endpoint. The trial was powered for sustained ROSC, not survival with good function. The mRS 0–3 outcome was 5/88 (6%) versus 3/91 (3%), and 30-day survival was 6/88 (7%) versus 6/91 (7%). This matters because cardiac arrest trials should ultimately prioritise survival, neurological function, and quality of life.⁸
• Visceral ischaemia remains a legitimate concern. A dedicated 2025 commentary highlighted the possibility that thoracic aortic balloon occlusion during low-flow cardiac arrest could worsen intestinal ischaemia, especially with prolonged occlusion time.⁹ In REBOARREST, intestinal ischaemia occurred in 1 intervention and 1 control patient, but harms were only captured in patients admitted to hospital and no autopsies were performed.
• Subsequent and parallel cardiac-arrest evidence is still developing. A separate multinational randomised trial protocol for REBOA in non-traumatic OHCA was published in 2024, indicating ongoing equipoise for differently timed or differently configured REBOA strategies.¹⁰
• Guideline position remains conservative. Current adult ALS guidance emphasises established resuscitation priorities and does not make prehospital REBOA a routine component of adult ALS algorithms.¹¹
• The broader REBOA literature urges caution. In exsanguinating trauma, the UK-REBOA randomised trial found worse outcomes with emergency department REBOA, underscoring that endovascular resuscitation may cause harm when implementation, patient selection, or timing are unfavourable.¹²
• Endovascular resuscitation is best viewed as a systems intervention. Contemporary expert review frames REBOA, selective aortic arch perfusion, and ECPR as complex interventions requiring vascular access skill, patient selection, governance, and pathway design rather than simple device adoption.¹³

Summary

• REBOARREST randomised 200 adults with non-traumatic OHCA to prehospital REBOA plus ALS or ALS alone; 179 were analysed after deferred consent was not obtained for 21.
• REBOA plus ALS did not significantly improve sustained ROSC: 25/88 (28%) versus 24/91 (26%); adjusted RD 1.8%; 95% CI -11 to 15; P=0.78.
• Patient-centred outcomes were also not significantly different: mRS 0–3 at 30 days was 5/88 (6%) versus 3/91 (3%), and 30-day survival was 6/88 (7%) versus 6/91 (7%).
• Actual occlusion was achieved in 51/88 intervention patients (58%), with median arrest-to-occlusion time of 47 minutes [41–54] and median randomisation-to-occlusion time of 14 minutes [11–16].
• The trial supports feasibility in expert prehospital teams but does not support routine prehospital REBOA deployment for non-traumatic OHCA outside research or highly selected endovascular resuscitation systems.

Overall Takeaway

REBOARREST is an important trial because it replaces uncontrolled physiological optimism with randomised pragmatic evidence in one of the most ambitious areas of prehospital resuscitation. It does not prove that aortic occlusion can never help in cardiac arrest, but it shows that late prehospital deployment of REBOA as an adjunct to ALS does not significantly improve sustained ROSC or 30-day outcomes in the tested population.

Bibliography

• 1.Paradis NA, Martin GB, Rivers EP, Goetting MG, Appleton TJ, Feingold M, et al. Coronary perfusion pressure and the return of spontaneous circulation in human cardiopulmonary resuscitation. JAMA. 1990;263:1106–1113.
• 2.Brede JR, Lafrenz T, Klepstad P, Skjærseth EA, Nordseth T, Søvik E, et al. Feasibility of pre-hospital resuscitative endovascular balloon occlusion of the aorta in non-traumatic out-of-hospital cardiac arrest. J Am Heart Assoc. 2019;8:e014394.
• 3.Brede JR, Skjærseth E, Klepstad P, Nordseth T, Krüger AJ. Changes in peripheral arterial blood pressure after resuscitative endovascular balloon occlusion of the aorta (REBOA) in non-traumatic cardiac arrest patients. BMC Emerg Med. 2021;21:157.
• 4.Jang DH, Lee DK, Jo YH, Park SM, Oh YT, Im CW. Resuscitative endovascular occlusion of the aorta (REBOA) as a mechanical method for increasing the coronary perfusion pressure in non-traumatic out-of-hospital cardiac arrest patients. Resuscitation. 2022;179:277–284.
• 5.Brede JR, Skulberg AK, Rehn M, Thorsen K, Klepstad P, Tylleskär I, et al. REBOARREST, resuscitative endovascular balloon occlusion of the aorta in non-traumatic out-of-hospital cardiac arrest: a study protocol for a randomised, parallel group, clinical multicentre trial. Trials. 2021;22:511.
• 6.Brede JR, Kramer-Johansen J, Rehn M. A needs assessment of resuscitative endovascular balloon occlusion of the aorta (REBOA) in non-traumatic out-of-hospital cardiac arrest in Norway. BMC Emerg Med. 2020;20:28.
• 7.Brede JR, Lafrenz T, Krüger AJ, Søvik E, Steffensen T, Kriesi C, et al. Resuscitative endovascular balloon occlusion of the aorta (REBOA) in non-traumatic out-of-hospital cardiac arrest: evaluation of an educational programme. BMJ Open. 2019;9:e027980.
• 8.Haywood K, Whitehead L, Nadkarni VM, Achana F, Beesems S, Böttiger BW, et al. COSCA (Core Outcome Set for Cardiac Arrest) in adults: an advisory statement from the International Liaison Committee on Resuscitation. Circulation. 2018;137:e783–e801.
• 9.Farbu BH, Brede JR. Balloon occlusion of the aorta during cardiac arrest — a death blow to the intestines? Scand J Trauma Resusc Emerg Med. 2025;33:24.
• 10.Kim HE, Chu SE, Jo YH, Chiang WC, Jang DH, Chang CH, et al. Effect of resuscitative endovascular balloon occlusion of the aorta in nontraumatic out-of-hospital cardiac arrest: a multinational, multicenter, randomized, controlled trial. Trials. 2024;25:118.
• 11.Soar J, Böttiger BW, Carli P, Carmona Jiménez F, Cimpoesu D, Cole G, et al. European Resuscitation Council Guidelines 2025 Adult Advanced Life Support. Resuscitation. 2025;215:110769.
• 12.Jansen JO, Hudson J, Cochran C, MacLennan G, Lendrum R, Sadek S, et al. Emergency department resuscitative endovascular balloon occlusion of the aorta in trauma patients with exsanguinating hemorrhage: the UK-REBOA randomized clinical trial. JAMA. 2023;330:1862–1871.
• 13.Adams TD, Barratt J, Manning JE, Rees P, Singer B, Barnard EBG. Endovascular resuscitation: an expert practice review. Emerg Med J. Published online 30 April 2026.