lnterservice/lndustfy Training, Simulation, and Education Conference (1/ITSEC) 2009

Task Specific Simulations for Medical Training:Fidelity Requirements Compared with Levels of Care

M. Beth H. Pettitt, Mr. Jack NorfleetUS Army Research Development andEngineering Command (RDECOM)

12423 Research ParkwayOrlando, Florida, USA 32826-3276

heth.h.pettitt@us.army.mil,jack.norfleet@us.army.mil

ABSTRACT

C. Ray Descheueaux, OD, FAAOOptometric Physician

2984 N. Alafaya Trail #1030Oviedo, Florida, USA 32765E-mail: craydx@yahoo.com

There are many virtual and constructive training systems that simulate injuries as well as degrading andimproving patient conditions. Most of these systems use simplistic models to mimic the physiologicalresponse to the injuries and the response to treatment. For example, when modeling bleeding from agunshot wound, many of the models are simply based on the amount of blood loss over time. An arbitrarytime limit is often established to indicate a failure to save the patient. Little to no consideration is given tothe munitions type or to the baseline physiology of the individual who is shot. Also, most of these gamesand simulations independently reinvent the math models and the physical models of these wounds.Proprietary nature of these diverse simulation platforms results in very little reuse.

This effort will explore how physiology is being represented in several simulation platforms. Targetedvirtual and constructive systems include: Pulse, STTC's Tactical Combat Casualty Care (TC3) game,STTC's!Forterra's OLIVE environment, and PEOSTRI's One Semi·Automated Force (OneSAF). Targetedmannequins include METI's Human Patient Simulators, Laerdals SimMan and Gaumards simulators. Ananalysis will be done on the level of fidelity currently included in each of these systems and on the pros andcons of how the physiological and the pharmacological responses are simulated. The analysis will alsoinclude a discussion on simulating versus replicating human physiology.

An initial hypothesis is that the higher fidelity medical simulations have interdependencies in themathematical models representing different physiological sub-systems, such as bleeding, heart rate, bloodpressure, etc. An attempt will be made to define a strategy for selecting the correct fidelity of humanphysiology models as well as ways to reuse existing models.

ABOUT THE AUTHORS

Beth Pettitt is the Division Chief for Soldier Simulation Environments (SSE), Simulation and TrainingTechnology Center, RDECOM. The SSE Division is leading the way in Medical Simulation Technologiesand Dismounted Soldier Simulation Technologies. Prior to this position, she was the Medical SimulationTechnologies Team Lead for the Simulation, Training and Instrumentation Command (STRICOM) whereshe was instrumental in establishing STRICOM's CTPS and Advanced Trauma Patient Simulation (ATPS)DTO programs. She has been actively involved in medical modeling and simulation for over ten years.Ms. Pettitt has twenty years experience in military modeling and simulation. She has worked on theMission to Mars Program, the National Air and Space Plane (NASP) Program, Modeling and Simulation ofthe Transportation Environment (MSTE) Program, and the Distributed Interactive Simulation (DIS)Program. Ms. Pettitt has a Bachelor of Science Degree in Mechanical Engineering, a Masters in BusinessAdministration and is pursuing a PhD in Modeling and Simulation.

Jack Norfleet is a Chief Engineer in the Soldier Simulation Environments division at the Army'sRDECOM-STTC. Currently, he is in responsible for the medical simulation research efforts at the STTC.Mr. Norfleet has 24 years of experience in developing military simulations for training medical skills, andforce on force skills. He has also worked on the development of various range instrumentation systems.

2009 Paper No. 9121 Page 1 of12

Report Documentation Page Form ApprovedOMB No. 0704-0188

Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering andmaintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, ArlingtonVA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if itdoes not display a currently valid OMB control number.

1. REPORT DATE 2009 2. REPORT TYPE

3. DATES COVERED 00-00-2009 to 00-00-2009

4. TITLE AND SUBTITLE Task Specific Simulations for Medical Training: Fidelity RequirementsCompared with Levels of Care

5a. CONTRACT NUMBER

5b. GRANT NUMBER

5c. PROGRAM ELEMENT NUMBER

6. AUTHOR(S) 5d. PROJECT NUMBER

5e. TASK NUMBER

5f. WORK UNIT NUMBER

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) US Army Research Development and Engineering Command(RDECOM),12423 Research Parkway,Orlando,FL,32826-3276

8. PERFORMING ORGANIZATIONREPORT NUMBER

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)

11. SPONSOR/MONITOR’S REPORT NUMBER(S)

12. DISTRIBUTION/AVAILABILITY STATEMENT Approved for public release; distribution unlimited

13. SUPPLEMENTARY NOTES Interservice/Industry Training, Simulation, and Education Conference (I/ITSEC) held in Orlando, FL on30 Nov - 3 Dec, 2009.

14. ABSTRACT

15. SUBJECT TERMS

16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF ABSTRACT Same as

Report (SAR)

18. NUMBEROF PAGES

12

19a. NAME OFRESPONSIBLE PERSON

a. REPORT unclassified

b. ABSTRACT unclassified

c. THIS PAGE unclassified

Standard Form 298 (Rev. 8-98) Prescribed by ANSI Std Z39-18

Interservice/lndustlY Training, Simulation, and Education Conference (I/ITSEC) 2009

Mr. Norfleet has a BSEE from UCF and an:MBA from Webster University. He has also trained as anEMT. He is currently enrolled in the Modeling and Simulation PhD program at UCF.

C. Ray Descheneaux, OD, FAAO is in private practice in Oviedo, FL. Dr. Descheneaux studied Bio­Medical Engineering at Western New England College. He then received his Doctorate in Optometry fromThe New England College of Optometry, Boston, MA. After a hospital-based residency at the Brocktonand West Roxbury VAMC, Dr. Deschcneaux moved to Florida to practice. He is a member ofthe FloridaOptometric Association, American Optometric Association, Central Florida Society of OptometricPbysicians, Beta Sigma Kappa Optometric Honor Society and is a Fellow of the AOA. He is currentlyenrolled in the Modeling and Simulation program at UCF pursuing a PhD.

2009 Paper No. 9121 Page 2 of12

lnterservicellndustry Training, Simulation, and Education Conference (l/ITSEC) 2009

Task Specific Simulations for Medical Training:Fidelity Requirements Compared with Levels of Care

M. Beth H. Pettitt, Mr. Jack NorfleetUS Army Research Development andEngineering Command (RDECOM)

12423 Research ParkwayOrlando, Florida, USA 32826-3276

beth.h.pettitt(a)us.army.mH,jack.norfleetra)us.armv.mil

BACKGROUND

Medicine is advancing at a dramatic rate. Newdiscoveries and techniques are demanding a more rapidand advanced medical education. However, medicaleducation and training has changed very little in morethan one hundred years, In the 1880's, Dr. WilliamHalstead influenced American medicine to follow thetraining he received in Europe. At the time, Americansurgeons were essentially self-taught, while theirGerman counterparts had direct instruction(Physiological Response to Infection, Trauma andSurgery, 2004), Mentored instruction like this evolvedinto the" ...see one, do one, teach one ..." model thathas become the standard for today's medical education(Carroll, 2008, p, 49), Unfortunately, thisapprenticeship model creates the situation where aninexperienced doctor learns new procedures (withsupervision) on live patients. The result is usuallypositive, however, "...errors that otherwise would notoccur are made as residents learn their skills" (Langhan,2008, p, 467), In 2000, an article in the Jonmal of theAmerican Medical Association reported that medicalerrors were the third leading cause of death in theUnited States (Starfield, 2000), Statistics like theseindicate that the medical training model should beupdated with the goal of improving patient safety,

The use of human cadavers ha.-', been a standard inmedical education for decades. Medical students andresidents utilize these cadavers for learning anatomyand performing procedures (Langhan, 2008), However,the cost of cadavers is increasing and the availability isdecreasing. Compounding the problem is the GlobalWar on Terror which has pushed limits on treatments toa point that soldiers and medics are performingtechniques that used to be reserved for surgeons.Increasing the skill levels of the basic medic hascertainly saved lives, but it ha.'i: also introduced over40,000 new students that require more advancedtraining. This huge spike in demand along with thegrowing scarcity of practice assets has forced the

2009 Paper No, 9121 Page 3 of 12

C. Ray Descheneaux, OD, FAAOOptometric Physician

2984 N. Alaraya Trail #1030Oviedo, Florida, USA 32765

craydx@yahoo.com

military to significantly increase the use of live animalsfor trauma and surgical training. While increasing livetissue training may be justifiable in the current conflict;the justification will fade as time passes~ To addressthis controversy, the military has mandated that the useof live tissue in training is to be greatly curtailed. TheDepartment of Defense has fonned the Use of LiveAnimals in Medical Education and Training Committeefor the sole purpose of finding ways to reduce oreliminate the need to train on goats and pigs.

Medical simulation has emerged as a leadingtechnology to address the problems mentioned above.Simulation offers the ability to practice procedures,receive feedback, document results and repeatprocedures, without negative repercussion to life orlimb. In response to the benefits of medical simulation,the military and industry have developed a wide varietyof simulations for medical training. The range anddepth of these simulators cover a large domain, fromvery simple task trainers to complex full bodysimulators. All of these simulators rely on some type ofrepresentation of human physiology. Some simulatorsuse simple computer~based algorithms to take anemergency medical services student through basic lifesaving techniques. Other simulators are much morecomplex, and include full mannequin human patientsimulators with integrated, self compensating heart rate,respiration, pulse, oxygen levels, and bleeding (Carroll& Messenger, 2008; Rosen, 2008; McLaughlin, Fitch,Goyal, Hayden, Kauh, Laack, et aL 2008),

THE PROBLEM

Training medical personnel is a very large problemspace that includes a wide range of cognitive levels andskill levels. Trainees include emergency medicalservices, nurses, physician assistants, and doctors. Thetraining that these groups need varies from basic firstaid through advanced techniques for saving lives. Thenext logical question is how advanced does thesimulator need to be to adequately train a particularindividual or team? Some of the fully functionalhuman patient simulators can easily cost $100,000 or

Interservice/lndustlY Training, Simulation, and Education Conference (I/ITSEC) 2009

BASIC HUMAN PHYSIOLOGY

Basic human blood flow

Hospital

(high." NumbtrofTl'llinees ,..Io'!!l

First EMT Para·ResPDnder medic

Figure 1: Hypothesis for the Aualysis

Assess

only determining factor, the final decision will take intoaccount other factors such as cost. Hopefully, theresearch and deductions herein can be used as a guideto aid in making an informed decision.

Arteries I IVeins

~vdJ

Treat

Diagnosis

II

Figure 2: Normal, Simplified HumanCardiovascular System

The human cardiovascular system is a complex networkof arteries and veins controlled by many intrinsicfactors. The core of the system is the heart. It acts asthe central hub for blood flow both toward and away.As depicted in Figure 2, the arteries carry blood awayfrom the heart and supply smaller and smaller vessels,until the end organ is reached. At this point there is anexchange of nutrients and, or waste. The blood is thencollected by a series of veins which becomeincreasingly larger as they progress back toward theheart. The major concept of this system is a 'closedloop' system. Blood is not simply left to wander orspill anywhere; it is contained within the organizednetwork of blood vessels (Anderson, 2006).

more. While high-end simulations can train a variety ofstudent types, the cost associated with these devices canbe difficult to justify. Langhan (2008) states, "...beforethe medical education community embraces moretechnologically advanced (and hence more expensive)simulators, the role of lower-cost, less technologicallyadvanced simulators must be evaluated" (p. 468)."What is the appropriate level of fidelity?" is thequestion that many within the medical training industryhave asked. Langhan (2008) also comments that,tl ...there are no empirical prospective studies thatsupport the use of more highly advanced modalitiesover lower-fidelity models" (p. 469). Langhan (2008)also suggests, II ...ultra high fidelity task trainers andvirtual reality computer-based programs, whileimpressive in their appearance, have not beendemonstrated to be superior to low or moderate fidelitytrainers" (p. 469).

Determining the fidelity needed for every training levelis a major undertaking, but for this discussion thequestion has been scaled to a manageable size.Specifically, this effort will investigate the availablesimulators and simulations that contain a bleedingcapability and evaluate their realism when compared tohuman physiology. This information will then becompared to the current groups of students that areusing or could utilize bleeding simulation in theirtraining in an attempt to identify what level ofhemorrhage simulation fidelity is needed for training atthe different levels of care.

Figure I graphically represents the basic hypothesis.By inspection and induction, a simple simulator isusually less expensive and therefore will moreeconomically train a higher number of students.However, the less expensive simulators are also lesssophisticated limiting the skill levels they can train.Conversely, the more sophisticated and generally moreexpensive simulators are designed with a tremendousnumber of clinical features and should be able to trainan individual to a much higher level of expertise. Therelationship between fidelity, cost and trainee skilllevels should be linear. At tlus time, there is no clearway to determine what fidelity is appropriate for eachlevel of required skilL This effort will attempt toanswer some of these questions and overlay simulatorcapabilities where they fit best, across the levels ofcare. To start this evaluation, basic human physiologywill be compared with current simulated physiology invarious devices. A rubric will then be created toevaluate each device against the user requirements.The skills required at each level of medical care willthen be examined and the rubric will be comparedagainst the learning objectives. Since fidelity is not the

2009 Paper No. 9121 Page 4 of12

Interservice/industlY Training, Simulation, and Education Conference (I/ITSEC) 2009

As depicted III Figure 3, when the normalcardiovascular physiology is disrupted, by anamputation of an extremity for example, there is amultitude of chemicals released for short term controland long term regulation. The simplified version of theprocess is conservation. Peripheral blood vessels areconstricted in order to decrease blood flow toextremities and increase central blood pressure. Theblood vessels supplying the heart and brain are dilated,allowing more hlood flow to these organs, The heartfills with more blood on each contraction to maximizeoutput to the essential organs, the heart and brain("Hypovolemic shock" n.d.; Kolecki, Menckhoff,2008).

Blood is selectively sent toheart and brain, sacrificingother organs

The heart fills more ateach contJ:action toincrease cardiacoutput

Peripheral bloodvessels constricted toreduce blood flow toextremities andincrease blood

Figure 3: Human Response to Shockfrom Blood Loss

PROCEDURESfTRAINING AT DIFFERENTLEVELS OF CARE

Exsanguinating hemorrhage was chosen as the scenariofor this evaluation because of its dire consequences inboth the civilian and military medical environments. Inthe civilian realms, hleeding still has a high mortalityrate. In the military, uncontrolled hleeding is thenumber one preventable cause of death on thebattlefield (Cain, 2008).. In order to focus the problemto a manageable size, bleeding control procedures wereresearched, and skills were identified at the pre-hospitaland bospitallevels of care.

At the pre-hospital level of care, skills of the firstresponder and the emergency medical technician(EM1) or paramedic were evaluated separately. The

2009 Paper No. 9121 Page 5 ~f12

first responder is limited in what he or she can legallydo, yet training these limited skills can save lives, Thefirst responder can administer direct pressure, dressings,and pressure points. EMTs and paramedics on the otherhand, have more treatments that they can administer,When confronted with life threatening hemorrhage, theescalation of treatment for an EMT/paramedic is asfollows: direct pressure, dressing, pressure dressing,and pressure points. If those methods fail, andprotocols allow, hemostatic agents and tourniquets arethe final options to stop the bleeding. EMTs andparamedics can also administer fluid replacementtherapy with saline-based volume enhancers and bloodexpanders. In either case, pre-hospital caregivers arelimited by the equipment available in their vehicle or intheir pack (Bledsoe, 1997; Buncome, 2008; Zeller,2008).

Once the patient reaches the hospital, many moretreatment options are available to the caregiver. At thislevel of care, skills of the doctor and surgeon wereexamined separately, The doctor follows the sameprogression of treatment as in pre-hospital care withadditional pharmacological and fluid replacementoptions, such as whole blood and natural bloodproducts, The surgeon expands care options withconventional surgery, cauterization, and embolization.Hospital caregivers also have aCCess to high techimaging capabilities whicb allow them to identify theexact problem area and focus their efforts on solvingthose problems. The hospital also has expandedcapabilities that monitor the major systems of the bodybut require a much greater understanding of theinterconnectivity of these biological systems. In thehospital, equipment limitations are typically not a majorproblem.

Even though many of the protocols appear to be thesame across the levels of care, the cognitive abilities ateach level of care differ significantly. In the pre­hospital arena, caregivers typically use patternrecognition to assess and treat the patient. Unless thecaregiver is. a physician, he or she does not diagnose apatient, but assess life threatening conditions and applytreatments to stabilize the patient. In the hospital, thedoctors and surgeons use results of advanced testing tocreate a more detailed understanding of what isoccurring throughout the body as a whole. Thisunderstanding of affected systems and possiblecomplications leads to detailed diagnoses and definitivecare. The hospitals also have the luxury of variouspharmacological treatments as well as time to observeand react to changes in the system. Therefore, therequirements for physiological and phannacologicalrepresentations in medical training systems should varysignificantly across the levels of care.

lnterservice/lndustry Training, Simulation, and Education Conference (1/ITSEC) 2009

PHYSIOLOGICAL/PHARMACOLOGICALREPRESENTATIONS IN MEDICAL

SIMULATIONS

Figure 5: Arthur Guytou's Computer Model of theCardiovascular System

Different levels of required training create the need fora variety of medical simulation and training tools. Thetools vary from simple representations to highlycomplex equations that have interdependentmathematical models to represent human physiology(Viceconti & Clapworthy, 2006), For example, thesimplest representations of bleeding, typically found infirst-person shooter games, dictate that if the patientreceives a pre-determined number of wounds,regardless of location, death occurs, or:

Death =2)Vumber of wounds t'l)

The next step in complexity is a simple time modelwhere, if the bleeding is not stopped within a certaintime, the patient dies. In another example, therepresentation of blood loss is driven by simplemathematical models where time is still a majorcomponent, but blood flow can vary or:

L'Blood loss =,' xdx

Where t = time

created a computer model (written in FORTRAN) thatbecame the basis for how blood pressure regulation andcardiac output were to be looked at for the future. Dr.Guyton utilized several hundred mathematicalequations to relate long-tenn blood pressure regulationand cardiac output. A circuit analysis was done usingpressure-flow and pressure:-volume mathematicalrelationships that were latcr developed into a graphicalanalysis of cardiac output and venous return to the heart(Taylor, 2004; Montani, 2008), A portion ofthis circuitis shown in Figure 5 (Hall, 2000),

(Darcy's law)

These two techniques are easy to understand andimplement, but are not particularly accurate. Bloodloss can be more accurately represented by Figure 4,where over time, one subsystem effects anothersuhsystem of the body,

Dr. Guyton is well known as a pioneer in the field ofmedicine, especially physiology (Taylor, 2004),Consequently, his models have become the backbone ofmany simulation products. Many current simulatorsutilize tbe Guyton physiological models and extract thecomponents they may need for their simulation product.(E. Fletcher, Executive Producer, Breakaway Ltd(Pulse!), personal communication, March 20, 2009),An example of a complex equation for bleeding isgiven helow ("Blood Flow"):

F=/',PR

The above equations represent a small example of thelevel of physiological complexity that might berequired in any of the more realistic simulation andtraining devices.

ih;bkJodlo%cOllwn!Oroo.sitval/GVef~rMli

Figure 4: Simple representation of a game-basedbleeding model.

For decades, scientists and researchers have beenexploring how to accurately model the complex andinterdependent physiological and pharmacologicalresponses of the human hody (Cobelli & Larson, 2008;Hoppensteadt & Peskin, 2002; Ottesen, Olufsen &Lnrsen, 2004), In 1972 Dr. Arthur Guyton andcolleagues published an article challenging the Currentview of cardiovascular physiology, By applying simpleengineering principles, Dr. Guyton and colleagues

Where:F ~ blood flowL~ length of tubeR = resistance

(Hagen-PoisenilJe equation)

v = fluid viscosityr = radius of tubeP = pressure

2009 Paper No, 9121 Page 6 of12

Interservice/lndustry Training, Simulation, and Education Conference (I/ITSEC) 2009

MEDICAL SIMULATIONS

Medical simulation has historically been a small nichewithin the training community. In fact, the militaryoften avoided accurate representations of casualtiesbecause it made their "real" simulations messy. Thissimplification was a disservice not only to the caregiverwho was not afforded the chance to train his or herskills while preparing for war, but to the tacticalcommanders who were not trained to deal with thedifficulties of managing casualties while stillcompleting the mission. In fact, the civilian communityhas often been the leader in implementing simulationinto medical training. Discussions with programmanagers at METI revealed that the company'sGovernment systems sales are still only about 20% ofoverall sales.

Medical simulations are categorized as follows:personal computer (PC)-based training systems, task­specific trainers, full-body mannequin training systems,and surgical simulations. PC and mannequin basedsystems were evaluated for this exercise because theytypically contain some type of physiological model.The evaluation of these technologies explores fourdifferent measures of simulation fidelity: visualrepresentations, behavioral responses, haptics, andphysiological models.

PC-based systems depict human patients by combiningvisual representations, physiological models, andbehavioral responses. Matll1equin-based systemsinclude the same qualities as PC-based systems and alsoinclude haptics.

Visual representations typically involve the fidelity ofthe system itself, as well as how accurately the woundand bleeding are depicted. Behavioral responses referto what the simulation does when the trainee intervenes.Does the system respond correctly and in a timely andclinically meaningful way? Haptics refers to the senseof touch and the realism of this interaction.

The PC-based systems are usually driven by a specificscenario and therefore have a limited scope. Thesesystems are totally ba'ied on visual and auditory cuesand typically have no haptic feedback. These are,however, very affordable and therefore lend themselveswell to training large numbers of students. Examples ofmedical PC game base systems include: OneSAF,Pulse!, TC3 and OLIVE.

OneSAF is a constructive war fighter simulation thatuses very simple models to represent wounding andtime to death, where time to death is determined by acombination of the number of wounds and time. Pulse,

2009 Paper No. 9121 Page 7 ~f12

TC3, and OLIVE were designed specifically formedical training and, therefore, have more complex andinterdependent mathematical models (H. Mall, VPEngineering. ECS (TC3), personal communication,February and March 2009). Most of these models areproprietary and therefore, a basic description is depictedin Figure 6 (Mall & Frolich, 2006).

HeaJ1Rate

Trachea EfficiencyLeft Lung EfficiencyRight Lung Efficiency

PainAnxiety

Figure 6: Representation of a Game~based

bleeding model.

Mannequin Based Medical Simulations

Full-body mannequin based training systems typicallydepict the most complex representations of humanphysiology. This family of simulators is particularlyuseful for training both cognitive and psychomotorskills because a patient outcome can be achieved. Inother words, the patient can improve, stabilize, or die.Full-body mannequin simulators are appropriate forboth partial task training as well as team training, andthey introduce the difficulties of taking care of apatient, not just a body part or computer screen. Thesesystems are often based on very complex physiologicalrepresentations and contain a great deal of internal andexternal computer hardware as well as various sensors,pumps, and other mechanical components. Thecomplex physiology and physical attributes tend tomake these systems very expensive. Some examples ofthese mannequins include: Laerdal's G3 and SimMan,

lnterservice/lndustry Training, Simulation, and Education Conference (l/ITSEC) 2009

Gaumards HAL, and METTs I-Stan and HPS, Themanufacturers of all of these systems claim to havephysiological models but the capabilities of thesesimulated physiologies vary significantly.

Two different models of Laerdal simulators wereexaminedl the venerable SimMan and the emerging G3.The SimMan can train bleeding contraIl butphysiological responses are the responsibility of theoperator. The G3 marketing material claimsphysiological modelsl but discussions with thecompany suggest similar functionality to the SimMan(,1, Hawkins, Soutbeast Regional Sales, Laerdal,personal communicationsl February 2009 and March19,2009),

Gaumard's newest human patient simulator is calledl1AL. ' HAL is a full-size human patient simulator thatallows participants the opportunity to practice variousprocedures. According to the company literaturel thissystem is tetherless, but offers a wireless tracking log ofup to six participants. HAL is advertised to have vitalsigns that respond to the current physiological conditionand participant interventions. The unit comes withpreprogrammed scenarios which can be modified ornew scenarios can be authored. Optional packagesinclude: armlleg stumps that will bleed appropriatelywhen compared to heart rate and blood pressurel traumapackagesl and wound modules. The model used as thebasis for HAL's physiology was not provided byGaumard (Gaumard, 2008),

METI has touted its complex physiology for years fromits early human patient simulator (BPS) to its newestsimulator, the I-Stan (Van Meurs, Good, & Lampotangl1997; Van Mems, Nikkelson & Good, 1998). Thesemodels are self-compensating and they realisticallyadjust the vital signs along the continuum from healthy,through shock, to death. Many years of civilian andmilitary use have validated these models which wereoriginally based on Guyton's work.

ANALYSIS

In order to assess the fidelity against the levels of care,standards of measurement were required. Table 1describes the rubric that was developed. The first fourcharacteristics (physiological, visual, behavioralresponse and haptic fidelity) address the specificobjective of this analysis. The others are given asexamples of the complexity of the issue and as samplesof other qualities they may be important when selectingmedical simulation and training tools. Cost, althoughvery important, was left out of this analysis and shouldbe treated as an independent variable.

2009 Paper No. 9121 Page 8 of12

Once the rubric was established, the fidelity of eachsimulation was evaluated. Grades were then assignedbased on various forms of available information.OneSAF, TC3, Olive, the Laerdal SimMan, MET! I­Stan and HPS were evaluated using technicaldocumentation and experience that was gained fromprevious developments. Pulse!, Gaumards HALl andLaerdal G3 were evaluated through marketing literatureand through interviews with company representatives.The grading results are represented in Table 2.

The next step in the analysis was applying a gradingfactor to each simulation factor based on the cognitiveand skills requirements of each caregiver at thedifferent levels of care. For example, physiologicalfidelity is much more important when training doctorsand surgeons because they operate at a much highercognitive level. The first responders and EMTs onlyrequire adequate signs and symptoms to assess a lifethreatening pattern so a complex physiology is notneeded.

CONCLUSIONS

Part of the original hypothesis stated that a simple, orlower fidelity, simulator is usually less expensive and,therefore, will be more economical to train manystudents. A more sophisticated and generally moreexpensive simulator can train to a more advanced level.The analysis showed that the fidelity of human medicalresponses involves many factors; includingphysiological, visual, behavioral, and haptic fidelity andthat the requirement for these features is quitedependent upon the needs of the training audience. Theanalysis also showed no direct correlation betweenoverall fidelity and levels of care. If cost were not anissue at medical training locations, a more accuratehuman representation would be desirable, since itmitigates the need for an operator to become an expertin the complex interactions of anatomy and physiology.

This analysis concludes with four recommendations toassist in solving the identified problemRecommendation 1 is to establish an open source,standardized library of complex physiological modelsto reduce the cost of simulators making them moreavailable for training all levels of care. While optimal,it may be unrealistic given the difficult and costly taskof maintaining configuration control while protectingintellectual property rights. A simpler and moreeconomical approach would be to standardize thedesired inputs and outputs and allow the market tomanage the details inside of each model. This wouldprotect intellectual property and spur innovation whilereducing the need for each new simulation technologyto start [Tom scratch.

lnterservicellndustry Training, Simulation, and Education Conference (1IITSEC) 2009

Table 1: Rubric for evaluating medical simulations

1 3 5

Physiological Time or state based: Semi-autonomous: Self compensating:Fidelity Little to no physiology. Physiology based on actual Physiology based on

Death based on pre- algorithm, but scenario pre- human physiology.detennined number of detennined by anthor. Systems interact. Outcomewounds or time after wound based on interventionwithout intervention

Visual Minimal to none Simple graphics; Cartoon, Life like animations,Fidelity Icons/colors non-life like representations; actions and skin

Not clinically meaningful Stiff, plastic, artificialBehavioral Operator induced Semi-Autonomous Timely, accurate andResponse Subjective Pass/Fail automated reaction toFidelitv interventionsHaptic None Tactile stimulation Tactile stimulation withFidelity forced feedback

Operator Skill Advanced training needed: Mild to moderate training Minimal training needed:Medical training and/or needed: Basic computer skills forcomputer skills needed to Basic medical background start-up and runningprogram and/or author with intermediate computer scenario, Little to noscenario. skills needed to program medical training needed.

and/or author scenario. Basic knowledge ofsimulator's system only.

Durability Fragile: Simulator confined Covered: Full field capabilities:to laboratory only due to Simulator able to travel out of Simulator able to withstandfragility of internal, external lab, but needs to be protected transportation and withoutparts. from elements. regard to elements.

Users One: Two-five: More than five:

I Single user only. Two to five person scenarios. Team based scenarios.

IRefresh i>lOmin: 3-10 min: 2 min or less:

Resetting simulation and/or Resetting simulation and/or Resetting simulation and/orsimulator takes more than simulator takes three to ten simulator takes twoten minutes. minutes. minutes or less.

After Action None: Pre-detennined: Detailed:Review No after action review After action review pre- After action review

available. determined by machine and/or customizable ba<;ed onauthor. Review based on yes author. Measures quantityor no, positive or negative. and quality along with time

to response.Availability Limited: Commercial: Open source:

Limited to government or Available too many users~ but Open to public witheducation affiliation. contains proprietary minimal limitations. Open

information retained by source physiology andmanufacturer and/or models.developer.

Distributed Fixed: Web based: Mobile:Training Simulator confined to lab Available on web globally. Completely mobile and

use only or Pc. Limited by web access. available to any location atanytime.

2009 Paper No. 9121 Page 9 qf12

lnterservicellndustry Training, Simulation, and Education COl~rerel1ce (l/ITSEC) 2009

Table 2: Analysis of Medical Simulations with respect to bleeding

PC Based Gaming

Environments Mannequin Based Systems

Criteria "A" "B~l "C" ltD" "E" ifF" "Gil "H" "J"Physiological Fidelity 3 3 3 1 3 2 3 4. 4

Visual Fidelity 4 3 3 1 3 3 3 S 4

Behavioral/ResponseFidelity 3 3 3 3 4 3 3 S S

Haptic Fidelity 1 1 1 1 S S S S S

Operator Skill 3 S 3 1 3 1 3 3 4Durability 1 1 1 1 3 1 3 4 3Users 1 1 S S 3 3 3 3 3Refresh S S 3 3 3 3 3 3 3After Action Review 3 3 3 3 2 2 3 3 3Availability 1 1 3 1 1 3 3 3 3Distributed Training 2 3 3 1 S 4 4 S 4

f Bl dOfCtLt t DOllT hi 3 Ra e : eQmremen s a 1 eren eve S 0 are or ee mg

I Levels of Care I1"

Criteria Responder EMT Doctor Surgeon

Physiological Fidelity 1 1 , S 4

Visual Fidelity 4 4 1 3 2

Behavioral/Response Fidelity 2 3 3 4

Haptic Fidelity 4 4 2 3

Operator Skill S S S S

Durability S 4 2 2

Users 1 2 3 S

Refresh S S 3 3

After Action Review 3 3 S S

Availability 3 3 3 3

Distributed Training 3 3 2 2

Recommendation 2 is for trainers searching for medicalsimulation solutions to perform a detailed and objectivetask and cognitive analysis on their programs ofinstruction, perform a market survey of possiblesolutions, and then prioritize and crosswalk theirtraining requirements against the available simulationcapabilities. This analysis would identify when higherfidelity models are necessary ba..ed on terminallearning objectives and would remove some of the

subjectivity that currently plagues simulationpurchasing decisions. By perfonning the legwork priorto the purchase decision, the proper fidelity can beapplied which will reduce cost and mitigate the need foran operator to become an expert in running simulatorsthat are too complex for the training task. Thisrequirements analysis model would be a fine start for anautomated analysis tool or expert system that wouldallow easier evaluations across different tasks, Too

2009 Paper No. 9121 Page 10 of12

lnterservice/lndustry Training, Simulation, and Education Conference (l/ITSEC) 2009

often, marketing literature or personnel convincetraining organizations that their "one size fits all"simulation is optimal for all training scenarios. The riskof this approach is in allowing salesmanship to overrideeducational need. One point this exercise has madevery clear is that there is no "best" medical simulationfor all levels of training.

Recommendation 3 is to develop a standardizedprotocol for the storage, recording, and transfer of data,including the development of a standard taxonomy. Inother words, create a common verbal and technicallanguage that covers medical simulation.

Although there are many factors driving simulation inmedicine, there is some resistance. One concern is thetransfer of knowledge. Experts have stated, "theknowledge translation from the simulated environmentto the clinical world has not been fully explored"(Langhan, 2008, p. 468). Recommendation 4 is toconduct further research in the area of validity. Oncethe transfer of knowledge utilizing a simulator can bevalidated, it should make the acceptance of simulationin medical education more mainstream (Kobayashi,Patterson, Overly, Shapiro, Williams, 2008). A smallsubset of medicine has already embraced thistechnology. The field of Anesthesiology, for example,uses simulation for training during education(Matveevskii & Gravenstein, 2008). Validation will notbe a trivial task; however, it is a challenge that deservesattention.

To further develop this analysis model, expert trainersat various levels of care should be consulted todetermine the fidelity issues they face every day. Theirinputs should be compared to the critical learning tasksof various medical curricula to validate and expand themodel. This expansion could easily result in an expertsystem that simplifies the answer to the question:"What fidelity simulator is needed to teach a specificmedical course?"

REFERENCES

Anderson, R. M. (2006). The gross physiology of thecardiovascular system. Retrieved fromhttp://www.cardiovascular.cx/March 13, 2009.

Bledsoe, B. E. (1997). Trauma Team EMS Protocols.http://www.ssgfx.com/CP2020/medtech/procedures/nrotocols.htm

"Blood Flow." (n. d.) Retrieved April 28, 2009, from:http://en.wikipedia.orgiwikilBlood_flow

Buncome, Madison, Yancey. (2008) Hemorrhagecontrol, hemostatic agent/tourniquet. EMS onlinetraining.

2009 Paper No. 9121 Page 11 of12

http://emsstaff.buncomhecounty.org/inhousetraining/protocol update 08/docs/Proccdure%2030.pdf

Cain, Jeffery S. (2008). Appropriate prebospitaltourniquet use,---Journal ofEmergency MedicalServices.

ar on trauma/appropriate prehospital tourniquet use.htmJ

Cobelli, C., Larson, E. (2008). Introduction toModeling in Physiology and Medicine. Boston:Elsevier.

Carroll, J. D., Messenger. J. C. (2008). Medicalsimulation; the new tool for training and skillassessment. Perspectives in Biology and Medicine,51 (1), 47-60.

Gaumard simulations for health care education. (2008).Gaumard Scientific, HAL product literature.

Good, M.L. (2003), Patient simulation for training basicand advanced clinical skills. Medical Education, 37,14-21.

Hall, J. E. (2004). Arthur Guyton's computer model ofthe cardiova.'icular system.

Am J Physiol Regul1ntegr Comp Physiol. 287, RI009­RIOII. doi:IO.1I52/classicessays.00007.2004.

Hoppensteadt, F . C., Peskin, C. S. (2002). Modelingand Simulation in Medicine and the Life Sciences (2nd ed.). New York: Springer.

Hypovolemic Shock. (n.d.) Retrieved February 19,2009, fromhttp://www.nlm.nih.gov.medlineplus/eney/article/OO0167.htm

Kolecki, P, Menchoff, C. R. (2008). Shock,hypovolemic. EMedicine Emergency Medicine.Retrieved February 19,2009 fromhttp://emedicine.medscape.com/articleI760145­overview

Kobayashi, L, Patterson, M. D., Overly, F. L., Shapiro,M. J., Williams, K. A., Jay, G. D. (2008).Educational and research implications of portablehuman patient simulation in acute care medicine.AcadEmergencyMedicine, 15 (11),1166-1174.

Langhan, T. S. (2008). Simulation training foremergency medicine residents: time to moveforward. The Journal qfthe Canadian Associationo(Emergency Physicians, 10 (5).467-469.

Mall, H., & Frolich, S. (2006). Modeling of SyntheticCa'iualties. ECS, Orlando, authors private collection.

Matveevskii, A. S., Gravenstein, N. (2008). Role ofsimulators, educational programs, and nontechnicalskills in anesthesia resident selection, education, andcompetency assessment Journal ofCritical Care(23),167-172.

McLaughin, S., Fitch, M. T., Goyal, D. G., Hayden, E.,Kauh, C. Y., Laack, T. A. et al. (2008).Simulation in graduate medical education 2008: a

Intersen1ice/lndustry Training, Simulation, and Education Conference (1/ITSEC) 2009

review for emergency medicine. Acad Emerg Med,15 (JI), 1117-1129.

Montani, J, P" Van Vliet, B. N. (2008). Understandingthe contribution ofGuyton~s large circulatory modelto long-tenn control of arterial pressure.Experimental Physiology. 94,382-388.

Ottesen, J. T., Olufsen, M. S., Larsen, J. K (2004).Applied Mathematical Models in HumanPhysiology. Philadelphia: SIAM.

Physiological Re~ponse to Infection, Trauma andSurgery. (n.d.) Retrieved January 15, 2009, fromhttp://sundcal­tutor.onl.uk/core/preop?/physiologv.htm.

Starfield, B. (2000). America's Healthcare System isThird Leading Cause ofDeath. Retrieved March22,2009, from http://www.health-carc­reform.net/causedeath.htm

Rosen, KR (2007). The history of medicalsimulation. Journal ofCritical Care (23),157-166.

2009 Paper No. 9121 Page 12 of12

Taylor, A. E. (2004). Arthur C. Guyton, M.D. (1919­2003); Guyton's legacy. Lymphology 37, 39-40.

Viceconti, M., Clapworthy, G. (2006). The virtualphysiological human: challenges and opportunities,812-815. Retrieved April, 8, 2009 from IEEEXplore.

Van Meurs, W.L., Good, M.L., & Lampotang, S.L.(1997). Functional anatomy of full-scale patientsimulators. Journal ofClinical Monitoring, 13, 317­324.

Van Mours, W.L., Nikkelson, E., & Good, M.L. (1998).Pharmacokinetic-phannacodynamic model foreducation simulation. IEEE Transactions onBiomedical Engineering. 45(5). 582-590.

Zeller, J. (2008). Beyond the battlefield: the use ofhemostatic dressings in civilian EMS. Journal ofEmergency Medical Services.http://www.journals.e1sevierhealth.com/periodicals!;ems/article/PIISO197251008700871/1,111text