Introduction of an Extracorporeal Cardiopulmonary Resuscitation Eligibility Protocol for Paramedics in Atlantic Canada: A Pilot Knowledge Translation Project.

Introduction There is currently no protocol for the initiation of extracorporeal cardiopulmonary resuscitation (ECPR) for out of hospital cardiac arrest (OHCA) in Atlantic Canada. Advanced care paramedics (ACPs) perform advanced cardiac life support in the prehospital setting often completing the entire resuscitation on-scene. Implementation of ECPR will present a novel intervention that is only available at the receiving hospital. Our objective is to determine if an educational program can improve identification of ECPR candidates by paramedics. Establishing paramedic competence will ensure rapid transfer of eligible patients for a potentially life-saving intervention. Methods An educational program was delivered to paramedics including a short seminar and pocket card coupled with simulated OHCA cases. A before-and-after study design using a case-based survey was employed. Paramedics were scored on their ability to correctly identify patients suffering OHCA who met the inclusion criteria for our ECPR protocol. A Wilcoxon matched-pairs signed rank test was employed to compare paramedics' scores before and after the education delivery. A six-month follow-up is planned to assess retention. Qualitative data was also collected from paramedics during simulation to help identify practical issues, potential barriers, and to refine inclusion and exclusion criteria prior to the implementation of our protocol in the prehospital setting. Results The median score pre-education was 10 (IQR: 9-10.5) compared to 14 (IQR: 13-15) after education delivery. The median difference between groups was 5. The Wilcoxon matched-pairs test demonstrated a significant improvement in the paramedics' ability to correctly identify ECPR candidates after completing our educational program z = -2.67, p = 0.0039. Conclusion Paramedic training through a didactic session coupled with a pocket card and simulation appeared to be a feasible method of knowledge translation. Six-month follow-up data will help ensure knowledge retention is achieved.

Introduction

In recent years there has been a paradigm shift towards advanced training for paramedics. Providers such as advanced care paramedics (ACP) and critical care paramedics (CCP) have acquired advanced skills including endotracheal intubation, administration of fluids, and delivery of various medications. These advanced skills have been shown to be advantageous in certain patient populations. For example, a study in Scotland demonstrated that prehospital thrombolysis delivered by paramedics shortened time to thrombolytic in patients with out of hospital ST-elevation myocardial infarction [1].

Alternatively, in out of hospital cardiac arrest (OHCA), it has been suggested that basic life support (BLS) with a focus on early defibrillation may improve outcomes [2]. Although several advanced prehospital treatments and interventions have been studied, they fail to reliably improve outcomes [3]. Traditionally in our region ACPs spend more time on scene than primary care paramedics (PCP), as they can deliver advanced interventions. A large multi-centre trial examined the effect of scene time on neurological outcome in patients surviving OHCA. Their findings suggest that an intermediate scene time of 4-7 mins results in the best neurologic outcomes in patients who could not be resuscitated in the field [4]. It is known that survival of OHCA improves with shorter emergency medical services (EMS) response times, EMS cardiopulmonary resuscitation (CPR), and early defibrillation [5]. Based on these findings it has been postulated that optimal treatment of patients suffering OCHA should continue to emphasize early response times, quality CPR and early defibrillation. What has remained less clear is the optimal care strategy for patients in whom these early interventions fail.

Emerging evidence suggests that extracorporeal cardiopulmonary resuscitation (ECPR) applied to specific patient populations in a timely manner can improve both survival, and survival with a good neurological outcome in OHCA [6-8]. Currently, at our institution and in our prehospital EMS system there is no protocol for initiation of ECPR for refractory OHCA. Our goal was to develop a feasible knowledge translation tool for paramedic triage of potential ECPR patients. This pilot project was limited to advanced care paramedics to allow for increased feasibility with the smaller sample of available ACPs in our region. If successful, the educational tool will be implemented to include all prehospital providers. This is of great importance in our prehospital system as there currently is not 24-hour coverage of ACPs in the community and they are not always available to respond to the OHCA calls. Should we wish to provide 24-hour ECPR coverage, all prehospital providers will have to be trained to appropriately identify ECPR candidates. The current study addressed the first phase of this potential protocol, patient selection, and attempted to identify if an educational package consisting of a didactic teaching session, an educational pocket card, and simulated OHCA cases can improve paramedic ability to correctly identify ECPR candidates.

Materials and methods

Study Design

This study used a standard before-and-after design which examined the performance of ACPs in appropriately identifying patients who met inclusion criteria for ECPR delivery as well as correctly excluding patients based on exclusion criteria. Performance was tested before and after the delivery of an educational training session. This educational program included a pocket card, and a didactic teaching session followed by three high fidelity simulations. A survey composed of clinical vignettes was performed using Google Forms to test the paramedics' ability to correctly triage patients (Appendix).

Setting

Saint John is a coastal city located in the Atlantic Canadian province of New Brunswick. It is the second largest metropolitan area in New Brunswick with a population of 126,202 persons [9]. The Saint John Regional Hospital is a community and teaching hospital with approximately 445 beds and 48,000 visits to the emergency department each year [10,11]. It is the only level 1 trauma centre in New Brunswick and is also the location of the New Brunswick Heart Centre, the only percutaneous coronary intervention (PCI) capable hospital in New Brunswick.

New Brunswick has a provincial EMS comprised of mostly PCPs, however ACPs have recently been introduced into the provincial prehospital care system on a trial basis. In cases of suspected OHCA the closest available paramedic team is dispatched to the scene. If the initial ambulance on scene consists of PCPs they will treat the patient according to their BLS treatment algorithm and an additional ACP unit will be dispatched when available. The ACP unit then either travels to the scene of the ongoing resuscitation or meets the PCP ambulance during transport to hospital. After arrival to the patient the ACP team then assumes the lead for the resuscitation according to their advanced life support (ALS) treatment pathway. Each treatment pathway has its own criteria for patient transport and termination of resuscitation. Data in the provincial ambulance database demonstrated a trend of increased scene time in OHCA when the treating paramedics were ACPs.

The ACP treatment protocol for OHCA was developed based on the assumption that ALS providers can perform initial resuscitation that is equivalent to an in-hospital resuscitation attempt. Therefore, longer scene times have been generally accepted. While this assumption was likely true at the time of its development, the introduction of an ECPR protocol makes available a new time sensitive in-hospital intervention. This novel intervention, extracorporeal membrane oxygenation (ECMO), could improve outcomes in certain patients suffering refractory OHCA.

Data Collection and Analysis

Ethical approval was sought prior to the collection of any data. Ethical approval was granted by the Horizon Health Research Ethics Board (REB File #: 2017-2487). ACPs were recruited on a voluntary basis to complete a pre-survey by email or in person prior to participation in training. A didactic teaching session on ECPR and the eligibility criteria for our centre were provided. A description of the study and a declaration of consent were provided on the initial page of the survey before any survey content was displayed. No personally identifiable data was collected. Participants were asked to create a unique identifier to avoid duplication of data. Two surveys were distributed to the ACPs. The pre-survey presented several clinical cases with patients randomly organized as either eligible or non-eligible for ECPR. The eligibility criteria can be viewed in Table 1. The same survey presented in a random order was then distributed after the paramedics completed their education. A Wilcoxon matched-pairs signed rank test was performed to compare the pre-education survey results to the post-education results to determine if the paramedics improved their ability to correctly identify patients eligible for ECPR. Undertriage and overtriage rates were calculated according to the formula proposed by Peng and Xiang [12].

Table 1

Eligibility criteria.

CPR: Cardiopulmonary resuscitation; ROSC: Return of spontaneous circulation; PEA: Pulseless electrical activity; VT: Ventricular tachycardia; VF: Ventricular fibrillation.

Inclusion Criteria	Exclusion Criteria
1. Witnessed out of hospital cardiac arrest	1. Unwitnessed cardiac arrest
2. Age 18-70	2. Asystole after initial resuscitation (at the time of transport decision)
3. No flow time < 10 mins (from arrest to initiation of CPR)	3. Suspected Etiology:
4. 10 minutes or 3 rounds of CPR completed (whichever comes first) without ROSC	A) Uncontrolled bleeding
5. PEA/VT/VF as heart rhythm	B) Irreversible brain damage
6. Mechanical chest compression device available	C) Trauma
7. Cardiac catheterization Lab open (7 am – 7 pm)	4. Comorbidities:
8. ≤20-minute transport time to hospital	A) Standing do-not-resuscitate order
	B) Undergoing end-of-life care
	C) Unable to fit mechanical chest compression device on patient
	5. Pregnancy

Results

A total of nine participants were included in this pilot project. Eight participants had the designation of advanced care paramedic and one participant held the designation of critical care paramedic. A summary of demographic information can be found in Table 2.

Table 2

Demographics

ACP: Advanced care paramedic; CCP: Critical care paramedic.

Demographics
	Male, n (proportion)	Female
Gender	7 (0.78)	2 (0.22)
	25-34	35-44	45-54
Age	2 (0.22)	6 (0.67)	1 (0.11)
	ACP	CCP
Level of Training	8 (0.89)	1 (0.11)

Paramedics scored a median of 10 (IQR: 9-10.5) in the pre-education survey. This was compared to the median score of 14 (IQR: 13-15) in the post-education survey. The median difference between the pre- and post-education survey was five. The primary outcome of interest, correct triage of ECPR eligible patients was significantly improved after the education delivery as demonstrated by the results of the Wilcoxon matched-pairs signed rank test, z = -2.67, p = 0.0039. A graphical representation of the paramedics’ performance through their quartiles is displayed in the whisker plot presented in Figure 1.

Figure 1

Whisker plot representation of the median, IQR, and range of paramedic performance both pre- and post-education delivery.

ECPR: Extracorporeal cardiopulmonary resuscitation; IQR: Inter-quartile range.

Aggregate paramedic responses to individual questions are demonstrated in Figure 2. After the education delivery, paramedic performance on individual scenarios improved or was neutral on 14 out of 16 scenarios (87.5%). In two scenarios where the patients were eligible for ECPR correct triage decreased after the education delivery. In the scenario with an unwitnessed cardiac arrest all paramedics correctly excluded the patient both before and after the education delivery. In the remaining eight scenarios with clear exclusion criteria, there was a higher proportion of correctly excluded patients after the education delivery.

Figure 2

Aggregate summary of paramedic performance on individual survey questions.

ECPR: Extracorporeal cardiopulmonary resuscitation

Table 3 utilizes a contingency table to represent the triage decisions made by the paramedics in each patient scenario.

Table 3

Performance of paramedic ECPR activation for individual patient vignettes.

Overtriage = a/(a + b)

Undertriage = d/(b + d)

ECPR: Extracorporeal cardiopulmonary resuscitation

	ECPR Non-Eligible Patient		ECPR Eligible Patient	Total
Protocol Activated	4 (a)		49 (b)	53
No Protocol Activation	77 (c)		14 (d)	91
Total	81		63	144
	Triage Performance
	Overtriage	Undertriage
	7.5%	22%

Qualitative data collected through feedback from the paramedics during the simulation sessions included perceived potential difficulty using the mechanical chest compression system (MCCS) in extractions with narrow passageways which are a common feature of our local architectures. The paramedics also noted the need to develop strategies to troubleshoot MCCS device errors during transport. Refining the inclusion and exclusion criteria was also suggested to give the paramedics clear direction for the inclusion or exclusion of pregnant patients and pediatric patients.

Discussion

The goal of this pilot study was to assess an EMS knowledge translation component required in the development of an ECPR protocol that will be implemented into our hospital and prehospital care systems. Specifically, we investigated whether an educational package consisting of a didactic lecture coupled with high fidelity simulation could improve paramedics’ ability to correctly triage patients who meet eligibility for implementation of ECPR. We found that prior to formal training paramedics were able to correctly triage appropriate ECPR patients over 60% of the time. This suggests that paramedics have a significant gestalt for identifying patients who are candidates for ECPR. Their performance improved to a correct triage rate of almost 90% after the delivery of our education package. Further testing is planned to assess paramedic retention in six months.

Review of the paramedics’ survey performance in the individual scenarios revealed a trend of improved triage in cases where clear exclusion criteria were present. Interestingly, in two scenarios where patients met the inclusion criteria, the paramedics’ performance showed a negative trend after the education delivery. In these scenarios, there was some ambiguous data in the vignette i.e.: it was unclear if adequate resuscitation (three cycles or 10 minutes of CPR) had been performed. Despite the ambiguity, there were no exclusion criteria present to make the patient ineligible for ECPR. This suggests that perhaps paramedics should only consider the exclusion criteria. This could increase the specificity of patient identification and avoid undertriage of patients who may be eligible. This would work well in the flow of our proposed protocol. An additional review of the inclusion and exclusion criteria would be performed by the attending physician at the receiving hospital before the patient would be placed on ECMO. Although there is no current consensus on acceptable overtriage and undertriage for ECPR, similar principles apply to the rational used for the accepted targets in trauma. The focus has been to minimize undertriage and the potential for associated morbidity and mortality. A goal of less than 5% undertriage and a range of 25-35% overtriage are the accepted standards in trauma [13]. The implementation of a simplified triage criteria for paramedics focusing on absolute contraindications should bring our undertriage and overtriage rates into better alignment with the published standards in trauma, although further testing will be required to confirm this hypothesis.

Simulation with paramedics allowed us to provide real time feedback during the simulated OHCA cases. In the simulations, the paramedic playing the role of ACP acted as the team leader. They coordinated the resuscitation and had the ECPR pocket card in hand during resuscitation. Closed loop communication was used between paramedics and a review of the eligibility criteria took place prior to the transport decision point. No measurable decrease in CPR performance was noted during the eligibility criteria check. The patient was attached to the MCCS device prior to the completion of three cycles of CPR or a total CPR time of 10 minutes. With the aid of the pocket card, paramedics correctly triaged patients in all simulated cases.

In the literature to date, it has been demonstrated that studies with a shorter time to initiation of ECMO have a higher percentage of patients with favorable outcomes [14]. Furthermore, implementation of an ECPR simulation program has been shown to reduce times to ECMO in real patients [15]. For these reasons it was essential to involve paramedics in the development of our local ECPR program. Having them simulate scenarios allowed us to better understand potential barriers for implementing the ECPR protocol. ECMO is a high acuity but low opportunity scenario. Building a malleable protocol that can evolve over time, involves appropriate stakeholders, and responds to any failures will be key to success [16]. We have implemented this strategy and continue to improve our protocol as we embark on our journey to bring this worthwhile service to the local population.

Limitations

The results of this pilot project must be interpreted with caution. It is a pilot project; it will contribute to the framework in developing a more robust training experience to providers participating in our ECPR program. It should not be interpreted in isolation but will provide context for the future research within our ECPR program. The numbers in this study are low and a formal power analysis was not performed, instead a convenience sampling was chosen. We attempted to involve as many local ACPs as possible and were successful in doing this, however, we were limited by the number of a local ACPs and a finite time period to complete this pilot project.

Conclusions

We sought to determine if an educational program can improve identification of ECPR candidates by paramedics. Paramedic training through a didactic session coupled with a pocket card and simulation appeared to be a feasible method of knowledge translation. Establishing paramedic competence will ensure rapid transfer of eligible patients for ECMO, a potentially life-saving intervention. Six-month follow-up data will help ensure knowledge retention is achieved.