JOURNAL OF ENDOUROLOGYVolume 19, Number 3, April 2005© Mary Ann Liebert, Inc.

Can Video Games be Used to Predict or ImproveLaparoscopic Skills?

BRADLEY H. ROSENBERG, M.D.,1 DOUGLAS LANDSITTEL,2 and TIMOTHY D. AVERCH, M.D.1

ABSTRACT

Background and Purpose: Performance of laparoscopic surgery requires adequate hand–eye coordination.Video games are an effective way to judge one’s hand–eye coordination, and practicing these games mayimprove one’s skills. Our goal was to see if there is a correlation between skill in video games and skill inlaparoscopy. Also, we hoped to demonstrate that practicing video games can improve one’s laparoscopicskills.

Subjects and Methods: Eleven medical students (nine male, two female) volunteered to participate. On day1, each student played three commercially available video games (Top Spin, XSN Sports; Project Gotham Rac-ing 2, Bizarre Creations; and Amped 2, XSN Sports) for 30 minutes on an X-box (Microsoft, Seattle, WA) andwas judged both objectively and subjectively. Next, the students performed four laparoscopic tasks (objecttransfer, tracing a figure-of-eight, suture placement, and knot-tying) in a swine model and were assessed fortime to complete the task, number of errors committed, and hand–eye coordination. The students were thenrandomized to control (group A) or “training” (i.e., video game practicing; group B) arms. Two weeks later,all students repeated the laparoscopic skills laboratory and were reassessed.

Results: Spearman correlation coefficients demonstrated a significant relation between many of the pa-rameters, particularly time to complete each task and hand–eye coordination at the different games. Therewas a weaker association between video game performance and both laparoscopic errors committed andhand–eye coordination. Group B subjects did not improve significantly over those in group A in any measure(P �0.05 for all).

Conclusion: Video game aptitude appears to predict the level of laparoscopic skill in the novice surgeon. Inthis study, practicing video games did not improve one’s laparoscopic skill significantly, but a larger studywith more practice time could prove games to be helpful.

372

INTRODUCTION

COINCIDING WITH THE INTRODUCTION and expan-sion of laparoscopy in the past 20 years has been a surge

in the interest in, and the complexity of, commercially avail-able video games. Many of today’s surgeons (especiallyyounger surgeons) grew up playing video games as a hobby.Casual banter among surgeons has suggested a link betweenvideo game playing and surgical skills, as most believe videogames can sharpen one’s hand–eye coordination. Intuitively,therefore, it seems that practicing video games could improveone’s surgical acumen. Much of today’s surgery has becomeminimally invasive, either endoscopic or laparoscopic, and

both types of procedures seem to mimic video games in manyways.

With the recent introduction of laparoscopy to most surgicalresidencies, a need for skill training has developed. Obviously,programs would prefer their residents to begin their laparo-scopic experience with less-challenging cases, but urology hasbeen hampered in laparoscopic skills acquisition by the lack ofa common “beginner” operation like cholecystectomy in gen-eral surgery. Therefore, programs seek out different models onwhich their residents may acquire basic laparoscopic skills.Inanimate trainers have been shown by other authors to improvehand–eye coordination, which could translate into greater lap-aroscopic skills.1 Different types of trainers are available, from

1Departments of 1Urology and 2Biostatistics, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania.

cadaver models to elaborate virtual reality (VR) simulators.These models aid the novice laparoscopist in acquiring the skillshe or she may find awkward compared with open surgery2–5

such as compensating for the diminished tactile feedback, op-erating on a fulcrum, handling laparoscopic instruments er-gonomically, and working in three dimensions while looking ata two-dimensional image.

It has recently been shown that video-game playing can im-prove visual skills, but to our knowledge, no one has looked atvideo games specifically and how they correlate with laparo-scopic surgical experience.6 The goal of our study was to takestudents with no surgical experience and see if video-game acu-men correlated with performance in laparoscopic surgery.Moreover, we sought to discover if practicing video gamescould positively impact one’s laparoscopic skill. If game play-ing does translate into laparoscopic success, it could be an im-portant resource for training programs. Not only could they bestchoose surgical candidates on the basis of video game-playinghistory, but also the games could serve as an excellent inex-pensive, readily available skill exercise.

SUBJECTS AND METHODS

Subjects

Eleven students (nine male, two female) with an average ageof 25.7 years (range 21–32 years) were recruited from the Uni-versity of Pittsburgh School of Medicine to take part in this In-stitutional Review Board-approved study. They completed aquestionnaire inquiring about their handedness, career aspira-tions (i.e., medical or surgical), experience with video games,and experience with laparoscopy. All but one was right-handed.Six aspired to surgical careers; the remainder were not sure. Allstudents had completed large-animal training as required by theInstitutional Animal Care and Use Committee (IACUC). Thestudy took place on 2 days 2 weeks apart, each day consistingof two sessions, one video game and one laparoscopic. Noneof the students had significant laparoscopy experience, and nonehad played the games used in this study.

Day 1: Video game session

After an introductory session to master the joystick and re-ceive instruction on the basics of each game, the students spentapproximately 30 minutes playing three commercially availablevideo games (Top Spin, XSN Sports; Project Gotham Racing 2(PGR2), Bizarre Creations; Amped 2, XSN Sports) on an X-box(Microsoft, Seattle, WA). Game settings were standardized to be-ginner levels. Each student played one set of tennis in Top Spin,one “Street Race” and one “Cone Challenge” on PGR2, and onetrip downhill on Amped 2. Their performances were evaluatedboth objectively by the computer (i.e., wins, crashes, time, points,etc.) and subjectively by an observer (either an attending urolo-gist or a urology resident), who graded the student 1 through 5on overall hand–eye coordination (1 � worst, 5 � best).

Day 1: Laparoscopic skills lab (session 1)

Next, the students proceeded to the animal laboratory andwere familiarized with laparoscopic instrumentation, suturing,

and knot-tying. They then performed a series of laparoscopictasks on one of two anesthetized pigs, as approved by theIACUC. The four tasks (object transfer, tracing a figure-of-eight, suture placement, and knot-tying) are described in detailin the Appendix. For each task, the students were evaluated forthe time it took to complete the task, the number of errors made,and hand–eye coordination (1 � worst; 5 � best) by either anattending urologist or a urology resident.

Day 2

The students were randomized into control (Group A: N �6) and training (Group B; N � 5) arms. Students in group Bwere asked to practice playing any type of video game over the2-week hiatus between study dates; the amount of time theyspent was recorded. These students returned on day 2, and eachplayed 2 hours worth of the same X-box video games theyplayed on day 1. Group A students were asked to refrain fromplaying any video games over the 2-week period. All 11 stu-dents returned to the laparoscopic skills laboratory (session 2),repeated the same four tasks, and were graded.

Statistical analysis

To determine if a relation exists between video-game apti-tude and baseline laparoscopic skills, the Spearman correlationcoefficient was utilized. For the video games, the analysis wasbased on the assessment of hand–eye coordination in Top Spin,Amped 2, and the “Cone Challenge” in PGR2 and place fin-ished in “Street Race.” We calculated correlations between eachof these summary measures and each quantitative variable fromthe four laparoscopic tasks. To look for a training effect on skilllevel between groups A and B, P values were calculated usingthe Wilcoxon rank-sum test. Significance for all variables wasdetermined by a P value of �0.05.

RESULTS

Video game experience was minimal, moderate, and exten-sive in 3, 5, and 3 students, respectively. Spearman coefficientsdemonstrated a statistically significant correlation betweenvideo game experience and video-game performance in two ofthe four exercises (Cone Challenge and Amped 2).

Association between video-game performance andlaparoscopic skills

Table 1 summarizes the correlations between video game andlaparoscopy performance at baseline. The denoted Spearmancorrelations (*) were statistically significant at P �0.05 or mar-ginally significant at P � 0.10 (†). The laparoscopic task mea-surement that best correlated with video game performance wastime to complete the objective. Five significant correlationswere identified between video games and time, along with twomore that were marginally significant. This difference was morepronounced in the simpler tasks (object transfer and figure-of-eight). The more-complex tasks (suturing and knot-tying)showed little difference (only one significant parameter). Interms of number of errors and hand–eye coordination, for alltasks and all games, there were only two correlations that weremarginally significant.

VIDEO GAMES AND LAPAROSCOPY 373

Effect of video-game training on changes inlaparoscopic skills

The baseline characteristics of groups A and B showed nosignificance in video-game performance or laparoscopic skill(Table 2; P values all �0.20). Students in group B spent an av-erage of 6.2 hours “training” (i.e., practicing video games) overthe 2-week period (range 2.5–9 hours). No student in group Aplayed any video games during that time. Table 2 lists the re-sults of the laparoscopic skills laboratory broken down by groupand session. In the group as a whole, the time taken to com-plete each task improved on average from session 1 to session2, but this change was significant only for object transfer andknot-tying. The number of errors committed and the hand–eyecoordination rating for the whole group did not change signif-icantly between the two sessions. The mean change for eachskill was calculated by subtracting the baseline measure fromthe follow-up measure. A positive number thus represents anincrease in the measure from baseline to follow-up; a negativenumber represents a decrease from baseline to follow-up. GroupB improved no more than group A (P � 0.25) in any measure.Only one measure showed significance, where the control groupactually showed an improvement over the training group inknot-tying errors committed.

DISCUSSION

With the current widespread use of laparoscopic and endo-scopic surgery, attention throughout the surgical literature hasturned to methods of training surgeons in these techniques. Cer-tainly, laparoendoscopic surgery does require a skill set muchdifferent than that of open surgery, and many surgeons may not

feel innately comfortable with these. Subramonian and associ-ates7 showed that laparoscopic-skill acquisition is in fact moredifficult than that for open surgery, further validating the needfor training. Laparoscopic surgery requires a surgeon to thinkin three dimensions while looking at a two-dimensional image.Additionally, he or she is operating on a fulcrum with limitedtactile feedback. Because these skills are so foreign to many,trainers have been employed in an attempt to reduce the riskassociated with initial laparoendoscopic surgical experience.

Trainers not only help with skills acquisition in current sur-gical trainees but also may aid programs in selecting residentcandidates. On that topic, Gettman and colleagues8 found thatbasic human performance resources could be used to predictlaparoscopic performance. In other words, a surgeon’s innatelaparoscopic ability could be predicted by measuring objectivessuch as visual information-processing speed, upper-extremitysteadiness, and isometric strength.

Many laparoendoscopic trainers are now available, includ-ing everything from cadavers to both low- and high-fidelitybench models and VR simulators. A ureteroscopy simulator(URO Mentor; Simbionix Ltd, Lob, Israel) has been found toresult in rapid skills acquisition for novice trainees.2 A VR sim-ulator (Minimally Invasive Surgical Trainer in Virtual Reality[MIST-VR]) has proven capable of distinguishing between ex-perienced surgeons and novices.3 It likewise provides for rapidacquisition of skills in many beginners. Matsumoto and cowork-ers4 determined that any hands-on training, whether it was ona low- or a high-fidelity model, improved performance morethan a simple didactic session. So it does appear conclusive thatstudents quickly improve their skills when operating on mod-els. But does that necessarily translate into an actual operativeprocedure? Few studies have looked at trainers and how theyimpact eventual live surgery, and so far, the results have been

ROSENBERG374

TABLE 1. SPEARMAN CORRELATIONS BETWEEN VIDEO GAME AND LAPAROSCOPY PERFORMANCE

Video game

Laparoscopy Coneoperation Top Spina Street Raceb Challengea Amped 2a

Object transferTime needed �0.73* 0.58† �0.53† �0.68*No. errors �0.005 0.13 0.12 0.39Hand–eye* �0.11 0.09 �0.23 �0.38

Figure-of-eightTime needed �0.64* 0.47 �0.50 �0.60*No. errors 0.0 0.32 0.16 0.0Hand–eye* �0.27 0.16 �0.51† �0.64†

SutureTime needed �0.27 0.23 �0.13 �0.36No. errors �0.40 0.47 �0.38 �0.49Hand–eye* 0.32 �0.45 0.42 0.54†

Knot tyingTime needed �0.40 0.26 �0.37 �0.63*No. errors 0.10 0.27 0.0 �0.20Hand–eye* 0.25 �0.45 0.04 0.13

aSkill assessed via hand–eye coordination rating: 1 (worst) to 5 (best).bSkill assessed via place finished: 1 (best) to 6 (worst).*p � 0.05.†p � 0.10.

mixed. Traxter et al9 concluded that in vivo training is whatmatters most. They found that students who practiced on aninanimate box trainer performed porcine laparoscopic nephrec-tomy no better than those without the extra training. On theother hand, Grantcharov and associates5 had better results us-ing the MIST-VR. They convincingly showed that surgical res-idents trained on a VR simulator performed live laparoscopiccholecystectomy better than a control group.5 Likewise, Whit-ted and coworkers10 found that formal laparoscopic training ofgynecologic residents led to better patient outcomes.

The “perceptual” learning that takes place with these mod-els tends to be specific to the task.6 The more realistic the model,supposedly, the more useful it is. That may also explain why aVR simulator would prove more important than an inanimatebox trainer. This result would imply, for instance, that playinga car-racing video game would not necessarily improve one’slaparoscopic suturing ability. Therefore, most of the models aredesigned to replicate laparoendoscopic surgery as closely aspossible. Commercial video games, on the other hand, wouldprovide perceptual learning, namely, improvement in hand–eyecoordination, rather than discrete task training like that providedby VR simulators. Despite their attractiveness as a model, inview of their widespread availability and low cost, video games

cannot train a surgeon as well as VR can. A VR simulation,however, is not without its own inherent weaknesses, namelythe current “smoothness” of the image, the lack of haptic feed-back, and certainly cost.11

Video games improve one’s visual skills. Green and Bave-lier6 showed that habitual video-game players possess enhancedattentional capacity. Moreover, those investigators showed thatnonplayers improved their visual attention after video-gametraining. This would suggest that game playing may improvesurgical skills, but to our knowledge, no one had yet appliedthis idea to laparoscopy specifically. We sought to discover ifvideo-game aptitude predicted performance in a live laparo-scopic model and whether the perceptual learning one gainsfrom playing video games can translate into improvement inlaparoscopic skills.

Our study did demonstrate some evidence of moderate tostrong correlations between video-game performance and skillin laparoscopy, which goes along with our prediction. This wasmost evident in the simpler laparoscopic drills (object transferand figure of eight). The more complex tasks proved dauntingto all the students, not just the weak video gamesters. Time tocomplete the task correlated best with video-game skill, moreso than the number of errors committed or hand–eye coordina-

VIDEO GAMES AND LAPAROSCOPY 375

TABLE 2. LAPAROSCOPIC SKILL RESULTS BROKEN DOWN BY GROUP AND SESSION

Mean times

Session 1 Session 2 ChangeSkill Group (t1) (t2) (t2 � t1) Range

Object transferTime (secs)a Control 61 30 �31 �78, �7

Training 47 27 �20 �44, �7Errors Control 0.7 0.3 �0.4 �1, 1

Training 1.4 0.8 �0.6 �4, �1Hand–eye Control 3.5 3.5 0 �1, 2

Training 3.8 3.8 0 �2, 3Figure-of-eight

Time Control 67 37 �30 �199, 13Training 40 20 �20 38, 2

Errors Control 0.2 0.5 0.3 �1, 2Training 0 0.6 0.6 0.0, 2.0

Hand–eye Control 3.5 2.7 �0.8 �3, 1Training 3.8 3.2 �0.6 �3, 1

SuturingTime Control 181 179 �2 �305, 283

Training 108 96 �12 �106, 155Errors Control 1.7 2.8 1.1 0, 2

Training 1.6 2.4 0.8 �2, 7Hand–eye Control 2.8 2.3 �0.5 �1, 0

Training 3.6 3.0 �0.6 �2, 1Knot-Tying

Timea Control 139 75 �64 �135, 48Training 116 68 �48 �140, 60

Errors Control 1.3 1.0 �0.3 �2, 0Training 0.6 1.8 1.2 0, 2

Hand–eye Control 3.7 3.3 �0.4 �1, 1Training 4.2 3.2 �1 �2, 1

aSignificant improvement (P � 0.05) by the Wilcoxon signed-rank test for the whole group in the given skill from session 1to session 2.

No measure of any task showed significant improvement for the training group over the control group.

tion. Lack of significant p values in many cases appears to bea consequence of the small sample size.

Our data show that practicing video games does not seem toimprove one’s laparoscopic skills significantly. Overall, therewas improvement in session 2 compared with session 1, mostnotably in terms of time, but no parameter showed significantimprovement for group B over group A. It was our hypothesisthat extensive video gaming would lead to significant im-provements in laparoscopy; however, both groups improvedequally in session 2, at least in terms of procedure time. Ad-mittedly, the sample size here is small, and the training periodwas short. Perhaps a larger study with more intense trainingwould demonstrate a significant difference between the twoarms. It appears more likely that actual laparoscopic experienceor highly realistic simulation of laparoscopy is the only way togain laparoscopic aptitude. This is consistent with the conclu-sions drawn by previous investigators.5,9

Certain limitations of this study should be discussed. Thesmall sample size limits statistical significance in many cases.Our goal was 24 subjects, which was determined to provide ad-equate power for the analysis, but we could recruit no morethan 11. Also, longer training is liable to create a greater “train-ing effect.” Naturally, it is difficult for busy students to spendmuch time playing video games during the academic year. Also,the laparoscopic observers were not blinded to the training sta-tus of the subject. This could have affected the hand–eye co-ordination assessment, but it should not affect either time re-quired or errors committed. In order to simplify comparisonsbetween video game performance and laparoscopic skill, onlyone variable from each video game was utilized (hand–eye co-ordination in three of four). Incorporating more measures fromthe video games would likely be more thorough, but we optedto use the one best summary measure for each.

CONCLUSIONS

Video games do improve one’s hand–eye coordination, andthey appear to predict inherent laparoscopic skill. Practicing,however, does not seem to offer a way to improve these skills.At this point, the best way to improve as a laparoscopist ap-pears to be by practicing laparoscopy itself or working on a re-alistic simulator. A larger-scale study is indicated to understandconclusively what if any role video games may play in laparo-scopic surgery.

APPENDIX

Task 1: Object transfer. An 11-mm round rubber washer ispicked up off the sigmoid colon with one grasping instrument,transferred into the grasper in the surgeon’s other hand, andplaced on the liver on a spot cauterized with an X.

Task 2: Cauterize a figure-of-eight. With the surgeon’s dom-inant hand, a figure-of-eight is cauterized on the pig’s abdom-inal sidewall.

Task 3: Suturing of bowel. Using free-hand laparoscopic su-

turing techniques, a suture needle is picked up off the bowel,adjusted appropriately within the needle driver, driven througha segment of small bowel, and pulled out the other side.

Task 4: Knot tying. Using the suture just placed through thebowel, a single knot is tied by the instrument-tie technique.

ACKNOWLEDGMENTS

We are grateful to the following individuals for their help andparticipation in this study: Hugh Perkin, Benjamin Davies, RobertStein, Zachary Zuniga, Danielle Sweeney, and Glenn Cannon.

REFERENCES

1. Tsai CL, Heinrichs WL. Acquisition of eye–hand coordinationskills for videoendoscopic surgery. J Am Assoc Gynecol Laparosc1994;1(4 part 2):S37.

2. Watterson JD, Beiko DT, Kuan JK, et al. A randomized, prospec-tive blinded study validating the acquisition of ureteroscopy skillsusing a computer based virtual reality endourological simulator. JUrol 2002;168:1928.

3. Gallagher AG, Lederman AB, McGlade K, et al. Discriminativevalidity of the Minimally Invasive Surgical Trainer in Virtual Re-ality (MIST-VR) using criteria levels based on expert performance.Surg Endosc 2004; 2004;18(4):660–665.

4. Matsumoto ED, Hamstra SJ, Radomski SB, et al. The effect ofbench model fidelity on endourological skills: A randomized con-trolled study. J Urol 2002;167:1243.

5. Grantcharov TP, Kristiansen VB, Bendix J, et al. Randomized clin-ical trial of virtual reality simulation for laparoscopic skills train-ing. Br J Surg 2004;91:146.

6. Green CS, Bavelier D. Action video game modifies visual selec-tive attention. Nature 2003;423:534.

7. Subramonian K, DeSylva S, Bishai P, et al. Acquiring surgicalskills: A comparative study of oepn versus laparoscopic surgery.Eur Urol 2004;45:346.

8. Gettman MT, Kondraske GV, Traxer O, et al. Assessment of ba-sic human performance resources predicts operative performanceof laparoscopic surgery. J Am Coll Surg 2003;197:489.

9. Traxter O, Gettman MT, Napper CA, et al. The impact of intenselaparoscopic skills training on the operative performance of urol-ogy residents. J Urol 2001;166:1658.

10. Whitted RW, Pietro PA, Martin G, et al. A retrospective study eval-uating the impact of formal laparoscopic training on patient out-comes in a residency program. J Am Assoc Gynecol Laparosc2003;10:484.

11. Kuo RL, Delvecchio FC, Preminger GM. Virtual reality: Currenturologic applications and future developments. J Endourol 2001;15:117.

Address reprint requests to:Timothy Averch, M.D.

Dept. of UrologyUniversity of Pittsburgh

3471 Fifth Ave., Suite 700Pittsburgh, PA 15213

E-mail: averchtd@upmc.edu

ROSENBERG376