Multiorganizational and multidisciplinary high-fidelity simulation of extracorporeal cardiopulmonary resuscitation for out-of-hospital cardiac arrest
Article information
Abstract
Objective
Extracorporeal cardiopulmonary resuscitation (ECPR) involves the initiation of venoarterial extracorporeal membrane oxygenation (ECMO) in patients in cardiac arrest. After implementation of a city-wide ECPR initiative, we brought together emergency medical services (EMS) personnel and two hospital-based teams (emergency medicine and ECMO) for a high-fidelity in situ ECPR simulation with a focus on identifying systems challenges and opportunities for improvement.
Methods
We performed a high-fidelity in situ simulation of cardiac arrest requiring ECPR including ambulance dispatch, on-scene resuscitation, transportation to receiving hospital and cannulation for ECMO. Timing metrics were quantified. Focused debriefing and thematic analysis were conducted to identify areas for system improvement.
Results
EMS arrived on scene within 3 minutes of dispatch with defibrillation provided 3 minutes later. Initial ECPR activation occurred 12 minutes after arrest with arrival to the emergency department 29 minutes after arrest. Successful cannulation to ECMO occurred 51 minutes after arrest. Thematic areas for systems improvement included communication, equipment, coordination of care and physical plant, and patient selection for ECPR.
Conclusion
In this novel multi-site, multiorganizational, and multidisciplinary high-fidelity ECPR simulation, we successfully simulated an ECPR arrest from dispatch to cannulation with a focus on systems factors and latent patient safety issues. Arrival to the emergency department and cannulation for ECPR were consistent with previously published studies of out-of-hospital ECPR.
INTRODUCTION
Extracorporeal cardiopulmonary resuscitation (ECPR) involves the initiation of venoarterial extracorporeal membrane oxygenation (ECMO) in patients in cardiac arrest [1]. ECPR provides greater perfusion during cardiac arrest resuscitation than conventional cardiopulmonary resuscitation (CPR) [2,3], and has been shown to increase the rates of return of spontaneous circulation [4], successful defibrillation [5], and potentially survival to hospital discharge in selected patient populations [6].
Despite its benefits, ECPR is an invasive, resource-intensive therapy that requires specialist involvement, and appropriate patient selection is critical [7,8]. Additionally, ECPR is a time-sensitive treatment modality [9] requiring close collaboration between prehospital teams, emergency department (ED) care, and ECMO teams. In the United States, most emergency medical services (EMS) are provided by emergency medical technicians and paramedics [10]. Although prehospital physician-initiated ECPR has been explored in the United States [11], most ECPR occurs in the hospital setting with EMS transportation to an ECMO-capable receiving ED [12].
In situ simulations are a simulation-based education tool that facilitates quality improvement, helps improve patient safety, and delivers medical simulation in the clinical setting, rather than at a dedicated training site [13]. Team-based and interdisciplinary simulation-based training has been shown to improve teamwork [14] and provide opportunities for disciplines to train together on high-risk, low-frequency events [15] without risk of patient harm. After implementation of a city-wide ECPR initiative, we brought together EMS personnel and two hospital-based teams (emergency medicine [EM] and ECMO) for a high-fidelity in situ ECPR simulation with a focus on identifying systems challenges and opportunities for improvement.
METHODS
Site
Boston EMS, a bureau of the Boston Public Health Commission, is the City of Boston’s municipal prehospital provider, one of three public safety services that respond to 9-1-1 calls. Serving a residential population of approximately 675,000, which increases to 1.2 million due to commuters, students, and tourists, Boston EMS responded to more than 138,000 calls for service in 2023. For cardiac arrest call types, there is a dual response by both basic life support (BLS) and advanced life support (ALS) ambulances, with pre-arrival instructions provided by Boston EMS call-takers regarding layperson initiation of CPR. Cardiac arrest calls also trigger a dual response by Boston Fire, a first responder agency. In February 2024, in coordination with a consortium of ECMO teams at medical centers throughout Boston, Boston EMS implemented a new ECPR protocol for select patients in refractory shockable cardiac arrest.
Brigham and Women’s Hospital (BWH) is a large academic medical center with a level 1 trauma designation, an active ECMO and cardiothoracic surgery program, and academic training programs in EM and cardiothoracic surgery. BWH has had an active ECMO program since 2013. The ECMO service comprises thoracic surgeons, cardiac surgeons, intensivists, and respiratory therapists who have specialized training in ECMO. The BWH ECMO Service performs between 70 and 100 ECMO runs per year.
Preparation
A core planning team of experts in patient safety (PSJ), medical simulation (AJE), and ECPR/ECMO (RRS) was assembled to lead the preparations. Three smaller working groups were also convened, with one focused on prehospital care, one focused on hospital operations, and one focused on the ECMO team.
The prehospital operations teams included the core planning team and a medical director, field supervisor, and training officer for Boston EMS. We arranged two dedicated ambulances (one BLS and one ALS) to be available for the duration of the simulation, to ensure that city-wide resources were not taken out of service. The medical director and field supervisor were also on-site for the simulation to aid with dispatching and briefing/debriefing of the EMS providers.
The hospital operations team included the core planning team, the ED medical director (DEB), and the ED nursing director. This team worked to ensure availability of the resuscitation room and ED nursing educators, to avoid conflicts with other scheduled training, and to assist with communications on the day of the simulation. The simulation was intentionally scheduled during a weekday morning at a time when the ED census was expected to be lower than usual and when typical staffing patterns were expected (e.g., not on an educational conference day for the residency).
The ECMO team included the hospital ECMO medical director (RRS) and the ECMO program manager. The purpose of this team was to ensure adequate testing, delivery, and function of the cannulation simulator to the ED, as well as availability of the training ECMO pump (Getinge).
Simulation
Our goal in the simulation was to test, with as high a degree of fidelity as possible, the ECPR activation in a multidisciplinary and multiorganizational fashion to identify latent systems issues and opportunities for improvement. A Megacode Kelly manikin (Laerdal Medical,) was utilized, which allowed for endotracheal intubation, defibrillation, intravenous and intraosseous medication administration, and application of a mechanical chest compression device (LUCAS, Stryker Medical), to ensure adequate fidelity for prehospital providers.
The simulation manikin was in an off-site administrative building two blocks from the main hospital, which allowed hospital simulation resources to be deployed at a location that would enable dispatch of the EMS crew to a realistic site that would not otherwise be served by hospital emergency response systems. A simulation specialist was utilized as an embedded participant confederate to perform bystander CPR, operate the manikin, and provide relevant history to the prehospital team.
The simulation scenario was designed to meet all criteria for activation of ECPR (Table 1). Briefly, the simulated patient was a 55-year-old man with no known medical history, medications, or allergies, who experienced a witnessed cardiac arrest at work with immediate bystander CPR. The manikin was programmed to stay in refractory ventricular fibrillation for the entire scenario. Pulse oximetry and capnography were verbally conveyed by the simulation specialist if requested.
Extracorporeal cardiopulmonary resuscitation criteria
The EMS providers staged one block away from the hospital and were dispatched directly to the scene by the on-site field supervisor. EMS was aware of the ECPR-based scenario and encouraged to follow all relevant standing medical control orders and protocols [16]. By design, the BLS crew arrived on scene first. They were provided with medical history by the embedded participant, implemented BLS interventions for out-of-hospital cardiac arrest (OHCA), and updated the ALS providers. ALS arrived shortly thereafter and assumed control of the resuscitation, following all ALS standing medical control orders and procedures for OHCA.
In Boston, the activation of ECPR at participating institutions is a two-step process. Once individuals are identified as meeting ECPR criteria, an initial “ECPR alert” is transmitted to the hospital by Boston EMS without any further patient information. The receiving hospital will send a group page and activate the ECMO team to assemble in the ED to prepare for cannulation. A subsequent “entry note” is provided by the regional Centralized Emergency Medical Dispatch (CMED) communications center and provides more detailed medical information about the patient, interventions performed thus far, and an estimated time of arrival to the receiving facility.
In our simulation, after recognition that the patient met all criteria for ECPR, the Boston EMS crew provided an initial “ECPR alert” to the hospital. This was the initial activation of the hospital-based team, which was not aware of the simulated nature of the scenario. Upon receiving the ECPR alert, the ED team activated the ECMO team according to the hospital protocol. Boston EMS continued on-scene interventions per protocol; once in the transporting ambulance, the secondary “entry note” was provided via CMED. Transportation to the receiving hospital was via ambulance; the use of lights and sirens was at the EMS team's discretion.
On arrival to the receiving ED, Boston EMS provided verbal hand-off to the ED team, which included an ED attending and resident physician, nurses, emergency service assistants, respiratory therapists, as well as any members of the ECMO team that had arrived. After the manikin was transferred to the ED stretcher, the ECMO cannulation inserts were attached to the manikin by an embedded participant (RRS) while resuscitative efforts continued. The ECMO cannulation inserts consisted of medical gel with silicone tubing placed through the gel inserts connected to a fluid bag to simulate the femoral vessels, modeled after Pang et al. [17]. As described above, the manikin was programmed to stay in a persistent ventricular fibrillation despite defibrillation attempts. The scenario continued until cannulation was achieved and the simulated patient was placed on full ECMO support. At that point, the simulation was stopped.
Debriefing and gathering of feedback
Focused debriefing was conducted as a large group, then with each individual group (ECMO team, ED team, and prehospital team) separately. Improvement suggestions were gathered by systematic role group questioning. Thematic analysis was performed on all suggestions. All participants received a survey to assess opinions about the program's effectiveness and to provide further targeted systems feedback.
RESULTS
Simulation and metrics
Upon scene arrival, the BLS crew obtained the relevant medical history, assumed control of the scene, and started CPR according to BLS protocols. An automated external defibrillator (AED) was applied, and a defibrillation was delivered. The time from dispatch to scene arrival for BLS was 3 minutes, and time to initial defibrillation via AED from BLS arrival was 3 minutes. The full timing is shown in Table 2.
Clinical metrics
ALS arrived on scene within 3 minutes of BLS and attempted subsequent defibrillation via the manual defibrillator 2 minutes after arrival on the scene. Endotracheal intubation was performed using direct laryngoscopy (Fig. 1). Amiodarone and epinephrine were administered per advanced cardiac life support protocols via intraosseous access. The decision between intravenous or intraosseous access is at the discretion of the paramedic in the Massachusetts EMS Statewide Protocols [16]. A LUCAS device was applied to perform automated chest compressions. Subsequent defibrillations were continued per protocol.
Emergency medical services personnel on scene, performing endotracheal intubation during mechanical chest compressions.
Prehospital notification of the potential ECMO-eligible patient was sent 7 minutes after EMS arrival (10 minutes after arrest) and received by the hospital 2 minutes later. Extrication and full prehospital radio report occurred 17 minutes after EMS arrival to the scene and 20 minutes post-arrest. Transport time to the hospital was 5 minutes and the patient was roomed 3 minutes after the ambulance's arrival to the ambulance bay.
In the ED, initial pulse/rhythm check and defibrillation occurred 3 and 4 minutes after arrival (respectively). Subsequent defibrillations occurred approximately every 4 minutes until the manikin was placed on ECMO. The initial groin puncture occurred 5 minutes after rooming (34 minutes post-arrest), with the second groin puncture 7 minutes subsequently (Fig. 2). The manikin was placed on full ECMO support 22 minutes after rooming in the ED, 51 minutes post-arrest.
Cannulation for extracorporeal cardiopulmonary resuscitation while chest compressions and resuscitation are ongoing.
Feedback and themes
Structured debriefing occurred as a group and then with individual teams. Several thematic elements emerged through the structured debriefing.
Communication
The receiving ED team was appreciative of the initial prehospital “ECPR alert” provided before hospital arrival as they were able to activate the ECMO team and gather critical resources before the patient arrived. As with the dynamic assembly of new teams in a high-stakes scenario, not all participants knew all members of the teams, particularly between the ECMO and ED teams. Formalized introductions and physical team member sticker labels, like those used during our institution’s trauma resuscitations, were recommended for improvement. Other identified areas for improvement included closed-loop communication and situational awareness. Additionally, when the ECMO team required supplies not typically stocked in the resuscitation room, they were unsure whom to ask for assistance. Crowd control and resource management were also raised as opportunities for improvement, particularly given the large composition of all teams that were assembled.
Although the ECMO team assembled quickly and arrived at the ED bay before the simulated patient arrival, they noted that dynamic team changes could delay the arrival. For example, the ECMO fellow may be called to participate in an organ transplantation and would not be immediately available; redundancies in the backup scheduling were recommended to ensure immediate ECMO availability. In addition, the ED pharmacist was not aware of the resuscitation. Adding members to the ECMO group page was suggested as a possible solution to ensure notification of all relevant team members.
Equipment
Several equipment issues were identified. Prehospital, an initial AED failed to deliver a defibrillation and was taken out of service for repairs. The LUCAS device did not deliver compressions initially and required repositioning, after which it functioned appropriately; in addition, the duration of the resuscitation necessitated a battery change for the LUCAS at the end of the scenario. As noted above, the ECMO team required additional supplies during the cannulation attempt, such as ultrasound probe covers, and needed to request this from the ED team. The ECMO team requested that a designated individual from the ED team be assigned as a runner to obtain needed supplies. A checklist of required ECMO supplies was suggested, as was a dedicated storage location in the designated resuscitation areas.
Coordination of equipment between the teams was another theme. Boston EMS noted that the initial LUCAS device was applied by their team and used for prehospital care and transport. To avoid interruptions to chest compressions, they suggested that the prehospital LUCAS remain on the patient until after cannulation while the EMS crew cleaned the ambulance and prepared to go back into service; in the case of a prolonged resuscitation or need to redeploy to the field, they suggested a discussion between the EMS and ED teams about the prehospital exchange for ED LUCAS. The ED teams and ECMO teams also both required ultrasonography during the resuscitation. Bringing a second ultrasound into the room to serve each team was also suggested as a possible solution.
Physical organization and human factors issues were raised as possible areas for improvement. For example, the EMS team requested that the receiving hospital stretcher be located on one side of the room to facilitate placement of the prehospital stretcher and patient transfer. After removal of the stretcher, the ECMO team requested a dedicated area at the foot of the bed to allow for set up and access of ECMO supplies near the cannulating physician and the nursing/respiratory therapist who was designated as the first assistant.
Candidacy for ECPR
Appropriate patient selection, candidacy, and timing of ECPR were raised as a major theme. Boston EMS reported reviewing the major criteria for ECPR before providing the prehospital alert, but noted that limited bystander information may be incomplete. They also noted a tension between the longstanding practice of Boston EMS of on-scene management of cardiac arrest and the shift for prompt transport to meet the predefined goal of arrest-to-hospital arrival of less than 30 minutes for ECPR cases. Finally, the hospital teams observed that there was no systematic and formalized review of ECPR criteria before cannulation was performed in the resuscitation bay; a checklist of all inclusion and exclusion criteria placed with the ECMO supplies in the resuscitation bay was identified as a possible solution.
DISCUSSION
In this study, we report the results of a multidisciplinary, multi-site, and multiorganizational trial of in situ simulation for ECPR with a focus on quality and system improvement. By coordinating between multiple agencies and hospital-based departments and service lines, we identified several latent systems issues to target for improvement.
Some of the identified systems issues, such as room layout and equipment stocking are applicable primarily to our institution. Closed-loop communication and crowd control may depend on the specific team assembled, their experience with each other, and continued iteration and development of the ECPR program. Other issues, such as equipment coordination between EMS and the ED (e.g., with the LUCAS device) are more broadly applicable to the prehospital and emergency community.
One major issue raised was the time from arrest to ECPR cannulation. Over time, many EMS agencies have moved from a philosophy of rapid transportation to the ED for resuscitation (colloquially called “scoop and run”) to a philosophy of on-scene management with focus on defibrillation, airway management, high-quality CPR, and treatment of reversible causes (“stay and play”). This philosophy shift recognized that many important aspects of resuscitation for cardiac arrest can be effectively delivered on scene. However, with the advent of time-sensitive interventions that are performed primarily in the hospital (e.g., ECPR), the focus of on-scene management may need to shift back to rapid transportation to a receiving ECPR-capable ED for patients who are candidates for ECPR.
The goal of arriving at the receiving facility in less than 30 minutes from arrest may be difficult to achieve, even with the limitation of on-scene efforts. For example, in our scenario, arrival after dispatch was only 3 minutes and transport to the hospital was 5 minutes, yet the time from arrest to rooming in the ED was 29 minutes post-arrest. This may in part have been exacerbated by equipment challenges (e.g., initial AED failure, LUCAS device requiring repositioning); however, these are realistic challenges that can be encountered in the field. While our time from arrest to initiation of ECMO was consistent with those seen in previously published studies of ECPR programs [12], this may have been artificially lowered by the close location of the simulated arrest to the ED. Additionally, the EMS team was aware of the nature of the ECPR candidacy, which may have further artificially lowered the transport time from that in a real patient scenario. Further quality improvement efforts should be targeted at improving dispatch to ED arrival times. Despite multiple factors biasing this scenario toward a short transport time, the goal of arrival to arrest of 30 minutes was barely met, suggesting that there is little room for error and the inclusion criteria may be insufficient for significant patient eligibility. Even with a “scoop and run” approach, 30 minutes may still be insufficient for dispatch, arrival and evaluation, and transport to the appropriate receiving hospital in a dense urban environment. In a geographic modeling study, only 1.68% of the US population was eligible for ECPR with a 45-minute window [18], which suggests that either the time towards ECPR candidacy should be extended or the factors associated with time to arrival should be shortened [19].
Additionally, the goal time from arrest to receiving ED arrival of less than 30 minutes may raise accessibility and patient selection concerns. The time of day and traffic patterns in Boston can create prolonged transport times. Thus, despite the number of ECMO-capable centers in Boston, there are still many areas of the city and surrounding suburbs that may require at least a half hour of transportation from ambulance dispatch to scene arrival to receiving hospital arrival, even without any on-scene resuscitative efforts. Further monitoring of transport times and geographic distribution of cardiac arrest calls will be necessary to better understand these challenges and identify potential solutions. Several ECMO programs have attempted to solve this issue by bringing the ECPR cannulation process out of the hospital, but this requires dedicated equipment, training, and personnel [20].
There are several limitations to this study. The first is that this is a single-site study, and the feedback and systems issues identified may not be applicable to all ECPR programs. However, the structure of the implementation between prehospital and hospital-based providers with structured debriefing should be expected to produce site-specific feedback if implemented at a different institution. Second, this was a planned scenario with dedicated EMS crews who were aware of the scenario beforehand. This undoubtedly shortened our time from dispatch to scene arrival and may have biased the performance of the EMS personnel. Furthermore, our simulation started close to the hospital, which further shortened the transport time. Because there was one set of individuals who participated in the scenario, it is possible that different combinations of team members would have achieved different performance benchmarks and identified different patient safety issues. It is also worth mentioning that in situ simulation in the clinical environment has the potential for patient harm by distracting clinicians away from patients presenting for care during the scenario period. Finally, it is possible that personnel behaved differently in the simulated resuscitation than they would have with an actual patient.
In conclusion, we conducted a multidisciplinary, multi-site, and multiorganizational trial of in situ simulation for ECPR. Close planning between multiple teams was essential for the success of the scenario. Our metrics were similar to previously published studies of ECPR, with a time from arrest to ED arrival of 29 minutes and arrest to full ECMO support of 51 minutes. Structured debriefing identified several systems issues for improvement including communication, equipment, candidacy for ECPR, and the role of on-scene EMS treatment. Hospitals and health care systems considering initiating an ECPR protocol should consider in situ simulation to identify latent patient safety threats specific to that institution, in addition to using simulation to train multidisciplinary teams in high-risk, low-frequency events.
Notes
FUNDING
None.
CONFLICT OF INTEREST
No potential conflict of interest relevant to this article was reported.
AUTHORS’ CONTRIBUTIONS
Conceptualization: PSJ, AJE, RRS; Data curation: SW, HS; Formal analysis: all authors; Investigation: all authors; Methodology: EAT, SW; Project administration: all authors; Resources: HS; Software: all authors; Supervision: HS; Validation: HS; Visualization: EAT, SW; Writing–original draft: PSJ; Writing–review & editing: all authors.