Publication

• Title: Expedited transfer from the scene for refractory out-of-hospital cardiac arrest in Australia: a prospective, multicentre, parallel, open label, randomised clinical trial
• Acronym: EVIDENCE
• Year: 2025
• Journal published in: The Lancet Respiratory Medicine
• Citation: Burns B, Marschner IC, Coggins A, Oliver M, Facer R, Pradhananga B, et al; EVIDENCE Trial Investigators. Expedited transfer from the scene for refractory out-of-hospital cardiac arrest in Australia: a prospective, multicentre, parallel, open label, randomised clinical trial. Lancet Respir Med. 2025;13(10):921-932.

Context & Rationale

• Background

  • Refractory out-of-hospital cardiac arrest (OHCA) has a low probability of neurologically favourable survival despite contemporary advanced life support (ALS) and post-arrest care.
  • Definitive therapies for potentially reversible causes (e.g., coronary occlusion, pulmonary embolus) require hospital diagnostics and interventions that are time-critical.
  • Extracorporeal cardiopulmonary resuscitation (ECPR) can provide circulatory support while definitive therapies are delivered, but depends on rapid access and careful candidate selection.
  • Intra-arrest transport might shorten time-to-definitive therapy but risks poorer chest-compression quality, interruptions during extrication/transport, safety issues, and inappropriate transport of patients who would otherwise be terminated on scene.
  • Before EVIDENCE, transport-under-CPR strategies were mainly supported by observational comparisons with substantial confounding and selection bias, and there was no definitive randomised evidence in refractory OHCA.
• Research Question/Hypothesis

  • Whether an EMS strategy of expedited transfer from scene (with ongoing resuscitation) to an appropriate receiving hospital improves survival with favourable neurological outcome compared with a strategy of more extended on-scene resuscitation with transport at clinician discretion.
  • Hypothesis: expedited intra-arrest transfer would increase survival with cerebral performance category (CPC) 1–2 at hospital discharge.
• Why This Matters

  • Tests whether “earlier transport” is a justified population-level EMS policy for refractory OHCA in an urban system, rather than a selectively applied pathway.
  • Clarifies the degree of clinical benefit (or harm) that could plausibly be achieved by transport strategy alone in the absence of a mandated 24/7 cath lab/ECPR pathway.
  • Informs health-system planning for cardiac arrest centres, mechanical CPR capability, ECPR availability, and termination-of-resuscitation policies.

Design & Methods

• Research Question: In adults with refractory OHCA, does an expedited intra-arrest transfer strategy to an appropriate receiving hospital improve neurologically favourable survival (CPC 1–2) at hospital discharge compared with standard (non-expedited) on-scene resuscitation with transport at clinician discretion?
• Study Type: Prospective, multicentre, parallel-group, open-label, randomised clinical trial with prehospital randomisation; stratified by initial rhythm (VT/VF vs PEA); conducted across two metropolitan Sydney regions with 15 receiving hospitals.
• Population:
  • Setting: NSW Ambulance-attended adult OHCA in metropolitan Sydney; enrolment required a mechanical chest compression device to be present at the incident.
  • Inclusion criteria (prehospital): age 18–70 years; witnessed arrest; initial rhythm ventricular tachycardia/ventricular fibrillation (or AED shock delivered) or pulseless electrical activity; bystander CPR started within <5 minutes and ongoing on ambulance arrival; refractory to initial professional resuscitation at the time of randomisation (no sustained ROSC after ~15 minutes/three complete resuscitation cycles).
  • Exclusion criteria (prehospital): traumatic cardiac arrest; asystole as first rhythm; terminal end-stage illness; current advance care directive limiting treatment; advanced cognitive impairment (e.g., dementia); paramedic judgement that enrolment was not in the patient’s best interests.
• Intervention:
  • Expedited transfer strategy: depart scene as soon as feasible after randomisation with ongoing resuscitation and mechanical chest compressions; protocol target was scene departure approximately 15 minutes after arrival of the first paramedic responder.
  • Destination and pre-notification: transport to the nearest appropriate receiving hospital with pre-notification; destinations included ventricular tachycardia/cath lab-capable hospitals and ECPR-capable centres.
  • Hospital pathway (pragmatic): during business hours, direct-to-catheterisation laboratory assessment could occur where initiation within ~1 hour of arrest was feasible; at ECPR-capable centres, ECPR team pre-notified and cannulation could be initiated for patients meeting local criteria, with coronary angiography after ECMO flow; after hours, invasive strategy decisions were per usual practice (ECPR not mandated due to non-resident teams).
• Comparison:
  • Standard (non-expedited) strategy: continued on-scene ALS per existing NSW Ambulance refractory OHCA protocols, including termination rules for non-shockable rhythms at approximately 20 minutes on-scene, and transport at clinician discretion.
  • Hospital care: if transported, in-hospital interventions (angiography, PCI, ECPR) were per usual practice without trial-mandated triggers.
• Blinding: Open-label for EMS and receiving teams; assessment of CPC was masked; follow-up neurological status was assessed via telephone follow-up by research personnel.
• Statistics: A total of 200 participants were required to detect an absolute increase in CPC 1–2 survival from 10% to 25% (RR 2.5) with 80% power at the 5% two-sided significance level; primary analysis was intention-to-treat using RR regression adjusted for rhythm stratum, with risk differences and 95% CIs reported; no multiplicity adjustment for secondary outcomes.
• Follow-Up Period: To hospital discharge (primary endpoint), plus 4-week and 6-month follow-up for survival and CPC (follow-up completion reported to Aug 29, 2024).

Key Results

This trial was not stopped early. Recruitment completed close to the planned sample size (197 analysed in the intention-to-treat population).

• Neurologically favourable survival at discharge was identical (15 events in each group), with an adjusted RR 0.95; 95% CI 0.5 to 1.8; P=0.87.
• Prehospital process separation was achieved (total time on scene 26 [21–34] vs 36 [28–44] minutes), but the time from arrest to hospital arrival was 55 (47–65) vs 61 (52–70) minutes, and definitive therapies were infrequent (coronary angiography 33% vs 29%; ECPR with ECMO flow 7% vs 5%).
• Based on the reported confidence distribution, there was 91% confidence that the absolute difference in the primary outcome was less than 10% in favour of either strategy.

Internal Validity

• Randomisation and Allocation:
  • Central randomisation occurred prehospital via a secure REDCap smartphone application using permuted blocks stratified by initial rhythm (VT/VF vs PEA).
  • Allocation concealment was maintained until randomisation; study investigators and other personnel did not have access to the randomisation module.
• Drop out or exclusions (post randomisation):
  • 206 patients were randomised.
  • 9 were excluded from the intention-to-treat analysis: randomisation application failure (n=1) and major ineligibility (n=8).
  • Intention-to-treat population: 197 (expedited n=102; standard n=95).
  • Among 37 survivors to discharge, 4-week follow-up was completed in 31, and 6-month follow-up in 33.
• Performance/Detection Bias:
  • Open-label delivery could influence termination decisions, destination selection, and escalation to cath lab/ECPR.
  • Assessment of CPC was masked, reducing detection bias for the primary neurological outcome.
• Protocol Adherence:
  • Randomisation to scene departure: 15 (10–21) minutes in the expedited group vs 24 (16–30) minutes in the standard group (median, IQR).
  • Total time on scene: 26 (21–34) minutes vs 36 (28–44) minutes (median, IQR).
  • Transport contamination was substantial: 86/95 (91%) standard-group patients were transported to hospital.
• Baseline Characteristics:
  • Groups were broadly comparable: age 58 (49–65) vs 57 (45–64) years; male sex 82% vs 81%; initial rhythm VT/VF 75% vs 76%; witnessed arrest 98% vs 97%.
  • PEA subgroup had very low favourable outcome (0/25 vs 1/23 [4%]) which limits power for interaction and dilutes any system-level effect.
• Heterogeneity:
  • Pragmatic delivery across 15 hospitals with variable ECPR capability and business-hours access to urgent cath lab workflows.
  • No adjustment for potential clustering at the EMS crew or hospital level was reported.
• Timing:
  • Arrest to hospital arrival: 55 (47–65) vs 61 (52–70) minutes (median, IQR), indicating modest separation at the key mechanistic timepoint.
  • Time-dependent targets for definitive therapies (e.g., cath lab or ECPR initiation within ~1 hour of arrest) were not guaranteed within the protocol and depended on service availability.
• Dose:
  • Coronary angiography: 34/102 (33%) vs 28/95 (29%).
  • ECPR with ECMO flow: 7/102 (7%) vs 5/95 (5%).
  • The low and similar exposure to definitive therapies limits inference about “expedited transport + definitive intervention” as an integrated bundle.
• Separation of the Variable of Interest:
  • Total time on scene: 26 (21–34) minutes vs 36 (28–44) minutes (median, IQR).
  • Randomisation to hospital arrival: 32 (26–41) minutes vs 40 (32–50) minutes (median, IQR).
  • Arrest to hospital arrival: 55 (47–65) minutes vs 61 (52–70) minutes (median, IQR).
• Outcome Assessment:
  • Primary outcome (CPC at discharge) is clinically meaningful but can be influenced by prognostication and withdrawal-of-care practices; masking of CPC assessment mitigates (but does not eliminate) this risk.
• Statistical Rigor:
  • Primary analysis used RR regression adjusted for rhythm stratum with prespecified subgroup analyses for rhythm and age.
  • Trial was powered for a large absolute effect (10% to 25%); the confidence intervals remain compatible with smaller benefits or harms.

Conclusion on Internal Validity: Moderate. Randomisation and masked primary outcome assessment were strong, and prehospital process separation was achieved, but open-label care, high transport in the control group, modest separation in arrest-to-hospital time, and low exposure to definitive hospital therapies constrain causal inference about the intended time-critical mechanism.

External Validity

• Population Representativeness:
  • Applies to an urban Australian EMS system with high bystander CPR rates, mechanical CPR availability, and short transport times.
  • Eligibility restricted to witnessed arrests with VT/VF or PEA, age 18–70, and excluded traumatic arrests and asystole, limiting applicability to broader OHCA populations.
• Applicability:
  • Most applicable to metropolitan systems where both strategies already result in high rates of transport and where cath lab/ECPR access is variable and not universally 24/7.
  • Less applicable to rural/remote regions with longer transport times, systems without mechanical CPR, or mature ECPR networks with standardised 24/7 activation and direct-to-ECPR centre routing.
  • Does not directly test selective transport of narrowly defined ECPR candidates, nor prehospital ECPR strategies.

Conclusion on External Validity: Moderate. Findings generalise well to similar urban EMS systems with comparable infrastructure and transport times, but may not translate to settings with different termination practices, longer prehospital intervals, or a fully standardised 24/7 cardiac arrest centre/ECPR pathway.

Strengths & Limitations

• Strengths:
  • Randomised evaluation of a systems-of-care question that is otherwise difficult to study without major confounding.
  • Pragmatic multicentre design across 15 receiving hospitals, reflecting real-world variation in capability and workflow.
  • Meaningful and patient-centred neurological primary endpoint (CPC 1–2) with masked assessment and follow-up to 6 months.
  • Clear prehospital process separation in scene time (26 [21–34] vs 36 [28–44] minutes) demonstrating feasibility of the expedited strategy in a metropolitan system.
• Limitations:
  • Substantial overlap/contamination: 86/95 (91%) standard-group patients were transported to hospital.
  • Modest separation at the mechanistic endpoint (arrest-to-hospital arrival 55 [47–65] vs 61 [52–70] minutes) and similar transport duration from scene departure (16 minutes in both groups).
  • Hospital component was pragmatic and not standardised; cath lab and ECPR pathways were not mandated and were constrained by business-hours availability, with low ECPR utilisation (12/197; 6%).
  • Open-label care could influence termination and downstream decisions, potentially affecting both transport exposure and neurological outcomes.
  • Powered for a large effect size; smaller clinically relevant effects cannot be excluded with high precision.

Interpretation & Why It Matters

• Transport strategy alone

  In this system, an “expedited transfer” policy shortened scene time but did not improve neurologically favourable survival; this suggests that earlier transport without a consistently delivered, time-critical definitive therapy pathway is unlikely to produce large population-level gains.
• Mechanism and systems design

  The intended mechanism (earlier access to cath lab/ECPR) was not strongly expressed in the delivered care (angiography 33% vs 29%; ECPR flow 7% vs 5%), highlighting that system capability and availability (including out-of-hours provision) are central determinants of any benefit.
• Implications for practice

  These data support continued emphasis on high-performance on-scene resuscitation and selective, capability-linked transport decisions rather than routine intra-arrest transfer for all refractory OHCA patients in similar systems.
• Implications for future trials

  Future evaluations likely need enrichment for high-likelihood “reversible” phenotypes and a fully integrated bundle (dispatch optimisation, on-scene performance, direct-to-capable centre routing, and guaranteed 24/7 definitive therapy delivery) to test the hypothesis that time-critical hospital interventions improve outcomes.

Controversies & Subsequent Evidence

• Intervention definition and “dose” of definitive therapy: The trial primarily tested a transport strategy embedded within pragmatic hospital practice, rather than a mandated cath lab/ECPR pathway; limited and variable access (particularly out-of-hours) may have reduced the capacity to detect benefit attributable to definitive therapies.¹
• Time-to-treatment separation: Despite earlier scene departure, arrest-to-hospital arrival was 55 (47–65) vs 61 (52–70) minutes, and transport time from scene departure was identical (16 minutes in both groups), limiting contrast at the key mechanistic timepoint.¹
• Pragmatism versus enrichment: Broad inclusion maximised generalisability but may have diluted effects in patients least likely to benefit from rapid reperfusion/ECPR (e.g., PEA subgroup primary outcome 0/25 vs 1/23 [4%]).¹
• Comparator contamination: High hospital transport in the standard group (86/95 [91%]) means the comparison approximated “earlier vs later transport” rather than a strict “transport vs stay-and-play” contrast, potentially biasing towards the null.
• Operational feasibility of hospital-based ECPR at scale: A post-hoc analysis combining the EVIDENCE RCT and a registry found that only a small proportion of refractory OHCA patients were truly eligible for hospital-based ECPR under common time and destination constraints, with exclusions driven by arrival >1 hour, non-ECPR destination, and out-of-hours limitations.²
• Syntheses of intra-arrest transport and ECPR pathways: Systematic reviews/meta-analyses highlight heterogeneity across EMS systems and programme maturity, with pooled estimates sensitive to candidate selection and time-to-ECMO, supporting cautious extrapolation between regions.³⁴⁵
• Relationship to other randomised evidence: Contemporary randomised trials of hyperinvasive/ECPR strategies have produced heterogeneous results, reinforcing that benefit (if present) likely depends on tightly time-coupled and standardised systems of care rather than transport timing alone.

Summary

• In 197 adults with refractory witnessed OHCA in metropolitan Sydney, expedited intra-arrest transfer did not improve neurologically favourable survival at hospital discharge (CPC 1–2: 15/102 [15%] vs 15/95 [16%]; RR 0.95; 95% CI 0.5 to 1.8; P=0.87).
• Survival to discharge and to 6 months were similar between groups (19/102 [19%] vs 18/95 [19%] survival to discharge; 19/102 [19%] vs 17/95 [18%] survival to 6 months).
• The expedited strategy achieved prehospital separation (total time on scene 26 [21–34] vs 36 [28–44] minutes), but arrest-to-hospital arrival was 55 (47–65) vs 61 (52–70) minutes.
• Definitive hospital therapies were infrequent and broadly similar (coronary angiography 33% vs 29%; ECPR with ECMO flow 7% vs 5%).
• The findings argue against routine expedited transport for all refractory OHCA in similar systems without a consistently delivered, time-critical definitive therapy pathway.

Overall Takeaway

EVIDENCE is a landmark, pragmatic prehospital randomised trial addressing whether expedited intra-arrest transport improves outcomes in refractory OHCA. Despite achieving shorter on-scene times, expedited transfer did not increase neurologically favourable survival and did not materially change the delivery of definitive hospital therapies (angiography/ECPR) at scale, underscoring that transport timing alone is unlikely to be transformative without an integrated, consistently deliverable cardiac arrest centre/ECPR pathway.

Bibliography

• 1Joiner A, Ong D. Continued resuscitation at the scene versus early transport for refractory out-of-hospital cardiac arrest. Lancet Respir Med. 2025;13:864-865.
• 2Greenberg M, Burns B, Lay P, et al. ECPR eligible refractory out of hospital cardiac arrests - post-hoc analysis of the EVIDENCE Randomised Controlled Trial and Registry. Resuscitation. 2026;206:110996.
• 3Burns B, Keijzers G, Smith K, et al. Expedited transport versus continued resuscitation at the scene for refractory out-of-hospital cardiac arrest: a systematic review and meta-analysis. Resuscitation Plus. 2023;17:100482.
• 4Heuts S, Taccone FS, Lorusso R, et al. Extracorporeal cardiopulmonary resuscitation for out-of-hospital cardiac arrest: a Bayesian meta-analysis. Critical Care. 2024;28:217.
• 5Ali B, Yount K, Morykwas M, et al. Extracorporeal cardiopulmonary resuscitation for out-of-hospital cardiac arrest: a meta-analysis of randomised trials. JACC Cardiovasc Interv. 2023;16(14):1825-1827.