Bringing doctors together

or

A free, private network for verified physicians

Leading voices

Dr. Stefan Wheat

General Emergency Medicine · UW

View

Dr. Ryan Stanton

General Emergency Medicine · ACEP

View

Dr. Jim Dahle

General Emergency Medicine · The White Coat Investor

View

Dr. Mohamed Hagahmed

General Emergency Medicine · University of Pittsburgh

View

Dr. Sorabh Khandelwal

General Emergency Medicine · OSU

View

Dr. Jennifer Wiler

General Emergency Medicine · UCD

View

Dr. Gita Pensa

General Emergency Medicine · Brown University

View

Dr. Heidi Best

General Emergency Medicine · Emergency Physicians of Tidewater

View

Dr. Kimberly Chernoby

General Emergency Medicine · GW

View

Dr. Amal Mattu

General Emergency Medicine · UMSOM

View
  • Discover your community

  • Stay current

  • A profile that represents you

  • Find your people

About RoonPrivacyTerms

    Overview

    International Consensus on Evidence Gaps and Research Opportunities in Extracorporeal Cardiopulmonary Resuscitation for Refractory Out-of-Hospital Cardiac Arrest: A Report From the National Heart, Lung, and Blood Institute Workshop.

    Mar 5, 2025DOI: 10.1161/jaha.124.036108Journal ArticleAHA

    Written by: Laurie J MorrisonElizabeth A HuntBrian GrunauThomas P AufderheideClifton CallawayJoseph E TonnaComilla SassonAudrey BlewerBryan F McNallyDemetris YannopoulosJan BelohlavekJason BartosAlain CombesAhamed IdrisRaina M MerchantLeith StatesEmily TinsleyRenee WongScott T YoungquistGeorge SopkoKarl B Kern

    Find these authors on Roon

    CS

    Dr. Comilla Sasson

    General Emergency Medicine · CU Anschutz Medical Campus

    SY

    Dr. Scott Youngquist

    General Emergency Medicine · UU

    APPROACH

    The NHLBI virtual workshop invited over 100 international multidisciplinary experts (Table S1) in resuscitation science to address 4 objectives related to ECPR for OHCA, which included: Review the current science and regulatory policy related to ECPR.Consider the ethics of implementing a sophisticated intervention that cannot be delivered everywhere.Discuss where ECPR best fits in the OHCA resuscitation system of care and the research and health care delivery workforce requirements.Provide a concise, prioritized list of research opportunities based on identified gaps achieved through a rank order derived by the primary authors reviewing all submitted section chair summaries, vetted across section chairs initially and ratified by consensus across authors.Review the current science and regulatory policy related to ECPR.Consider the ethics of implementing a sophisticated intervention that cannot be delivered everywhere.Discuss where ECPR best fits in the OHCA resuscitation system of care and the research and health care delivery workforce requirements.Provide a concise, prioritized list of research opportunities based on identified gaps achieved through a rank order derived by the primary authors reviewing all submitted section chair summaries, vetted across section chairs initially and ratified by consensus across authors.This workshop addresses 3 key NHLBI (NIH) research priorities32: Turning discovery into cardiovascular healthAddress the inequities in careAddress the plateau in cardiovascular outcomesTurning discovery into cardiovascular healthAddress the inequities in careAddress the plateau in cardiovascular outcomesThis workshop aligns with the NHLBI Strategic Objectives 3 through 8 and, also with 2 of 8 recommendations (namely recommendations 6 and 7) from the 2015 Institute of Medicine Report33 to accelerate research on the pathophysiology, new therapies, and translation of cardiac arrest science, as well as studies on the evaluation and adoption of cardiac arrest therapies.

    SYNTHESIS

    A search of Medline, Embase, and Evidence‐Based Medicine Reviews for randomized clinical trials and observational studies published before May 22, 2018, identified 25 observational studies pertaining to adult OHCA (15), adult in‐hospital cardiac arrest (IHCA; 7) and pediatric IHCA (3). The authors concluded that the studies had a critical risk of confounding potentially limiting internal validity, and the heterogeneity across studies precluded a meaningful meta‐analysis. The review illustrated the need for rigorous randomized clinical trials or alternative study designs, minimizing bias to clarify the role of ECPR in cardiac arrest.18 Based on this review, the International Liaison Committee on Resuscitation currently suggests that ECPR may be considered as a rescue therapy for selected patients with cardiac arrest when conventional CPR is failing in settings in which ECPR can be implemented (weak recommendation, very‐low‐certainty evidence).34
    The ARREST17 trial was a single‐center, open‐label, randomized trial truncated after 30 patients were recruited from an unknown denominator of total patients with OHCA treated during the trial. Randomized patients were identified upon emergency department (ED) arrival as having a low probability of a favorable outcome. The ARREST trial was the first trial to show a significant improvement of survival or neurological outcomes at hospital discharge in patients with OHCA treated with ECPR as part of a in‐hospital care strategy for initial rhythm VF cases resistant to OHCA resuscitation presenting to the ED in addition to standard care, which typically included CPR, defibrillation, intubation, intravenous medications, angiography, and revascularization if indicated and postarrest care compliant with current guidelines. The authors called the strategy; ECPR‐facilitated resuscitation. The Data and Safety Monitoring Board recommended stopping the ARREST trial on the basis of a substantial improvement in the primary outcome attributed to ECPR‐facilitated resuscitation after randomization. The ARREST trial primary end point, survival to hospital discharge, was reported in 6 of 14 (43%) in the intervention arm versus 1 of 15 (7%) in the standard arm (risk difference, 36% [3.7%–59.2%]; 0.9861 posterior probably of ECMO superiority; Fishers' exact testing statistic=0.0352; P < 0.04). Neurological outcome was assessed using a modified Rankin Scale; a 6‐point disability scale with possible scores ranging from 0 to 5, and a category 6 for patients who die. All 6 ECPR survivors improved and achieved a favorable modified Rankin Scale score (≤3) at 6‐month follow‐up. The lone survivor in the control arm died before 3 months. No unanticipated serious adverse events were observed with ECPR (Table 1).
    Variables of Interest17, 29, 30
    ACLS indicates advanced cardiac life support; ARREST, Advanced Reperfusion Strategies for Refractory Cardiac Arrest; CPC, cerebral performance category; CPR, cardiopulmonary resuscitation; ECPR, extracorporeal cardiopulmonary resuscitation; INCEPTION, Early Initiation of Extracorporeal Life Support in Refractory OHCA; NA, not available; Neuro, neurological function; and OHCA, out‐of‐hospital cardiac arrest.
    *Non‐White ECMO arm; standard arm; total: Black or African‐American 1 (7%); 1 (7%); 2 (7%); Native American or Native Alaskan1 (7%); 0; 1 (3%); Not specified by family 9 (60%); 11 (73%); 20 (67%).
    Published after the NHLBI workshop, the prospective single‐center, open‐label, randomized Prague OHCA trial29 compared ECPR with standard care in patients with refractory OHCA presenting with shockable and nonshockable initial rhythms. Standard care was resuscitation in the prehospital setting for the majority of patients with prehospital termination for unsuccessful resuscitations. Patients were randomized in the prehospital setting to compare standard advanced cardiac life support on site versus early intra‐arrest transport for immediate in‐hospital ECPR implementation. Both treatment and control received postarrest care as per institutional guidelines. As recommended by the Data and Safety Monitoring Board, the trial was stopped early for futility after showing no significant improvement in the primary end point of favorable neurological survival at 180 days achieved in 39 of 124 (31.5%) patients in the invasive arm and 29 of 132 (22.0%) in the standard arm (odds ratio [OR], 1.63 [95% CI, 0.93–2.85]; P=0.09) where favorable neurological survival was defined as Cerebral Performance Category 1 or 2, a scale based on the degree of cognitive disability and consequent physical and functional limitations. The secondary end point of favorable neurological survival at 30 days was achieved in 38 of 124 (30.6%) versus 24 of 132 (18.2%) (OR, 1.99 [95% CI, 1.1–3.57]; P=0.02). The trial has also demonstrated several important observations in this randomized refractory OHCA population. The trial was conducted in a well‐functioning prehospital care system with bystander CPR in 99% of cases, telephone‐assisted CPR in 75% of cases; and early hospital resuscitation with immediate access to ECPR <60 minutes from CPR start. These performance metrics could be challenging to replicate in other care systems. The trial design made it challenging to show a significant effect of the intervention as (1) patients with refractory OHCA, regardless of initial rhythm, were enrolled, resulting in 49% enrollment of patients with nonshockable initial rhythms; (2) 10 prognostically favorable standard‐arm cases were actually treated with ECPR, of whom 40% achieved favorable 180‐day outcomes (in comparison with invasive‐arm cases treated with ECPR, of whom 20% achieved favorable 180‐day outcomes, highlighting the impact on between‐group comparisons); and (3) neurological function at 180 days was defined as the primary outcome rather than survival to discharge or 30‐day outcomes29 (Table 1).
    In early 2023, results were published from a multicenter randomized trial in the Netherlands, the INCEPTION trial (NCT03101787; 10 sites and 12 EMS systems of care) testing ECPR (70 patients) versus conventional CPR in OHCA (64 patients).30 It was designed to achieve an absolute difference of 22% increase in the primary outcome Cerebral Performance Category score of 1 or 2 at 30 days for the intervention group compared with controls. The primary end point at the prespecified sample size for this adaptive pragmatic trial design was reported in 10 of 64 (16%) with conventional treatment and 14 of 70 (20%) with ECPR (OR, 1.40 [95% CI, 0.5–3.50]; P=0.52). Serious adverse event rates were similar for both cohorts. It is noted that the duration of intensive care following ECPR initiation was short, consisting of a median of only 1 day (interquartile range, 1–4 days)30 (Table 1).
    The ARREST trial17 provides clinical evidence that implementing an ECPR‐facilitated resuscitation program for out‐of‐hospital refractory cardiac arrest can improve outcomes with field identification. The question remains how many other communities can achieve similar results following the same strict and demanding protocol‐driven approaches as the ARREST trial and whether alternative approaches could provide similar successful results. The Prague OHCA trial29 and the INCEPTION trial30 provide evidence that not all ECPR protocols will be equally successful. Together, these 3 randomized trials show the importance of: A foundation of excellent basic life support (including layperson participation) and advanced cardiac life support programs within the community, and medical community agreement and cooperation regarding the implementation and support of ECPR.Careful patient selection for ECPR.Critical time limitations for successful ECPR.Importance of provider training, experience, and expertise.The need for total commitment to long‐term intensive care unit care of patients treated with ECPR, including a protocol‐driven approach to angiography and revascularization if indicated and withdrawal of life‐sustaining therapy.Administrative support to achieve the above requirements.
    A foundation of excellent basic life support (including layperson participation) and advanced cardiac life support programs within the community, and medical community agreement and cooperation regarding the implementation and support of ECPR.
    Careful patient selection for ECPR.
    Critical time limitations for successful ECPR.
    Importance of provider training, experience, and expertise.
    The need for total commitment to long‐term intensive care unit care of patients treated with ECPR, including a protocol‐driven approach to angiography and revascularization if indicated and withdrawal of life‐sustaining therapy.
    Administrative support to achieve the above requirements.
    The EROCA (Extracorporeal Cardiopulmonary Resuscitation for Refractory Out‐of‐Hospital Cardiac Arrest) study35 tested the feasibility of selecting patients with OHCA for expedited transport and initiating ECPR in the ED. Of 151 OHCA 911 calls, 15 subjects (10%) were randomized in a 4:1 ratio to expedited transport (12) versus standard care (3). Five of 12 patients randomized to expedited transport had an ED arrival time of ≤30 minutes (overall mean, 32.5 minutes [SD, 7.1]). Five patents were eligible for and treated with ECPR, and of these, 3 ECPR‐treated subjects had flow initiated in ≤30 minutes of ED arrival (overall mean, 32.4 minutes [SD, 10.9]). No subject in either group survived with a good neurological outcome, and the trial did not meet predefined thresholds for feasibility outcomes.
    The pragmatic, parallel BC (British Columbia) study36 compared the outcomes of ECPR‐eligible OHCAs in regions with and without an ECPR protocol over 35 months. The predicted sample size for the ECPR region was 70 subjects with OHCA (35 estimated to receive ECPR) and a comparison group of 350 subjects with OHCA during the same time interval. This cohort study was prematurely terminated for changes in hospital protocol and COVID‐19 after 58 patients were enrolled in the ECPR group (a total of 13 patients were managed with ECPR: 11 in the ECPR region and 2 in control regions). There was no difference in the primary outcome of survival or favorable neurological outcomes at hospital discharge between the groups. With such a small number of cases in the ECPR‐treated group, the premature study termination, and the higher‐than‐anticipated outcomes in the control group, the study was underpowered. Like the Prague trial, cases were identified in the prehospital setting. The BC study reported a 60% return of spontaneous circulation rate and 30% survival to discharge with conventional CPR. The authors surmised that the higher survival in the control group made it more difficult for ECPR to demonstrate a significant improvement in outcomes.
    Reynolds et al37 conducted a secondary analysis of patients enrolled in the Resuscitation Outcomes Consortium Epistry–Cardiac Arrest to identify potential ECPR candidates and described baseline and functional survival, as well as resuscitation duration. Only 11% of the patients were potential ECPR candidates. Of these, the conventional standard care survival‐to‐hospital discharge rate was 38%, and 30% of the patients were discharged with a favorable neurological outcome.
    In 2022, the ELSO (Extracorporeal Life Support Organization) investigators published an ECMO interim guideline statement.38 They concluded that there are limited robust data on patient selection criteria and recommended locally derived inclusion criteria, where resources are wisely used on patients most likely to benefit.
    When inclusion and exclusion criteria are locally derived and applied to a population of cardiac arrest cases in randomized and implementation studies, few patients meet ECPR criteria (Table 2),17, 29, 30, 35, 36, 37, 38, 39 and these criteria to date are inconsistently applied across studies and trials such that the results are not comparable. Given that rigorous data are lacking, the authors by consensus recommended a universal approach to inclusion criteria for future research to address this knowledge gap. Table 2 lists the recommended inclusion criteria for future research by consensus.
    Inclusion Criteria for ECPR Trials and Feasibility Studies17, 29, 30, 35, 36, 37, 38, 39
    CPC indicates Cerebral Performance Category; CPR, cardiopulmonary resuscitation; DNR, do not resuscitate or advance directive; ECPR, extracorporeal cardiopulmonary resuscitation; ED, emergency department; ELSO, Extracorporeal Life Support Organization; EMS, emergency medical services; EROCA, Extracorporeal Cardiopulmonary Resuscitation for Refractory Out‐of‐Hospital Cardiac Arrest; ETco 2, end‐tidal CO2; GIS, Geographic Information System; LTC, long‐term care; OHCA, out‐of‐hospital cardiac arrest; RCT, randomized controlled trial; ROSC, return of spontaneous circulation; and TOR, termination of resuscitation universal rule for out‐of‐hospital cardiac arrest.
    In 2023, a meta‐analysis and trial sequence analysis was published reviewing the evidence for ECPR for both in‐hospital and out‐of‐hospital adult patients in cardiac arrest.40 Compared with conventional CPR, ECPR decreased the mortality rate in patients with IHCA (157/185 [84.9%] versus 129/185 [69.7%]; OR, 0.42 [95% CI, 0.25–0.70]; P=0.0009) but not in patients with OHCA (3485/4332 [80.4%] versus 3263/4330 [75.4%]; OR, 0.76 [95% CI, 0.54–1.07]; P=0.12). The authors noted their results do not preclude the possibility that ECPR could be effective in OHCA if the appropriate patient population and timing of ECPR initiation could be defined.

    REGISTRY DATA

    Registries collect data for surveillance, quality improvement, and research purposes. By design, data burden is least with surveillance and greatest with research. Understanding ECPR impact at the hospital, community, regional, state, and national levels requires a thoughtful approach to goals and sustainability based on a clear understanding of funding commitment over time and baseline data. Registries are informative in several scenarios for clinical practice and with regulatory agencies. For example, a registry can provide valid evidence for an off‐label indication to be transitioned to on‐label. However, cohort studies have limitations for using causal inference including missing data and limited data quality oversight generic to all registries. Specific to ECPR, it is important to note that inclusion of only subjects receiving ECPR will limit the ability to infer any ECPR harm or benefit, and such inference would rely only on comparisons with historical data or data from other sources that might not have the same detail. There is also a potential bias of clinicians selecting ECPR, whereby unmeasured features of the patients might make them qualitatively different from comparators in a way that is invisible to the registry. Therefore, a registry that includes all ECPR‐eligible patients and ensures complete capture of ECPR cases would be informative to allow case estimation and regional use over time.
    The largest international registry of >200 000 patients of all ages undergoing ECMO, including ECPR, is maintained by ELSO, an international nonprofit consortium of health care institutions, researchers, patients, and industry partners dedicated to research, patient advocacy, education, and supporting industry partners in their development and monitoring of extracorporeal life support therapies. ELSO supports observational research and embedded prospective studies. ELSO collects predominantly inpatient data from over 543 self‐reporting centers on 6 continents and across 56 countries. ELSO also includes data from 369 US registered member hospitals, systems, or institutions, enrolling both OHCA and IHCA and reporting outcomes. ELSO has recently updated ECPR data collection, but it does not currently capture information on outcomes after hospital discharge. Participating institutions pay a subscription based on country of origin from $75 to $2250 per annum. As of October 2022, 12 125 adult cardiac arrest ECPR cases have been entered with a 30% survival rate to hospital discharge.41
    There are other national and regional registries reporting ECPR data. The EXCEL (A Comprehensive Binational Registry on the Treatment and Outcomes of Patients Requiring ECMO; NCT03793257) registry provides detailed data about ECMO use and patient outcomes in Australia and New Zealand up to 12 months after discharge for all adult patients42 and supports embedded trials and research. EXCEL submits data to the ELSO registry on behalf of ELSO member centers in that region.
    The Sudden Death Expertise Center prospective registry captures data from Paris and 3 of its suburbs (Hauts‐de‐Seine, Seine‐Saint‐Denis, and Val‐de‐Marne); an area covering 762 km2 with a population of 6.8 million. Seven years of data published in 2019 reported 21 800 OHCA cases with presumed cardiac cause, of which 13 191 were treated with advanced life support, and of these, 525 were treated with ECPR. ECPR was applied in the prehospital setting in 33% of cases with application failure in ≈10%. Survival to hospital discharge was not significantly different (8.4% ECPR versus 8.6% conventional, propensity matched; P=0.41). Favorable neurological outcome at discharge was significantly better with conventional resuscitation (84% ECPR versus 96% conventional, P=0.001).43
    Started in 2016, the Extracorporeal Resuscitation Consortium Emergency Department ECMO Collaborative44 at the University of Utah has 10 contributing institutes enrolling patients with OHCA with ECPR in the United States and Canada. The Extracorporeal Resuscitation Consortium is built off the variables of the ELSO registry, offers free registration, and encourages linked membership between the registries. A 2020 publication reports 75 cases over 5 years with a 26.5% survival rate.44
    There are 2 existing US cardiac arrest registries: the CARES (Cardiac Arrest Registry to Enhance Survival) OHCA registry data set45 and the AHA's GWTG‐R (Get With The Guidelines–Resuscitation)46 for IHCA resuscitation data.
    CARES45 has a catchment area of >175 million people, who are served by >2300 EMS agencies and >2500 hospitals in 46 states, generating an annual call volume of >150 000. CARES collects 24 essential prehospital and 5 hospital data points but does not include any ECPR data points. A population‐based state subscription fee ($17000–25 000 per annum) covers all hospitals and EMS agencies annually. The ongoing 7‐year NIH‐funded multisite, pragmatic, cluster‐randomized, phase III RACECARS (Randomized Cluster Evaluation of Cardiac Arrest Systems)47 trial is testing a strategic, tailored OHCA system of care optimization intervention that can be adjusted to the needs and abilities/resources of EMS regions. The intervention is tailored county by county, involving nearly 60 counties with 4 main elements: (1) optimized medical EMS‐dispatch performance with rapid recognition of cardiac arrest and dispatch of emergency response, (2) enhanced EMS‐dispatch telephone coaching of bystander CPR, (3) improved first responder performance with automated external defibrillator use, and (4) comprehensive public training of CPR and automated external defibrillator use. The trial uses the CARES registry to capture consecutive OHCAs and hospital follow‐up, enrolling essentially the entire region target population. The trial (NCT04660526) collects the Cerebral Performance Category score at hospital discharge and modified Rankin Scale score and quality of life measures up to 12 months after OHCA. Collecting data from the prehospital setting to hospital admission and discharge is challenging, as EMS delivers patients to the nearest hospital on the basis of EMS Global Positioning System position, potentially increasing data capture challenges across multiple hospitals. Congress passed the Cardiovascular Opportunities Legacy Act in late 2022 providing appropriation for the CDC CARES Expansion and Modernization Grant to 2027. Grant goals include CARES expansion to all 50 states, dedicated quality improvement activities, a focus on addressing disparity in at‐risk populations, and the development of a Next‐Generation CARES software platform. This will allow for a modernized data experience and easier identification of all patients receiving ECPR nationally.
    The AHA's GWTG‐R46 quality improvement program has >500 participating hospitals nationwide and 5 hospitals in the Kingdom of Saudi Arabia, collectively contributing >35 000 cases of IHCA annually to a comprehensive cardiopulmonary arrest data set. Additionally, GWTG‐R gathers data on in‐hospital rapid response/medical emergency teams, in‐hospital acute respiratory compromise events, and post–cardiac arrest care for both IHCAs and OHCAs. Participating sites can compare their data against blinded benchmark groups such as academic centers, rural sites, and pediatric‐only facilities, and geographically related sites. They also receive a personalized annual Risk Standardized Survival to Discharge report. An optional ECMO data set includes 49 discrete ECMO fields to capture details on its use during an in‐hospital cardiopulmonary arrest. These fields capture cannulation specifics, flow, and cerebral imaging, as well as neurological injury during or after ECMO up to discharge or in‐hospital death. ELSO patient record numbers may be collected for record cross‐referencing, quality improvement purposes, and investigator analysis. GWTG‐R is scheduled for a significant platform update in 2024, which will incorporate all ECMO and ECPR data elements into the core data set. Contributing GWTG sites pay an annual subscription fee based on their discharge volume.
    The US Food and Drug Administration presented a regulatory perspective at the NHLBI workshop. The reclassification of medical devices for ECMO from class III (high risk device) to class II (intermediate‐risk device) provided manufacturers with the 510(k) regulatory pathway to obtain marketing clearance for the indications for use identified in the final order48 (ie, “… assisted extracorporeal circulation and physiologic gas exchange of the patient's blood in patients with acute respiratory failure or acute cardiopulmonary failure, where other available treatment options have failed, and continued clinical deterioration is expected or the risk of death is imminent”). The US Food and Drug Administration's Final Order also identified health risks and subsequent special controls necessary to address these risks, as well as provided a reasonable assurance of safety and effectiveness when used as intended. With the regulatory pathway and special controls identified, sponsors can submit their in vitro, in vivo, and clinical data in a 510(k) application in support of marketing clearance.
    The investigational device exemption and premarket approval regulatory pathways are also available for situations when sponsors are interested in investigating the use of devices (not yet cleared for ECMO indications) in patient populations who may benefit from ECMO for a proposed use where sufficient clinical information does not exist to fall within the defined class II regulation or where additional clinical data may be needed to address the special controls.
    Thus, the path forward to ECPR clinical adoption most likely requires consistent eligibility criteria and an international registry for surveillance, best practices definition, and performance monitoring through a common data set. Site participation in future ECPR clinical trials evaluating long‐term outcomes and cost effectiveness should depend on achieving and maintaining preset performance requirements for the system of care.
    Health equity49 is providing care that does not vary in quality because of personal characteristics such as sex, race, ethnicity, geography, immigrant status, and socioeconomic status. Health equity is achieved when every person has the opportunity to attain their full individual health potential regardless of social position or other socially determined construct.50 There are known inequities in OHCA treatment and survival outcomes by race, sex, geography, and socioeconomic status. For nearly 10 years, it has been known that patients who are Black or Hispanic individuals, women, those who lack private insurance, or live in rural settings have lower rates of good neurological recovery and survival to hospital discharge when compared, respectively, with patients who are White individuals, men, privately insured, or live in urban settings.51, 52
    If left unaddressed, these inequities are heightened in the context of ECPR, which is more time intensive, requires specialized training, and can be impacted by selection and implicit biases. For example, in the ARREST trial, eligibility criteria included an initial shockable rhythm. As neighborhood race and ethnicity (Black and Hispanic) and lower socioeconomic status are associated with lower initial shockable rhythm rates, fewer patients from these groups might be eligible for ECPR with such criteria.53 The low rates of bystander‐initiated CPR and automated external defibrillator use also remain unaddressed in these communities.54, 55 Thus, the minimum criteria for ECPR implementation is health equity, such that the incidence rate of VF and survival and functional survival after OHCA are comparable to the control arms in the trials, through an optimized system of care for OHCA resuscitation. Models potentially addressing this issue are discussed below.
    Validation of the highly successful ARREST trial17 methodology across diverse neighborhoods and regions may be challenging due to in‐hospital inequities. The 3‐member study team performing ECPR were highly trained and skilled with significant clinical ECPR experience for years. A successful community program for cardiac arrest treatment was organized and functional for nearly a decade before ARREST began. The physician operators cannulated for ECPR in the catheterization laboratory, performed any subsequent coronary interventions, and followed each enrolled patient to discharge or death with institution‐supported aggressive postresuscitation multidisciplinary intensive care using strict formalized guidelines for withdrawal of life‐sustaining therapy. These physicians were supported institutionally for their roles and time commitment within the ECPR program. The ARREST27 and Prague OHCA29 investigators concluded that the survival benefit is derived from the successful implementation of ECPR as a system of care for OHCA resuscitation, not just cannulating and initiating ECPR.
    Adding ECPR to a system of care for OHCA resuscitation (Figure 2) has the potential to both amplify current inequities or create new ones. ECPR‐specific potential causes of inequity span from factors related to the patient, prehospital, hospital/provider, and posthospital/rehabilitation, as well as variability in selecting and reporting outcomes (Figure 1). Patient factors: Patient characteristics have been studied in OHCA, but given the limited patient populations who have received ECPR, it is not clear what impact these factors will have on receipt of the intervention.Prehospital factors: Prehospital providers must recognize potential ECPR candidates in the field and then transport them to a hospital providing ECPR. There is potential for selection and implicit bias as to who is eligible for ECPR based on prehospital arrest factors (eg, bystander CPR, automated external defibrillator accessibility and use, presenting rhythm, quality of prehospital resuscitation care, and number of providers responding to each call).Hospital factors: Selection and implicit bias can also occur in the hospital setting, impacting ECPR patient eligibility. Most hospitals do not currently have the equipment or trained providers to implement an ECPR program for OHCA patients, and significant resources are required to initiate such a program. As a result, ECPR availability is limited to high‐resource settings. There are also potential inequities in the prognosis of futility and the decision to limit life‐sustaining therapy after ECPR initiation.56 Addressing hospital factors is important as ELSO registry data suggest that the quality of in‐hospital care postcannulation for ECPR is associated with survival.57, 58, 59Posthospital/rehabilitation factors: Focused, intensive rehabilitation care is integral to improving long‐term outcomes for patients with OHCA.60 There are known inequities in accessing rehabilitation services (inpatient and outpatient), which may amplify post‐ECPR outcome variability.61, 62Variation in reported outcomes: Standardized patient‐reported outcomes at hospital discharge and 3‐, 6‐, and 12‐month follow‐up are not uniformly captured and reported in the United States,63, 64 making it more difficult to measure and address health inequities.
    Patient factors: Patient characteristics have been studied in OHCA, but given the limited patient populations who have received ECPR, it is not clear what impact these factors will have on receipt of the intervention.
    Prehospital factors: Prehospital providers must recognize potential ECPR candidates in the field and then transport them to a hospital providing ECPR. There is potential for selection and implicit bias as to who is eligible for ECPR based on prehospital arrest factors (eg, bystander CPR, automated external defibrillator accessibility and use, presenting rhythm, quality of prehospital resuscitation care, and number of providers responding to each call).
    Hospital factors: Selection and implicit bias can also occur in the hospital setting, impacting ECPR patient eligibility. Most hospitals do not currently have the equipment or trained providers to implement an ECPR program for OHCA patients, and significant resources are required to initiate such a program. As a result, ECPR availability is limited to high‐resource settings. There are also potential inequities in the prognosis of futility and the decision to limit life‐sustaining therapy after ECPR initiation.56 Addressing hospital factors is important as ELSO registry data suggest that the quality of in‐hospital care postcannulation for ECPR is associated with survival.57, 58, 59
    Posthospital/rehabilitation factors: Focused, intensive rehabilitation care is integral to improving long‐term outcomes for patients with OHCA.60 There are known inequities in accessing rehabilitation services (inpatient and outpatient), which may amplify post‐ECPR outcome variability.61, 62
    Variation in reported outcomes: Standardized patient‐reported outcomes at hospital discharge and 3‐, 6‐, and 12‐month follow‐up are not uniformly captured and reported in the United States,63, 64 making it more difficult to measure and address health inequities.
    Provided with permission from collaborator, Dr Sarah Perlman's presentation to the National Heart, Lung, and Blood Institute of the US National Institutes of Health convened a virtual workshop on August 23 and 24, 2021; URL:https://www.nhlbi.nih.gov/events/2021/extracorporeal‐cpr‐ecpr‐out‐hospital‐cardiac‐arrest. bCPR indicates; CABG, coronary artery bypass graft; ECMO, extracorporeal membrane oxygenation; ECPR, extracorporeal cardiopulmonary resuscitation; bCPR, bystander CPR, EMS, emergency medical services; ICD, implantable cardioverter‐defibrillator; SES, socioeconomic status; and WLST, withdrawal of life‐sustaining therapy.

    MODELS OF EQUITABLE ACCESS IN THE UNITED STATES

    When considering ECPR implementation strategies to promote equity, much can be learned from other acute emergent conditions, which require a continuum of care. For example, regionalization of ST‐segment–elevation myocardial infarction (STEMI) care increased certain sex and racial disparities,65 but implementation of comprehensive STEMI protocols by standardizing eligibility, selection, and transport criteria have reduced these differences.66 Many interventions have been proposed to address existing disparities along the continuum of stroke care. These included community outreach and addressing underlying health risk disparities. Yet disparities in stroke care continue to persist. It is apparent that identifying successful implementation strategies to address and reduce inequities has proven challenging and will require addressing many research gaps to design effective interventions.67

    HEALTH CARE DELIVERY WORKFORCE REQUIREMENTS

    The workshop also focused on extending the reach of ECPR systems of care for OHCA resuscitation from large academic centers to community, rural, and remote sites to improve equitable access to this advanced therapy. Whether the success of the ARREST trial can be replicated in other communities remains to be proven. It is recognized that ECPR‐facilitated resuscitation, which was used in the ARREST trial, is a complex, resource‐intensive system of care.
    To be efficacious, the health care system structure must be built upon a well‐functioning and effective comprehensive cardiac arrest care system alongside excellent EMS dispatch recognition of cardiac arrest with CPR instructions, high‐quality EMS care, and evidence‐based hospital care. Observational data support a volume/outcome relationship for best results with both conventional resuscitation of OHCA68 and ECMO for all causes.69 Thus, the best community outcomes are likely to occur when experience is concentrated in hub hospitals that draw from a catchment area with a sufficient case count to create continuous expertise and skill maintenance. Referral models from low‐resource settings (spokes) to hubs have been developed for trauma care and may offer insight. Similar to prevailing dogma in trauma systems, the use of ECPR in a system of care for OHCA resuscitation appears to possess an optimal implementation time interval. Observational data support the concept that once 60 minutes has elapsed between collapse and ECPR support initiation, functional survival declines rapidly,70 and the best outcomes were associated with a <30‐minute time interval.71 This limit on survivability from refractory OHCA reduces the geographic reach of centralizing ECPR at hubs.68 However, there may be alternative models for patient selection, cannulation location, and technologies that can help overcome this time barrier. The hub‐and‐spoke model may be most appropriate for ECPR, and this implementation strategy has worked well in supporting hierarchical care at trauma, stroke, acute myocardial infarction, and cardiac arrest regional centers. For other high‐risk medical conditions, hospital mortality decreases as hospital volume increases to >100 cases per year.72 Post–cardiac arrest treatment has better outcomes at experienced hubs.69, 73 For ECPR, similar relationships are harder to demonstrate, presumably because more complex and higher‐risk cases are more likely to be transferred to high‐volume, more experienced centers. This practice pattern might create paradoxical lower survival, longer stay lengths, and higher costs at more experienced sites, as was described for the treatment of severe cardiopulmonary failure with ECMO, raising a question of potential selection bias, which deserves further investigation.74 For sites providing ECMO, better outcomes are associated with volumes of at least 12 per year per site.69 Accurate program evaluations and system comparisons require thoughtful, risk‐adjusted modeling.
    Future payment models and ECPR quality assessment must also be careful not to penalize centers for rendering care to more dire cases. Concern about public reporting of catheterization laboratories' mortality rates without adequate consideration of the population treated has been raised in the past when such reporting has not acknowledged the expected differences in postangiographic mortality rates between patients with STEMI with and without associated OHCA. These expected rates differ by 10‐fold (30%–50% mortality with cardiac arrest and 3%–5% mortality for those without cardiac arrest). The expected mortality rate is even higher in those with STEMI associated with refractory OHCA such as the population currently being treated with ECPR. This fact may impede some medical centers from adopting this emerging treatment modality in that such adoption could result in a public perception of higher mortality and inferior care at that institution, when in fact it may be providing superior care to this high‐risk patient population.75, 76, 77, 78, 79
    For other time‐critical, complex illnesses, transfer to specialty hubs can be direct from EMS or indirect via stabilization at a local hospital followed by transfer to a hub. These models are constrained by geography, as hubs tend to be urban and may be some distance from rural patients or may be inequitably distributed within regions.80, 81 Similar geographic disparities have been noted for cardiac emergencies, including cardiac arrest.24 Mission Lifeline is the AHA national initiative to advance the system of care for patients with acute, high‐risk time‐sensitive life and/or quality of life‐threatening disease states, such as STEMI, non‐STEMI, stroke, and OHCA within a defined region. Mission Lifeline is focused on supporting hospitals, EMS agencies, regions, and communities while forming and enhancing collaborative, efficient, and effective systems of care82 (Figure 3). The number of experienced ECPR centers has increased each year since 2010, accelerated by the pandemic, which may help reduce this geographic disparity in access to potential ECPR hubs. Future workforce development will need to choose between training a large number of providers at many hospitals to initiate ECPR versus intensively training a smaller number of providers at centralized hubs, which receive more cases. The Minnesota model has yielded amazing results.17 Training was set with a single cardiologist doing >100 cases and then supervising a second cardiologist doing >100 cases. This volume of mentorship and experience is most likely the model that is required for successful implementation of this challenging intervention. Institutions with lower volumes may have to develop a shared experience with high‐volume sites to achieve similar mentorship and high‐quality outcomes.
    All system data, including coverage area, are self‐reported data. Adapted from Mission: Lifeline Program,89 copyright © 2017, American Heart Association, Inc. The Creative Commons license does not apply to this content. Use of the material in any format is prohibited without written permission from the American Heart Association, Inc. STEMI indicates ST‐segment–elevation myocardial infarction.
    Insights were shared on the importance of team training for high‐acuity/low‐frequency events, using pediatric cardiac arrest care as an example. Pediatric cardiac arrests that occur both inside and outside of the hospital setting present unique challenges to the health care system and serve as a good exemplar of the need for and strategies to address training for low‐frequency events. There is additional cognitive load associated with the manner in which resuscitation guidelines, equipment sizes, and medication dosing varies on the basis of the age and weight of a child.83 A prospective, multicenter, observational study at 50 EDs observed interprofessional teams managing standardized, simulated cardiac arrest cases for a 5‐year‐old mannequin.84 The basic life support adherence scores were lowest for the lowest‐volume EDs, while EDs embedded in teaching hospitals with trauma designation and only serving children had the highest overall adherence to AHA pediatric resuscitation guidelines. Yet overall adherence to resuscitation guidelines was poor for all hospital groups. Walsh et al describe a model in which academic medical centers partner with community hospitals to provide in situ simulation training.85 Data support that simulation‐based training can improve subsequent performance in both simulated and real pediatric trauma events.86, 87 An additional challenge presented to EMS teams managing pediatric OHCA relative to an adult event, is that there is no US Food and Drug Administration–approved mechanical CPR device for children aged <18 years. If a child has a cardiac arrest and needs transport to a community ED and/or from a community hospital to a tertiary care children's center, there has been concern about whether high‐quality CPR can be delivered. Noje et al measured resuscitation performance of pediatric transport teams and confirmed that with training and commitment it is possible to deliver high‐quality CPR during transport, even in the absence of a mechanical CPR device.88
    The application of the hub‐and‐spoke model for pediatric cardiac arrest varies, depending on many factors. A child in cardiac arrest may be picked up in the field, and an adult‐sized adolescent can be placed on a mechanical CPR device and taken via flight or ambulance directly to a tertiary care children's hospital. Depending on the features (ie, witnessed arrest, immediate bystander CPR, short transport time), it is possible the hospital will consider activating their ECPR system and cannulate the child onto ECPR. Another situation would be a child who is critically ill and taken to a community hospital where they have a cardiac arrest witnessed by the health care team with immediate CPR. In that case, there are currently children's hospitals that do a rapid dispatch of their transport team, pick up the child, and transport them back to the children's hospital. The ECPR team will be prepared to attempt to put the child on ECPR if the scenario meets their existing protocols (ie, if elapsed time permits the possibility of getting a child on ECPR within 60 minutes). A third model currently existing is for a children's hospital to send a team from their hub to the community hospital (ie, spoke), including a surgeon and perfusionist, who can cannulate the child onto ECPR at the outside hospital. A final model is for the community hospital to cannulate the child for ECPR, and the transport team will leave the children's hospital to pick up the child at the outside hospital and transport them on ECPR. All these scenarios require frequent and regular interdisciplinary, interinstitutional training to even consider any of these options.
    Research Priorities
    ECPR indicates extracorporeal cardiopulmonary resuscitation; ELSO, Extracorporeal Life Support Organization; IHCA, in‐hospital cardiac arrest; and OHCA, out‐of‐hospital cardiac arrest.
    Gap: There are fundamental inequities in access to high‐quality resuscitation systems of care across US neighborhoods and regions.Gap: Defined performance metrics are needed to successfully address access inequities, optimized treatment (EMS, ED, intensive care unit, cardiology, neurology) and overall outcomes after OHCA in each community (metropolitan, urban, rural, and remote). It is important to note that not every region will achieve the same performance. Continuous reporting and optimizing the metrics in a community may lead to public health or system of care interventions that are unique to the community and result in a better survival.Gap: There is a need to develop community‐based implementation strategies to optimize out‐of‐hospital systems of care with performance metrics to ensure equity in the quality of resuscitation, survival outcomes, and access to ECPR where appropriate. These implementation strategies should be guided by implementation science theories and frameworks and include qualitative interviews to better understand the attitudes and beliefs pertaining to ECPR in diverse communities before implementation. Community strategies are also needed for reporting and optimizing CPR metrics, Utstein variables, and survival outcomes after cardiac arrest.
    Gap: There are fundamental inequities in access to high‐quality resuscitation systems of care across US neighborhoods and regions.
    Gap: Defined performance metrics are needed to successfully address access inequities, optimized treatment (EMS, ED, intensive care unit, cardiology, neurology) and overall outcomes after OHCA in each community (metropolitan, urban, rural, and remote). It is important to note that not every region will achieve the same performance. Continuous reporting and optimizing the metrics in a community may lead to public health or system of care interventions that are unique to the community and result in a better survival.
    Gap: There is a need to develop community‐based implementation strategies to optimize out‐of‐hospital systems of care with performance metrics to ensure equity in the quality of resuscitation, survival outcomes, and access to ECPR where appropriate. These implementation strategies should be guided by implementation science theories and frameworks and include qualitative interviews to better understand the attitudes and beliefs pertaining to ECPR in diverse communities before implementation. Community strategies are also needed for reporting and optimizing CPR metrics, Utstein variables, and survival outcomes after cardiac arrest.
    Research opportunity (RO) Rank 1: Establish core performance and equity metrics and ensure the reporting of these metrics by sex, gender, and race across all regions (rural, urban, out‐of‐hospital, low resource versus high resource) in all existing US cardiac arrest registries (OHCA and IHCA). 4Gap: Data infrastructure with ECPR data points and common data elements for OHCA surveillance and research is lacking. Building on existing OHCA and IHCA registries and contributing to the international ELSO registry and collaborating with the ELSO investigators will enable US cardiac arrest surveillance and research studies and facilitates achieving the sample size required to answer important research questions early in implementation.
    Gap: Data infrastructure with ECPR data points and common data elements for OHCA surveillance and research is lacking. Building on existing OHCA and IHCA registries and contributing to the international ELSO registry and collaborating with the ELSO investigators will enable US cardiac arrest surveillance and research studies and facilitates achieving the sample size required to answer important research questions early in implementation.
    RO rank 2: Expand existing US cardiac arrest registries (OHCA and IHCA) to identify all ECPR cases, as well as collect and report the ELSO data variables and quality‐of‐life outcomes after discharge.
    RO rank 3: Share ELSO data variables from these 2 expanded US registries with ELSO through a partnership agreement to enhance the North American contribution to this international registry. 5Gap: There is a lack of reliable data on the number of people experiencing cardiac arrest who may benefit from ECPR‐facilitated resuscitation.6Gap: Defined resuscitation techniques and diagnostic tools are needed to increase the number of identified patients who may benefit from ECPR.
    Gap: There is a lack of reliable data on the number of people experiencing cardiac arrest who may benefit from ECPR‐facilitated resuscitation.
    Gap: Defined resuscitation techniques and diagnostic tools are needed to increase the number of identified patients who may benefit from ECPR.
    RO rank 4: Evaluate and refine the recommended consensus population of patients with OHCA (Table 2) most likely to benefit.
    RO rank 5: Identify and validate innovative tools beyond Utstein characteristics for identifying patients with OHCA without severe neurological injury who may be excellent ECPR candidates.
    RO rank 6: Develop and test innovative ways to measure CPR‐driven perfusion to adjust CPR technique and minimize or avoid neurological injury during CPR before ECPR initiation. 7Gap: Well‐designed clinical efficacy trials in a variety of communities with optimized basic life support and advanced cardiac life support systems to evaluate ECPR‐facilitated systems of care are needed. Whether this should include a large multicenter trial was controversial, with some in favor believing more patient data were desirable, and others suggesting a multicenter trial was unnecessary given the results of the 3 published randomized trials.20, 32, 33
    Gap: Well‐designed clinical efficacy trials in a variety of communities with optimized basic life support and advanced cardiac life support systems to evaluate ECPR‐facilitated systems of care are needed. Whether this should include a large multicenter trial was controversial, with some in favor believing more patient data were desirable, and others suggesting a multicenter trial was unnecessary given the results of the 3 published randomized trials.20, 32, 33
    RO rank 7: Establish core performance metrics, time intervals, and treatment guidelines for adding ECPR to a system of care for OHCA and IHCA resuscitation.
    RO rank 8: Conduct well‐designed efficacy trials across multiple institutions and systems of care where the control arm of OHCA is established and can maintain a regionally optimized survival rate for this community during the trial. Outcomes should include long‐term quality of life, patient‐reported outcomes, cost effectiveness measures, and equity metrics.
    RO rank 9: Evaluate existing international training programs and strategies used in published trials of ECPR to test readiness for more generalizable implementation in additional efficacy trials. 8Gap: Basic and clinical research are needed to examine the development and progression of new ECPR‐related discoveries and address bench‐to‐bedside implementation challenges.9Gap: Optimal hardware and software configurations are lacking for prolonged ECPR use when applied early (ie, initially in the prehospital setting).10Gap: A knowledge gap on the circulatory and hemostatic physiology and ECPR biology/physics during prolonged CPR exists.
    Gap: Basic and clinical research are needed to examine the development and progression of new ECPR‐related discoveries and address bench‐to‐bedside implementation challenges.
    Gap: Optimal hardware and software configurations are lacking for prolonged ECPR use when applied early (ie, initially in the prehospital setting).
    Gap: A knowledge gap on the circulatory and hemostatic physiology and ECPR biology/physics during prolonged CPR exists.
    RO rank 10: Address challenges in ECPR technology and implementation through basic, clinical, simulation, and human factors research, including deriving and investigating technological innovations to facilitate first‐pass success for percutaneous femoral vessel cannulation.

    CONCLUSIONS

    Based on small studies, ECPR offers promising outcomes; however, optimal patient characteristics and operating logistics have not been determined. The international expert panel from the NHLBI workshop suggests there are significant inequities in OHCA care that need to be addressed such that the outcomes are optimized for each target region prior to implementing ECPR in a system of care for OHCA resuscitation. The body of science pertaining to ECPR would benefit from the United States contributing uniform data to existing registries and sharing data with the international ELSO registry, to optimize the sample size for observational research. Additional well‐designed efficacy trials recruiting across different regions of care evaluating long‐term follow‐up, including patient‐reported outcomes, cost effectiveness measures, and equity metrics would contribute significantly to the body of science. The workshop participants defined the population of patients with OHCA most likely to benefit from ECPR. ECPR eligibility. This includes patients aged 18 to 75 years who are functioning independently without comorbidities (ie, chronic obstructive pulmonary disease, heart failure, dialysis, liver failure, or active malignancy) before suffering a witnessed OHCA without any obvious cause of the cardiac arrest; presenting in a shockable rhythm and without return of spontaneous circulation after 3 advanced life support rounds; and transported with mechanical CPR to an ECPR‐capable setting within 30 minutes. ECPR implementation would benefit from basic, clinical, and implementation (preclinical translational, simulation, human factors) research to address the technological challenges unique to implementation of ECPR within a system of care for OHCA resuscitation.

    Sources of Funding

    This work was supported by the National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD.

    Disclosures

    Dr Hunt is a member of Zoll Inc. speaker's bureau. Thomas Aufderheide has research grant support from the NHLBI, National Institute of Neurological Disorders and Stroke, research equipment support from Zoll Inc., and is a consultant to Medtronic. Dr Callaway has research grant support from the NIH and the National Aeronautics and Space Administration. Dr Tonna has research grant support from the NHLBI and serves as the Chairman of the ELSO registry. Dr Blewer has research grant support from the NIH, NHLBI, AHA, and the Laerdal Foundation. Dr McNally is Executive Director of the CARES program, as well as research support from the CDC CARES Expansion and Modernization Grant. Dr Yannopoulos has research grant funding from the NIH. Dr Belohlavek is a consultant for Getinge and Abiomed, and serves on a speaking bureau for Resuscitec. Dr Youngquist has research grant support from Department of Defense, Zoll Inc., and serves as a consultant to CoLabs, and on the speaker bureau for Zoll Inc. Dr Kern is a consultant to LivaNova and Stryker Inc. The remaining authors have no disclosures to report.

    Supporting information

    Table S1
    CS

    Dr. Comilla Sasson

    Emergency Medicine · General Emergency Medicine
    University of Colorado Anschutz
    comilla