|Home | About | Journals | Submit | Contact Us | Français|
What I read I forget
What I see I remember
What I do I understand
The use of manikins in emergency department medical education is well established. Early models enabled training in hands‐on cardiopulmonary resuscitation. Later improvements in manikin technology provided the realistic simulation of intubation and defibrillation.
The educational advantages of realistic low‐level simulators in scenario‐based teaching have been previously described1 (box 1). Indeed, they have a key role in adult and paediatric life support group courses. However, the realism in each scenario is limited by the technology. The trainee's need for a progress report (“what's the patient doing now?”) is familiar to seasoned instructors—skilled facilitation is essential.
But what if the technology allowed the manikin to breathe for itself? What if it had palpable pulses (or did not)? What if the instructor could control the respiratory rate, oxygen saturation and blood pressure in addition to heart rate and rhythm? What if the manikin could actually speak? And what if the model could physiologically respond automatically and appropriately to any given intravenous agent? Could you then not remove the facilitator completely from the room, allowing the trainee to totally engage in the scenario?
In this paper, we describe and compare the two types of high‐level simulators currently in use in the UK. We give an account of our 2‐year experience in educating trainees at the Bristol Medical Simulation Centre, Bristol, UK. We report on how close current high‐level simulators are to a proposed ideal for scenario‐based teaching—the remote‐facilitator run scenario. Finally, we outline other current educational uses of simulators in the emergency department setting and hint at future roles.
The candidate is called into the simulated resuscitation room by a telephone or tannoy call “registrar to resus”. The call is made by a staff nurse who presents a 12‐lead electrocardiogram (ECG) to the trainee on arrival. The patient is a 64‐year‐old man with chest pain suggestive of ischaemic heart disease. The patient is gowned and receiving 2 l of oxygen via a mask. He is visibly tachypnoeic, displays eye opening and is talking. His oxygen saturation, heart rhythm and blood pressure are continuously displayed on a monitor. The nurse reports that the paramedics have sited a cannula, and given the patient aspirin and glyceryl trinitrate, with little effect. The ECG shows an acute inferior myocardial infarction. The candidate is able to take history from and examine the patient. One of the faculty provides the voice of the patient via a remote microphone. The candidate may elect to prescribe a thrombolytic agent.
Various options available to the faculty in the scenario design may include a relative contra‐indication such as marked hypertension, hypotension following initiation of thrombolysis, ventricular tachycardia, etc. The faculty meanwhile observe the scenario from behind a one‐way mirror. They have control over the parameters listed in table 11,, depending on the type of simulator used. Two‐way communication with a faculty “plant” within the scenario (the nurse) is achievable through headsets and microphone. This should enable the smooth running of the scenario.
Facilitated peer group feedback with or without the use of a video follows completion of the scenario. Peers will have observed the scenario either directly through a one‐way mirror or through a video link. Points for consideration in analysis of candidate performance may include adherence to established protocols, clinical knowledge, communication skills and team leadership.
In January 2000, we secured funding for teaching emergency medicine trainees at the Bristol Medical Simulation Centre. We ran five study days in 2000 targeted at medical emergencies with themes as detailed in table 22.. We used a Meti model (Human Patient Simulator System 5, Model C Manikin, Medical Education Technology Incorporated, Gainsville, Florida, USA) but were able to revisit some of the scenarios in 2002 with Laerdal SimMan (System 1, Laerdal Medical Limited, Orpington, Kent, UK). Each day consisted of four workshops and four scenarios. The trainees, supplemented by clinical fellows and staff grades, were divided into two groups of, typically, six. Each participant was sent precourse learning suggestions. Each of the specialist registrars had an opportunity to act as a candidate in one of the scenarios.
Trainee evaluation was led by two of the trainees. Feedback from all 10 trainees who had attended simulator training was sought through a questionnaire. This was analysed, and the results have been presented nationally. Trainer evaluation was sought by oral and written feedback.
From 2001 we have run four Scottish Airway and Ventilation Emergency (SAVE) courses a year using the Meti model. SAVE essentially centres around six typical resuscitation room scenarios in which there is a need for rapid sequence induction. Medical and trauma scenarios are included. Feedback was again sought from trainees and trainers who we selected to put through the course.
The trainees believed that overall, the simulator offered valuable training experience. Using a purpose‐built centre remote from the hospital environment enabled focused training without distraction. The structure of the day (alternating scenarios and workshops) was commended.
The advantages and disadvantages of the two types of simulators, as experienced by both trainers and trainees, are summarised in box 2. It is worth emphasising that the Meti simulator was prone to crashing. A complete crash required a tiresome wait while rebooting the control computer. A partial crash was also problematic. A sudden loss of the oxygen saturation facility might prompt the trainee into seeking facilitation from the nurse, seen to be wearing the headphones. By inadvertently reverting to typical advanced life‐support course behaviour, the advantages of the remote facilitator Laerdal model was lost.
In general, the better‐simulated scenarios with either model were those associated with cardiac arrest or coma or those with a cardiac theme. Scenarios were more realistic if every effort was made to recreate a resuscitation room rather than an anaesthetic room environment. In addition, having all participants in uniform added to the realism.
The training day required at least four trainers. Two were needed for each scenario, so that typically each trainer saw only half the scenarios and half the workshops. This limited the continuous professional development value of the day. Any trainees new to the simulator also needed an introduction (“normal” breath sounds, its airway anatomy, etc). Some of the trainees found the simulated scenarios threatening despite attempts at empathetic facilitation. Finally, the trainers believed that the trainee's “value” of the education was not necessarily based on the high‐level simulator, but rather on the effort that the trainers had put into preparing the day.
Overall, the feedback was highly positive and similar to that from trainers elsewhere.3 Trauma scenarios could also be further enhanced by use of supplementary plastic make‐up. The SimMan model can be exchanged for the Meti in this field—indeed, we had to do so when the Meti developed major problems. The scenarios benefited from its better airway and breathing features, and scenario control was easier. Its lack of eye features limited the realism of any simulated awake or semiconscious patient, however.
Senior emergency department nurses from the Bristol Royal Infirmary, Bristol, UK, have successfully run national thrombolysis study days for emergency department nurses at the Bristol Medical Simulation Centre since 2000. One of us instructs on a course on transport of the critically ill patient, a potential role for the speciality. Emergency medicine crisis resource management courses have been run in North America3 and Australia since 1998. The uncertainty, complexity and rapidly changing priorities of simultaneous patients in a resuscitation room are reproduced. The training has largely been based on aviation industry models. Leadership and team dynamics are explored. Laerdal SimMan's portability also gives trainers the option of simulating scenarios in the real rather than the virtual resuscitation room—for example, analysing trauma team performance. The Bristol Medical Simulation Centre also has the technology to enable satellite‐linked educational conferencing so that other groups of trainees and trainers can observe scenarios and offer feedback.
Future improvements are eagerly awaited for both simulators described. A simulated colour change (pink/blue/pale) and beads of sweat over the forehead will surely add realism in many scenarios. Replication of its competitor's eye features will enable Laerdal to produce a hugely competitive model given its user‐friendly interface, its better airway and breathing simulation, and its reliability, portability and comparable cost. However, the METI HPS system 6 incorporates a Macintosh control system that, although slow to load, produces a more intuitive interface and allows several virtual patients to be run simultaneously. These software patients can be superimposed in turn on one or more manikins from a single wireless control computer. The new METI “emergency care simulator”, on the other hand, trades some physical features such as realistic gas exchange for increased portability, and can function on its own battery supply for several hours in the back of an ambulance.
Are high‐level simulators the emergency medicine clinical assessment tool of the future? Perhaps. Clearly, the same scenario can be reproduced, giving each candidate the same test. High‐level simulator assessment is also likely to become more feasible as an increasing number of centres acquire them. More doctors are therefore likely to be exposed to teaching with these simulators, and therefore testing on a novelty manikin is less likely.
However, high‐level simulation as an assessment tool scores poorly at present in terms of validity, particularly in comparison with direct observation of the critically ill patient. Their proposed improvements, as outlined, may enable more valid assessment in future.
High‐level simulators have proved a valuable addition to specialist registrar training in emergency medicine in the south west. The Meti model is the more realistic, but is prone to technological problems. It also takes considerable effort in learning how to use it. The Laerdal model has better airway features and much better breath sounds. It is also simple to use. It is, however, limited by its eye features (upgrade awaited). The Meti model functions well for rapid‐sequence induction training purposes. Although medical emergencies are not universally well created in either model, proposed technological improvements should enhance realism. Nevertheless, the remote facilitator‐run scenario is possible if well prepared. Total engagement of the candidate in the scenario has been witnessed, enabling more realistic assessment of a trainee's performance. Their use in summative assessment (see box 34) is likely in time. We encourage high‐level simulator use as an educational tool in other regions, with the recommendations highlighted (box 4).
We thank the support of our fellow trainers in the Southwest. We also acknowledge the help of Andrew McIndoe and Alan Jones of the Bristol Medical Simulation Centre; Daranee Boon and Jason Louis, who led the SpR evaluation; and Phil Davies, Ian Higginson, Sarah Lloyd, Chris McLauchlan, Cliff Mann and Andrew McIndoe for offering comments on the paper. SAVE is the creation of Neil Nichol, Ursula Mackintosh and colleagues based at the Scottish Simulation Centre in Stirling.
ECG - electrocardiogram
SAVE - Scottish Airway and Ventilation Emergency
Funding: With thanks from the Medical and Dental Educational Levy via the Southwest Deanery.
Competing interests: None.