Title: Acute Stress Impacts Psychomotor Bimanual Performance During Simulated Tumor

Resection Task

Khalid Mohammed A. Bajunaid, MD

Department of Experimental surgery

McGill University, Montreal, Quebec, Canada

May 2015

A thesis submitted to McGill University in partial fulfillment of the requirements of the degree of

Master of Science in Experimental surgery

© Khalid Bajunaid 2015

A
Table of contents

Abstract ……………………………………………………………………………………… 5

Résumés ……………………………………………………………………………………… 7

Acknowledgment …………………………………………………………………………… 9

Preface & authors contribution …………………………………………..………………….. 11

Abbreviations ………………………………………………………………………………… 12

1. Introduction and background ……………………………………………………………… 13

a.
Safety and training in the surgical environment ……….……….…………… 13

b.
Stress Concept ………………………………………………………………… 14

c.
Simulation and Stress ………………………………………………………… 14

d.
NeuroTouch Virtual Reality Platform ……………………………………… 16

e.
Psychomotor Skills Metrics …………………………………..…….……… 16

2. Manuscript ………………………………………………………………………………. 17

II.
Introduction ………………………………………………………………………… 18

III.
Methods ………………………………………………………………………………. 20

a.
The State Trait Anxiety Inventory (STAI) ……………...…………………… 20

b.
Simulated Virtual Reality Scenarios ………………………………………….. 20

c.
Simulated Operative Resection Procedures …………………………………. 21

d.
NeuroTouch Metrics ……………………………………………………….... 22

e.
Psychophysiological Assessment …………………………………………… 23

f.
Statistical Analysis ………………………………………………………… 23

i.
Statistical Model ………………………………………………...... … 24

B
IV.
Results ………………………………………………………………………………… 26

a.
Demographics and STAI …………………………………………………….. 26

b.
Acute Stress and Neurosurgical Performance ………………………………… 26

c.
Psychomotor Performance Immediately Following a Stress Inducing Event … 28

d.
Group Analysis ……………………………………………………………….. 29

V.
Discussion …………………………………………………………………………….. 29

a.
What Differentiates ‘Expert’ From Novice/Resident Performance During Acute

Stress? ………………………………………….…………………………….. 30

b.
Psychomotor Performance Following Acute Stress ……………………….… 32

VI.
Strengths and Limitations …………………………………………………………... 33

VII.
Overall Conclusion and Future Directions …………………………………………. 35

a.
Conclusion ……………………………………………………..…………….. 35

b.
Future directions ……………………………………………………………… 35

VIII.
References …………………………………………………..……………………….. 38

IX.
Tables ………………………………………………………………………………… 45

a.
Table 1: State Trait Anxiety Inventory (STAI) Results of each group pre and post

experiment ………………………………………………………………….... 46

X.
Figures ………………………………………………………………………………… 47

a.
Figure 1: NeuroTouch scenario and operator using the platform ……………. 48

b.
Figure 2: Tier 1 metrics of participant groups ……………………………….. 49

c.
Figure 3: Tier 2 metrics of participant groups ……………………………….. 50

d.
Figure 4: Advanced tier 2 metrics of participant groups …………………….. 51

e.
Figure 5: Statistical differences of advanced tier 2 metrics of participant groups .52

C
f.
Figure 6: Heat map of statistically significant affected metrics when baseline and

stress scenario values were compared ………………………………………… 53

XI.
Appendices …………………………………………………………………………….. 54

a.
Appendix 1: Personal Data Form ……………………………………………... 55

b.
Appendix 2: Evaluation Form ………………………………………………… 58

c.
Appendix 3: Validated State Trait Anxiety Inventory Questionnaire ……..….. 60

D
Abstract

Background: Severe bleeding during neurosurgical operations may result in acute stress affecting operator bimanual psychomotor performance leading to surgical error and adverse patient outcome.

Objective methods to assess the influence of acute stress on neurosurgical bimanual psychomotor performance have not been developed. Virtual reality simulators, such as Neuro-Touch, allow the testing of acute stress on psychomotor performance in risk free environments. Thus, the purpose of this study was to explore the impact of a simulated stressful virtual reality tumor resection scenario, utilizing NeuroTouch, to answer two questions: 1) What is the impact of acute stress on bimanual psychomotor performance during resection of simulated tumors? 2) Does acute stress influence bimanual psychomotor performance immediately following the stressful episode?

Methods: Study participants included 6 neurosurgeons, 6 senior and 6 junior neurosurgical residents, and 6 medical students. Participants resected a total of 6 simulated tumors, one of which

(tumor 4) involved uncontrollable ‘intraoperative’ bleeding resulting in simulated patient cardiac arrest, and thus, providing the acute stress scenario. Tier 1 metrics assessed included blood loss, percentage tumor resected and ‘normal’ brain volume removed. Tier 2 metrics included simulated sucker and ultrasonic aspirator total tip path length, in addition to the sum and maximum forces utilized by these instruments. Advanced tier 2 metrics evaluated included efficiency index, coordination index, ultrasonic aspirator path length index, and ultrasonic aspirator bimanual forces ratio. All metrics were assessed before, during, and after the stressful scenario.

Results: The stress scenario caused significant increases in blood loss in all groups. Decreased tumor resection and brain volume removed occurred in the junior resident and medical student groups. Sucker total tip path length increased in the neurosurgeon group while senior residents increased sucker forces. Psychomotor performance in advanced tier 2 metrics was altered during

E
the stress scenario in all participant groups. Performance in all advanced tier 2 metrics returned to pre-stress levels in post stress scenario tumor resections.

Conclusions: Results demonstrated that acute stress initiated by simulated severe intra-operative bleeding significantly decreases bimanual psychomotor performance during the acute stressful episode. The simulated intra-operative bleeding event had no significant influence on the advanced tier 2 metrics monitored during immediate post stress operative performance.

F
RÉSUMÉs :

Historique : Le saignement sévère durant les opérations neurochirurgicales peut résulter en stress aigu affectant la performance psychomotrice bi-manuelle de l’opérateur ce qui pourrait emmener à l’erreur chirurgicale et à des résultats nuisibles chez le patient. Les méthodes objectives n’ont pas

étés développées pour évaluer l’influence du stress aigu sur la performance psychomotrice bi- manuelle neurochirurgicale. Des simulateurs en réalité virtuelle, tels que le Neuro-Touch, permettent de tester le stress aigu de la performance psychomotrice qui se tient dans des environnements libres de facteurs de risque. Ainsi, l’objectif de cette étude était d’explorer l’impact d’une résection tumorale virtuelle de nature stressante tout en utilisant Neuro-Touch pour ensuite répondre à deux questions :

1)
Quel est l’impact du stress aigu sur la performance psychomotrice bi-manuelle durant une

résection de tumeur; et

2)
Est-ce que le stress aigu joue un rôle d’influence sur la performance psychomotrice bi-

manuelle immédiatement suivant l’épisode de simulation stressante?

Méthode : Cette étude a inclut 6 neurochirurgiens, 12 résidents en neurochirurgie (dont 6 séniors et 6 juniors) et 6 étudiants médicaux. Ces participants ont réséqué un total de 6 tumeurs simulées, dont une d’elles (tumeur 4) impliquait un saignement intra-opératif incontrôlable qui a provoqué une crise cardiaque simulée. Ceci a servi comme scénario de simulation de stress aigu. Palier 1 : les paramètres de mesure évalués ont inclus la perte de sang, le pourcentage de tumeur réséquée et la partie du cerveau normal retirée. Palier 2 : les paramètres de mesure évalués ont inclus la longueur du trajet de la pointe totale d’un extracteur et d’un aspirateur ultrasonique, en plus de la force

G
(somme et maximum) utilisée par ces instruments. Palier 2 avancé : les paramètres de mesure

évalués ont inclus les index suivants : l’efficience, la coordination, la longueur du trajet de l’aspirateur ultrasonique et le ratio de force bi-manuelle de l’aspirateur ultrasonique. Les paramètres de mesure ont été évalués avant, pendant et après le scénario de simulation de stress.

Résultat : Le scénario de stress a démontré des augmentations significatives de perte de sang dans tous les groupes. Un niveau diminué de résections tumorales et de volume de cerveau étaient présents dans les groupes de résidents juniors et des étudiants médicaux. La longueur de la pointe totale de l’extracteur était élevée dans le groupe de neurochirurgiens tandis que les résidents séniors utilisaient une force augmentée de l’extracteur. La performance psychomotrice dans le palier 2 avancé était modifiée dans tous les groupes pendant le scénario de simulation de stress. La performance dans tous les paramètres de mesure du palier 2 avancé, est retournée au niveau précédant le scénario de simulation de stress dans les résections de tumeurs.

Conclusions : Les résultats ont démontré que, le stress aigu initié par une simulation de saignement sévère intra-opérative, diminue significativement la performance psychomotrice bi-manuelle pendant la simulation de scénario de stress aigu. L’épisode de saignement intra-opératif simulé n’avait pas d’influence significative dans les paramètres de mesure du palier 2 avancé. Ces paramètres ont été surveillés après la performance opérative immédiatement suite au scénario de simulation de stress.

H

Acknowledgements

To my thesis supervisor, Professor Rolando Del Maestro, I give my utmost appreciation and gratitude for his sincere and continuous effort to advance my career. During the time I was nourished and educated under his advisement I witnessed myself becoming a different person, researcher, and clinician. Thank you Professor Del Maestro for your great mentorship and may our relationship continue for years to come.

I would like to deliver my appreciation to my lab colleagues at the Neurosurgical Simulation

Research Center at the Montreal Neurological Institution and Hospital. Special thanks goes to Dr.

Fahad Alotibi, Dr. Alexander Winkler-Schwartz, Dr. Gmaan Alzahrani, Dr. Abdulrahman Sabbagh,

Dr. Ibrahim Marwa along with Jawad Fares and Marta Baggiani for their very essential research input and continued support during the experimental trial and the manuscript revision. Additionally,

I would like to especially thank and acknowledge Mr. Muhammad Mullah for his guidance and effort during the statistical analysis. In like manner I would like to thank Dr. Penny Werthner, Dean

Faculty of Kinesiology University of Calgary and Mrs. Sommer Christie from University of

Calgary, Faculty of Kinesiology for their support and contribution during the trial.

Sincere thanks goes also to the Simulation and Digital Health Group-National Research

Council Canada NeuroTouch development team, including special thanks to Dr. Robert DiRaddo

Group Leader and his team members including Denis Laroche, Valérie Pazos, Nusrat Choudhury,

Patricia Debergue and Linda Pecora along with many others members for their support in the development of the scenarios utilized in my studies. I would also like to thank Amy Haddlesey for her help with the manuscript.

I
Also, I would like to acknowledge the neurosurgical staff and residents at the Montreal

Neurological Institute and Hospital for their participation in the study, without them it would have been impossible to conduct these studies.

Special thanks goes to Professor Saleh Baeesa, Division of Neurosurgery, King Abdul-Aziz

University, Jeddah, Saudi Arabia, for his continuous encouragement and support to pursuit research endeavours and career advancement.

To my friend, companion, and wife Nadia, thank you for being there during my research journey. Without you and our daughter layla, the road would have been more difficult to walk.

Finally, to the two who gave me life, my parents, I would have never been where I am today without your sacrifices and support over the years. Both of you taught me to be resilient and persistent to achieve my goals, and here I am becoming the man you will always be proud of.

A@

Preface & contribution of authors

This Thesis was structured in manuscript-based manner. The Journal of Neurosurgery accepted the original manuscript. The manuscript was edited by adding more comprehensive methodology, results, and discussion in line with the requirement of thesis submission by McGill

University Faculty of Graduate Studies.

The Candidate functioned as the principle investigator for the study, including data collection, data analysis, data integration and writing of the scientific manuscript.

Mr. Muhammad Mullah MSc, Department of Epidemiology, Biostatistics and Occupational

Health, McGill University was instrumental in developing statistical models to carry out the analysis.

Dr. Fahad Alotibi MD. MSc, Hamed Azarnoush PhD, and Gmaan Alzahrani MD MA, helped develop the tier 1, tier 2 and advanced tier 2 metrics.

Dr. Alexander Winkler-Schwartz MD, was essential in the recruitment of medical students, helping to carry out the trial and manuscript revision.

Jawad Fares and Marta Baggiani helped in running the trial.

Dr. Penny Werthner, PhD, Dean Faculty of Kinesiology and Mrs. Sommer Christie BSc

MHK, from University of Calgary were essential in the carrying out of the physiological monitoring during the trial. These results are not a component of my thesis and will be the part of other communications.

AA
Abbreviations

BVR = Brain Volume Removed

MFA= Maximum Force Applied

NRC = National Research Council Canada

PLI = Path Length Index

SFU = Sum of Forces Utilized

STAI = State Trait Anxiety Inventory

SD = Standard deviation

TTPL = Total Tip Path Length

TPR = Tumour Percentage Resected

VR = Virtual reality

AB
Introduction:

Safety and training in the surgical environment

To all surgeons, patient safety and operative efficiency are fundamental objectives during operative procedures. The achievement of these objectives is dependent on the training physicians receive. In the report called “To Err is Human”, the authors suggest that poor surgical training contributed to surgical errors.[1] The introduction of working hour restrictions for residents, by decreasing the number of working hours per week will ultimately lead to decreased surgical experience and training received by residents.[2] As a consequence, this will affect surgical expertise, which is a complex concept that includes both technical and non-technical skills. This expertise develops over time with training and exposure to a multitude of problems along the continuum of learning.[3] In the current system of training, which is based on novice/expert apprenticeship model residents observe the teachers perform operations before they themselves are allowed to operate.[4] However, this method of training decreases surgical efficiency leading to as much as a 35% increase in operating time.[5] The availability of methods that would allow for repetitive yet safe and economically reasonable training would allow trainees to both practice to improve their performance and acquire expertise in non-threatening and patient safe environments.[6, 7]

Surgery as a speciality and neurosurgery as a subspecialty, are highly demanding and stressful careers. Exposure to daily stressors being an integral part of training and every day practice.[8, 9] The advancement in the simulation technology along with the developing a better haptics, more realistic virtual reality scenarios in conjunction with validated metrics has opened the door to more intense surgical skills training and understanding of the set of acquired skills by the trainees.

AC
Stress Concept:

The initial model of stress theory was introduced by Lazarus in 1966 and underwent multiple revisions including the 1986 modification by Lazarus and Folkman.[10, 11] According to their revised stress theory, there are two main concepts of stress: Appraisal and coping. Appraisal is defined as the individual’s specific perception of the stressors, while coping is the effort to manage specific demand. In this model stress is considered to be a relational notion between the individual and the environment.[10, 12] In one definition of stress, stress is defined as “relationship with the environment that the person appraises as significant for his or her well-being and in which the demands tax or exceed available coping resources”.[11] It is also important to understand that acute stress can be real or implied. In the case of surgeons they may not be in direct personal danger but they still can exhibit physiological and psychological reactions in stressful situations when their patents are at risk.[13, 14] It is important to differentiate acute stress from chronic stressors such as burnout and fatigue. The effect of fatigue has been extensively explored in multiple specialities.[15-

18] The studies carried out by Ganju et al to assess the effect of fatigue on neurosurgical residents’ skills found a marginal decrease in task proficiency in the post-call period.[18] The focus of the investigations in this research was on furthering our understanding of the influence of acute stress that surgeons may experience during the resection of cerebral tumors on their psychomotor function.

Simulation and Stress:

Multiple industries recognised the importance of continuous training and assessment in a safe environment using virtual reality platforms. The aviation industry is one of the most advanced in regards to virtual reality training.[19, 20] Simulation technology is also employed in the welding, military, and auto industries.[20-22] Virtual reality platforms allow training in a safe environment

AD
while maintaining the ability to continuously measure, assess and modify participant performance.

In some industries, such as the military, it was shown that operator performance can be affected by stress.[23-25] Training methods, including crisis simulation, focused on improving operator performance and stress coping mechanisms have been implemented.[24] Studies on stress and implementation of crisis resource management principals are not widely available in healthcare.

Anesthetists were one of the first medical disciplines to introduce stress management in training.

This speciality has developed and employed Anaesthesia Crisis Resource Management principals to train their residents on how to manage stress and cope with acute stressors.[26]

The study by Moorthy et al. found that exposure to stressors such as background noise, time constraint, verbal mathematical task or combination of stressors would impair the operator dexterity during the performance of laparoscopic task.[27] These findings were echoed in the laparoscopic surgery task assessment by Hassan et al.[28] and Arora et al.[29]. In the field of neurosurgery, one of the highly stressful fields of surgery, no study has objectively assessed surgeon’s bimanual psychomotor performance during and/or following an acute stress episode.

The impact of acute stress on performance has not been extensively explored in any surgical setting. This is due to the complexity of stress and the unstandardized definition of what constitutes stress along with the difficulty of inducing acute stress in surgical environment while maintaining patient safety. Multiple factors have been identified as causes of stress in the operating room. These factors includes technical factors such as bleeding and learning of new operative technique in addition to non-technical factors such as time constraints, distractions and operating room team members.[27, 30] These stressors if not managed with proper coping mechanism especially among novice surgeons may lead to deleterious effect on surgical performance.[14, 27, 28].

AE
NeuroTouch Virtual Reality Platform:

The National Research Council of Canada (NRC) in collaboration with a group of neurosurgeons developed the NeuroTouch Virtual Reality (VR) Simulator. NeuroTouch is a computer based simulation system that utilizes haptic feedback and finite element mechanics to simulate brain deformation.[31, 32] During the use of the NeuroTouch, participants perform the surgical procedure through the binoculars of a surgical microscope giving the participant a similar

‘feel’ to the microsurgical operative environment. Multiple tools can be used and interchanged to train the individual on a wide range of surgical skills, including: bipolar instruments, micro-scissors and suction aspirator. Surgical instruments are connected to the simulator through a haptic system, which allows the tracking of the movements and delivers the feeling of the tissue the participant is interacting with, based on the programmed mechanical properties.[31-33]

Psychomotor Skills Metrics

Multiple validated metrics were developed at the Neurosurgical Simulation Research Center at Montreal Neurological Institute and Hospital.[6, 34, 35] Psychomotor metrics included tier 1 metrics that are basic metrics that deal with safety and quality related skills such as amount of blood loss, brain volume removed and tumor percentage resected. Tier 2 metrics also deal with safety and include maximum forces applied by the instruments and sum of forces utilized by the instruments.

Advanced tier 2 metrics were developed that deal with efficiency and bimanual interactive skills of the participant. These metrics include: instrument total tip path length, efficiency index, coordination index and ultrasonic aspirator bimanual force ratio.[7, 34, 35]

The availability of the NeuroTouch simulator along with the development of validated assessment metrics that measure bimanual psychomotor performance allows the testing of technical skills under a variety of conditions and circumstances including acute stress.

AF

Manuscript: Acute Stress Impacts Psychomotor Bimanual Performance During Simulated Tumor Resection Task Submitted to Journal of Neurosurgery

The preceding work were augmented with additional material and discussion to reflect the requirement of thesis submission for Master of Science in Experimental Surgery

Manuscript accepted for publications in Journal of Neurosurgery (Manuscript Accepted on 21 May 2015)

AG
Introduction

Severe, difficult to control, intra-operative bleeding is considered by surgeons as one of the most common causes of acute stress in the operating room.[30, 36, 37] Acute stress may be defined as a “relationship with the environment that the person appraises as significant for his or her well- being and in which the demands tax or exceed available coping resources.”[10] It is important to note that acute stress can be real (i.e., threat to personal well-being) or implied (i.e., self-doubt, fear of failure), therefore although surgeons may not be in immediate danger personally, they would exhibit similar physiological and psychological reactions in a stressful situation (i.e., perceived significant risk to his/her patient).[13] During neurosurgical operations on tumors significant bleeding may occur secondary to injury to venous sinuses, major arterial vessels and the multiple small vessels associated with highly vascular tumors. In the majority of cases this bleeding can be brought under control by direct pressure using patties, coagulation of the vessel(s) involved and the use of other hemostatic techniques. Removal of the patties at times is associated with recurrent hemorrhage and the need for further measures to deal with the bleeding. On occasion the bleeding can be so significant that patients develop bradycardia and cardiac arrest may occur. It is difficult to accurately measure the range of psychomotor skills employed by the ‘expert’ in the operating room when faced with such acute stress-inducing events and his/her skills after experiencing such stress.[38] Assessing and/or imparting these ‘expert’ skills to the novice/resident during such complex brain tumor resections may prevent errors that may influence patient safety.

In the military, aviation industry, and high performance sports acute stress has been recognized as a significant factor affecting performance.[19, 39, 40] These fields have implemented training methods and strategies to improve performance and stress coping mechanisms.[19, 25, 40,

41] Anesthetists employ Anesthesia Crisis Resource Management principles to instruct their

AH
trainees to better cope with acute stress encountered during operative procedures.[26] The ability to objectively measure a neurosurgeon’s technical skills during and after acute stress inducing bleeding events in the operating room may lead to a better understanding of how to mitigate the adverse effects of stress on neurosurgical performance.

The NeuroTouch virtual reality (VR) simulator has haptic feedback, uses finite element mechanics to assess brain deformation while at the same time generating real time metrics monitoring tissue removal, blood loss and psychomotor performance. Using this VR platform we have developed and validated a series of technical skills assessment tools, which can objectively measure bimanual psychomotor performance skills.[3, 6, 7, 31, 32, 34, 35, 38] This system allowed us to develop a series of simulated scenarios in which we could assess the effect of acute stress secondary to severe bleeding during brain tumor resection on bimanual psychomotor performance before, during and immediately following the stress inducing event.[7, 34, 35]

The availability of the NeuroTouch VR simulator and validated bimanual performance metrics allowed us to address two questions: 1) What is the impact of acute stress on bimanual psychomotor performance during resection of simulated tumors? 2) Does acute stress influence bimanual psychomotor performance immediately following the stressful episode? The answers to these two questions allowed us to begin developing a framework outlining the specific psychomotor skills that ‘experts’ use during stress-inducing events associated with significant bleeding.

Our initial study hypothesis were that acute stress will have a negative impact on bimanual psychomotor performance during and after the acute stress episode

AI
Methods

All participants in the study signed an approved McGill University Ethical Review Board consent. Including 24 individuals, 6 Board Certified neurosurgeons, 6 senior neurosurgical residents (years 4-6), 6 junior neurosurgical residents (years 1-3), and 6 senior medical students from the same institute. Collected demographic data (appendix 1) included age, sex, handedness, resident training level, hours of video games and musical instruments played weekly.

The State Trait Anxiety Inventory (STAI):

Stress perception was assessed using the validated short form of the State Trait Anxiety

Inventory (STAI) questionnaire adapted to this study.[42, 43] The questionnaire was completed immediately before and after the simulated tasks. The STAI questionnaire consisted of six items

(calm, tense, upset, relaxed, content and worried) as adapted for surgical trials measured on a four point Likert scale (Appendix 3). The minimum achievable STAI score that can be achieved is 6 and the maximum is 24, where higher scores imply greater anxiety perception.

Simulated Virtual Reality Scenarios

To address the study questions 3 simulated brain tumor scenarios involving 2 tumors each,

6 tumors in total, with identical visual appearance and stiffness (Young’s modulus 9kPa) and random bleeding points based on previous tumor scenarios were developed.[7, 31, 35] The color and circular tumor structure chosen for each of the 6 tumors was an identical simulated glioma-like brain tumor appearance derived from an actual patient’s malignant glioma image (Figure 1A).[35]

Each tumor was imbedded in a simulated cortical surface (Young’s modulus 3kPa) which differentiated visually (through different coloration) from the deep white matter like background

B@
(Young’s modulus 3kPa) providing a distinct border interface between the tumor (Young’s modulus

9kPa) and surrounding structures and thus more realism (Figure 1A). To accurately represent the range of human brain and tumor stiffness, a tactile cue in our scenarios, we assessed multiple samples from eleven different human glial tumors immediately after operative removal and measured their brain tumor stiffness (Young’s modulus).[7, 31, 32, 38]

In scenario 1 both tumors were identical and during all resections the participant hears the heartbeat at a regular 60 beats per minute. The first tumor in scenario 2 was similar to those in scenario 1 but the second tumor was specifically designed to be the stress scenario. In this acute stress scenario rapid and uncontrollable bleeding occurred as soon as the tumor was attempted to be removed by the simulated aspirator and/or simulated suction device (sucker). At 75 seconds into the attempted resection and control of bleeding of this tumor the audible heart rate dropped to 30 beats per minute, at 90 seconds to 15 beats per minute and at 105 seconds cardiac arrest was simulated for the remaining 15 seconds of the scenario. Immediately following this event, participants were asked to resect two further tumors scenarios identical to the first three tumors which allowed us to assess bimanual psychomotor performance following an acute stress event.

Simulated Operative Resection Procedures

This study was conducted utilizing the previously described NeuroTouch platform (Figure

1C).[3, 6, 7, 31, 32, 34, 35, 38, 44-46] All operators were acquainted with the system since they had enrolled in a previous trial involving resection of simulated tumors utilizing NeuroTouch.[6]

Each individual who was to participate in the trial was asked if he/she had any knowledge of the trial to be conducted. Two individuals with foreknowledge of the trial’s purpose were excluded.

None of the participants who completed the trial knew that the purpose of the study was to assess

BA
the influence of acute stress, the sequence of the scenarios or had heard of information related to this trial. Each individual who finished the trial was asked not to relate any information related to the trial to any other possible participants. The goal was defined both verbally and in writing as the resection of 6 simulated brain tumors with minimal removal of background tissue, representing simulated “normal” brain tissue. This was accomplished using a bimanual technique utilizing a simulated sucker to control bleeding sites in the non-dominant hand and a simulated ultrasonic aspirator for tumor removal with the dominant hand (Figure 1B, C). The simulated sucker and aspirator were set at functional levels, which resulted in adequate bleeding control and tumor removal as determined in previous studies and could not be modified. Using a predefined sequence, participants began by resecting the left tumor followed by the resection of the right in scenario 1 and followed this sequence for scenarios 2 and 3. In previous studies 3 minutes was allowed for complete resection of a simulated tumor of this shape and size.[6, 7, 35] In this study, all participant were instructed to resect each tumor within 2 minutes time limit to further increase stress and participants could not revisit the left tumor once commencing resection of the right.[6, 7, 35]

NeuroTouch Metrics

Safety, quality and efficiency of simulated operative procedures can be assessed utilizing tier 1 and tier 2 metrics as we have previously described.[6, 7, 35] Independent hand motor skills can also be measured with tier 2 metrics and advanced tier 2 metrics were specifically designed and validated to assess motor and cognitive neurosurgical bimanual skills interaction while achieving the simulated tumor resection goal.[34, 35] Assessed tier 1 metrics included amount of blood loss, tumor percentage resected and brain volume removed. Tier 2 metrics involved total tip path lengths, maximum force utilized and sum of forces utilized by instruments. Advanced Tier 2 metrics

BB
assessed were efficiency index, coordination index, ultrasonic aspirator path length index and ultrasonic aspirator bimanual forces ratio.[34, 35]

Psychophysiological Assessment

The operative environment as well as the resident training environment can be extremely stressful resulting in both physiological and cognitive alterations.[8, 47] Physiological (heart rate, respiration rate, heart rate variability, electromyography, and peripheral body temperature) and neurological (electroencephalography) responses were measured at baseline, during the resection of each tumor, and during rest periods between each scenario (Figure 1C).[48-50] The results of these studies will be the focus of further communications.

Statistical Analysis

Demographics and STAI data have been summarized in terms of mean and standard deviation (SD). Since we assessed 24 participants and each participant performed several operations on similar types of tumor, the repeated outcomes from each subject were considered to be correlated. The correlated outcomes were analyzed using linear mixed effects models.[51] Outcome variables were transformed appropriately (e.g. taking log, square root, etc.) to satisfy the normality assumption of the residuals for fitted models. Results were considered statistically significant only when p-value < 0.05.

BC
Statistical Model

In order to evaluate our objectives, we considered the linear mixed effects models. Since we have 24 participants and each participant performed several operations on similar types of tumor, we thought the repeated outcomes from each subject may be correlated. Therefore, to respect the correlated nature of the data, we adopted linear mixed effects models. We transformed the outcome variables suitably (e.g., log, square root) to justify the normality assumption of the residuals for chosen models.

More precisely, we fit five different types of model of the forms:

Model 1: Yij = (β0 + bi ) + β1 Stress ij + εij

Model 2:

Y = (β + b ) + β Stress + β D + β D + β D ij 0 i 1 ij 21 1ij 22 2ij 23 3ij (D *Stress ) (D *Stress ) (D *Stress ) + β31 1ij ij + β32 2ij ij + β33 3ij ij + εij

Model 3: Yij = (β0 + bi ) + β1 After Stress ij + εij

Model 4:

Yij = (β0 + bi ) + β1 After Stressij + β21D1ij + β22D2ij + β23D3ij

(D * After Stress ) (D *After Stress ) (D *After Stress ) + β31 1ij ij + β32 2ij ijij + β33 3ij ij + εij

Model 5: Yij = (β0 + bi ) + β11D1ij + β12D2ij + β13D3ij + εij where i = 1, …, 24 (participant); j = 1, 2, 3 (tumor 1B, 2A, 2B) for Model 1 and 2: j = 1, 2, 3, 4

(tumor 1B, 2A, 3A, 3B) for Model 3 and 4; j = 1, 2 (tumor 1B, 2A or 3A, 3B) for Model 5; β0, β1,

2 β11, β12, β13, β21, β22, β23, β31, β32, β33 are regression coefficients, bi ~ N(0, σb ) are random effects that induce correlation among responses Yij,

Stressij = 1, if the measurements are taken on tumor 2B

0 , if the measurements are taken on tumor 1B or 2A

BD
D1ij = 1, if participant is a SENIOR resident

0, otherwise

D2ij = 1, if participant is a JUNIOR resident

0, otherwise

D3ij = 1, if participant is a STUDENT

0, otherwise so that STAFF/CONSULTANT serves as reference category,

After Stressij = 1, if the measurements are taken on tumor 3A or 3B

0, if the measurements are taken on tumor 1B or 2A.

Model 1 was used to assess the overall stress effect by comparing the performance on tumor 1B,

2A (pre stress scenarios) versus 2B (stressful scenario). Model 2 was used to evaluate the effect of stress on the participant groups (staff, senior resident, junior resident, student) using (comparing) the data from tumor 1B, 2A to 2B.

Model 3 was used to evaluate performance of the participants before (tumor 1B, 2A) and after

(tumor 3A, 3B) operating the stressful tumor (2B). Model 4 was used to asses the performance of the participant groups (staff, senior resident, junior resident, student) using (comparing) the data from tumor 1B, 2A to 3A, 3B.

Model 5 was fitted to the data measured on tumor 1B and 2A and again to the tumor 3A and 3B to assess the performance of the senior, junior and students as compared to the stuff. All models were fit in ‘R’ using ‘lme’ package.

Each of the above models was fit to analyze the data obtained for 13 metrics (tumor percentage resected, blood loss, brain volume removed, total tip path length for both sucker and aspirator, sum

BE
forces utilized, maximum force applied, efficiency index, suction coordination index, ultrasonic aspirator path length index, and ultrasonic aspirator bimanual force ratio)

Results

Demographics and STAI:

The mean age of neurosurgeons was 42.2 (7.3) compared to 31.5 (2.1) for senior, 28.5 (1.9) for junior residents and 24.8 (3.6) for medical students. Males accounted for 83.3 % of the participants and 83.3% of participants were right handed. Three participants (12.5%) played musical instruments while 8 (33%) reported playing video games distributed among the 4 groups.

Average sleep time on the night before participating in the study was 5.9 (1.3) hours suggesting that the participants were relatively well rested.

Participants showed a statistically significant increase in mean STAI score when baseline (10.2) and post-trial values (12.3) were compared. Neurosurgeons demonstrated the highest mean scores

(12.5) on the baseline STAI questionnaire. However, the only statistically significant higher values were seen between neurosurgeon and the student (8.3) groups. Statistically significant increases in the post study stress perception scores were only seen in junior resident and students groups (Table

1).

Acute Stress and Neurosurgical Performance:

To assess each of our objectives, different linear mixed effects models were used one per metric. Results for the first tumor were not included in the final data since it was considered a rehearsal scenario allowing the participants who had previously used the NeuroTouch platform to reacquaint themselves with the system. Metric measurements from scenario 2 and 3 were combined

BF
to provide baseline (pre-test) performance values before the acute stress scenario 4 and scenario 5 and 6 were combined to assess immediate post-stress operative performance.

Neurosurgeon baseline tier 1 blood loss was significantly lower and baseline tumor percentage resected was significantly higher than that of other groups (Figure 2). Although neurosurgeons had higher mean values of brain volume removed than other groups it was only significantly higher than the student group. As expected during the stress scenario, a significant increase in blood loss was seen in all groups (Figure 2). Tumor percentage resected and brain volume removed decreased significantly in the junior resident and student groups. This trend was seen in the neurosurgeon and senior resident groups but did not reach statistical significance.

The baseline tier 2 mean values for total tip path length sum and maximum force utilized for both instruments were not statistically different among groups. Neurosurgeons had a statistically significant increase in sucker total tip path length during the stress scenario consistent with the concept that they were moving the sucker more vigorously in an attempt to control bleeding (Figure

3). Although one might consider that neurosurgeons may decrease sucker movement to lessen the possibility of injury related undesired sucker contact with brain tissue our results are consistent with increased sucker movement during significant bleeding. However, this was not associated with a statistical change in sucker sum or maximum forces utilized. Senior residents, on the other hand, did not significantly increase sucker total tip path length but did significantly increase their sum and maximum forces utilized with this instrument. Junior residents significantly decreased aspirator total tip path length and both junior residents and students significantly decreased aspirator sum of forces utilized during the stress scenario. This was consistent with their removal of the aspirator from the simulated operative field.

BG
A significant drop in efficiency index, a measure of cognitive-motor skills interaction, calculated as the percentage of time spent actively resecting each tumor divided by total time allowed for the task, was noted during the stressful scenario in all groups (Figure 4). Baseline neurosurgeon efficiency index was significantly higher than all other groups (Figure 5).

Neurosurgeon baseline suction coordination index, a measure of the coordinated use of both instruments (sucker and ultrasonic aspirator) simultaneously, was also significantly higher than the other groups (Figure 5). Suction coordination index declined during the stressful scenario among all participant groups and reached statistical significance in the neurosurgeon, junior resident and student groups. Baseline ultrasonic aspirator path length index, defined as the percentage of ultrasonic aspirator total tip path length interacting directly with the tumor divided by the overall ultrasonic aspirator total tip path length, was significantly higher in the neurosurgeon group when compared to the junior resident and student group (Figure 5). Neurosurgeons showed no significant change in this index during the stress scenario while significant declines were seen in the senior, junior resident and students groups. No significantly different mean baseline ultrasonic aspirator bimanual force ratio, the average forces applied by ultrasonic aspirator and sucker divided by the average forces applied by the aspirator alone, a quality of two-hand interaction, were seen.

Neurosurgeons and senior residents significantly increased their bimanual force ratios during the stress scenario. This is consistent with the concept that they were actively using both instruments to control the bleeding.

Psychomotor Performance Immediately Following a Stress Inducing Event

For the neurosurgeon group, the tier 1 metric, amount of blood loss after the stress-inducing event is the only metric in which a statistical improvement in performance was seen. A significant

BH
decrease in the amount of blood loss was also seen in the senior resident group, as was a significant increase in the percentage of tumor resected. An increase in the tier 1 metrics of percentage tumor resection and brain volume removed were the only statistically significant changes seen in the junior resident group. No significant differences were seen in any of the metrics studied when pre and post stress performance were compared for the student group.

Group Analysis

The stress scenario was designed to prevent the participants from controlling bleeding and a significant increase in this tier 1 metric was seen for all groups. Baseline blood loss was significantly lower and percentage tumor removed was significantly higher in the neurosurgeon group consistent with their ‘expert’ designation. A number of the baseline advanced tier 2 metrics clearly differentiate the neurosurgical group from the other groups, which is consistent with the validated metrics that we have previously reported in Alotaibi et al.[35] Regardless of the group assessed advanced tier 2 metrics bimanual psychomotor performance were predominately the ones significantly altered during the stress scenario. Comparing pre and immediate post stress results only tier 1 metrics were increased suggesting that the majority of the metrics assessed are not significantly changed by stress associated with simulated uncontrollable bleeding in this tumor resection scenario.

Discussion

The NeuroTouch platform[7, 34, 35] with the multiple validated metrics previously developed, along with our ability to develop simulated scenarios in which bimanual psychomotor performance before, during and after a stress scenario can be measured allowed us to examine

BI
specific ‘expert’ technical skills behaviour. The stress scenario was modeled on the intraoperative experience of neurosurgeons subjected to acute stress by severe blood loss, which can result in audible patient bradycardia and very uncommonly cardiac arrest.

What Differentiates ‘Expert’ From Novice/Resident Performance During Acute Stress?

Dissecting neurosurgeon psychomotor performance during this acute stress scenario has multiple interactive aspects but this can be broken down into specific components. Overall, participants showed a statistically significant increase in STAI score when baseline and post-trial values were compared. Neurosurgeons had the highest mean values on the baseline STAI questionnaire suggesting that participating in these simulated scenarios does cause them subjective stress. This may be related to the fact that they know that all aspects of their performance are being assessed and that their performance will be compared to resident groups. Interestingly neurosurgeons and senior residents were the only groups to comment on the questionnaire that they did not experience more stress after the trial compared to before it. In a study by Sexton et al, surgeons were more likely to deny the effect of fatigue on performance compared to pilots and this denial may help to explain these findings in our study.[19]

Figure 6 provides a summary heat map of significant results of tier 1, 2 and advanced tier 2 metrics for each of the groups during the stress scenario and some patterns emerge. For tier 1 metrics all groups have significantly increased blood loss as expected but only the junior residents and students show significant decreases in percentage tumor resected and brain volume removed associated with the bleeding. Neurosurgeon baseline mean percentage tumor resected was 94(6.9) while for senior and junior residents this value was 67(24.3) and 68(17.8) respectively. This fell to

82(19.4) during the stress scenario for neurosurgeons, including two neurosurgeons who

C@
completely resected the tumor despite the bleeding and 54(28.8) and 36(27.5) for the senior and junior residents respectively. Although these falls in resection performance values are somewhat equivalent, neurosurgeons perform better during the stress episode than residents at baseline. To improve resident performance during such bleeding episodes more focus on their ability to carry out resections both before and during stress scenarios would appear warranted.

For tier 2 metrics the major significant differences are found between neurosurgeons and senior residents. Neurosurgeons significantly increase sucker total tip path length in an attempt to decrease blood loss, improve visibility, and resect tumor while not significantly increasing their sucker sum and maximum forces utilized. Senior residents, on the other hand, do not significantly increase sucker total tip path length but significantly increase sucker sum and maximum forces utilized. While operating on a patient’s tumor the neurosurgeon has no method to know the forces he/she or a resident being supervised is applying with the operating instruments. Brain injury resulting from the utilization of too much force over time or the improper application of excessive force can result in increased patient morbidity. If ‘expert’ psychomotor behaviour in this simulated bleeding scenario is equivalent to ‘expert’ operating room performance, it suggests that neurosurgeons have learned to increase sucker movement and control sucker forces during bleeding situations. This improves visualization and tumor resection while mitigating the risk of instrument excessive force on brain tissues that may impact patient safety. Using these insights, residents can train to increase sucker movement and modulate sucker forces when controlling bleeding utilizing simulators like NeuroTouch that provide haptic feedback. Visual readouts of a resident’s instrument movement and force application compared to ‘expert’ proficiency performance benchmarks could be continuously provided during simulated operative procedures.[7, 35] This would allow learners to immediately modify his/her instrument behaviour during the simulated task or latter with

CA
instructor assessment and guidance. Junior residents use both significantly less aspirator total tip path length and sum of forces utilized suggesting decreased maximization of aspirator usefulness during bleeding, a simulation addressable issue.

Advanced tier 2 metrics results help explain several tier 1 and 2 results. Neurosurgeons had significantly higher baseline efficiency, suction coordination and ultrasonic aspirator path length indices (Figure 4) resulting in their ability to remove higher tumor volumes at baseline, during and after the stress event. The stress scenario is also associated with a significant increase in ultrasonic aspirator bimanual force ratios index that assesses the quality of two hand interaction in the neurosurgeon and senior resident groups. This suggests that the coordinated use of the aspirator and sucker are important component of ‘experts’ trying to deal with the bleeding in the stress scenario.

These activities are also amendable to VR simulator training.

Psychomotor Performance Following Acute Stress

Individuals after experiencing an acute stress event can decrease their psychomotor performance.[8, 13, 14, 39] Alternatively, they can be more attentive and focused resulting in improved psychomotor performance immediately following the stress-inducing episode.[14] Acute stress may also have no significant influence on subsequent performance.[14] Scenarios 5 and 6 were specifically designed to assess these questions. An alternate reason for increased performance involving a series of similar scenarios could be a learning effect of repetition.[52]

Significant improvements in neurosurgeon and senior resident group bimanual psychomotor performance after the stress scenario were seen in the tier 1 metric of blood loss. Neurosurgeons and senior residents being unsure of but anticipating increased amounts of bleeding, in subsequent scenarios, may have focused on the bleeding control aspect of their performance. No psychomotor

CB
bimanual performance metrics assessed in this study demonstrated a statistical significant decrease in any group when baseline values were compared to values immediately after the stress-inducing episode of simulated bleeding and patient arrest. Therefore this stress scenario is not associated with deteriorated participant bimanual psychomotor performance. Senior and junior residents had improved percentage of tumor resected after the acute stress episode which was associated with significant increases of brain volume removed for junior residents. Although it is not possible to eliminate learning effect as a reason for these improvements the fact that these occurred only in 2 specific tier 1 metrics, only in 2 groups, rather than in multiple metrics is more consistent with their improved performance after stress. However for all tier 2 and advanced tier 2 metrics no significant change in psychomotor performance after the stress was seen in any group.

Strengths and Limitations

NeuroTouch technology has allowed us to address the two questions proposed by this study and provided some insights into how ‘expert’ (neurosurgeons) and novices (residents) actually perform simulated brain tumor operations during and after a stress-inducing event. The baseline tier 1 metrics of blood loss and of total percentage tumor resected differentiated the neurosurgeon and resident groups and provides construct validity for these metrics utilizing the NeuroTouch simulator. These metrics were not validated in two previous studies but this may relate to changes in tumor structure, surrounding normal tissue and resection sequence in the present study.[34, 35]

The development of validated metrics in our previous study for advanced tier 2 metrics allowed the comparison of results and further validation for the NeuroTouch platform using data from this study.[35] The fact that 2 of 3 baseline tier 1 metrics differentiate neurosurgeon and resident groups and collaboration of the interesting finding that for 3 of 4 advanced tier 2 metrics studied the major

CC
improvement in operative skills occurs after residency training helps to confirm the utility of these bimanual metrics in simulation studies.[35] Serially tracking residents during training and after graduation is necessary to understand psychomotor skills acquisition sequence during residency and modification of these skills during neurosurgical practice. Guided by these validated metrics it seems reasonable to propose the development of a series of proficiency based benchmarks helping to develop training curriculum and self-assessment programs to maximize resident performance during the resection of cerebral tumors using the NeuroTouch simulator.[34, 35, 38, 42]

Caution is needed in the interpretation of our results. First, this study was focused on the effect of acute stress on psychomotor bimanual performance and should not be generalized to the influence of chronic stressors like fatigue and other conditions.[7, 48, 53] Second, operators were only allowed to use a simulated sucker and ultrasonic aspirator to control significant bleeding during the simulated tumor resection. This is not representative of the multiple complex interactive skills and techniques available to address this problem during patient tumor resections. The task quality, difficulty and short duration of the stress scenario may not adequately discriminate performance among the limited number of operators studied. These issues along with the type of metrics employed may have resulted in our inability to find more significant differences. These metrics are being assessed for their usefulness in other simulated neurosurgical operations and by other surgical specialties that may aid the universality of their application.[3, 33, 44, 45] Third, utilization of neurosurgeons, residents and students from only one institution may have resulted in our inability to find more significant differences between groups. Fourth, the performance of each participant was not videotaped and therefore was not available for assessment by a standardized scale since no such validated scale has been developed for simulated procedures in neurosurgery.[54] Fifth, the relationship of psychomotor performance to stress induced

CD
physiological changes is important in the dissection of operator response to anxiety inducing situations such as severe bleeding and will be addressed in further communications. But, despite the use of multiple stressors in the trial “e.g. time limitation, simulated acute bleeding, simulated patient cardiac arrest” these simulated acute stressors will not fully represent the anxiety level of real life situation where surgeons are aware they will not be able to hit the restart button and repeat the scenario again.[31, 50] This study was focused on the assessment of the influence of acute stress on the bimanual psychomotor performance but the utility of VR simulators like NeuroTouch will be limited unless it can be demonstrated that they enhance resident operating room performance.[38, 42, 53]

Overall Conclusion and Future Directions

Conclusion:

Our study is the first to demonstrate that acute stress initiated by simulated severe intra- operative bleeding during the resection of a simulated brain tumor significantly decreased bimanual psychomotor performance during the acute episode. The simulated intra-operative bleeding event had no significant influence on tier 2 and advance tier 2 metrics monitored during immediate post stress operative performance but significantly improved tier 1 control of bleeding in the neurosurgeon and senior resident group.

Future Directions

With the advancement in the virtual reality simulation field, it is becoming more critical to understand and objectively analyse ‘expert’ response and performance in multiple simulated settings resembling real life situations. As this study demonstrates the complexity of surgical

CE
bimanual performance and the multiple factors that could affect performance including acute stress and level of expertise have to be assessed if we are to further our understanding of surgical

‘expertise’.[3] The ultimate aim of any surgical training model and curriculum should be to prepare

‘trainees’ with both improved coping strategies and systems to advance their technical and non- technical skills in safe, non-threatening, and reproducible environments. Development of patients’ specific simulation scenarios accompanied by validated objective assessment metrics and multiple scenarios with a variety of simulated complications appears to be an obtainable goal. It will be essential to show that these virtual reality scenarios not only improve resident performance in simulation training but translate into enhanced resident performance in the operating room and thus have concurrent validity. The Neurosurgical Simulation Research Centre is collaborating on a NIH

R18 grant application on the development of objective measures of intraoperative neurosurgical performance that is being developed by the Dr. Josh Bederson team at Mount Sinai Hospital in New

York. This study in addition to the availability of validated simulation scenarios with validated metrics developed by the Neurosurgical Simulation Research Centre at McGill will provide the underpinning necessary to establish concurrent validity.

The results of studies such as described in this thesis can form the essential background for developing proficiency performance benchmarks [6] which then can be the foundation for creating curriculum to improve residency technical skills. Such curriculum will need to be tested in multiple centers worldwide. There are fifteen members of the NeuroTouch Consortium spread across three continents and an important issue facing these centers is the development of curriculum based on standardized and validated performance metrics.[32] After accomplishing these goals, it will be important to continuously modify these curriculum incorporating new proficiency performance benchmarks in an ongoing dynamic process based on updated research. For these goals to become

CF
reality, further understanding of the ‘expert’ and ‘novice’ performance under different simulations is required and the NeuroTouch simulator and the NeuroTouch consortium may provide one avenue to achieve these objectives.

CG
References

A8
(!'5
885
88
(++" '5
'
88
('%,('5
$
''
(
*"#2
* ! #

'
!)

.()"8
B@@@7
-"('%
&",
+,,8

B8
''25
885
-
%85
"%)
$
)
=;4$*'
,$'
,
$#
'( #)
"'#.
$%') +

-%' #3
&+"'
(.+'%
(
.+ +25
B@@E8
;F<7
)8
IDG:I8

C8
%"',:!'. 5
8
'
88
%
,-+(5
*' !
-%') (
#
#*'$(*''.2
#)') #

)$'.
#)$
%') 3
.+(,.+ +25
B@AC8


7
)8
C@:H8

D8
3'"$5
8 85
 #
#
)() #
)# !
( !!(3
!
&+"'
(.+'%
(
.+ +25

AIIC8
 ;C<7
)8
CEH:CFA8

E8
"'.5
8 85
-
%85

9$()9
$
$%') +
)' # #
$'
(*' !
'( #)(3
+!"/,
(

.+ +25
B@@D8
;D<7
)8
CFF:I6
",.,,"('
CFI:G@8

F8
%!+'"5
85
-
%85
'$  #.
'$'"#
#"' (
$'
"$+!
$
 "*!)
' #

*"$'(
( #

 ')*!
! ).
 "*!)$'
*'$$*3
(.+'%
(
.+ "%
.-"('5

B@AE8

G8
3+'(.,!5
85
-
%85
*'$(*' !
+ ')*!
'! ).
( "*!) $#
")' (
)$
((((

%(.$"$)$'
( !!(
*' #
' #
)*"$'
'() $#3
'-+'-"('%
(.+'%
(
(&).-+

,,",-
"(%( 2
'
.+ +25
B@AD8

H8
%'5
885

)(
$
*)
()'((
$#
%'$'"#2
"%! ) $#(
$'
!)

%'$(( $#(
*) $#3
&"
""'5
B@@I8
 ;A@
.))%<7
)8
BE:CC8

I8
%&'5
885
88
(.'5
'
8 8
-#+5
$,
)$
#) .
)

$

! 
#
)
' 2

*'#$*)
#
#*'$(*''.3
(+%
.+(,.+ +25
B@AC8
;E<7
)8
AAA:C8

CH
A@8
3+.,5
8
'
8
(%$&'5
$# ) +
$' (
$
)'((
#
)
((*
$
 '*!' ).5
"'

.#" (
$
)'((5
8
))%2
'
8
+.&.%%5
"-(+,8
AIHF5
)+"' +
8
)8
FC:H@8

AA8
3+.,5
885
$% #
)$'.
#
'('2
%()1
%'(#)1
#
*)*'3
,2!(,(&-"

""'
29 .'5
AIIC8
;C<7
)8
BCD:BDG8

AB8
+(!'5
885
)'((
#
$% #
$' (5
"'
#)'#) $#!
#.!$% 
$
)
$ !
:

+ $'!
 #(5
8 888
%-,5
"-(+8
B@@A5
+ &('7
1 (+8
)8
AEAFC:AEAG@8

AC8
0'5
88
'
88
"' "%5

$#%)
$
!!$()( (
#
 $!$.
#
 $"  #3

(+&(',
'
!/"(+5
B@@C8
;A<7
)8
B:AE8

AD8
+(+5
85
-
%85

"%)
$
()'((
$#
(*' !
%'$'"#2

(.()") 
'+ ,
$
)

! )')*'3
.+ +25
B@A@8
;C<7
)8
CAH:C@5
CC@
A:F8

AE8
!(%5
85
-
%85
)
$
) *
$#
%(.$"$)$'
#
$# ) +
( !!(3
!
&+"'

(.+'%
(
.+ +25
B@@H8
 ;B<7
)8
AIE:B@D8

AF8
+'-!+(/5
885
-
%85
%'$($% 
%'$'"#
)'
$#
# )
$#
!!
#

(*' !

%')"#)2
%'$(%) +
()*.3
 5
B@@A8
;GCBC<7
)8
ABBB:ABBC8

AG8
'",(5
885
88
+.&&('5
'
88
+/',-"'5

)
$
) *
$#
)

%'$'"#
$

( "*!)
#()) 
"$# )$' #
)( 3
(.+'%
(
%"'"%
('"-(+"' 5

AIHG8
;A<7
)8
BB:BD8

AH8
'#.5
85
-
%85

)
$
!!
$#
#*'$(*''.
'( #)(8
( !!(2
"%! ) $#(
$'
%$! .

'' #
'( #)
!!
%' $(3
(.+'%
(
.+(,.+ +25
B@AB8
;C<7
)8
DGH:DHB8

AI8
1-('5
885
8 8
!(&,5
'
8 8
%&+"!5
''$'1
()'((1
#
)",$'
#
"  #
#

+ ) $#2
'$((
() $#!
(*'+.(3
 5
B@@@8
;GBCG<7
)8
GDE:I8

CI
B@8
%%5
8 8
'
8 8
 5
+!*) #
)
) +#((
$
! )
( "*!)$'(
$'
)' # #

$")
( !!(2

'+ ,3
!
"'-+'-"('%
#(.+'%
(
/"-"('
),2!(%( 25
AIIH8
;C<7
)8

BBC:BDB8

BA8
"' +'5
8:85
 # )
!"#)
"$! #
#
( "*!) $#
$
,! #
%')
<2
#'(

$"%!- ).3
(.+'%
(
-!+&%
,-+,,,5
B@@A8
;B<7
)8
ADA:AIB8

BB8
!(&$5
8 85
 "*!) $#1
!'# #
#
:
%'$'"#2
+ #
'$"
*)$"$) +

+!$%"#)3
,+!
(%"25
AIIH8
;A<7
)8
EE:GD8

BC8
+",$%%5
885
-
%85
)'((
-%$(*'
)' # #2
#
+#)4(
%%'$3
+ (+&'

.'+
,-+,,5
B@@H7
)8
BGA:BHF8

BD8
+",$%%5
88
'
8
%,5
+'$" #
)
)(
$
()'((
$#
" ! )'.
%'$'"#2

*"#
)$'(1
)' # #1
#
(!) $#
()') (3
'(($
(
&"%"-+2
),2!(%( 25
AIIA7

)8
AHC:AIC8

BE8
(/(5
85
8
(($5
'
8
+,('5
! )'.
'* )
' # #2
#
'#
$'
)'((4$% #

 !!(5
"'
)'((
*) $#
#
'+#) $#5
8
"!'.&
'
8
+&$(5
"-(+,8

AIHI5
)+"' +
8
)8
CGG:DAH8

BF8
(%3&'5
885
-
%85
#()( 
' ( (
'($*'
"#"#)2
!4! 
( "*!) $#
)' # #

#
$%') #
'$$"
' ((3
(.+'%
(
%"'"%
',-!,"5
AIIE8
;H<7
)8
FGE:FHG8

BG8
((+-!25
85
-
%85

)
$
()'((4 #* #
$# ) $#(
$#
)
%'$'"#
$


!%'$($% 
)( 3
.+ "%
'(,()2
'
-!+
'-+/'-"('%
!'"*.,5
B@@C8
;I<7

)8
ADHA:ADHD8

D@
BH8
,,'5
85
-
%85
) +
()'((4$% #
()') (
"$#
#$+ (
#
(*''.
$''!)

, )
%$$'
+ ')*!
!%'$($% 
%'$'"#3
+"-",!
(.+'%
(
.+ +25
B@@F8
 ;AB<7
)8

AEED:I8

BI8
+(+5
85
-
%85
)'((
"% '(
%(.$"$)$'
%'$'"#
#
#$+ 
!%'$($% 
(*'$#(3

.+ "%
'(,()25
B@A@8
;A@<7
)8
BEHH:IC8

C@8
+(+5
85
-
%85
# #
#)'$%') +
()'((2
,)
$
(*'$#(
,#)
'$"

' ( (

)' # #
%'$'"0
&+"'
(.+'%
(
.+ +25
B@@I8
;D<7
)8
ECG:DC8

CA8
%(+&5
85
-
%85
*'$$*2

%.( (4(
+ ')*!
( "*!)$'
$'
'# !

" '$#*'$(*''.
)' # #3
.+(,.+ +25
B@AB8
 ;A
.))%
)+-"/<7
)8
CB:DB8

CB8
!(.!.+25
85
-
%85
*#"#)!(
$
*'$(*''.2
 ')*!
! ).
( (
$'
' # #

#
+!*) $#
$
# !
 !!(3
(+%
.+(,.+ +25
B@AC8
;E<7
)8
I:AI8

CC8
(,,.5
85
-
%85

+!$%"#)
$

+ ')*!
( "*!)$'
$'
)' # #
#*'$(*'$#(
)$

%'$'"
#
%')
#$($% 
#$#(!
)'#((%#$ !
(*''.3
.+(,.+ +25
B@AC8



7
)8
HE:IC8

CD8
%(-""5
885
-
%85
*'$(*' !
(((("#)
$
)' (
#!* #
*"#)
#

-)' ).
( #
)
 ')*!
! ).
 "*!)$'
*'$$*
6 
)' (73
.+ "%

''(/-"('5
B@AE8

CE8
%(-""5
885
-
%85
(((( #
 "#*!
'$'"#
#
' #
*"$'
() $#
 )

*'$$*1

 ')*!
! ).
 "*!)$'3
.+(,.+ +25
B@AE8


7
)8
HI:IH8

CF8
+(+5
85
-
%85
)$'(
$"%'$" ( #
().
#
(*''.2
()'((*!
+#)(
#
)
$%') #

'$$"3
&+"'
(.+'%
(
.+ +25
B@A@8
;A<7
)8
F@:E8

DA
CG8
-3%5
885
-
%85

)(
$
()'((
$#
(*' !
%'$'"#3
&+"'
(.+'%
(

.+ +25
B@@F8
 ;A<7
)8
E:A@8

CH8
%"',:!'. 5
85
-
%85
(((( #
%'$'"#
#
' #
)*"$'
'() $#
*( #

#$+!

+ ')*!
'! ).
( "*!)$'3
'-+'-"('%
(.+'%
(
(&).-+
,,",-
"(%( 2
'

.+ +25
B@AD8
;A<7
)8
A:I8

CI8
/",5
85
88
"&5
'
8
(+-,('5
%$')
%(.$%.( $!$.
#
%
%'$'"#

%%! ) $#(
$
()'((
"#"#)3
+"'")%,
'
)+-"
(
,-+,,
&' &'-5
B@@G7

)8
FAE:FCG8

D@8
(.%5
8
'
8
2'+5
(.$!$ !
%'%') $#
$'
)
!."% 
"(3
(.+'%
(

)(+-,
"',5
B@@I8
;AC<7
)8
ACIC:D@H8

DA8
.%%+5
85
).
!(($#(
) #
'$"
)
 '! #(3
+"-",!
(.+'%
(
.+ +25
B@@D8
 ;D<7

)8
CIC:D8

DB8
+(+5
85
-
%85

"%' !
)'((
(((("#)
$$!
6 72

( !1
'! !
#
+! 

%%'$
)$
"(*' #
()'((
#
)
$%') #
'$$"3
(+%
(.+'%
(
.+ +25
B@A@8

;H<7
)8
AGEF:FC8

DC8
+-.5
88
'
8
$$+5

+!$%"#)
$

( -4 )"
($')4$'"
$
)
())
(!

$
)
% !''
))5' )
#- ).
#+#)$'.
6 73
+"-",!
(.+'%
(
%"'"%

,2!(%( 25
AIIB8
;C<7
)8
C@A:C@F8

DD8
+,!'25
85
-
%85
+!$%"#)
$
)
 !!
( "*!)$'
$'
#$($% 
( #*(
(*''.2


#,
 4 ! ).
+ ')*!
'! ).
( "*!)$'
$'
#$($% 
( #*(
(*''.3
&+"'
#(.+'%

(
+!"'(%( 2
=
%%+ 25
B@AD8
;D<7
)8
CC@:D8

DB
DE8
+,!'25
85
-
%85

 !!
( "*!)$'
$'
#$($% 
( #*(
(*''.
6 72


+! ) $#
()*.3
(.+'%
(
-(%+2' (%( 2
:

=
$
.+ +25
B@AD8
;A<7
)8
D@8

DF8
(%%(025
85
-
%85
%')$'
-%' #
)'" #(
%'$'"#
#

( "*!)

$"%*)'4(
' #
)*"$'
'() $#
)( 3
'-+'-"('%
(.+'%
(
(&).-+
,,",-

"(%( 2
'
.+ +25
B@AE8

DG8
%!5
885
88
+",!% 5
'
88
!' %-5
)'((
#
*'#$*)
"$#
(*'$#(2

#'()# #
#
"# #
)
(.#'$"
#
+$  #
)
+'(
$#(&*#(3

+!"/,
(
.+ +25
B@@I8
;D<7
)8
CGA:CGF8

DH8
+ +-
.)
'
''2
+-!'+5
# #
)
)'((
(%$#(2

(
$

 $
#
*'$
, )
!."% 
)!)(3
"( $5
B@AA8
;C<7
)8
IB:

ID8

DI8
(,5
85
-
%85
%) " / #
" '$(*' !
( !!(
, )

#*'$ 3


.+(,"'5
B@@I8
7
)8
HG8

E@8
0+-!'+5
85
8
!+",-"5
'
8
.)5
*'$
#
 $
)' # #
, )

!."% 
)!)(3
.+((''-"(',5
B@AC;.&&+<7
)8
CB:CG8

EA8
"+5
88
'
8 8
+5
#$"4)(
"$!(
$'
!$# )* #!
)3
"(&-+",5

AIHB8
;D<7
)8
IFC:GD8

EB8
())+5
885
8 8
&",('5
'
88
0",5
'# #
*'+(
#
(*' !
%') 3

(,- +.-
"%
(.+'%5
B@@G8
 ;IHF<7
)8
GGG:GGI8

EC8
"+$&'5
885
-
%85

*(
$
( "*!) $#
#
#*'$(*' !
*) $#
#
)' # #3

(.+'%
(
.+(,.+ +25
B@AD8
;B<7
)8
BBH:DF8

DC
ED8
"+$&2+5
885
-
%85
*' !
( !!
#
$"%! ) $#
')(
)'
' )' 
(*''.3
0

' %'
(.+'%
(
""'5
B@AC8
;AE<7
)8
ADCD:DB8

DD

Tables

DE

Table.1 State Trait Anxiety Inventory (STAI) Results of each group pre and post experiment

STAI pre STAI post experiment experiment Staff 12.5 (4.6) 13.5 (2.1) Seniors 10 (2.3) 10.8 (3.9) Juniors 10 (2.7) 13.6 (3.1)* Students 8.3 (1.9) 11.5 (5.1)* Overall 10.2 12.4* Shown results are Means (SD) * Indicate statistically significant result (P value <0.05)

Table 1: State Trait Anxiety Inventory (STAI) Results Results are Means (SD) for n=6 for each group * Indicates statistically significant result (p<0.05).

DF

Figures

DG

Figure 1: A) View of the operating scene. (B) Mannequin head with two haptic devices, which provide force feedback for the simulated sucker and ultrasonic aspirator. (C) View of left handed operator using NeuroTouch with a simulated ultrasonic aspirator in the left hand and sucker in the right along with the physiological monitors utilized during this study.

DH

Figure 2: Tier 1 metrics of participant groups Tier 1 mean (+/-SD) in cubic centimeters of neurosurgeon (n=6), senior (n=6), junior resident (n=6) and medical student (n=6) groups at baseline, during and immediately following the stress scenario. Statistical differences between baseline, stress and post stress values are represented as *p<0.05 and **p<0.01. Left upper: Arrows indicate which groups are statistically different from other groups (p<0.05).

DI

Figure 3: Tier 2 metrics of participant groups Tier 2 mean (+/-SD) of neurosurgeon (n=6), senior (n=6), junior resident (n=6) and medical student (n=6) groups at baseline, during and immediately following the stress scenario are outlined in millimeters for total path length and in Newtons for forces. Statistical differences between baseline, stress and post stress values are represented as *p<0.05 and **p<0.01.

E@

Figure 4: Advanced tier 2 metrics of participant groups Advance tier 2 mean (+/-SD) of neurosurgeon (n=6), senior (n=6), junior resident (n=6) and medical student (n=6) groups at baseline, during and immediately following the stress scenario are outlined. Statistical differences between baseline, stress and post stress values are represented as *p<0.05 and **p<0.01.

EA

Figure 5: Statistical differences of advanced tier 2 metrics of participant groups Advance tier 2 mean values of neurosurgeon (n=6), senior (n=6), junior resident (n=6) and medical student (n=6) groups at baseline, during and immediately following the stress scenario are outlined. Arrows indicate which groups are statistically different from other groups (p<0.05).

EB

Figure 6: Heat map of statistically significant affected metrics when baseline and stress scenario values were compared (p<0.05).

EC

Appendices

ED

Appendix 1 Personal Data Form

This form will remain confidential and will not be made public. The data entered here will serve for group stratification during the analysis of the data.

Name: ______Age: ______

Sex: M □ F □ Handedness: Left □ Right □ Ambidextrous □

1. You are a: ☐ Medical Student a)
Year in medical school: ☐1 ☐2 ☐3 ☐4 b)
Are you interested in surgery? ☐Yes ☐No ☐Not sure a.
If yes, please indicate what type of surgery you are interested in (Please indicate if you are not sure): ______c)
Have you done a surgical rotation before? ☐Yes ☐no

☐ Neurosurgical resident a)
Level of training: PGY ______b)
Fellow: Year of fellowship: ______

☐ Staff Neurosurgeon a)
Years in practice: ______b)
Approximate number of meningioma cases done: ______

2. On average, how many hours per week do you play a musical instrument? (Write 0 if you don’t play an instrument): ______

3. On average, how many hours per week do you play video games (write 0 if you don’t play video games): ______

a. Please specify the type of video game you play the most:

EE
-First-person-shooter (i.e. Call Of Duty)
-Sports (i.e. NHL 2011)
-Real-time strategy (i.e. Starcraft)
-Role playing (i.e. World of Warcraft)
-Wii / XBOX Kinect/ PS3 move games
-Other (specify): ______

4. Please indicate: a. Approximate number of hours of sleep you had last night: ______b. On average how many hour of sleep per night have you had over the last week: ______

5. Are you currently “on call”? ☐Yes ☐No a. If yes, for how long have you been “on call”? ______

6. How would you describe your current workload? ☐Manageable ☐Moderate ☐Overwhelming Please describe what your workload details (e.g. academic, private practice, education): ______

7. Are you currently experiencing any physical discomfort (i.e., pain, injury, muscle tension, headache etc.)? ☐Yes ☐No a. If yes, please describe: ______

8. Is anyone in the room responsible for evaluating your performance, teaching you, or otherwise related to you (friend, classmate and/or colleague)? ☐Yes ☐No a. If yes, please describe the relationship: ______

9. In general what is your role for surgical procedures (lead surgeon, assistant)? ______

10. How familiar are you with the procedure for brain tumor removal? ☐ Little to no experience ☐ Some experience ☐ Very experienced

EF
11. Have you ever used the NeuroTouch simulator before? ☐Yes ☐No a. if yes, please indicate when, how long, and for what purpose: ______

11. Have you ever used the any neurosurgical simulator before? ☐Yes ☐No a. if yes, please indicate when, how long, and for what purpose: ______

12. Have you heard anything about what is involved in this trial other than that we will be measuring your performance on the neurosurgical simulator? Yes No a. If yes, please explain what you have heard: ______

EG
Appendix 2 Evaluation Form

The goal of this trial was to assess the effect of stress on surgical performance during the removal of simulated brain tumor

1)
On a Scale from 1-5, please rate the difficulty of each challenge (1- very easy, 5- very hard) a.
Scenario #1, tumour #1 1 2 3 4 5 b.
Scenario #1, tumour #2 1 2 3 4 5 c.
Scenario #2, tumour #1 1 2 3 4 5 d.
Scenario #2, tumour #2 1 2 3 4 5 e.
Scenario #3, tumour #1 1 2 3 4 5 f.
Scenario #3, tumour #2 1 2 3 4 5

2)
On a scale from 1-5, please rate the realism of the simulated stressful experience (1- very easy, 5- very hard) 1 2 3 4 5

3)
On a scale from 1-5, please rate the visual realism of the CUSA simulation task (1-completely unrealistic, 5-completely realistic) 1 2 3 4 5

4)
On a scale from 1-5, please rate the sensory realism (the feel of the different tissues) of this CUSA simulation task (1-completely unrealistic, 5-completely realistic): 1 2 3 4 5

5)
On a scale from 1-5, please rate your overall satisfaction with this CUSA simulation task (1-completely unsatisfied, 5-completely satisfied):

1 2 3 4 5

6)
If this simulator was available in my department, I would use it for maintaining my technical skills. (1-completely disagree, 5-completely agree):

1 2 3 4 5

EH

7)
If this simulator was available in my department, I would encourage my residents to use it to improve and assess their technical skills.

1 2 3 4 5

8)
How would you improve this CUSA simulation task? Please explain: ______

9)
Other comments concerning the trial. ______

EI
Appendix 3 Validated State Trait Anxiety Inventory Questionnaire Post- Test

A number of statements which people have used to describe themselves are given below. Read each statement and then circle the most appropriate number to the right of the statement to indicate how you feel right now, at this moment. There are no right or wrong answers. Do not spend too much time on any one statement but give the answer which seems to describe your present feeling best.

Not at all Somewhat Moderately Very much

1-
I feel calm 1 2 3 4

2-
I am tense 1 2 3 4

3-
I feel upset 1 2 3 4

4-
I am relaxed 1 2 3 4

5-
I feel content 1 2 3 4

6-
I am worried 1 2 3 4

Please make sure that you have answered all the questions.

F@