Skip to main content

Foundational Trials

European Epinephrine Study

Image: Shutterstock / ChaNaWiT

Publication

  • Title: A Comparison of Repeated High Doses and Repeated Standard Doses of Epinephrine for Cardiac Arrest Outside the Hospital
  • Acronym: None (European Epinephrine Study Group)
  • Year: 1998
  • Journal published in: The New England Journal of Medicine
  • Citation: Gueugniaud PY, Mols P, Goldstein P, et al. A comparison of repeated high doses and repeated standard doses of epinephrine for cardiac arrest outside the hospital. N Engl J Med. 1998;339:1595-1601.

Context & Rationale

  • Background
    • Epinephrine (adrenaline) was embedded in advanced life support algorithms, intended to increase aortic diastolic pressure and coronary perfusion pressure via α-adrenergic vasoconstriction, thereby improving defibrillation success and return of spontaneous circulation (ROSC).
    • Experimental studies and small clinical series suggested that higher doses might augment perfusion pressures and increase ROSC, particularly in prolonged arrests or refractory rhythms.
    • Earlier clinical trials of high-dose epinephrine reported higher initial resuscitation rates without consistent improvement in survival to discharge, and many regimens used only a single high dose rather than repeated dosing during ongoing resuscitation.
  • Research Question/Hypothesis
    • In adult out-of-hospital cardiac arrest (OHCA) patients requiring epinephrine during advanced life support, repeated high-dose epinephrine (5 mg every 3 minutes) would improve successful resuscitation (ROSC and hospital admission) and translate into improved survival (and neurological outcome) compared with repeated standard-dose epinephrine (1 mg every 3 minutes).
  • Why This Matters
    • OHCA carries very high mortality; even small absolute improvements in survival could yield substantial population benefit.
    • Clarifying whether dose escalation improves patient-centred outcomes (survival with good neurological function) helps avoid practice driven by surrogate endpoints (ROSC/admission) alone.
    • The trial tested a pragmatic “real-world” prehospital strategy in physician-led European emergency medical systems, directly informing resuscitation algorithms and subsequent trial design.

Design & Methods

  • Research Question:
    • Among adults with OHCA who met rhythm-based criteria for epinephrine during advanced cardiac life support, does repeated high-dose epinephrine (5 mg per dose) improve outcomes compared with repeated standard-dose epinephrine (1 mg per dose)?
  • Study Type:
    • Prospective, multicentre, randomised, double-blind, parallel-group trial.
    • Prehospital setting (out-of-hospital cardiac arrest) across 12 emergency medical systems in France and Belgium; September 1994 to September 1996.
  • Population:
    • Adults (≥18 years) with non-traumatic OHCA treated by mobile emergency teams.
    • Eligibility triggered by initial rhythm and early response:
      • Asystole or pulseless electrical activity, or
      • Ventricular fibrillation persisting after three shocks.
    • Key exclusions:
      • Traumatic cardiac arrest.
      • Obvious irreversible death (e.g., decapitation, rigor mortis, extensive burns).
      • Epinephrine given before enrolment.
      • Age <18 years.
  • Intervention:
    • High-dose epinephrine: 5 mg per dose in 5 mL ampoules; administered IV (peripheral or central) or endotracheally.
    • Dosing schedule: repeated every 3 minutes; maximum 15 doses per patient (maximum cumulative dose 75 mg).
  • Comparison:
    • Standard-dose epinephrine: 1 mg per dose in identical 5 mL ampoules; administered via the same routes and at the same 3-minute intervals; maximum 15 doses.
    • Co-interventions (defibrillation, airway/ventilation strategy, other drugs) followed contemporaneous advanced life support practice within participating systems.
  • Blinding:
    • Double-blind: identical ampoules and packaging; clinicians and receiving hospitals were unaware of assigned dose.
    • Implications: reduces risk of performance bias for decisions to continue resuscitation and for post-ROSC care intensity; early endpoints were largely objective.
  • Statistics:
    • Power calculation: Not reported in the index manuscript.
    • Two-sided significance level: 0.05.
    • Effect reporting: differences in proportions with 95% confidence intervals for key outcomes.
    • Analysis approach: described “true intention-to-treat” analysis (all treated patients; N=3907) and a “final analysis” after exclusions (N=3327).
  • Follow-Up Period:
    • Early: ROSC and hospital admission.
    • Later: survival assessed at 24 hours, 1 week, 1 month, and 1 year; neurological status assessed at discharge (and survival tracked to 1 year).

Key Results

This trial was not stopped early. It completed enrolment over September 1994 to September 1996.

Outcome High-dose epinephrine (5 mg/dose) Standard-dose epinephrine (1 mg/dose) Effect p value / 95% CI Notes
Return of spontaneous circulation (ROSC) 677/1677 (40.4%) 600/1650 (36.4%) Risk difference +4.0% 95% CI 0.7 to 7.3; P=0.02 Final analysis cohort (N=3327)
Admission to hospital 444/1677 (26.5%) 390/1650 (23.6%) Risk difference +2.9% 95% CI −0.1 to 5.9; P=0.05 Final analysis cohort
Admission after a single epinephrine injection 62/444 (14.0%) 43/390 (11.0%) Risk difference +3.0% 95% CI −1.2 to 7.1; P=0.16 Conditional on admission
Survival at 24 hours 237/1677 (14.1%) 229/1650 (13.9%) Not reported P=0.89 All patients (final analysis cohort)
Survival to hospital discharge 38/1677 (2.3%) 46/1650 (2.8%) Risk difference −0.5% 95% CI −1.5 to 0.6; P=0.34 Primary patient-centred outcome as reported
Survival at 1 year 35/1677 (2.1%) 43/1650 (2.6%) Not reported P=0.32 All patients (final analysis cohort)
Favourable neurological status at discharge (CPC 1–2) 29/38 (76.3%) 33/46 (71.7%) Not reported P=0.64 Among discharged survivors
Discharge without neurological impairment (CPC 1) 26/38 (68.4%) 26/46 (56.5%) Not reported P=0.95 Among discharged survivors
  • High-dose epinephrine increased ROSC (40.4% vs 36.4%; P=0.02) and admission to hospital (26.5% vs 23.6%; P=0.05), but did not improve survival to discharge (2.3% vs 2.8%; P=0.34) or survival at 1 year (2.1% vs 2.6%; P=0.32).
  • Rhythm subgroup signals were discordant: in asystole, ROSC was 36.9% vs 32.2% (P=0.01) and admission 25.2% vs 20.3% (P=0.004), whereas in coarse ventricular fibrillation ROSC was numerically lower with high dose (55.9% vs 64.3%; P=0.19).
  • Potential harm signals in early post-resuscitation course were reported: in-hospital mortality during the first 24 hours was 38.7% in the high-dose group vs 32.4% in the standard-dose group (P=0.06).

Internal Validity

  • Randomisation and Allocation:
    • Central randomisation with prepackaged study boxes (15 identical ampoules); boxes distributed in sets of 10 (5 high-dose, 5 standard-dose) and used sequentially.
    • Allocation concealment during resuscitation was credible because ampoules were visually identical and teams/hospitals were blinded to dose.
  • Drop out or exclusions (post randomisation):
    • 3946 study boxes were assigned; 39 were excluded because of incorrect use of the assigned package.
    • “True intention-to-treat” analysis included 3907 treated patients (standard-dose 1938; high-dose 1969).
    • Final analysis excluded 580 patients (345 traumatic arrests; 45 not meeting eligibility; 190 with essential data not available or lost to follow-up), leaving 3327 (standard-dose 1650; high-dose 1677).
    • The large post-randomisation exclusion fraction risks attrition bias relative to a strict ITT framework, particularly if exclusions were not independent of outcomes.
  • Performance/Detection Bias:
    • Double-blinding reduced the risk that clinicians’ expectations influenced continuation of resuscitation or post-ROSC management.
    • Primary early outcomes (ROSC, admission, survival timepoints) were objective; neurological outcomes relied on clinical documentation (CPC at discharge), which may be less standardised.
  • Protocol Adherence:
    • Drug separation was substantial: mean total epinephrine dose 29.0 ± 18.5 mg (high-dose) vs 6.1 ± 3.7 mg (standard-dose).
    • Mean number of epinephrine doses was similar (5.8 ± 3.7 vs 6.1 ± 3.7), consistent with comparable resuscitation duration/effort.
    • Route distribution was similar: peripheral IV 87.2% vs 87.0%; central IV 8.8% vs 9.1%; endotracheal 4.1% vs 3.9%.
  • Baseline Characteristics:
    • Groups were broadly comparable for key arrest features (witnessed arrest 46.8% vs 47.8%; bystander CPR 10.3% vs 9.3%).
    • Initial rhythm distribution was similar: asystole 74.7% vs 74.6%; pulseless electrical activity 8.4% vs 8.4%; coarse ventricular fibrillation 8.5% vs 8.5%; fine/medium ventricular fibrillation 8.4% vs 8.5%.
    • Age differed statistically (64.5 ± 14.9 vs 66.7 ± 14.6 years), with unclear clinical relevance for these outcomes.
  • Heterogeneity:
    • Trial spanned 12 centres across two countries; centre practice varied, including active compression–decompression CPR use (647/1677 vs 645/1650).
    • Because CPR modality was not randomised, centre-level practice differences may confound subgroup/interaction findings.
  • Timing:
    • Epinephrine was administered late relative to collapse in both groups (20.6 ± 14.1 minutes vs 20.7 ± 12.9 minutes), which may constrain any achievable effect on neurological survival.
  • Dose:
    • High-dose strategy used 5 mg every 3 minutes (up to 15 doses), creating large cumulative exposure compared with standard dosing; physiological plausibility exists for increasing ROSC, but the trial did not demonstrate improved survival to discharge or 1-year survival.
  • Separation of the Variable of Interest:
    • ROSC: 40.4% (high-dose) vs 36.4% (standard-dose).
    • Admission: 26.5% vs 23.6%.
    • Discharge: 2.3% vs 2.8%.
  • Outcome Assessment:
    • Endpoints were prespecified and largely objective; longer-term survival was tracked to 1 year.
    • Neurological outcome was summarised using CPC at discharge; among discharged survivors, CPC 1–2 occurred in 76.3% vs 71.7% (P=0.64).
  • Statistical Rigor:
    • Two-sided α=0.05 and 95% confidence intervals for differences in proportions were reported for key outcomes.
    • No reported power calculation; the very low event rate for discharge survival (2–3%) limits precision for patient-centred endpoints.

Conclusion on Internal Validity: Overall, internal validity appears moderate: allocation concealment and blinding were strong and treatment separation was large, but substantial post-randomisation exclusions, centre-level heterogeneity (including non-randomised CPR modality), and late drug delivery complicate causal inference for survival and neurological outcomes.

External Validity

  • Population Representativeness:
    • Adult OHCA population in France/Belgium, with a high proportion of asystole (~75%) and low bystander CPR (~10%), reflecting substantial delays to effective circulation/defibrillation in many cases.
    • Care was delivered by physician-led mobile emergency teams, which differs from paramedic-led systems in many countries.
  • Applicability:
    • The tested regimen (5 mg every 3 minutes, up to 15 doses) is not contemporary standard practice; direct translation to current dosing is therefore limited.
    • Findings remain highly relevant for the conceptual question of dose escalation: improving ROSC/admission does not necessarily translate into improved survival or neurological outcome.
    • Systems with earlier high-quality bystander CPR and earlier defibrillation may have different rhythm distributions and therapeutic windows, potentially modifying the balance of benefit and harm from vasopressors.

Conclusion on External Validity: Generalisability is moderate for comparable European-style advanced prehospital systems, but limited for modern systems with earlier CPR/defibrillation and for current guideline dosing; the central message about surrogate versus patient-centred outcomes remains widely applicable.

Strengths & Limitations

  • Strengths:
    • Large, multicentre, double-blind, pragmatic prehospital randomised trial in adult OHCA.
    • Clear separation of exposure: mean total epinephrine dose 29.0 ± 18.5 mg vs 6.1 ± 3.7 mg.
    • Objective early outcomes with longer follow-up, including survival to 1 year and neurological status at discharge.
  • Limitations:
    • No reported power calculation; survival to discharge was very low (2–3%), limiting precision for patient-centred outcomes.
    • Substantial post-randomisation exclusions (580/3907) and missing data in the final analysis cohort may bias effect estimates.
    • Low bystander CPR rate and delayed epinephrine administration (~20 minutes) may reduce relevance to current systems prioritising early CPR/defibrillation.
    • Active compression–decompression CPR was used in a sizeable minority without randomisation, complicating interpretation of interaction analyses.

Interpretation & Why It Matters

  • ROSC is not enough
    High-dose epinephrine improved ROSC (40.4% vs 36.4%) and admission (26.5% vs 23.6%) but did not improve survival to discharge (2.3% vs 2.8%) or 1-year survival (2.1% vs 2.6%), underscoring the risk of equating physiological success with meaningful recovery.
  • Dose escalation is not a solution
    The strategy of repeated 5 mg dosing (mean total 29.0 mg) did not translate into better neurological outcomes among survivors (CPC 1–2: 76.3% vs 71.7%), supporting the view that escalation beyond standard dosing is unlikely to improve patient-centred outcomes in unselected OHCA populations.
  • Trial design lessons for resuscitation research
    Very low discharge survival rates and late drug delivery (~20 minutes) highlight the importance of designing trials around early, system-level interventions (bystander CPR, defibrillation, minimising no-flow/low-flow time) and around patient-centred endpoints with adequate statistical power.

Controversies & Subsequent Evidence

  • Adopting high-dose epinephrine based on improved intermediate endpoints (ROSC/admission) was challenged because survival to discharge was numerically lower with high dose (2.3% vs 2.8%), emphasising that surrogate endpoints can mislead when neurological survival is unchanged.1
  • Interaction findings were debated because CPR modality (standard versus active compression–decompression) reflected centre practice rather than randomisation, yet discharge survival appeared to “cross over” by CPR type (standard ACLS: 1.7% high-dose vs 3.6% standard-dose; ACD-ACLS: 3.1% vs 1.6%); these observations are hypothesis-generating rather than definitive.1
  • Subsequent placebo-controlled evidence for epinephrine (standard dosing) and subsequent syntheses consistently demonstrate increased ROSC and hospital admission, with only modest (or absent) improvement in longer-term, patient-centred survival outcomes, reinforcing that vasopressor strategies should be evaluated against meaningful survival and neurological recovery rather than ROSC alone.234
  • Large observational datasets have reported time-dependent associations in which delayed prehospital epinephrine is linked to worse neurological outcomes, providing a plausible explanation for why late vasopressor administration may increase ROSC without improving discharge survival or long-term outcomes.56
  • Contemporary guidelines recommend standard-dose epinephrine (1 mg IV/IO every 3–5 minutes) and do not recommend routine high-dose epinephrine; the dose-escalation strategy tested in this European trial is therefore generally viewed as unsupported by patient-centred evidence.78910

Summary

  • In a large, multicentre, double-blind prehospital trial (France/Belgium), repeated 5 mg epinephrine increased ROSC and hospital admission compared with repeated 1 mg dosing.
  • There was no improvement in survival to discharge (2.3% vs 2.8%) or survival at 1 year (2.1% vs 2.6%).
  • Neurological outcomes among survivors were similar (CPC 1–2: 76.3% vs 71.7%).
  • Subgroup signals differed by rhythm (benefit for ROSC/admission in asystole, no benefit in ventricular fibrillation), but interaction inferences were limited by centre-level practice heterogeneity.
  • The trial helped shift emphasis from surrogate resuscitation endpoints to meaningful survival and neurological recovery when evaluating vasopressor strategies.

Further Reading

Other Trials

Systematic Review & Meta Analysis

Observational Studies

Guidelines

Notes

  • Key Results table reports the “final analysis” cohort (standard-dose n=1650; high-dose n=1677) unless otherwise stated; the index manuscript also presented an earlier “true intention-to-treat” analysis (n=3907) with similar qualitative conclusions for survival to discharge.

Overall Takeaway

This European Epinephrine Study was a landmark, double-blind, multicentre prehospital trial that rigorously tested whether escalating epinephrine dose during adult OHCA improves outcomes. It showed that repeated 5 mg dosing increases physiological “success” (ROSC and admission) but does not improve meaningful survival or neurological recovery, reinforcing modern resuscitation’s focus on early CPR/defibrillation and patient-centred endpoints rather than drug-driven surrogate outcomes.

Overall Summary

  • Repeated high-dose epinephrine increased ROSC and admission but did not improve survival to discharge or 1-year survival.
  • Neurological outcome among survivors was similar despite large dose separation (mean total 29.0 mg vs 6.1 mg).
  • The trial helped de-adopt high-dose epinephrine and sharpened emphasis on patient-centred endpoints in resuscitation trials and guidelines.

Bibliography