Visuospatial and psychomotor aptitude predicts endovascular performance of inexperienced individuals on a virtual reality simulator

EDUCATION CORNER

From the Society for Clinical Vascular Surgery

Visuospatial and psychomotor aptitude predictsendovascular performance of inexperiencedindividuals on a virtual reality simulatorIsabelle Van Herzeele, MD, PhD,a,b Kevin G. L. O’Donoghue, BSc,a

Rajesh Aggarwal, MA, MRCS, PhD,a Frank Vermassen, MD, PhD,b Ara Darzi, KBE, MD, FRCS,a andNicholas J. W. Cheshire, MD, FRCS,a,c London, United Kingdom; and Gent, Belgium

Objectives: This study evaluated virtual reality (VR) simulation for endovascular training of medical students to determinewhether innate perceptual, visuospatial, and psychomotor aptitude (VSA) can predict initial and plateau phase oftechnical endovascular skills acquisition.Methods: Twenty medical students received didactic and endovascular training on a commercially available VR simulator.Each student treated a series of 10 identical noncomplex renal artery stenoses endovascularly. The simulator recordedperformance data instantly and objectively. An experienced interventionalist rated the performance at the initial and finalsessions using generic (out of 40) and procedure-specific (out of 30) rating scales. VSA were tested with fine motordexterity (FMD, Perdue Pegboard), psychomotor ability (minimally invasive virtual reality surgical trainer [MIST-VR]),image recall (Rey-Osterrieth), and organizational aptitude (map-planning). VSA performance scores were correlated withthe assessment parameters of endovascular skills at commencement and completion of training.Results: Medical students exhibited statistically significant learning curves from the initial to the plateau performance forcontrast usage (medians, 28 vs 17 mL, P < .001), total procedure time (2120 vs 867 seconds, P < .001), and fluoroscopytime (993 vs. 507 seconds, P < .001). Scores on generic and procedure-specific rating scales improved significantly (10vs 25, P < .001; 8 vs 17 P < .001). Significant correlations were noted for FMD with initial and plateau sessions forfluoroscopy time (rs � �0.564, P � .010; rs � �.449, P � .047). FMD correlated with procedure-specific scores at theinitial session (rs � .607, P � .006). Image recall correlated with generic skills at the end of training (rs � .587, P � .006).Conclusions: Simulator-based training in endovascular skills improved performance in medical students. There weresignificant correlations between initial endovascular skill and fine motor dexterity as well as with image recall at end of thetraining period. In addition to current recruitment strategies, VSA may be a useful tool for predictive validity studies.

From the Department of Biosurgery and Surgical Technologya and theRegional Vascular Unit,c Imperial College Healthcare NHS Trust, Lon-don; and the Department of Thoracic and Vascular Surgery, UniversityHospital Ghent, Gent.b

This work was supported by funding from the Imperial College HealthcareBiomedical Research Centre, a Clinical Doctoral Grant from the Fund forScientific Research Flanders (FWO), Belgium, and a grant from Mentice,Gothenburg, Sweden, and Boston Scientific Corporation, Natick, Massand Nanterre, France. They did not, however, interfere in the studydesign, data collection, analysis, or interpretation of the data, or in thedecision to submit the manuscript for publication.

Competition of interest: none.Winner of the Poster Session at the Society of Vascular Surgery Vascular

Annual Meeting, San Diego, Jun 7, 2008.Correspondence: I. Van Herzeele, Department of Thoracic and Vascular

Surgery 2K12IC, University Hospital Ghent, De Pintelaan 185, 9000Gent, Belgium (e-mail: [email protected]).

The editors and reviewers of this article have no relevant financial relationshipsto disclose per the JVS policy that requires reviewers to decline review of anymanuscript for which they may have a competition of interest.

0741-5214/$36.00Copyright © 2010 by the Society for Vascular Surgery.

The introduction of catheter-based minimally invasiveinterventions has revolutionized the management of vascu-lar disease.1,2 Vascular surgeons and patients have acceptedthe role of endovascular therapy due to reduced procedur-ally related discomfort, faster recovery, and decreased hos-pital length of stay.3 These advantages are reduced whenendovascular procedures are done by inexperienced orpoorly supervised physicians, which may lead to increasedcomplications and poorer patient outcomes.4,5 Inexperi-enced practitioners need to learn how to maneuver endo-vascular tools carefully through blood vessels by manipu-lating endovascular tools within a 3-dimensional (3D)environment whilst viewing only a 2D fluoroscopic image.The operator must also work with reduced tactile feedback.

Until recently, core endovascular skills were acquiredby performing diagnostic angiograms. However, the in-creased use of noninvasive imaging techniques and bud-getary constraints in the operating room have restrictedtraining opportunities,6 and the implementation of the

1035

nds

JOURNAL OF VASCULAR SURGERYApril 20101036 Van Herzeele et al

countries has resulted in a marked reduction in the hoursavailable for training (15,000 hours vs 35,000 hours 10years ago).7,8 Patients have also become more demandingand less tolerant toward surgical errors and expect theirprimary operator to be proficient.9 Surgical educators havetherefore been compelled to search for more effective andcreative means of teaching surgical skills.10

Simulation has been used extensively in other high-riskprofessions, such as aviation, allowing the acquisition ofessential skills before the person is required to function inreal-life situations.11,12 Virtual reality technology has the po-tential to improve patient safety by allowing inexperiencedindividuals to train away from the patient in a nonpressuredenvironment, to be trained at their own pace, and to learnfrom their mistakes without putting patients at risk.13

Previous studies have used simulators to examine thepossibility of predicting surgical performance in laparo-scopic or general surgical procedures by testing innatevisuospatial abilities.14-17 To our knowledge, however, nostudies have attempted to relate innate abilities to endovas-cular performance on a virtual reality (VR) simulator. Thetwo aims of this study were to determine (1) if medicalstudents could acquire the appropriate endovascular skillsto perform a renal artery angioplasty and stent procedureon a VR simulator and (2) if differences in performancesbetween naïve individuals could be due to variability intheir psychomotor and visuospatial abilities (VSA).

METHODS

Participants. The study recruited 20 medical studentsfrom a university teaching hospital. Excluded were individ-uals with prior endovascular experience, surgical simula-tion, or visuospatial testing. All participants gave informedconsent before enrollment in the study.

Cognitive skills training. The endovascular tools,skills, indications, and protocol required to safely perform arenal artery dilation and stenting procedure were describedand clarified during a half-hour interactive presentation.13,18

Visuospatial abilities testing. Five validated tests ofVSA and psychomotor ability were chosen after an exten-

Table I. Brief description of visuospatial and psychomoto

Test Descript

Rey-Osterrieth Complex Figure Image is copied and theinstantly and after a d

Perdue Pegboard Small pins are placed inand right hands indivsimultaneously

Grooved Pegboard Grooved pins are placedusing only the domin

Map Planning Participant must find thbetween two points omap whilst avoiding r

Minimally Invasive Surgical Trainer Laparoscopic surgical siparticipants completeboth left and right ha

sive literature search.16,19-22 It was the intention for each

test to measure skills relevant to endovascular procedures(Table I; Fig 1). Each student was given written instruc-tions about the task and a mock question to determine ifthe task was understood. All tasks were timed and com-pleted as instructed in the literature.19-21

The simulator. The simulator used in the study wasthe Vascular Intervention Simulation Trainer (ProcedicusVIST, Mentice, Gothenburg, Sweden) and has been exten-sively described previously.11,23-25 At the commencementof the study, a standardized brief introduction to the VRsimulator was delivered to all students before the endovas-cular treatment of a right-sided, nonostial lesion of therenal artery was demonstrated. A protocol for endovasculartreatment of renal artery stenosis, correct handling of en-dovascular tools, and major errors possible during each stepof the intervention were explained and demonstrated in astandardized fashion.13,26,27 Before each assessment on theVR simulator, every participant had to explain the protocolfrom memory to ensure retention of cognitive skills.

Task performed. All participants treated an identicalsimulated left-sided nonostial renal artery lesion 10 times, amodule that has been validated in other studies.23,28 Thepatient’s case summary showing the renal lesion was ex-plained. Passive assistance was provided during the simulatedprocedure. Appropriate endovascular tools were selectedwhen asked for, and orientation of the C-arm was modified asrequested. A virtual ruler was available but could not bemoved. In line with similar experiments in the literature, amaximum of two sessions per day were allowed with a mini-mum of 1 hour between sessions.23 Verbal advice was pro-vided only when a student performed the same error threetimes.

Performance evaluation. Visuospatial and psychomo-tor aptitude were evaluated using the scoring systems pro-vided with the tests. The VIST automatically provides aprocedure report for each session containing two types ofinternal assessment parameters. Procedure time, contrastvolume, and fluoroscopy time were the quantitative metricsrecorded for this study. Qualitative metrics are also regis-tered by the VIST and include the clinical parameters of

s

st Skills tested

wn from memory Visuospatial constructional ability, visualmemory, and executive function

rd by using leftly and also

Fine motor dexterity

grooved holesand

Fine motor dexterity

rtest routechematic of a citylocks

Organizational aptitude

or in whichact tasks with

Ability to interact in a 3D environmentfrom 2D screen

r test

ion te

n draelaya boaidual

intoant he shon a soadbmulatabstr

placement accuracy of stent or balloon (mm), lesion cover-

JOURNAL OF VASCULAR SURGERYVolume 51, Number 4 Van Herzeele et al 1037

age by the balloon or stent (%), balloon/vessel ratio (range,0-1), residual stenosis (%), and stent/vessel ratio.5,29 Astent/vessel ratio of 1 is ideal, with a value of �1 indicatingan undersized stent and �1 an oversized stent for thevirtual lesion. Error scoring was not available for this mod-ule at the time the study was undertaken and errors did not

Fig 1. A, The Rey-Osterrieth Complex Figure, is first copied, thendrawn from memory instantly, and drawn again after a delay of 10minutes. B, Individuals who take the Perdue Pegboard test mustperform a variety of single-and double-handed fine motor tasks.Performance on this test correlated well with initial performance.

cause complications, for example, over-dilating a vessel did

not result in rupture of the artery. At the end of each task,a procedure report was automatically generated that in-cluded all recorded simulator metrics.

During the first and tenth procedure, an experiencedendovascular interventionalist (external assessment) usedtwo rating scales to assess the candidates. The genericrating scale was based on the global rating scale used inObjective Structured Assessment of Technical Skills(OSATS)30-33 and evaluates basic and generic endovascularskills (maximum score, 40). The procedure-specific ratingscale (maximum score, 35; Fig 2) was developed by expe-rienced endovascular practitioners who analyzed andscored the different steps and skills required to safely com-plete a renal artery stenting procedure 29 After the initialsession, the observer provided standardized, formativefeedback to each medical student by identifying the areas ofweakness in renal artery stenting based mainly on thedefinitions provided in the procedure-specific rating scale.

Statistical analysis. Data were analyzed with non-parametric tests using SPSS 15.0 software (SPSS Inc, Chi-cago, Ill). Learning curves were analyzed using a Friedman(nonparametric repeated measures analysis of variance) test.Sequential comparisons were made to identify plateau levelsfor all significant variables. Both initial and plateau perfor-mances were analyzed for correlation on the visuospatial tests.The Spearman rank test was used to examine correlations forcontinuous data, and a Kendall tau B test was used for cate-goric data. The Wilcoxon signed rank test was used to inves-tigate skills improvement from the first to tenth session, asmeasured by the rating scales. Data results with a value of P �.050 were considered statistically significant.

RESULTS

Demographics. Eight men and 12 women participated,consisting of 10 third-year and 10 fourth-year medical stu-dents who were a median age of 22 years (range 20-26 years).All were right handed, except for one. All students successfullycompleted each of the VSA tests before commencing thetraining program. Although all students finished the required10 sessions on the simulator in groups of five �2 weeks, 31 ofthe 200 procedure reports (15.5%) were incomplete becauseof technical difficulties with the VIST.

Learning curves. Medical students exhibited statisti-cally significant learning curves for total procedure time(medians, 2120 vs 867 seconds, P � .001, Fig 3), fluoros-copy time (993 vs 507 seconds, P � .001), contrast usage(28 vs 17 mL, P � .001), and number of recorded angio-grams (6.5 vs 4, P � .003). The plateau phase for thelearning curve occurred at the seventh session for totalprocedure time, the sixth session for fluoroscopy time, thefourth session for contrast volume, and the eighth sessionfor number of angiograms. The qualitative metrics of theVR simulator (eg, stent/vessel ratio) did not improve sig-nificantly when the initial and final sessions were compared(medians, 1.06 vs 1.04, P � .673). However, the experi-enced observer scored the final performance significantly

exter

JOURNAL OF VASCULAR SURGERYApril 20101038 Van Herzeele et al

(median scores, 10 vs 25, P � .001) and procedure-specific(8 vs 17, P � .001) rating scales (Fig 4).

Initially, the group demonstrated a large variability

Fig 2. A procedure-specific rating scale was used for liveinitial and final session.

in performance, but as their training progressed, this led

to increased consistency. This becomes clear when thefinal session is compared with the initial session for thevalid internal metrics (interquartile range of the first

nal assessment by an experienced interventionalist at the

session vs the final session for procedure time was 783 vs

sent the outliers.

sent the outliers.

JOURNAL OF VASCULAR SURGERYVolume 51, Number 4 Van Herzeele et al 1039

332 seconds, Fig 3). Fig 5 illustrates how the proceduretimes for three different students diminished during the10 sessions, illustrating individual learning patterns.

Correlation between visuospatial ability and initialsimulator performance. The medical students who scoredhigher on the Perdue Pegboard pressed the fluoroscopy pedalfor a shorter duration (rs � �0.5640, P � .010) and usedcontrast less liberally (rs � 0.511, P � .021). Higher scores onthe Perdue Pegboard also tended to lead to higher procedure-specific scoring at the initial session (rs � .607, P � .006). Theother test of fine motor dexterity, the grooved pegboard, alsocorrelated with initial endovascular performance rated by theprocedure-specific scale (rs � �0.564, P � .011). The Rey-Osterrieth Complex Figure (ROCF) test, map planning, andthe minimally invasive virtual reality surgical trainer (MIST-VR)did not correlate with initial performance during virtual renalartery stent procedures (Table II).

Correlation between visuospatial ability and endperformance. The correlation between fine motor dexter-ity (Perdue Pegboard) and the performance on the VRsimulator at the plateau phase was less apparent than withthe initial performance (fluoroscopy time: rs � �0.449,P � .047). Those medical students with superior imagerecall (ROCF) tended to press the fluoroscopy pedal for ashorter amount of time (rs � 0.489, P � .029).

Performance on the ROCF correlated with the externalassessment of generic endovascular skills (time to copyROCF, rs � �0.587, P � .006). The Perdue Pegboard(rs � 0.674, P � .001), MIST-VR (rs � �0.588, P �.006), and map planning (rs � 0.579, P � .007) were alsoobserved to correlate with the generic rating scale at thefinal session. The grooved pegboard test, however, did notcorrelate with end performance (rs � �0.286, P � .221;Table II).

DISCUSSION

This study has revealed that naïve individuals can ac-

Fig 5. The procedure times are illustrated for three differentparticipants. The triangles represent a student in the 95th percen-tile for initial procedure time, squares represent a student withapproximately a median value for the initial procedure time, andcircles represent a student in the fifth percentile.

Fig 4. Box and whisker plot represents the improvement fromsession 1 to 10 during external assessment using generic andprocedure-specific rating scales. (Wilcoxon signed rank test, P �.001). The white boxes illustrate scores at the initial session andhatched boxes scores for the final session. The horizontal line in themiddle of each box indicates the median, the top and bottom bordersof the box mark the 75th and 25th percentiles, respectively, thewhiskers mark the 90th and 10th percentiles, and the circles repre-

Fig 3. Box and whisker plot represents the total procedure timenecessary to complete the virtual renal artery stenting procedurefor each of the 10 sessions (Friedman test, P � .001). The plateauphase is represented by the grey boxes. The horizontal line in themiddle of each box indicates the median, the top and bottom bordersof the box mark the 75th and 25th percentiles, respectively, thewhiskers mark the 90th and 10th percentiles, and the circles repre-

quire endovascular skills to perform a renal artery stent

ell rep

JOURNAL OF VASCULAR SURGERYApril 20101040 Van Herzeele et al

procedure on the basis of VR simulator metrics and expertratings. Furthermore, VSA and psychomotor testing correlatewith initial and end performance during simulated complexendovascular interventions. FMD correlated clearly with ini-tial performance on the VR simulator based on the internalquantitative metrics and the external ratings of the interven-tionalist. For example, students who scored higher on thePerdue Pegboard test pressed the fluoroscopy pedal for ashorter duration at the initial session. This suggests that themore dextrous students had the innate ability to maneuverthe wires more gently and to keep the endovascular toolssteadier, allowing a more economic fluoroscopy use at theirfirst endovascular intervention. The quality of their proce-dure specific skills was also scored higher by the expertobserver and may imply that fine motor testing is able topredict initial performance during a complex endovascularintervention and reflect the handiness of naïve individuals.

We also wanted to establish whether any of these testscould predict the performance at the end of training. The

Table II. Correlations between visuospatial and psychomointerventionalists based assessment (Spearman rank test)

Initial session

Quantitative simulator metricsIn

Internal metrics Ex

TestProcedure

timeFluoroscopy

timeContrast

volProc

spe

Fine motor testsGrooved Peg Board, rs

a �0P

Perdue PegboardDom hand, rs 0

PNon-dom hand, rs �0.564 0

P .01L � R � both, rs 0.48 0

P .032Perdue Pegboard

Assembly, rs 0.511P .021

Visuospatial Ability TestsROCF

Copy time (s), rsP

Imm recall time, rsP

Del recall score, rsP

Del recall time, rsP

Map planning, rsP

MIST-VR time, righthand, rs 0.55

P .012

Del, Delayed; Imm, immediate; L, left hand score; R, right hand score; MIComplex Figure.aCorrelation coefficients (rs) values are presented with P values. An empty c

ROCF results did indeed show significant correlation with

fluoroscopy time at the end of training. This suggests thatcandidates with a poorer ability to recall images might needto press the fluoroscopy pedal more frequently to visualizethe renal lesion, thus exposing the virtual patient and theinterventional team to more radiation.

The link with the expert rating of generic endovascularskill at session 10 suggests that the quality of technicalperformance of medical students with greater VSA and psy-chomotor abilities during a complex endovascular procedurewas higher. The ROCF test is believed to test executivefunction along with visuoconstructional ability.21 This couldfurther explain its correlation to end performance because itmay be able to identify those candidates with superior frontallobe aptitude. These students may have a greater ability tolearn and understand the complex task of endovascular renalartery stenting. Stefanidis et al16 did an analogous study in thelaparoscopic domain and observed significant correlations be-tween time to reach proficiency and the ROCF test scores.Wanzel et al17 demonstrated that visuospatial abilities corre-

ability scores with internal simulator metrics and external

Plateau session Final session

tionalistings Quantitative simulator metrics

Interventionalistratings

l metrics Internal metrics External metrics

-Generic

Proceduretime

Fluoroscopytime

Contrastvol

Procedure-specific Generic

�0.497.026

�0.449.047

0.502.028

0.484 0.674.03 .001

�0.587.006

�0.486.035

0.494.027

0.489.029

0.579.007

�0.491 �0.588.033 .006

, virtual reality minimally invasive surgical trainer; ROCF, Rey-Osterrieth

resents a nonsignificant correlation (P � .050).

tor

tervenrat

terna

edurecific

.564

.011

.502

.028

.607

.006

.46

.048

ST-VR

lated with surgical performance for dental students but not for

JOURNAL OF VASCULAR SURGERYVolume 51, Number 4 Van Herzeele et al 1041

surgical residents or staff surgeons, suggesting surgical expe-rience eventually becomes more important than innate VSA.This study, however, showed that at least in the early phase oflearning, VSA is an important correlate of technical endovas-cular skill.

Among the study limitations are that the study protocoldid not require the medical students to reach previouslyestablished benchmark levels of skill in the simulation lab.Likewise, this study did not investigate the transfer ofmotor skills from the simulator lab to the operating roomor angiography suite.

An identical lesion was treated during all 10 simula-tions; hence, the improvement in simulator metric perfor-mance could be attributed to student familiarity with themodule, resulting in automated tool selection and C-armpositioning. However, this would not solely explain thesignificant improvement in quality of performance asscored by the observer, although we must note that theassessor was not blinded to the session number.

Technical difficulties occurred on 15.5% of simulations,which potentially affected the results in this study but also dem-onstrates the significant reliability limitations with this simulator.

Finally, although several positive associations werenoted between the students’ VSAs and their initial and endperformances during a complex endovascular intervention,several others did not correlate. Further research is re-quired, including more participants, to investigate this in-termittent and complex relationship between the psy-chomotor and visuospatial abilities of naïve individuals.

Future work ought to focus on the effect that naturalvariations in VSA might have on long-term skills acquisi-tion, because this study did not investigate if the higherperformance levels were sustained (retention).

There is a perception that these tests may have a futurerole in selecting candidates for surgical training, but theirvalidity in this context needs to be further proven beforepractical application, particularly as this study was under-taken using a VR simulator as a surrogate for technicalendovascular skill in an interventional suite.14-17

In the current training environment of reduced train-ing hours, it is essential to move away from the classicalHalstedian method of training and to modernize educationto ensure appropriate access to surgical simulation as anintegral part of the training experience.6-8,32 Once traineeshave reached a predefined set of proficiency criterion in theskills laboratory, cognitive and motor behavior can be usedand shaped in the real angiography suite. VR simulation hasbeen shown to enable surgical trainees to reach an encour-aging level of proficiency before entering an interventionalsuite to treat actual patients.13

The VIST simulator used in this study currently costs£120,000; in addition, substantial maintenance costswould be incurred to address technical difficulties if thesesimulators were routinely used in residency training pro-grams. Although it is unlikely that many institutions wouldbe able to afford such expensive resources, these simulatorsshould be available to all trainees in specialist regional or

Although not the focus of this study, individuals werenoted to demonstrate various learning patterns (Fig 5).VSA testing may be helpful not only to select future endovas-cular physicians but also to assist in tailoring the trainingprograms to the needs of the trainees. It may identify thosewho may acquire technical skills more rapidly and those whomay need additional training to reach proficiency. Furtherstudies are in progress to explore this finding in more detail.Aptitude testing may also be used to highlight specific techni-cal tasks that trainees find more challenging and that mayrequire supplementary instructions and mentoring.

CONCLUSIONS

This study has demonstrated that novice trainees canimprove their technical endovascular skills during a virtualreality complex endovascular intervention. More impor-tant, visuospatial and psychomotor abilities have been iden-tified as good predictors of initial and end performance in aVR endovascular simulator training schedule. These testsmay find a new role in trainee selection or in identifyingtrainees’ strengths or weaknesses to allow more tailored andefficient approaches to training programmes.

We thank Dimitrios Stefanidis, MD, PhD, for his valuableassistance in the design of this study, and we also thank all ofthe medical students who agreed to participate in the study.

AUTHOR CONTRIBUTIONS

Conception and design: IV, KOAnalysis and interpretation: IV, KO, RA, FV, NCData collection: IV, KOWriting the article: IV, KOCritical revision of the article: IV, RA, FV, AD, NCFinal approval of the article: IV, KO, RA, FV, AD, NCStatistical analysis: IV, KO, RAObtained funding: AD, NCOverall responsibility: IV

REFERENCES

1. Endovascular aneurysm repair versus open repair in patients with ab-dominal aortic aneurysm (EVAR trial 1): randomised controlled trial.Lancet 2005;365:2179-86.

2. Suding PN, McMaster W, Hansen E, Hatfield AW, Gordon IL, WilsonSE. Increased endovascular interventions decrease the rate of lower limbartery bypass operations without an increase in major amputation rate.Ann Vasc Surg 2008;22:195-9.

3. Nasr MK, McCarthy RJ, Hardman J, Chalmers A, Horrocks M. Theincreasing role of percutaneous transluminal angioplasty in the primarymanagement of critical limb ischaemia. Eur J Vasc Endovasc Surg2002;23:398-403.

4. Seymour NE, Gallagher AG, Roman SA, O’Brien MK, Bansal VK,Andersen DK, et al. Virtual reality training improves operating roomperformance: results of a randomized, double-blinded study. Ann Surg2002;236:458-63; discussion 63-4.

5. Van Herzeele I, Aggarwal R, Choong A, Brightwell R, Vermassen FE,Cheshire NJ. Virtual reality simulation objectively differentiates level ofcarotid stent experience in experienced interventionalists. J Vasc Surg2007;46:855-63.

6. Lamont PM, Scott DJ. The impact of shortened training times on thediscipline of vascular surgery in the United Kingdom. Am J Surg

JOURNAL OF VASCULAR SURGERYApril 20101042 Lawrence

7. Wang L, Wang X, Laird N, Zuckerman B, Stubblefield P, Xu X.Polymorphism in maternal LRP8 gene is associated with fetal growth.Am J Hum Genet 2006;78:770-7.

8. Kochunov P, Lancaster JL, Glahn DC, Purdy D, Laird AR, Gao F, et al.Retrospective motion correction protocol for high-resolution anatom-ical MRI. Hum Brain Mapp 2006;27:957-62.

9. Nguyen TN, Silver D, Arthurs B. Consent to cataract surgery performedby residents. Can J Ophthalmol 2005;40:34-7.

10. Moulton CA, Dubrowski A, Macrae H, Graham B, Grober E, ReznickR. Teaching surgical skills: what kind of practice makes perfect?: arandomized, controlled trial. Ann Surg 2006;244:400-9.

11. Chaer RA, Derubertis BG, Lin SC, Bush HL, Karwowski JK, Birk D, et al.Simulation improves resident performance in catheter-based intervention:results of a randomized, controlled study. Ann Surg 2006;244:343-52.

12. Patel AA, Gould DA. Simulators in interventional radiology trainingand evaluation: a paradigm shift is on the horizon. J Vasc Interv Radiol2006;17:S163-73.

13. Gallagher AG, Ritter EM, Champion H, Higgins G, Fried MP, MosesG, et al. Virtual reality simulation for the operating room: proficiency-based training as a paradigm shift in surgical skills training. Ann Surg2005;241:364-72.

14. Gettman MT, Kondraske GV, Traxer O, Ogan K, Napper C, Jones DB, etal. Assessment of basic human performance resources predicts operativeperformance of laparoscopic surgery. J Am Coll Surg 2003;197:489-96.

15. Rosenthal R, Gantert WA, Scheidegger D, Oertli D. Can skills assess-ment on a virtual reality trainer predict a surgical trainee’s talent inlaparoscopic surgery? Surg Endosc 2006;20:1286-90.

16. Stefanidis D, Korndorffer JR Jr, Black FW, Dunne JB, Sierra R, Tou-chard CL, et al. Psychomotor testing predicts rate of skill acquisitionfor proficiency-based laparoscopic skills training. Surgery 2006;140:252-62.

17. Wanzel KR, Hamstra SJ, Anastakis DJ, Matsumoto ED, CusimanoMD. Effect of visual-spatial ability on learning of spatially-complexsurgical skills. Lancet 2002;359:230-1.

18. Van Herzeele I, Aggarwal R, Neequaye S, Darzi A, Vermassen F, CheshireNJ. Cognitive training improves clinically relevant outcomes during simu-lated endovascular procedures. J Vasc Surg 2008;48:1223-30, 30e1.

19. Boltuch J, Sabeti S, Amighi J, Dick P, Mlekusch W, Schlager O, et al.Procedure-related complications and early neurological adverse eventsof unprotected and protected carotid stenting: temporal trends in aconsecutive patient series. J Endovasc Ther 2005;12:538-47.

20. Yu F, Sabeti PC, Hardenbol P, Fu Q, Fry B, Lu X, et al. Positiveselection of a pre-expansion CAG repeat of the human SCA2 gene.

during the residency. The article supports an approach where

21. Shin MS, Park SY, Park SR, Seol SH, Kwon JS. Clinical and empiricalapplications of the Rey-Osterrieth Complex Figure Test. Nat Protoc2006;1:892-9.

22. Aggarwal R, Grantcharov TP, Eriksen JR, Blirup D, Kristiansen VB,Funch-Jensen P, et al. An evidence-based virtual reality training pro-gram for novice laparoscopic surgeons. Ann Surg 2006;244:310-4.

23. Aggarwal R, Black SA, Hance JR, Darzi A, Cheshire NJ. Virtual realitysimulation training can improve inexperienced surgeons’ endovascularskills. Eur J Vasc Endovasc Surg 2006;31:588-93.

24. Hsu JH, Younan D, Pandalai S, Gillespie BT, Jain RA, Schippert DW, etal. Use of computer simulation for determining endovascular skill levelsin a carotid stenting model. J Vasc Surg 2004;40:1118-25.

25. Neequaye SK, Aggarwal R, Brightwell R, Van Herzeele I, Darzi A,Cheshire NJ. Identification of skills common to renal and iliac endovas-cular procedures performed on a virtual reality simulator. Eur J VascEndovasc Surg 2007;33:525-32.

26. Fried MP, Satava R, Weghorst S, Gallagher AG, Sasaki C, Ross D, et al.Identifying and reducing errors with surgical simulation. Qual SafHealth Care 2004;13 Suppl 1:i19-26.

27. Reznick RK, MacRae H. Teaching surgical skills–changes in the wind.N Engl J Med 2006;355:2664-9.

28. Berry M, Lystig T, Reznick R, Lonn L. Assessment of a virtual inter-ventional simulator trainer. J Endovasc Ther 2006;13:237-43.

29. Van Herzeele I, Aggarwal R, Malik I, Gaines P, Hamady M, Darzi A, etal. Validation of video-based skill assessment in carotid artery stenting.Eur J Vasc Endovasc Surg 2009;38:1-9.

30. Hislop SJ, Hsu JH, Narins CR, Gillespie BT, Jain RA, Schippert DW, etal. Simulator assessment of innate endovascular aptitude versus empir-ically correct performance. J Vasc Surg 2006;43:47-55.

31. Martin JA, Regehr G, Reznick R, MacRae H, Murnaghan J, HutchisonC, et al. Objective structured assessment of technical skill (OSATS) forsurgical residents. Br J Surg 1997;84:273-8.

32. Reznick RK. Teaching and testing technical skills. Am J Surg 1993;165:358-61.

33. MacRae HM, Satterthwaite L, Reznick RK. Setting up a surgical skillscenter. World J Surg 2008;32:189-95.

34. Van Herzeele I, Aggarwal R, Neequaye S, Hamady M, Cleveland T,Darzi A, et al. Experienced endovascular interventionalists objectivelyimprove their skills by attending carotid artery stent training courses.Eur J Vasc Endovasc Surg 2008;35:541-50.

Submitted Jul 27, 2009; accepted Nov 1, 2009.

INVITED COMMENTARY

Peter F. Lawrence, MD, Los Angeles, Calif

For surgery residents, the challenges of achieving technicalexcellence have become greater as the number of hours available totrain has been reduced and the range of procedures performed bya vascular surgeon has increased. Medical students spend less timeon surgery rotations, so the process of selecting for vascular surgeryresidencies students who have the potential to develop into excel-lent technical surgeons has become more difficult.

This article by Van Herzeele et al assesses the variation in medicalstudents’ performance on nonmedical technical exercises and corre-lates their psychomotor skill level with performance of a renal angio-plasty on a simulator. Their initial proficiency in nonmedical visuospa-tial and psychomotor tests was predictive of a better performance ofrenal angioplasty on a simulator. After six sessions of simulator train-ing, most students reached a plateau in improvement.

This is important information, primarily because it demon-strates that medical students arrive at a residency training programwith variable levels of technical facility, so that a “one size fits all”approach to education, which currently characterizes most surgicaleducation, is unlikely to lead to optimal technical skills acquisition

residents achieve maximal technical proficiency on a simulator bypracticing about six procedures, and then advance to patients, inaddition to serving for a prescribed time on each rotation.

The article is also important for what it does not claim. Thereis no claim that those with excellent initial psychomotor skills willbecome better technical surgeons—“isoperformance,” or transferof skills to the operating room, depends on many other factors andmay or may not be correlated with the initial skill level. In addition,the authors do not attempt to determine whether higher initial skilllevels are inherent or acquired by practicing other motor skillsbefore medical school. Ericsson has shown that “expert perfor-mance” in other fields, particularly athletics and music, is as depen-dent on the number of hours of practice and that at least 10,000hours are required, even of a prodigy, to achieve excellence.

Further research in the acquisition of technical excellence iscritical to our technique-oriented specialty. The acquisition of tech-nical skill must become as important an area of assessment as cognitiveability for surgeons in training. We need to learn how to initially assessskill levels and then develop more effective methods of skills training