Skip to main content

Foundational Trials

PARAMEDIC2

Image: Shutterstock / Dragon Images

Ampoule of adrenaline being lifted from a drug tray

Publication

  • Title: A Randomized Trial of Epinephrine in Out-of-Hospital Cardiac Arrest
  • Acronym: PARAMEDIC2
  • Year: 2018
  • Journal published in: The New England Journal of Medicine
  • Citation: Perkins GD, Ji C, Deakin CD, et al. A randomized trial of epinephrine in out-of-hospital cardiac arrest. N Engl J Med. 2018;379(8):711-721.

Context & Rationale

  • Background
    • Adrenaline (epinephrine) has been embedded in adult advanced life support algorithms for decades, largely for its α-adrenergic vasoconstrictor effect (augmenting aortic diastolic pressure and coronary perfusion pressure during CPR).
    • Before PARAMEDIC2, evidence for clinically meaningful outcomes (longer-term survival with good neurological function) was limited and conflicted, despite consistent physiological improvements (ROSC and hospital transport).
    • Observational data were hard to interpret because adrenaline is preferentially given in longer, more refractory arrests (confounding by indication), and quality of CPR and post-resuscitation care vary substantially.
    • Clinicians and guideline panels faced a genuine equipoise problem: adrenaline could increase ROSC and short-term survival but plausibly worsen neurological outcome by prolonging low-flow states and aggravating post-resuscitation microcirculatory dysfunction.
  • Research Question/Hypothesis
    • In adults with out-of-hospital cardiac arrest treated by paramedics, does standard-dose IV/IO epinephrine (1 mg boluses) compared with placebo improve survival at 30 days?
    • Key secondary hypothesis: any survival benefit should translate into more survivors with favourable neurological outcome (modified Rankin Scale [mRS] 0–3), rather than increasing survival with severe disability.
  • Why This Matters
    • The intervention is ubiquitous, inexpensive, and delivered at population scale; even small effect sizes have major public health and system-level implications.
    • A blinded, placebo-controlled prehospital RCT directly addresses longstanding uncertainty and provides a higher-causal-inference benchmark than registry analyses.
    • By pairing survival endpoints with neurological status and resource-use signals, the trial forces explicit consideration of what constitutes “benefit” in resuscitation science.

Design & Methods

  • Research Question: Among adults with out-of-hospital cardiac arrest in whom paramedics attempted resuscitation and opened a trial drug pack during ALS, does epinephrine (1 mg IV/IO boluses) versus placebo improve 30-day survival?
  • Study Type: Pragmatic, individually randomised, multicentre, double-blind, placebo-controlled, prehospital trial across five UK NHS ambulance services (England and Wales); investigator-initiated; 1:1 allocation using minimisation; deferred consent model.
  • Population:
    • Setting: Out-of-hospital advanced life support delivered by paramedics; randomisation occurred at the point a study drug pack was opened during ongoing resuscitation.
    • Inclusion criteria: Adult out-of-hospital cardiac arrest with resuscitation attempted by participating paramedics; trial pack opened for IV/IO study-drug administration during ALS.
    • Key exclusion criteria: Known/suspected age <16 years; known pregnancy; arrest due to anaphylaxis or life-threatening asthma; adrenaline administered before ambulance arrival; other protocol-specified ineligibility identified at screening or (occasionally) after pack opening (e.g., “do not attempt resuscitation” status, ROSC before drug administration).
  • Intervention:
    • Drug and dose: Epinephrine 1 mg (10 mL of 1:10,000) IV or IO (prefilled syringes).
    • Timing (ALS-aligned): Non-shockable rhythms: as soon as IV/IO access achieved; shockable rhythms: after the third shock.
    • Repeat dosing: 1 mg every 3–5 minutes while resuscitation continued, up to 10 doses (single study pack).
  • Comparison:
    • Placebo: 0.9% saline in identical prefilled syringes and identical dosing schedule (including a 10-dose cap per pack).
    • Co-interventions: All other resuscitation and post-resuscitation care per contemporaneous UK ALS guidance and local hospital practice.
  • Blinding: Double-blind (paramedics, receiving clinicians, outcome assessors, and trial team); allocation concealed via identical packs/syringes; emergency unblinding only if clinically essential.
  • Statistics: A total sample size of 8000 was planned to detect a 2% absolute difference in 30-day survival (6% vs 8%) with 93% power at a two-sided 5% significance level; primary analysis used a modified intention-to-treat population (randomised and confirmed to have received the assigned intervention), estimating odds ratios using regression models (unadjusted and prespecified adjusted analyses).
  • Follow-Up Period: Primary endpoint at 30 days; neurological outcomes at hospital discharge and 3 months (mRS); longer-term follow-up (including 6 and 12 months) reported subsequently.

Key Results

This trial was not stopped early. Recruitment continued to the planned sample size (8014 analysed), with pre-specified interim data monitoring and no stopping boundary reached.

Outcome Epinephrine Placebo Effect p value / 95% CI Notes
Survival at 30 days (primary) 130/4012 (3.2%) 94/3995 (2.4%) OR 1.39 95% CI 1.06 to 1.82; P=0.02 Adjusted OR 1.47; 95% CI 1.09 to 1.97
Survival until hospital admission (sustained ROSC to admission) 947/3973 (23.8%) 319/3982 (8.0%) OR 3.59 95% CI 3.14 to 4.12; P=Not reported Adjusted OR 3.83; 95% CI 3.30 to 4.43
Survival to hospital discharge 128/4009 (3.2%) 91/3995 (2.3%) OR 1.41 95% CI 1.08 to 1.86; P=Not reported Adjusted OR 1.48; 95% CI 1.10 to 2.00
Survival to discharge with favourable neurological outcome (mRS 0–3) 87/4007 (2.2%) 74/3994 (1.9%) OR 1.18 95% CI 0.86 to 1.61; P=Not reported Adjusted OR 1.19; 95% CI 0.85 to 1.68
Survival at 3 months 121/4009 (3.0%) 86/3991 (2.2%) OR 1.41 95% CI 1.07 to 1.87; P=Not reported Adjusted OR 1.47; 95% CI 1.08 to 2.00
Survival at 3 months with favourable neurological outcome (mRS 0–3) 82/3986 (2.1%) 63/3979 (1.6%) OR 1.31 95% CI 0.94 to 1.82; P=Not reported Adjusted OR 1.39; 95% CI 0.97 to 2.01
Severe neurological impairment among survivors at discharge (mRS 4–5) 39/126 (31.0%) 16/90 (17.8%) Not reported Not reported Reported among survivors; illustrates disability burden in the intervention arm
Median ICU length of stay among survivors (days) 7.5 (IQR 3.6–14.4) 7.0 (IQR 3.7–19.1) Not applicable Not reported Among non-survivors: 2.0 (0.9–3.8) vs 3.0 (1.6–5.6)
  • Epinephrine increased 30-day survival from 2.4% to 3.2% (OR 1.39; 95% CI 1.06 to 1.82; P=0.02), corresponding to a number needed to treat of 112 (95% CI 63 to 500) for one additional 30-day survivor.
  • The physiological/early pathway signal was large (ROSC 1457/4015 [36.3%] vs 468/3999 [11.7%]; sustained ROSC to hospital admission 23.8% vs 8.0%), but there was no statistically significant increase in favourable neurological survival at discharge (2.2% vs 1.9%) or at 3 months (2.1% vs 1.6%).
  • Among survivors, severe disability at discharge (mRS 4–5) was more frequent with epinephrine (31.0% vs 17.8%), sharpening the survival–disability trade-off implicit in the primary result.

Internal Validity

  • Randomisation and Allocation
    • Individual allocation via concealed trial packs opened during resuscitation; 1:1 allocation using a minimisation algorithm.
    • Allocation concealment was operationally robust because treatment identity was embedded in indistinguishable packs/syringes.
  • Drop out or exclusions (post-randomisation)
    • Trial packs opened: 8103.
    • Excluded after randomisation: 87 (plus 2 with unknown allocation), leaving 8014 analysed (4015 epinephrine; 3999 placebo).
    • Primary outcome denominators indicate near-complete ascertainment (4012 and 3995 for 30-day survival; 99.9% in each group).
  • Performance/Detection Bias
    • Double-blinding across prehospital and in-hospital phases reduces co-intervention and ascertainment bias risk for survival outcomes.
    • Primary outcome (30-day survival) is objective; neurological outcome (mRS) is more vulnerable to assessment variability but remained prespecified and standardised.
  • Protocol Adherence
    • Median emergency call-to-trial agent time was similar: 21.5 minutes (IQR 16.0–27.3) vs 21.1 minutes (IQR 16.1–27.4).
    • Mean number of trial-drug doses administered was similar: 4.9 ± 2.5 vs 5.1 ± 2.3.
    • Resuscitation duration (ALS initiation to cessation) differed modestly: 47.5 minutes (IQR 35.1–64.0) vs 43.1 minutes (IQR 33.5–56.1).
  • Baseline Characteristics
    • Age: 69.7 ± 15.6 years vs 69.8 ± 15.5 years; male sex: 65.0% vs 64.6%.
    • Initial shockable rhythm: 19.2% vs 18.7%; bystander CPR: 59.3% vs 58.7%.
    • Presumed medical cause: 91.1% vs 92.3%.
  • Heterogeneity
    • Delivered across five ambulance services with pragmatic care pathways; outcomes were analysed using regression models and pre-specified subgroup assessments.
    • Subgroup interaction signals were not compelling (e.g., shockable vs non-shockable interaction P=0.15 for the primary endpoint subgroup analysis), supporting a broadly consistent treatment effect direction.
  • Timing
    • Emergency call-to-ambulance arrival: 6.7 minutes (IQR 4.3–9.7) vs 6.6 minutes (IQR 4.2–9.6).
    • Call-to-trial drug (median ~21 minutes) reflects real-world EMS intervals; the design did not mandate earlier drug delivery beyond guideline timing.
  • Dose
    • Standard-dose epinephrine (1 mg IV/IO every 3–5 minutes; maximum 10 doses) reflects contemporaneous guideline dosing and supports construct validity.
    • Exposure was substantial (mean ~5 boluses), consistent with the high proportion of non-shockable rhythms and prolonged resuscitations.
  • Separation of the Variable of Interest
    • ROSC: 1457/4015 (36.3%) vs 468/3999 (11.7%).
    • Transported to hospital: 2041/4015 (50.8%) vs 1227/3999 (30.7%).
    • Survival until hospital admission (sustained ROSC): 23.8% vs 8.0% (OR 3.59; 95% CI 3.14 to 4.12).
  • Key Delivery Aspects
    • Pragmatic delivery embedded within ALS workflows reduces “trial effect” inflation and supports fidelity to real-world practice.
    • Deferred consent was used, reducing selection bias at enrolment and supporting inclusion of the sickest, time-critical population.
  • Crossover
    • Not reported.
  • Adjunctive therapy use
    • Amiodarone administration: 14.5% vs 9.2% (likely reflecting rhythm evolution and greater ROSC/transport in the epinephrine arm).
    • Number of defibrillation shocks after randomisation: 1.4 ± 2.4 vs 0.9 ± 2.0.
  • Outcome Assessment
    • Primary survival outcome is objective and near-complete.
    • Neurological outcome used mRS at discharge and 3 months; although mRS is widely used, it may be less sensitive to cognitive and psychosocial domains relevant after cardiac arrest.
  • Statistical Rigor
    • Planned sample size was achieved (8014 analysed), meeting the design’s power target.
    • Primary analysis was prespecified and statistically significant (P=0.02), with concordant adjusted analyses.

Conclusion on Internal Validity: Overall, internal validity appears strong given robust allocation concealment, double-blinding, pragmatic protocol delivery, and near-complete ascertainment of an objective primary endpoint; internal validity is modestly tempered by post-randomisation exclusions and the modified intention-to-treat definition.

External Validity

  • Population Representativeness
    • Typical adult OHCA case-mix (mean age ~70 years), with a predominance of non-shockable rhythms (PEA/asystole ~78%) and bystander CPR in ~59%.
    • Exclusions (children, pregnancy, anaphylaxis/life-threatening asthma, and those receiving adrenaline before ambulance arrival) limit applicability to those specific populations and to systems where pre-EMS adrenaline is common.
  • Applicability
    • Findings generalise well to EMS systems delivering paramedic-led ALS with similar response intervals and drug timing constraints.
    • Extrapolation to physician-led systems with substantially earlier drug delivery, or to in-hospital cardiac arrest, is biologically plausible but not directly tested.
    • The comparator was placebo (no epinephrine), making the result a direct estimate of epinephrine’s incremental value versus withholding the drug rather than versus alternative vasopressor strategies.

Conclusion on External Validity: External validity is generally good for adult out-of-hospital cardiac arrest managed by paramedic-led ALS in high-income EMS systems, but is limited for excluded subgroups and for settings with materially different timing and structure of prehospital care.

Strengths & Limitations

  • Strengths:
    • Large sample size (n=8014) with high event ascertainment for the primary endpoint.
    • Rare prehospital placebo-controlled, double-blind design, directly addressing confounding by indication.
    • Pragmatic delivery embedded within routine ALS pathways across multiple ambulance services.
    • Clinically meaningful outcome hierarchy: 30-day survival plus functional neurological outcomes (mRS) at discharge and 3 months.
  • Limitations:
    • Modified intention-to-treat analysis and post-randomisation exclusions introduce a small but non-zero risk of bias versus strict ITT.
    • Absolute survival benefit was small (0.8% at 30 days), and estimates for favourable neurological survival were imprecise due to low numbers of survivors.
    • Median drug delivery time (~21 minutes from call) reflects real-world systems but limits inference about earlier administration strategies.
    • mRS provides a broad disability signal but may not capture cognitive and psychosocial morbidity typical after cardiac arrest.

Interpretation & Why It Matters

  • Efficacy pathway clarity
    Epinephrine’s principal effect is upstream (ROSC and hospital admission), with only a small downstream gain in 30-day survival and no clear increase in favourable neurological survival, reframing adrenaline as a “ROSC drug” with uncertain patient-centred benefit.
  • Patient-centred trade-off
    The trial quantifies a pragmatic survival–disability tension: more survivors overall, but a higher proportion of survivors with severe neurological impairment (mRS 4–5) at discharge, requiring explicit value judgements in practice and policy.
  • System impact
    Because epinephrine substantially increases transport and short-term survival pathways, it is likely to increase downstream ICU/hospital utilisation even when the net number of neurologically favourable survivors changes little.
  • Methodological landmark
    PARAMEDIC2 demonstrates that large-scale, blinded, prehospital drug trials with deferred consent are feasible and can challenge long-standing guideline assumptions using high-causal-inference evidence.

Controversies & Subsequent Evidence

Summary

  • In adult OHCA managed by UK paramedic-led ALS, epinephrine modestly increased 30-day survival (3.2% vs 2.4%; OR 1.39; 95% CI 1.06 to 1.82; P=0.02).
  • Epinephrine produced a large physiological signal (ROSC 36.3% vs 11.7%; survival to hospital admission 23.8% vs 8.0%) but did not clearly increase survival with favourable neurological outcome at discharge or 3 months.
  • Among survivors, severe neurological impairment at discharge (mRS 4–5) occurred more frequently in the epinephrine arm (31.0% vs 17.8%).
  • Internal validity is strong overall (concealed allocation, double-blind design, near-complete follow-up), with modest concerns from post-randomisation exclusions and modified ITT.
  • Long-term follow-up shows small absolute differences persist at 6 and 12 months, with borderline-to-modest adjusted effects on survival and favourable neurological outcome.1

Overall Takeaway

PARAMEDIC2 is a landmark because it delivered a definitive, blinded, placebo-controlled estimate of epinephrine’s net effect in out-of-hospital cardiac arrest at real-world scale. It established that epinephrine strongly increases ROSC and hospital admission, yields a small absolute survival gain, and does not clearly increase neurologically favourable survival—forcing resuscitation science and guidelines to confront the survival–disability trade-off with quantified evidence.

Overall Summary

  • Epinephrine improves short-term survival pathways (ROSC/admission) and modestly increases 30-day survival, but does not clearly improve favourable neurological survival and increases the proportion of severely disabled survivors at discharge.

Bibliography