The Leiden anaesthesia simulator

British Journal of Anaesthesia 1994; 73: 287-292

CLINICAL INVESTIGATIONS

The Leiden anaesthesia simulator V. CHOPRA, F. H. M. ENGBERS, M. J. GEERTS, W. R. FILET, J. G. BOVILL AND J. SPIERDIJK

Summary We describe an anaesthesia simulator capable of simulating all possible situations during anaesthesia. The Leiden anaesthesia simulator (LAS) may be used with most commercially available anaesthesia equipment and monitors, which are connected to the simulated patient as they are to a patient. A commercially available intubation manikin attached to an electromechanical lung model represents the patient. The lung allows both spontaneous and mechanical ventilation. Compliance, resistance, tidal volume and ventilatory frequency may be altered by a controlling computer. Carbon dioxide production and oxygen uptake are simulated. Physiological signals (ECG, arterial, pulmonary arterial and central venous pressure waveforms) generated by a signal generator under software control provide input to the monitors. All types of ECG disturbances may be simulated. There are facilities for simulating non-invasive arterial pressure measurement and pulse oximetry. A series of physiological models is being developed to control interactions between the cardiovascular and respiratory variables. During a simulation session, a pre-defined scenario is presented to the trainee. The task of the trainee is to diagnose and treat the problem as if in real life. The simulator experiences on the LAS were judged as highly realistic by 28 subjects. This simulator is currently being used for teaching and training of anaesthetists, trainees and anaesthesia personnel and for research. (Br. J. Anaesth. 1994; 73: 287-292) Key w o r d s Computers. Model, computer simulation.

The practice of anaesthesia involves routine activities which have the potential of developing into critical incidents. In this respect, anaesthesia has much in common with other industries, such as aviation and the nuclear power industry. In these industries, simulators and training devices have been used for many years to maintain the efficiency and proficiency of their operators. In recent years there has been an increasing interest in the development and use of anaesthesia simulators. The Comprehensive Anesthesia Simulation Environment (CASE) developed by Gaba and DeAnda [1]- and the Gainesville anesthesia simulator (GAS) developed in the University of Florida [2] reproduce the anaesthesia work environment. CASE is now available commercially as Virtual Anesthesiology training simulation system

(CAE-Link Corporation, Binghamton, NY, USA). These devices are actively being used for teaching, training and research in anaesthesia [3-11]. We describe the Leiden anaesthesia simulator (LAS) which is capable of simulating all possible situations during anaesthesia. The Leiden anaesthesia simulator The LAS is designed to interface with most commercially available anaesthesia equipment and monitors (figs 1 and 2). The simulator consists of a commercially available head and thorax manikin (Laerdal Airway Management Trainer, Laerdal Medical, Stavanger, Norway) with an artificial arm (Adam Rouilly London Ltd, Sittingbourne, Kent, England). The manikin has an anatomically correct airway that can be manipulated and intubated. Laryngospasm and gastric regurgitation may be simulated mechanically. Breath sounds may be auscultated on both sides of the chest. The simulated veins on the artificial arm may be cannulated, fluids may be infused and drugs injected. Urine output is simulated by volumetrically controlled infusion of coloured fluid of varying concentration into a urine measurement system. ARTIFICIAL LUNG

The artificial lung is a modified electromechanical device (LSI 500, Dragerwerk AG, Liibeck, Germany) that allows both spontaneous and mechanical ventilation. Compliance and resistance of this device may be varied by electrical controls either manually or, as used in the LAS, under computer control. During spontaneous ventilation, tidal volume and ventilatory frequency may also be varied. Carbon dioxide production and oxygen uptake are simulated by computer-controlled flows of carbon dioxide and nitrogen to the lung by roller pumps. Because airway compliance and resistance can be varied, changes in airway pressure, spirometry and the capnogram can be simulated realistically. PHYSIOLOGICAL SIGNALS

Cardiovascular signals are generated by commercially available signal generators, the outputs of which are modified under software control to provide V. CHOPRA, MB, BS, FRCA, F. H. M. ENGBERS, MD, M. J. GEERTS, BSC, W. R. FILET, MD, J. G. BOVILL, MD, PHD, FFARCSI, J. SPIERDIJK, MD, PHD, FRCA(HON), Department of Anaesthesiology,

University Hospital Leiden, P.O. Box 9600, 2300 RC Leiden, The Netherlands. Accepted for publication: February 15, 1994.

British Journal of Anaesthesia

288

Figure 1 The Leiden anaesthesia simulator. 486 PC & dataacquistion board

RS232

Roller pumps

Artificial lung

RS232

RS232

Figure 2 Schematic representation of the Leiden anaesthesia simulator. Signals from commercial physiological signal generators, suitably modified under software control, provide input to real monitors. SAP = Systemic arterial pressure; CVP = central venous pressure; PAP = pulmonary arterial pressure; SaQ2 = arterial oxygen saturation

suitable inputs to the monitors (fig. 3). The MedSim 300 patient simulator (Dynatech Nevada, Carson City, NV, USA) generates ECG, systemic arterial, pulmonary arterial and central venous pressure waveforms. The ECG signals generated by MedSim 300 have a variable heart rate from 30 to 300 beat min~' and are coupled to all dynamic waveforms. The amplitude and axis of the ECG signals and the height of the ST segment may be varied. A variety of supraventricular and ventricular arrhythmias and conduction defects can be selected via an RS232

serial input from the computer. As the invasive arterial pressure outputs of MedSim 300 are of fixed amplitude (for example, 120/80 mm Hg for systemic arterial pressure), we developed a computercontrolled analogue modulator to achieve beat-tobeat amplitude adjustment of these outputs. Non-invasive arterial pressure (NIAP) measurement is simulated by connecting a Cufflink non-invasive arterial pressure analyser (Dynatech Nevada, Carson City, NV, USA) to the tubing of a standard pressure cuff applied to the artificial arm. This instrument is designed to provide accurate and repeatable dynamic arterial pressure waveforms for calibrating and testing automatic NIAP devices. In the LAS, the output of Cufflink is controlled by the computer via an RS232 interface. An electronic circuit with a computer controllable variable resistance simulates core and peripheral temperatures. Pulse oximetry is simulated by exposing the detector of a standard pulse oximeter finger probe to artificially generated light impulses synchronous with the ECG signal. The pulse oximeter probe is placed on an opaque artificial "finger" which has a photodiode on one side and a light emitting diode (LED) on the other (fig. 4). The light impulses generated by the pulse oximeter probe are detected by the photodiode and are fed into an electronic circuit (pulse splitter). This circuit generates two synchronous alternating electrical signals that correspond to the infrared (IR) and red (R) light impulses of the pulse oximeter probe. These impulses modulate two signals, one with a fixed amplitude and the other with a computer controlled variable amplitude, derived from the MedSim 300. The controlling voltage from the computer is determined by the preset oxygen saturation set either by a scenario or by the instructor. The modulated waveforms provide input to the LED driver which in turn leads to the generation of light signals by the LED. These light signals are detected by the pulse oximeter probe. This system can be used with most commercially

The Leiden anaesthesia simulator

289 THE SOFTWARE

The software is written in Turbo Pascal, version 6.0 (Borland International Inc., Scotts Valley, CA, USA). Where applicable, the source code is object oriented. All simulated variables and the progress of the simulated scenarios are displayed graphically on the instructor's computer screen. Fifteen different variables may be simulated (table 1) and adjusted independently of each other to generate a variety of scenarios (table 2), which may be run either in a manual or an automatic mode. In the manual mode, the instructor can adjust each variable from the keyboard. Each variable is translated via a translation table to the correct value for the D—A convertor drivers. A script file editor and interpreter have been developed which allow pre-programming of the scenarios according to a pre-defined sequence of events. More than one scenario can be superimposed to run in parallel. The scenarios may be frozen and replayed at any moment during a session. In the future, interactions between the cardiovascular and respiratory variables will be controlled by a series of physiological models. A pharmacological model will also be incorporated. The physical layout of the simulator is such that the instructor's computer screen is not visible to the person undergoing simulator training.

* 80

Figure 3 Photograph taken from the screen of an anaesthesia monitor showing some of the physiological signals generated and modified by the Leiden anaesthesia simulator. From the top, these are: ECG (with ventricular extrasystole) and heart rate, core temperature, radial arterial pressure waveform with corresponding pressures, central venous pressure waveform with corresponding pressures, the capnogram and arterial oxygen saturation.

available pulse oximeters and can simulate blood oxygen saturation in the range 40-100% with a resolution of 1 %. In addition, it produces a plethysmogram waveform.

SIMULATION SESSIONS

A typical simulation session consists of three parts, namely the briefing, the simulation and the debriefing. During the briefing, a general introduction to the simulator and information about the history, relevant physical findings and laboratory data of the simulated patient are presented. During the simulation session the trainee is presented with a predefined scenario. Depending upon the actions and responses of the trainee, additional scripts are superimposed on the basic scenario. If necessary, the instructor (a qualified anaesthetist) can override the automatic mode and modify the script during the session. The session is recorded on videotape. During debriefing after the simulation sessions, this tape is used to emphasize important training aspects arising from the simulation sessions. This provides

COMPUTER HARDWARE AND ELECTRONIC INTERFACES

The system uses a personal computer with a Cyrix 486 DLC microprocessor with a clock speed of 33 MHz (Royal Information Electronics Co. Ltd, Taipei Hsien, Taiwan) and a data acquisition card (Analog Devices, Norwood, MA, USA). Coordination of the sub-systems is achieved via 11 D-A and four A-D converters (Analog Devices, Norwood, MA, USA) and an eight-channel RS232 serial communication board (Digiboard, Eden Prairie, MN 55344, USA). Simulation of temperature and pulse oximetry and modification of three arterial pressure signals and lung variables are regulated by five specially developed electronic circuits.

Pulse splitter JUL_ Pulse 1 IR R detector

Photo- „ ^ diode From Pulse Artificial oximeter •finger' probe To

i

JL_' _JL From MedSim 300

Fixed gain amplifier

|/VJV.

[AJV Voltage controlled amplifier

i •

LED driver Electronic switch

—•

LED

JlTL

•1

-— Electronic switch

Controlling voltage from the computer

Figure 4 Schematic representation of the pulse oximeter simulator used in the Leiden anaesthesia simulator. IR = Infrared; R = red; LED = light emitting diode.

British Journal of Anaesthesia

290 Table I Simulated variables by the Leiden anaesthesia simulator

Table 3 The questionnaire used for evaluation of the Leiden anaesthesia simulator

Cardiovascular

Respiratory

Other

Systolic arterial pressure Diastolic arterial pressure Heart rate Central venous pressure ST segment changes Cardiac arrhythmias and conduction defects

Ventilatory frequency Tidal volume Compliance Airway resistance Capnography Pulse oximetry Oxygen uptake

Temperature (central and peripheral) Urine output

Please give a number on a scale from 0 to 10 as answers to the following questions. Do you think that the Leiden anaesthesia simulator represents reality with respect to the following? 1. The anaesthesia machine 2. The anaesthetic monitors 3. The simulated parameters 4. General appearance of the manikin 5. The physiological responses of the manikin 6. The reactions to administered drugs 7. The simulated scenarios 8. The surroundings How important do you think the role of anaesthesia simulators could be in? 9. teaching in anaesthesia 10. training in anaesthesia 11. continuing education in anaesthesia Anv other comments:

Table 2 Some of the scenarios simulated by the Leiden anaesthesia simulator Category

Scenarios

Patient-related

Light anaesthesia Hypertension Hypotension Acute haemorrhage Various cardiac arrhythmias, including ventricular fibrillation and ST segment changes Conduction defects Cardiac tamponade Myocardial ischaemia Myocardial infarction Left ventricular failure Cardiac arrest Endobronchial intubation Oesophageal intubation Venous air embolism Tension pneumothorax Bronchospasm Gastric regurgitation and aspiration Anaphylactic shock Malignant hyperthermia Total electricity failure Pipeline oxygen failure Total mechanical failure of the ventilator Failure of fresh gas flow Patient circuit disconnection Total airway obstruction Failure of monitoring devices: pulse oximeter, cardiovascular and gas monitors Obstructed or loose gas sampling line Pharmacodynamic effects of the administration of a variety of anaesthetic induction and inhalation agents, opioids, neuromuscular blocking drugs, inotropic and vasoactive agents, and other drugs

Equipment-related

Drug-related

necessary feedback to the trainee. The simulations are carried out in a dedicated room modified to resemble an operating theatre. Evaluation of the simulator Twenty eight anaesthetists and anaesthesia trainees with experience in anaesthesia ranging from 1 to 15 yr attended simulator sessions as part of a study designed to quantify the role of simulators in anaesthesia training (see accompanying article [12]).

Table 4 Evaluation scores (mean (SD)) of the Leiden anaesthesia simulator Category

Score (n = 28)

Anaesthesia machine Anaesthetic monitors Simulated variables General appearance of the manikin Physiological responses Reactions to the administered drugs Simulated scenarios The surroundings Role of simulators in Teaching in anaesthesia Training in anaesthesia Continuing education in anaesthesia

9.86 (0.55) 9 56 (0 71) 8.80(1.29) 4.12 (2.15) 7.60(1.53) 8.28 (1.21) 8.56(1.12) 6.76(1.45) 9.08 (1.07) 9.08 (1.07) 8.72 (1.40)

The subjects were presented with a scenario describing either anaphylactic shock or malignant hyperthermia during general anaesthesia. At the end of the session, each participant was asked to complete a questionnaire on the realism of the simulator and its possible role in teaching, training and continuing education in anaesthesia (table 3). The items on the questionnaire were scored using a scale 0-10, where 0 = "extremely poor" or "no correspondence with realism" and 10 = "excellent" or "unable to differentiate from reality". All participants returned the completed questionnaires. Apart from the general appearance of the manikin, all other features were judged to be realistic (table 4). Inability to observe patient-related physical signs was found to be the most important limitation of the simulator. All participants reported that their experience during the simulation sessions was very instructive and they considered it an important addition to the present training methods. The simulator was judged to be an excellent tool for teaching, training and continuing education in anaesthesia. Discussion The Leiden anaesthesia simulator is conceptually and functionally similar to CASE and GAS [1,2]. The use of real anaesthesia equipment enhances the

The Leiden anaesthesia simulator

realism of the system and introduces an element of stress during the simulator sessions. Computer screen simulation devices, such as the Anesthesia Simulator Consultant developed by Schwid and O'Donnell [13], lack the ability to simulate manmachine interface aspects of simulation. However, low cost and wider availability of these devices is an advantage. As in CASE [1], the manikin used in our simulator was judged to be the least realistic in simulating reality. Skin and temperature changes, eye signs and patient movements are not possible to simulate at present. A more realistic manikin, capable of simulating some of the patient responses would add considerably to the realism of the simulator. The current version of LAS uses a script system for generation of scenarios. An advantage of the script system is that it is relatively easy to generate scenarios with this system. Scenarios can be generated based on hypothetical incidents, from cases reported in the literature or from clinical experience. Moreover, the use of a script editor requires minimal training and the scenarios are easy to test. However, a qualified anaesthetist is needed to operate the system in order to intervene when the situation deviates from the planned scenario. Another qualified anaesthetist is required to give instructions to the trainee during the simulation session. An alternative to the script system is to develop a fully integrated system which incorporates physiological and pharmacological computer models. The advantage of using such a system is increased flexibility. However, the development of realistic models is extremely time consuming. For the future version of LAS, a series of physiological and pharmacological computer models are being implemented. These will eliminate the requirement of a qualified anaesthetist to operate the system during the simulation sessions. The instructor can then concentrate on giving instructions to the trainee. This should improve the efficiency of the system considerably. Three types of simulator sessions, analogous to those being used in aeroplane simulators during the training of pilots, have been developed and are being implemented: sessions for new trainees, recurrent sessions for continuing education and proficiency check sessions for the evaluation of the performance of anaesthetists. Each new trainee attends five formal simulator sessions of 1 h each, where the basic principles of anaesthesia are taught. Each successive simulator session is built on previous sessions. This is similar to the initial simulator sessions being given on GAS [7]. Senior trainees and qualified anaesthetists are exposed to recurrent training sessions, where they are expected to diagnose and treat not only the more frequently occurring critical incidents but also the rarely occurring life threatening events during anaesthesia. Proficiency check sessions are being used as an adjuvant to the formal periodic departmental evaluations of the performances of trainees. Although anaesthesia simulators have been developed primarily as teaching and training tools,

291

they can also be used for research. The CASE has been used extensively for research on the role of human factors in anaesthesia-related critical incidents [3-6, 8-11]. We have used LAS to provide anaesthesia personnel with experience with new anaesthetic equipment before its introduction into clinical use. Evaluation of ergonomic design and performance of anaesthetic equipment, especially during emergency situations, is another potential use of simulators. General interest and awareness about the role of anaesthesia simulators is increasing [14, 15]. As yet there is no evidence to suggest that the use of simulators is better than the current methods of training anaesthetists and helping them to maintain skills. Further studies are needed to quantify the role of simulators in teaching, training and research in anaesthesia.

Acknowledgements We acknowledge the financial assistance of the Ministry of Welfare, Public Health and Cultural Affairs of the Netherlands for the development of the Leiden anaesthesia simulator. We thank Drager Nederland BV and Datex Medical Electronics BV, The Netherlands for supplying necessary equipment and financial assistance. We are also grateful to Captains J. H. C. de Bruijne and J. MS. van Sliedregt of KLM Royal Dutch Airlines for their support and helpful discussions. Ms T. L. Kwiecien prepared the photographs for figures 1 and 3.

References 1. Gaba DM, DeAnda A. A comprehensive anesthesia simulation environment: recreating the operating room for research and training. Anesthesiology 1988; 69: 387-394. 2. Good ML, Gravenstein JS. Anesthesia simulators and training devices. International Anesthesiology Clinics 1989; 27: 161-166. 3. Gaba DM, DeAnda A. The responses of anesthesia trainees to simulated critical incidents. Anesthesia and Analgesia 1989; 68: 444-451. 4. DeAnda A, Gaba DM. Unplanned incidents during comprehensive anesthesia simulation. Anesthesia and Analgesia 1990; 71: 77-82. 5. DeAnda A, Gaba DM. Role of experience in the responses to simulated critical incidents. Anesthesia and Analgesia 1991; 72: 308-315. 6. Gaba DM. Dynamic decision making in anesthesiology: cognitive methods and training approaches. In: Evans DA, Patel VL eds. Advanced Models of Cognition for Medical Training and Practice. Berlin: Springer Verlag, 1992; 123-147. 7. Good ML, Gravenstein JS, Mahla ME, White SE, Banner MJ, Carovano RG, Lampotang S. Anesthesia simulation for learning basic anesthesia skills. Journal of Clinical Monitoring 1992; 8: 187-188. 8. Botney R, Gaba DM, Howard SK, Jump B. The role of fixation error in preventing the detection and correction of a simulated volatile anesthetic overdose. Anesthesiology 1993; 79: A1H5. 9. Botney R, Gaba DM, Howard SK. Anesthesiologist performance during a simulated loss of pipeline oxygen. Anesthesiology 1993; 79: All 18. 10. Holzman RS, Cooper JB, Small S, Gaba DM. Participant responses to realistic simulation training in Anesthesia Crisis Resource Management (ACRM). Anesthesiology 1993; 79: A1112. 11. Gaba DM, Fish KJ, Howard SK. Crisis Management in Anesthesiology. New York: Churchill Livingstone, 1994.

292 12. Chopra V, Gesink BJ, de Jong J, Bovill JG, Spierdijk J, Brand R. Does training on an anaesthesia simulator lead to improvement in performance? British Journal of Anaesthesia 1994; 73: 293-297. 13. Schwid HA, O'Donnell D. The anesthesia simulator recorder: a device to train and evaluate anesthesiologists' responses to critical incidents. Anesthesiology 1990; 72: 191-197.

British Journal of Anaesthesia 14. Good ML. Anaesthesia simulators and training devices. Anaesthesia 1990; 45: 525—526. 15. Gaba DM. Improving anesthesiologist's performance by simulating reality. Anestheswlogy 1992; 76: 491^194.