Construction of a reusable, high-fidelity model for simulation of extracorporeal cardiopulmonary resuscitation

There is increasing interest in the potential use of veno-arterial extracorporeal membrane oxygenation (VA-ECMO) in the prehospital setting to stabilise patients in refractory cardiac arrest.¹ VA-ECMO involves draining blood through a cannula sited in a large vein and passing it through a pump and a membrane oxygenator before returning it under pressure through another cannula sited in a large artery. It can therefore be used to temporarily replace the function of the heart and lungs. When VA-ECMO is used to manage refractory cardiac arrest, it is termed extracorporeal cardiopulmonary resuscitation (ECPR).

Out-of-hospital cardiac arrest patients have poor outcomes,^{2 3} not least because it takes time to package and transport them to heart attack centres. By bringing VA-ECMO into the prehospital environment, it is hoped that vital organ perfusion can be re-established more quickly in a refractory cardiac arrest, preventing irreversible hypoxic organ damage and potentially improving out-of-hospital cardiac arrest survival rates that remain very low.

Prehospital ECPR is fraught with difficulties and potential complications,⁴ many of which can be prepared for by simulation. Unfortunately, there are no cost-effective, commercially available manikins designed for ECPR. Previous attempts at simulating prehospital ECPR have used separate manikins for the dichotomous elements of cardiopulmonary resuscitation (CPR) and extracorporeal membrane oxygenation (ECMO); however, this significantly reduces the fidelity of the exercise. In-hospital ECMO manikins often have to forego the use of mechanical chest compression devices to prevent damage to the manikin. This again significantly reduces the fidelity and the physical challenge of cannulating vessels during active mechanical compressions. We describe a method of constructing a bespoke high-fidelity manikin for a prehospital ECPR simulation.

Methods

In preparation for a prehospital simulation exercise, we were keen to find a manikin that could be intubated and ventilated, receive chest compressions and be cannulated for ECMO. Any tubing simulating cannulated vessels had to be able to accommodate up to 50 cm of 25 Fr gauge cannula in a straight line in order to allow ‘venous drainage’. We also wanted a free-standing manikin with no external connections or cables to simulate with high fidelity the packaging and transport of a patient once commenced on VA-ECMO. We experimented with a Laerdal Resusci Anne QCPR model to see if adaptions could be made to accommodate our needs; however, we were unsuccessful in creating a workable model as there was minimal scope for adapting the manikin without causing significant damage.

We were able to source a first-generation Laerdal SimMan model, which was surplus to simulation requirements and had several areas of damage making it unsuitable for regular use. We made a simple adaption to the manikin by opening the abdomen and inserting a spare reservoir bag sourced by a cardiac perfusionist (figure 1, panel A). Silicone piping from the bag was channelled though the groin for cannulation (figure 1, panel B), and the circuit was completed and looped around the chest of the manikin before being secured in place with cable ties. The abdominal and chest walls were replaced and secured in place using tape (figure 1, panels C and D) before a thin patch of flesh coloured silicone was secured on top to conceal the pipes. We did not require the use of the SimMan’s electronics, motors or monitors as part of the scenario. We used a separate, portable monitor to display vital signs (predominantly electrocardiography and capnography).

Figure 1.

(A) Laerdal SimMan opened up and overlaid with reservoir bag and perfusion tubing circuit. Tubing directed up sides and across clavicles to avoid damage during chest compressions. (B) Silicone tubing channelled though both groins of the manikin for cannulation and secured in place. (C) Manikin with chest wall replaced. (D) Manikin with abdominal wall replaced. The reservoir bag was filled with red liquid prior to use to simulate blood.

The manikin was tested for intubation and CPR as well as the application of a Physiocontrol LUCAS CPR device prior to the simulation. Although there is a risk of damaging the damping spring for CPR with this device, our manikin survived the simulation in good condition and can be used again. Before commencing the simulation, the fluid reservoir was filled with artificial blood.

Outcomes

We successfully ran an out-of-hospital ECPR simulation, during which the team were able to intubate, perform chest compressions, percutaneously cannulate the femoral vessels and run a portable ECMO machine throughout the scenario. The complete exercise lasted almost 2 hours, which included the packaging, transferring the manikin down a staircase and into an ambulance and transporting it to a heart attack centre. Once there, it was transferred via an elevator onto a simulated catheter lab table without decannulation or malfunctioning of the model.

The manikin was more difficult to cannulate than previous models, as the vessels could not be easily palpated, although this is similar to the situation for patients in cardiac arrest. Additionally, the manikin was not echogenic, so the normal practice of using ultrasound, to visualise the vessels for cannulation or the location of wires in the simulated inferior vena cava, was not possible. There was minimal leakage from the cannulation sites on the manikin. Following our simulation, the ECMO cannulae were left in situ, and the manikin has been reused several times for teaching.

Discussion

The manikin we created was fit for purpose for our simulation and is a cheap and reproducible way of running a high-fidelity simulation for ECPR. Furthermore, it provides a useful teaching aid for ECMO novices. As mentioned, there was a small amount of leakage from the cannulation sites. This was not significant enough to impair the function of the ECMO machine as it ran for over an hour, but it is not known whether this was sufficient to impair the manikin’s electronics; we did not require the use of the manikin’s electronics for our scenario.

Future attempts to improve the manikin will focus on creating an echogenic medium enclosing the vessels, allowing the use of real-time ultrasound scanning for vascular puncture. We would also look at inserting a dual circulation with multiple vessels for differentiation between artery and vein. We would recommend our approach to others seeking to perform an ECMO simulation and ECPR in particular.

In conclusion, this is a cheap modification to existing, widely available technology that resulted in a very high-quality, high-fidelity simulation scenario. The manikin was robust enough to withstand a number of challenging transfers, without any disruption of the ECMO circuit.