effect of short-term pretrial practice on surgical proficiency ......effect of short-term pretrial...

EPTK

T“fwohcctbi

D

RFTiPnCM9

©P

ffect of Short-Term Pretrial Practice on Surgicalroficiency in Simulated Environments: A Randomizedrial of the “Preoperative Warm-Up” Effectanav Kahol, PhD, Richard M Satava, MD, FACS, John Ferrara, MD, Marshall L Smith, MD, PhD

BACKGROUND: Surgery is a skill-driven discipline. While other high-stake professions with comparable cogni-tive and psychomotor skill requirements often use warm-up exercises for achieving betterproficiency, the effects of such practice have not been investigated sufficiently in surgical tasks.

DESIGN: Subjects performed standardized exercises as a preoperative warm-up, after which the standard-ized exercises were repeated in a randomized order. In a variation to investigate the generaliz-ability of preoperative warm-up, the experimental group was allowed to warm-up with thestandardized exercises, after which a different task (electrocautery simulation) was performed.To investigate the effect of warm-up on fatigue, participants were involved in eight sessions(four before night call, four after night call), after which the tasks were repeated. Results wereanalyzed using ANOVA to plot differences between warm-up and followup condition.

RESULTS: All outcomes measures demonstrated statistically significant improvements after all of thepost�warm-up exercises (p � 0.01), and were seen in all groups with differing experience levels.In addition, the simple warm-up exercises led to a significant increase in proficiency in followupelectrocautery task for the experimental group when compared with the control group (p �0.0001). There was also significant improvement in performance of the fatigued group toapproximately baseline performance (p � 0.05), although they were not able to reach theiroptimal potential performance.

CONCLUSIONS: Preoperative warm-up for 15 to 20 minutes with simple surgical exercises leads to a substantialincrease in surgical skills proficiency during followup tasks. ( J Am Coll Surg 2009;208:
255–268. © 2009 by the American College of Surgeons)
butatbtit

pisfoeoscp

he Oxford English Dictionary defines the term warm-up asThe act or process of ‘warming up’ for a contest, and soorth, by light exercise or practice.”1 More generically,arm-up is “The act or process of raising the temperaturef an engine, electrical appliance, and so forth, to a leveligh enough for efficient working,”1 which captures theoncept of activating and preparing to a high level of effi-iency before beginning an activity. In sports and sports-raining literature, there are several published articles andooks highlighting the importance of warm-up in improv-ng performance and avoiding errors. The stretching hand-

isclosure Information: Nothing to disclose.

eceived April 6, 2008; Revised September 4, 2008.rom Department of Biomedical Informatics, Arizona State University,empe, AZ (Kahol), Department of Surgery, University of Washington Med-

cal Center, Seattle, WA (Satava), Phoenix Integrated Surgical Residency,hoenix, AZ (Ferrara), and Simulation Education and Training Center, Ban-er Health, Phoenix, AZ (Smith).orrespondence address: Richard M Satava, PhD, University of Washingtonedical Center, BB430, Box 356410, 1959 NE Pacific St, Seattle, WA

H8195-6410. Email: [email protected]

2552009 by the American College of Surgeons

ublished by Elsevier Inc.

ook published by the Stretching Institute2 says: “Warmingp before any physical activity does a number of beneficialhings, but primarily its main purpose is to prepare the bodynd mind for more strenuous activity . . .” it’s importanto start with the easiest and most gentle activity first,uilding on each part with more energetic activities, untilhe body is at a physical and mental peak. This is the staten which the body is most prepared for the physical activityo come.

Although the positive effect of warm-up on strenuoushysical activity is wellknown, research has also shown pos-tive effects of warm-up on cognitive skills3 in physicalports. Research has also shown that subjects tend to per-orm better at cognitive exercises in followup tasks. Theverall effect is seen as a cognitive arousal phenomenon thatnables subjects to more fully concentrate their resourcesn the task at hand. As such, cognitive arousal has beenhown to cause increased somatic and cortical activity4 andan even aid in resisting sleep. The effect of short-termractice is not limited to sports or physical activity only.
igh-stake, high-performance professions with substantial
ISSN 1072-7515/09/$36.00doi:10.1016/j.jamcollsurg.2008.09.029

mailto:[email protected]

psptpitc

aticm

esaihsvarctspsoosp

ttogptblbdsptbilw

pttcsatintwct

oshwm

efarfqwitEetcanf

HlTgmpeet

HvId

256 Kahol et al Preoperative Warm-Up Effect J Am Coll Surg

sychomotor skills proficiency, such as dancers, musicians,culptors, and painters, have, for centuries, used short-termractice or warm-up as a method of getting ready for theask at hand. A common thread among all activities thatromote use of warm-up includes strenuous physical activ-ty; strenuous mental activity with requirements of cogni-ive arousal; and ability to perform both within requiredoordination and task performance constraints.

Although warm-up has a documented positive effect onctivities that involve these requirements, it is also knowno be necessary to avoid deleterious consequences, such asnjury to self and others in the environment and to triggerognitive arousal that can substantially enhance perfor-ance ability.Modern-day surgery certainly qualifies as a high-stakes,

xpertise-driven field.5 The advent of minimally invasiveurgery, although revolutionary for patients and the oper-ting room, poses specific challenges for surgeons and res-dents. It requires surgeons to perform procedures withard-to-manipulate tools that impose undesirable con-traints on the movements of a surgeon. Although the ad-ent of robotic surgery and more intuitive manipulatorsddress some of these issues, modern-day surgery, and lapa-oscopic surgery in particular, still requires considerableognitive and psychomotor proficiency. There is substan-ial research that has shown that surgery requires bothtrenuous mental and cognitive activity.6 This raises an im-ortant question: if performing an operation involvestrenuous physical and mental activity, then would a “pre-perative warm-up” (or simply “warm-up” for the purposesf this article) activity that involves surgeons performingurgical exercises before the main task improve surgicalerformance in the main task at hand?Some earlier work has been reported to address a few of

hese issues. Do and colleagues7 used a laparoscopic “boxrainer” simulator to study the effect of warm-up exercisesn followup tasks. Participants performed a task involvingrasping pill-like plastic elements from a petri dish andlacing them into a bud vase. Time elapsed to complete theask and the number of pills successfully transferred to theud vase were used as measures of performance. Overall

aparoscopic performance was significantly improved foroth residents (all years combined) and the medical stu-ent control group (p � 0.0001). Although the studyhowed encouraging results of warm-up being used beforeerforming the main exercise, the study had certain limi-ations because, to a certain extent, it failed to distinguishetween learning effect and warm-up effect. The increasen proficiency might have occurred because subjects wereearning and adapting to the task. To isolate the effect of
arm-up on surgical proficiency from the improvement in i
sychomotor activation and cognitive arousal, it is importanto assimilate data across multiple trials with sufficient itera-ions to allow learning to occur to account for the “learningurve.” Also, their study did not include sophisticated mea-ures of surgical proficiency, such as hand-movement analysisnd tool-movement analysis, which can quantify psychomo-or proficiency in detail. In addition, tasks used in their exper-ment did not include a cognitive dimension, and thereforeot allowing for the study of the effect of warm-up on cogni-ive arousal. Finally, as the study used the same exercise asarm-up and followup exercise, it does not sufficiently indi-

ate that the increase in proficiency will carry over to otherasks, especially those with more clinical relevance.

This research embarks on defining a systematic method-logy to study the effect of short-term pretrial practice onurgical proficiency. We use simulation tasks targeted toone both psychomotor and cognitive skills to providearm-up exercises that will enhance the performance of theain surgical procedures.Several key questions need to be addressed to study the

ffect of warm-up on surgical proficiency and present a caseor including preoperative warm-up as an advantageousctivity for surgeons. We have developed a series of inter-elated hypotheses that address the issue systematically.Thelowchart in Figure 1 links the hypotheses with the keyuestions and variables they address. The first question thate address is whether warm-up does indeed affect the qual-

ty of performance in the followup tasks (the classic ques-ion of whether training translates to improved outcomes).xperiment 1 aims to establish the effect of warm-upxercises on followup tasks by looking at the global ques-ion (Hypothesis 5) and the main individual variables thatan be the reason for the measured effect. These variablesre experience of the operator, learning, fatigue, and theature of followup tasks. They are studied through theollowing hypotheses.

ypothesis 1: Warm-up is not dependent on theevel of experience of a surgeonhis hypothesis aims to study if the experience level of a sur-eon affects warm-up required. It is plausible that warm-upight only be required for less experienced surgeons or for

rocedures a surgeon is not familiar with. In sports, it has beenstablished that warm-up is necessary regardless of the level ofxperience of a player. We aim to investigate the relation be-ween experience and warm-up in surgery.

ypothesis 2: Warm-up and its intensity do notary with short-term practicet could be hypothesized that warm-up is required onlyuring training stages, and once a necessary level of learn-

ng has occurred for a particular task, there is no addi-

t“fpqsw

HeRdWem

HiFttp“t

i

HpTpablstp

MEBc

dTpcssect

mew

257Vol. 208, No. 2, February 2009 Kahol et al Preoperative Warm-Up Effect

ional requirement for warm-up. In essence, warm-up islearned” and does not provide benefit after an initial timerame. To test this hypothesis, experiments that track theerformance of subjects during several sessions are re-uired. It is a priori that no matter how experienced aurgeon might be, there is value added to performing aarm-up before every surgical procedure.

ypothesis 3: Warm-up is effective in reducingrrors resulting from fatigueesearch has shown that fatigue from night call and sleepeprivation can lead to decreased surgical proficiency.8

hether warm-up would help in alleviating some of theffects of fatigue is an intriguing question, with many prag-atic and positive implications.

ypothesis 4: Effectiveness of warm-up isndependent of the followup task to be performedrom a scientific point of view, this is an important investiga-ion that will reveal whether effects produced by generic prac-ice and exercises can carry over to more complex tasks; from aragmatic standpoint, this study will reveal whether a singlestandard” warm-up is sufficient or will multiple differentypes of warm-ups sessions need to be developed.

The final investigation lies in studying the nature of

Figure 1. Conceptual fra

ncrease in proficiency. f

ypothesis 5: Warm-up affects both cognitive andsychomotor skillshe study aims to establish if the increase in proficiency isurely psychomotor, helping increase dexterity, or does it alsoffect cognitive abilities of the surgeon. This will be measuredy including both attention and memory tasks, such as high-ighting series of pegs for a few seconds, which requires theubject be attentive enough to locate each peg and then, afterurning off the series, the subject must remember the order oflacement of the rings (see Methods section).

ETHODSarlier approval was obtained by the Institutional Reviewoard at Banner Good Samaritan Medical Center beforeonducting any experiment.

The present study focuses on a series of experimentsesigned to systematically explore the effect of warm-up.he experimental design extends the methodology pro-osed by Do and colleagues7 by using exercises in warm-upondition and followup condition. Subjects across varyingpecialties and experience levels (including senior traumaurgeons) were involved in multiple sessions performingxercises. These sessions were held in both precall and post-all condition, allowing for study of warm-up in alleviatinghe effect of night call on both cognitive and psychomotor

ork of the experiments.

atigue. In an additional experiment, different exercises

wqes

smtcCs(“pSdeiamstbopactts

twi

tttavtwrTtm

OTdcpttfdtsms

actu


ere used for warm-up and followup tasks. This alloweduantification of carry-over effect (generalization) of warm-upxercises. The study design and controls will be discussed inubsequent sections of this article.

A series of virtual simulations were developed that mea-ured both psychomotor and cognitive skills in a controlledanner. A simulation was designed for the virtual ring

ransfer task that is a part of validated basic laparoscopicourse offered by ProMIS Simulator and the Americanollege of Surgeons Certified Fundamentals of Laparo-

copic Surgery curriculum. In the virtual ring transfer tasksee Fig. 2), residents were tasked with grasping a series ofvirtual” rings and placing each on randomly highlightedegs on a board. The simulation was implemented using theensable haptic joystick, which allows for generation of threeegrees of force feedback in response to events in the virtualnvironment. OpenHL programming applied programmingnterface was used to design the simulation. The simulationllows for measurement of the tool tip in the virtual environ-ent. Additionally, while performing the simulated tasks,

ubjects wore the Cyberglove and Polhemus Liberty Trackerhat allowed for capture of hand movements (see Fig. 2). Theasic task involved 10 rings. After the participant places a ringn a highlighted peg, another peg is randomly chosen for thearticipant to put the ring on.This is repeated until all 10 ringsre correctly placed. This basic ring transfer task is a psy-homotor task used and validated in many simulators to honeool-manipulation skills. A cognitive error is marked for everyime the participant attempts to place a ring on the wrong peg,

Figure 2. Ring-transfer task implemented using the Senpick the ring and then place it on the highlighted peg. MoA subject using the system. It should be noted that thesimulator and can potentially be used for evaluation in

ignifying error in judgment by the subject. It should be noted b

hat the simulation does not allow placement of a ring on therong peg, and the participant is required to continue select-

ng pegs to put the ring on until the correct peg is chosen.This basic validated laparoscopic exercise was modified

o include cognitive variations, such as attention, visiospa-ial tracking, and intermodal transfer. Eight variations ofhe game were designed and are described in detail in Kaholnd colleagues9. These newly developed simulations werealidated through controlled experiments that showed thathe developed exercises can suitably replicate the realisticork environment of surgeons, offering exercises that

equire both psychomotor and cognitive dimensions.able 1 summarizes the exercises and cognitive modalityhey address. For detailed definitions on the cognitiveodality please see Kahol and colleagues.9

bjective proficiency evaluation measureshis section on methodology defines the hand-movementata-capture system and variables, and the tool-movementapture system and variables. For measuring laparoscopicroficiency, we used a combination of hand movement andool movement. Tool movement measured as movement ofhe tool tip in virtual environment is a validated measureor surgical proficiency.10 Kahol and colleagues11 intro-uced hand movement measured through Cyberglove andhe Polhemus Liberty Tracker as an effective measure ofurgical proficiency. Both tool movement and hand move-ent are representative of economy of motion, overall

moothness in execution, and their construct validity has

e Haptic joystick. Simulation requires the participant toent of the tool and hands are measured in the process.d-movement data-capture gloves can be worn with anyal operations.

sablvemhan

een established. Tool acceleration was calculated for the

etttcsnuT

H

TCaojmtcpaTmm(tocgwptbLiicmmeg

ETcsdgtscwwwcwge

EFPOGtt2e

EEwrswTvviamww

T

Cp

SWMPI


ntire duration of a task and normalized in a range of 0hrough 1. Smoothness of tool movement as predictedhrough this measure is 1 when overall acceleration is closeo 0. This is generally the case in well-executed motion withontrolled accelerations. On the other hand, jerky motionshow higher normalized acceleration and lower smooth-ess. The smoothness of tool movement was calculatedsing the following formulas:ool-movement smoothness

� 1 � normalized (tool acceleration) (1)

and-movement smoothness

� 1 � normalized (wrist acceleration) (2)

he data-capture setup shown in Figure 2 depicts the wirelessyberglove glove and the Polhemus Liberty tracker. The wrist

cceleration is calculated through tracking of the sensor placedn the wrist. With regard to tool-movement smoothness,erky hand motions lead to less smoothness, while controlled

ovements lead to increased smoothness. For every simula-ion exercise, time required to complete a task is recorded;ognitive errors are recorded as the number of times the ring islaced on the wrong peg. These four objective measures arelso supplemented with gesture-level proficiency measure.ask decomposition has emerged as a validated method toeasure surgical proficiency.11 In this approach, hand move-ent or tool movement is decomposed into smaller gestures

such as in, out, grasping and rotation). Each individual ges-ure is analyzed and, based on its similarity to the optimalccurrence of a gesture as determined by setting a benchmarkriteria from an expert group’s performance of that gesture, isiven a proficiency rating. For this article, hand movementas used for task decomposition. The algorithm for this pur-ose was described by Kahol and colleagues11 and was showno correlate highly with subjective proficiency ratings obtainedy surgeon teachers. The scoring method used is a standardikert scale between 0 and 10 for an entire exercise. Zero

mplies least proficiency in accomplishing the task and 10mplies highest proficiency.This measure is estimated throughombination of time elapsed and kinematic analysis of handotion. These five measures (gesture-level proficiency, hand-ovement smoothness, tool-movement smoothness, time

lapsed, and cognitive errors) provide a broad framework for

able 1. Exercises and the Primary Cognitive and Psychomo

ognitive andsychomotor faculty

Sensorimotorcoordination

Two-dimensionaltracking

T

ensorimotor coordination X Xorking memory Xovement planning X

reparatory attention Xntermodal transfer

lobal assessment of proficiency. d

xperimental designwo experiments were performed. The first experimentonsisted of eight sessions and the second experiment con-isted of a single session of data capture. Experiment 1 wasesigned to evaluate the overall effect of warm-up on sur-ical proficiency and its relation to experience level, short-erm practice, fatigue level, and cognitive and psychomotorkills. In Experiment 1, every subject acted as their ownontrol, wherein their performance was measured duringarm-up and then during followup trials. Experiment 2as designed to test if warm-up conducted in basic skillsas able to effect performance in a complex surgical task. A

ontrol group of subjects who performed a complex taskithout warming up was compared with an experimentalroup of subjects who warmed up with simple laparoscopicxercises before performing a complex task.

xperiment 1: Subjectsorty-six surgeons participated in the study. FourteenGY1 OB/GYN and general surgery residents, 10 PGY2B/GYN and general surgery residents, 11 PGY3 OB/YN and general surgery residents, and 10 attending

rauma surgeons were involved in the experiments. Nine-een women and 27 men comprised the subject pool, with1 subjects of OB/GYN specialty and 25 subjects of gen-ral surgery specialty.

xperiment 1: Designach participant was involved in eight sessions during 4eeks. Four of these sessions were held precall before the

esidents performed their night call and the remaining fouressions were held postcall. In each session, three exercisesere performed after filling in the fatigue questionnaire.hese three exercises were randomly chosen from the eightariations of the ring transfer exercises described here pre-iously. Each exercise was repeated two times. The firstteration was marked as warm-up trial and the second iter-tion was marked as the followup exercises. This experi-ent permitted an in-subject design wherein every subjectas their own control. During each session the subjectore Cyberglove and the Polhemus Liberty tracker on their

Faculty EngagedExercises

dimensionalacking Orientation

Preparatoryattention

Workingmemory

Visiohaptictransfer

X X X X XX X XX X XX X X X X

X X

tor

hree-tr

ominant hand. In the first session, the glove was cali-

bscmmadpeeti

tmMmpwt

wds

ww

iPfTfs

gflcgwi

faHgpp


rated to a participant’s hand and the calibration wastored. For every session performed by a participant, thealibrated glove was used to accurately record hand move-ents and wrist movements. Each session lasted approxi-ately 15 to 20 minutes. Exercises in the precall condition

nd the postcall condition were not matched. This wasone to control for learning effect that could have beenroduced for a particular exercise. During the course ofight sessions, each subject performed six iterations of eightxercises as described by Kahol and colleagues.9 Three ofhese iterations were for warm-up and the remaining threeterations were marked as followup condition.

Data captured in the eight sessions on hand movement,ool movement, time elapsed, and errors were processed as perethodology defined in the Objective Proficiency Evaluationeasures section to calculate the five objective proficiencyeasures for each exercise iteration. Statistical analysis was

erformed to compare performance of surgeons duringarm-up and followup exercises. ANOVA was used to study

he difference between warm-up and followup conditions.In the first level of analysis, all trials performed during

arm-up were grouped together and compared with all trialsuring followup. This analysis provided statistical data totudy the global differences between followup and

Figure 3. ANOVA plots comparing proficien

arm-up conditions to investigate questions of whetherarm-up improves performance of a subsequent task.In the second level of analysis, data were grouped accord-

ng to experience level of the surgeons (PGY1, PGY2,GY3, and senior surgeons) and the groups’ warm-up and

ollowup exercise iterations were compared using ANOVA.his analysis allowed study of Hypothesis 1 regarding dif-

erences of the effect of warm-up on surgical proficiency ofurgeons who had different levels of experience.

In the third type of analysis, exercise iterations wererouped according to sessions in which they were per-ormed. Overall change in proficiency measures in fol-owup condition and warm-up condition was plotted as aurve across the iterations of exercises and sessions. Thisraph helped provide insight into Hypothesis 2, abouthether the need for warm-up alters with short-term learn-

ng of surgical skills and exercises in eight sessions.The fourth type of analysis grouped precall warm-up and

ollowup sessions and compared them to postcall warm-upnd followup sessions. This comparison allowed for studyingypothesis 3, whether warm-up has a positive effect on sur-

ical proficiency during postcall fatigue state, thus improvingerformance of fatigued surgeons. In previous studies, we haveerformed detailed analysis on fatigue and its effect on profi-

cy in warm-up and followup condition.

clrcwsaCttp

aptesfsa

EEaepti

gt

ERetptg(iCadqffwiwgfwnaww


iency.9 Overall methodology is detailed in Kahol and col-eagues,9 but a few key points about the methodology andesult need to be elucidated. Exercises in the precall and post-all session were not matched and postcall session exercisesere chosen from the pool of exercises not used in precall

ession. Overall, results showed that a surgeon’s cognitive skillsnd psychomotor skills were negatively affected by fatigue.ognitive skills were more substantially affected. Results from

he fatigue experiments showed a very high correlation be-ween reported levels of fatigue in the questionnaire and actualerformance.

In the fifth type of analysis, warm-up and followup iter-tions of sensorimotor coordination exercise were com-ared with warm-up and followup iterations of all the otherypes of exercises grouped together. Table 1 shows that allxercises, except sensorimotor coordination exercise, had aubstantial cognitive component. This analysis was per-ormed for Hypothesis 4 to reveal if warm-up had a con-iderable effect on psychomotor skills and cognitive skillsnd to compare the performance differential.

xperiment 2: Subjectsight PGY2 and four PGY3 residents (six OB/GYN residentsnd six general surgery residents) participated in the secondxperiment designed to study effect of warm-up on surgicalroficiency in complex surgical procedures. Participants wereaken from the same pool as in Experiment 1, but this exper-ment was performed after completion of Experiment 1. Both

Figure 4. ANOVA plots comparing proficiency in warm-u
to experience level. SS, senior surgeon.
roups had equal exposure to and education of the ring-ransfer exercises before they began Experiment 2.

xperiment 2: Designesidents were divided into two groups of six residentsach. The control group performed only the surgical elec-rical diathermy task on the ProMIS simulator (with noreliminary warm-up exercises). The surgical diathermyask mimics electrosurgery skills required of excision of theall bladder during laparoscopic cholecystectomy. The taskwhich has been validated) requires the surgeon to performncision, removal of the cyst using diathermy, and suturing.omplete details of the operation are available from Bass

nd colleagues.13 Traditionally, the operation is performeduring earlier stages of the residency program and re-uires medium skill level. The experimental group per-ormed the simulated surgical diathermy task after per-orming warm-up exercises. Three warm-up exercisesere randomly chosen from the eight exercises defined

n Table 1 and by Kahol and colleagues.9 Each exerciseas repeated twice by the experimental group. The sur-eons wore Cyberglove and Ascension Tracker while per-orming the diathermy simulation. The tool movementas available from ProMIS simulator but no direct cog-itive errors could be recorded in the ProMIS simulators it does not allow for any such recording. This analysisas performed for Hypothesis 5 to determine whetherarm-up improves performance independent of the

ndition and followup condition data grouped according
p co

top

RFpwswtchse

HsFsp

clso

agaTosTcpl

HvFtopFdslncma


ype of followup task (or procedure). The performancef the control group and experimental group was com-ared on the diathermy simulation task using ANOVA.

ESULTSigure 3 shows the plot for comparison of the objectiveroficiency measures evaluated in the experiment inarm-up and followup condition. Each of the measures

howed a statistically significant difference betweenarm-up and followup conditions (p � 0.005). The

ime required for completion of tasks and number ofognitive errors decreased and gesture-level proficiency,and-movement smoothness, and tool-movementmoothness increased considerably. Overall reduction inrrors was 33%.

ypothesis 1: Warm-up is not dependent on level ofurgeon experienceigure 4 shows plots of the five objective proficiency mea-ures, grouped according to level of experience of thearticipants.Plots show that before warm-up, statistically significant

orrelation did not appear between the groups. In the fol-owup condition, each of the groups show a statisticallyignificant difference (p � 0.001). Consistent with resultsf overall analysis, the time required for completion of tasks

Figure 5. ANOVA plots comparing change in pr

nd number of cognitive errors decreased, althoughesture-level proficiency, hand-movement smoothness,nd tool-movement smoothness increased considerably.he percentage reduction in errors was calculated for eachf the groups. PGY1, PGY2, PGY3, and resident surgeonshowed 31%, 30%, 44%, and 29% reduction in errors.hese reductions show that each subject group showed

onsistent and substantial improvement supporting the hy-othesis that warm-up is needed for subjects at different

evels of expertise.

ypothesis 2: Warm-up and its intensity do notary with short-term practiceigure 5 shows improvement in objective proficiency inhree sessions of the exercises. Data were grouped in termsf each session of the exercise performance and the im-rovement in proficiency was calculated for each session.igure 5 shows that improvement in proficiency intro-uced because warm-up is shown in all three sessions con-istently and does not increase or decrease with short-termearning. Only gesture proficiency showed a statistically sig-ificant difference (p � 0.005) between the three sets of exer-ise performance, but there was not a continuous improve-ent (eg, learning curve), but rather simply differences in

mount of improvement from one set to the next.

oficiency in three iterations of exercises.

HrFwipawjwfocemtp(alfmpdi

apon

HiFc2fatws0Twpm

HpF

wup


ypothesis 3: Warm-up is considerably effective ineducing errors resulting from fatigueigure 6 shows plots of objective proficiency measures inarm-up and followup conditions with data being grouped

nto precall and postcall. Data shows that warm-up doesositively affect proficiency during postcall condition char-cterized by fatigue and sleep deprivation12 to a pointhere the warm-up returns performance in a fatigued sub-

ect to the same level as a precall subject who has notarmed up. Proficiency obtained in the postcall fatigue

ollowup condition is less than the proficiency measuresbtained in the precall followup group. These results indi-ate that warm-up does help in countering some of theffects of fatigue and sleep deprivation by returning perfor-ance in the fatigued subject to the equivalent of a nonfa-

igue baseline (with no warm-up). It appears that fatiguerevents enhancing beyond baseline performance of a freshprecall) subject. This interesting phenomenon deservesdditional, more deliberate investigations than these pre-iminary trials. It is important to emphasize that even in aatigued subject, every single parameter improved (hand-otion smoothness improved 198%, tool motion im-

roved 23.8%, gesture proficiency improved 103%, timeecreased by 32%, and errors decreased 34%), demonstrat-

Figure 6. ANOVA plots comparing proficiency in warm-up and follo

ng that substantial performance improvement can be p

chieved even in a fatigued individual, and that the im-rovement can return performance to approximately thatf the baseline performance of a rested individual who hasot warmed-up.

ypothesis 4: Effectiveness of warm-up isndependent of the followup task to be performedigure 7 shows the objective proficiency measures (exceptognitive errors) for the two groups involved in Experiment. The control group that did not warm-up before per-orming the diathermy task on the ProMIS simulatorchieved inferior proficiency scores when compared withhe experimental group. The experimental group thatarmed up with the exercises designed for this experiment

howed statistically significant better performance (p �.01) except for time elapsed where p � 0.08 was noted).his demonstrates, at least preliminarily, that simplearm-up tasks can be generalized for more than one type ofrocedure, but this trial is quite limited and needs muchore exploration.

ypothesis 5: Warm-up affects both cognitive andsychomotor skillsigures 8 and 9 show, respectively, the plots of objective

condition. Data were grouped into precall and postcall condition.

roficiency measures (means) in warm-up and followup

ctggsotefnats

DRscseGt

siwwr

fu0ts0tsbeirpsW

xper


onditions for the sensorimotor coordination (psychomo-or skills) exercise and the cognitive exercises (grouped to-ether), although Figure 10 shows t-statistic measure foresture proficiency and cognitive error for each exerciseeparately. The t-statistic value represents the change inbjective proficiency in the warm-up and followup condi-ion and shows how warm-up affects skills associated withach of the exercises separately. The highest value was notedor orientation exercise, although the lowest values wereoted for the memory exercises with four pegs to remembernd put the ring on. These results support the hypothesishat warm-up does lead to improvement in both types ofkills.

ISCUSSIONesults of this study provide substantive evidence that

hort-term practice (warm-up) for 15 minutes with exer-ises designed to target both psychomotor and cognitivekills that are involved in surgical procedures can greatlynhance skill proficiencies during a followup procedure.esture-level proficiency, movement smoothness, and

Figure 7. ANOVA plot comparing performance of control group and e

ool-movement smoothness are favorably enhanced with f

imple practice. Errors are substantially reduced, whichs arguably the most important positive effect ofarm-up exercises. Time required to complete a task,hich is one important indicator of efficiency, is also

educed substantially.Participants of varying experience levels all can benefit

rom warm-up. It is important to note that before warm-p, there was no statistically significant difference (p �.69) among the different levels of experience (althoughhere was a tendency for the more senior levels to have alightly higher, though not statistically significant (p �.04), baseline performance capability in the hand- andool-movement tasks). After warm-up exercises, there areubstantial differences within each of the expertise levelsetween warm-up and followup tasks, with approximatelyqual improvement for all levels of expertise. Because themprovement is so strong among all levels (25% to 40%eduction in error), it is fair to conclude that warm-uprovides a quantifiable performance improvement for sub-equent tasks regardless of experience level or expertise.

hat cannot be concluded is whether improved task per-

imental group (warmed-up) on diathermy task on ProMIS simulator.

ormance will transfer to improvement in the operation

rtsm

wtsftusbtBfmIiwil

citt

tfs

fttnvbsmpbmide

cpaapop

tion a


oom (virtual reality to operating room) and, more impor-ant, whether improvement in the operating room will re-ult in improved overall clinical outcomes. These are manyore longterm studies that must be pursued.Another possible explanation for the favorable effect of

arm-up exercises on followup exercises could be “short-erm familiarization effect.” The key element in clarifyinguch an explanation requires that warm-up exercises andollowup exercises are exactly the same. When the sameask is performed before and after warm-up, the resultsnequivocally demonstrate improvement. Results alsohow that the group that warmed-up with simple exercisesefore a complete procedure performed considerably betterhan the control group that directly performed the task.ecause improvement was demonstrated in both the same

ollowup task and a different followup task, the improve-ent is not a result of familiarization or the learning curve.

n a similar fashion, the results demonstrate that themmediate onset (first trial) of improvement from thearm-up exercises implies that this rapid improvement

n skill cannot be attributed to familiarization effect orearning effect.

Analysis of data collected during precall and postcallondition indicated that warm-up can improve proficiencyn the fatigued individual, and can bring their performanceo a level equivalent to their baseline nonfatigued condi-

Figure 8. ANOVA plots comparing proficiency in warm-up condi

ion. Warm-up cannot take a fatigued person and improve r

heir performance to their potential best performance (ie,ollowup performance after warm-up in a nonfatiguetate).

In trying to understand the reasons for improved per-ormance, analysis of the portfolio of tests demonstrateshat warm-up exercises enhance performance not onlyhrough basic psychomotor skills (sensorimotor coordi-ation), but also by intensifying cognitive faculties. Pre-ious research suggests that four concurrent tasks mighte close to the limit of working memory,14 beyond whichkill deteriorates; this was also suggested by the studiesentioned in this article that demonstrated a decrease in

erformance when the subject was required to remem-er four (rather than three) sequential pegs. Improve-ent caused in proficiency might only be a result of

mprovements in psychomotor skill (sensorimotor coor-ination) and some cognitive skills, but not a result ofxpanding or enhancing the available working memory.

Additional analysis indicates that warm-up exercisesan allow for surgeons to achieve cognitive arousal cou-led with exercises invoking motor circuits in the brainnd warming-up of hand and shoulder muscles tochieve sensorimotor coordination. This is based on im-rovements shown in all types of exercises, including thenes with substantial cognitive components. This hy-othesis is consistent with findings in the sports domain15 as

nd followup condition for sensorimotor coordination exercise.

elated to warm-up. Such effects last until a followup task is

csewgw

nianpw

(wmtstwmtadc

mioteatft

eppepdd

poTtbc

ion a


ompleted. Warm-up decrement has been shown to occur inports domain when a period of rest is introduced betweenxercises and the task. Future studies will investigatehether such effects are also a factor in laparoscopic sur-ery, or whether waiting for a prolonged time afterarm-up negates the benefit of warm-up.From a very pragmatic standpoint, it is encouraging to

ote that improvement in performance with warm-up isndependent of the type of task that follows (Experiments 1nd 2). Should additional research confirm these prelimi-ary results, it will be possible to develop a few, simplereoperative warm-up exercises that will be sufficient forhatever procedure will follow.Although the focus of the study has been the effect

both cognitive and psychomotor) of a preoperativearm-up caused by physical stimulation (ie, perfor-ance of a specific task), there is a possible confound in

he study—the psychological anticipation to perform aurgical procedure. This confounding factor might behat surgeons do not require such preoperative warm-uphen conducting actual operations because the environ-ent and expectation lead to required cognitive arousal,

uning of sensorimotor coordination, and preparatoryttention. Additional studies need to be designed to ad-ress the issue of psychological anticipation in a real

Figure 9. ANOVA plots comparing proficiency in warm-up condit

linical environment.

The basic nature of these laboratory-based experi-ents will need confirmation and validation by other

ndependent sources, and determination of whether lab-ratory results will transfer to the clinical domain. Fu-ure work in this domain would include conductingxperiments on the effect of pretrial warm-up beforectual operation. The current study also did not explorehe effect of different types of exercises on proficiency inollowup tasks. This issue will be explored through con-rolled experiments.

There are a few fundamental precepts that havemerged. First and foremost, the results confirm the ariori impression and observations of other high-riskrofessions that performing a preoperative warm-up ex-rcise before a surgical procedure can improve operativeerformance—in terms of more efficient motion, re-uced operative time (procedure efficiency), and re-uced errors (patient safety).Second, improvement appears to occur from both

sychomotor and cognitive skill enhancement, althoughverall working memory does not seem to be increased.his improvement is completely independent of exper-

ise of the surgeon, and the improvement will occurefore every operative procedure—there is no learningurve after which warm-up is not helpful.

nd followup condition for cognitive exercises grouped together.

Third, preoperative warm-up can improve performance

imtdaf

sdtballrw(mewab

bj

“htnwttbci

A

SAA

exerc


n a fatigued surgeon, although the performance improve-ent will only be to the surgeon’s baseline level and not to

heir optimal potential level. This finding will require ad-itional validation and investigation into other areas, suchs a longer warm-up period and quantifying the amount ofatigue . . . period and quantifying . . . fatigue.

There are substantial policy implications from these re-ults. The most important one is—if additional indepen-ent validation studies confirm these findings—whetherhere should be a required preoperative warm-up periodefore every surgical procedure (in the name of efficiencynd patient safety). Will there be a specific, quantifiableevel of proficiency the surgeon should obtain (on a simu-ator) before beginning a surgical procedure? Will it beequired that the surgeon documents that the preoperativearm-up was complete before starting the procedure

thereby documenting that the surgeon was at their opti-al performance)? Will it be necessary to require that op-

rating suites have readily available access to preoperativearm-up stations? If a surgeon does not warm-up and an

dverse event occurs, will the surgeon be liable? These are

Figure 10. Comparison of t-statistic value for each

ut a few of the possible implications, all of which need to D

e thoroughly investigated and discussed before any knee-erk regulations are put into place.

Once again, health care is beginning to confront acommon sense” practice that is routine in many otherigh-stakes professions. Results of this study are in-ended to begin to form the scientific foundation for aew approach to surgical practice: the preoperativearm-up. Just as the “see one, do one, teach one” tradi-

ion of surgical education is yielding to quantitativeraining and assessment of technical skills and criterion-ased benchmarks to demonstrate proficiency, so tooan preoperative warm-up become a new standard formproving operative performance and patient safety.

uthor Contributions

tudy conception and design: Kahol, Satava, Smith, Ferraracquisition of data: Kahol, Satava, Smith, Ferraranalysis and interpretation of data: Kahol, Satava, Smith,

Ferrara

ise for gesture-level proficiency and cognitive error.

rafting of manuscript: Kahol, Satava, Smith, Ferrara

R

1

1

1

1

1

1


EFERENCES

1. The Oxford English Dictionary, 2nd ed, 1989. Available at: http://ed200ed.com/cgi/entry/00280931?single/2&query_type/word&queryword/warm-up&first/1&max_to_show/10. Accessed April4, 2008.

2. Stretching I. Stretching handbook. Available at: http://www.thestretchinghandbook.com/archives/warm-up.php. AccessedNovember 8, 2007.

3. Anshel MH, Wrisberg CA. Reducing warm-up decrement in theperformance of a tennis serve. J Sport Exerc Psychol 1993;15:290–303.

4. Valck E, Cluydts R, Pirrera S. Effect of cognitive arousal on sleeplatency, somatic and cortical arousal following partial sleep de-privation. J Sleep Res 2004;13:295–304.

5. Aggarwal R, Moorthy K, Darzi A. Laparoscopic skills trainingand assessment. Br J Surg 2004;91:1549–1558.

6. Hall JC, Ellis C, Hamdorf J. Surgeons and cognitive processes.Br J Surg 2002;90:10–16.

7. Do AT, Cabbad M, Kerr A, et al. A warm-up laparoscopic exer-cise improves the subsequent laparoscopic performance of ob-gyn residents: a low-cost laparoscopic trainer. JSLS 2006;10:297–301.

8. Leyba M, Kahol K, Smith M, et al. The effect of fatigue on

psychomotor and cognitive skills of surgical residents. Presentedat the Annual Meeting of American College of Surgeons, LasVegas, NV, April 2007.

9. Kahol K, Leyba MJ, Deka M, et al. Effect of fatigue on psy-chomotor and cognitive skills. Am J Surg 2008;195:195–204.

0. Gallagher AG, Richie K, McClure N, McGuigan J. Objectivepsychomotor skills assessment of experienced, junior and novicelaparoscopists with virtual reality. World J Surg 2001;25:1478–1483.

1. Kahol K, Balasubramanium VN, Krishnan CN, et al. Measuringmovement expertise in surgical tasks. In: 14th Annual ACMInternational Conference on Multimedia. Santa Barbara, CA:IEEE; 2006:719–722.

2. Deaconson TF, O’Hair DP, Levy MF, et al. Sleep deprivation andresident performance. J Am Med Assoc 1988;260:1721–1727.

3. Bass LS, Treat MR. Laser tissue welding: a comprehensivereview of current and future. Lasers Surg Med 1995;17:315–349.

4. Cowan N. The magical number 4 in short-term memory: areconsideration of mental storage capacity. Brain Behav Sci2000;24:87–185.

5. Anshel MH, Wrisberg CA. ED283791: The effect of arousal
and focused attention on warm-up decrement. 1987.
http://www.ed200ed.com/cgi/entry/00280931?single/2%26query_type/word%26queryword/warm_up%26first/1a%20max_to_show/10



http://www.thestretchinghandbook.com/archives/warm_up.php

http://www.thestretchinghandbook.com/archives/warm_up.php

effect of short-term pretrial practice on surgical proficiency ......effect of short-term pretrial...

Documents