An international census of robotic urological practice and training

Archie Hughes-Hallett, BMBS1, Erik K Mayer, PhD1, Philip J Pratt, PhD1,2, Alex Mottrie, MD3, Ara Darzi, FRS1,2, Justin Vale, MS1

1. Department of Surgery and Cancer, Imperial College London2. The Hamlyn Centre for Robotic Surgery, Imperial College London3. OLV Clinic, Aalst, Belgium

Corresponding Author

Erik MayerDepartment of Surgery and Cancer, Imperial College London, St Marys Hospital Campus, London, W2 1NYe.mayer@imperial.ac.uk

Key words:

Word count manuscript text: 2,310

Word count abstract: 300

Word count abstract + manuscript = 2,603

Target Journal: European Urology (IF 10.476)

Key words: Robotic surgery, training, partial nephrectomy, prostatectomy, cystectomy

Take home message

Significant numbers of urologists do not feel their robotic training needs have been met. It is time to review robotic training to assess how best to train the robotic surgeons of the future.

Abstract

Background: The landscape of urological practice has changed with the increasing use of robots in urological surgery and has altered the training needs of the modern urological surgeon.

Objective: To determine the current state of robotic urological practice, to establish the way in which robotic training has been delivered and to ascertain whether this training was felt to be adequate.

Design, Setting and Participants: A questionnaire was emailed to robotic surgeons utilising the EAU robotic urology section mailing list.

Outcome measures and statistical analysis: Outcomes were subdivided into three groups; demographics, exposure and barriers to training, and delivery of training. A comparative analysis of the experiences of trainees and independently practising robotic surgeons was performed. Results and Limitations: The survey was distributed to 1423 surgeons. 239 of the 828 participants who opened the email completed the survey giving a response rate of 28.9%. The majority of respondents performed Robotic Assisted Laparoscopic Prostatectomy (90.6%) and were undertaking more than 50 robotic cases a year (55.6%). Overall 66.3% of respondents felt their robotic training needs had been met. When trainees and independently practising robotic urologists were analysed separately, trainee satisfaction was significantly lower (51.6% versus 71.6%, p=0.01). When a subgroup analysis of trainees was performed examining the relationship between regular simulator access and satisfaction, simulator access was a positive predictor of satisfaction, with 87.5% of those with regular access compared to 36.8% of those without access being satisfied (p<0.01).

Conclusions: This study has revealed that a significant number of urologists do not feel that their training needs have been met. Increased access to simulation, as part of a structured curriculum, appears to improve satisfaction with training and would, simultaneously, allow for a proportion of a surgeon’s learning cure to be removed from the operating room, potentially improving patient safety and reducing cost.

2

1. Introduction

Since the introduction of telerobotic surgical systems at the end of the last century, the growth of robotic surgical practice in urology has been exponential [1]. This growth in the use of robotic platforms has occurred without a simultaneous shift in the manner by which training takes place, and at a time when there are increasing time constraints for training with the 80-hour working week in North America and the European working time directive.

Concerns exist both over the quality of robotic surgical training [2–4] and the effect robotic practice has had on urological training in general [5]. Training in laparoscopic surgery has evolved since its inception in the 1980s with the creation of multiple simulation centres where surgeons can acquire skills outside of the operating room. This has been augmented with numerous published validated laparoscopic skills [6,7] and procedure-specific curricula [8]. Although some robotic skills (both virtual reality [9–11] and real world [12,13]) and procedure-specific [14] training tasks and curricula [15,16] have been validated, the extent of their implementation remains unknown.

The objectives of this study were to undertake a census of contemporary robotic practice and training, and secondarily to establish whether there is a correlation between satisfaction with robotic training and the manner in which it had been delivered.

2. Methods

2.1 Survey design and administration

An anonymous web based questionnaire (Appendix 1) was distributed to the mailing list of the robotic section of the European Association of Urology (EAU). The survey was designed in accordance with the available recommendations for web-based surveys [17,18] and constituted part of a larger survey of robotic urologists. Prior to distribution, it was trialled by both native and non-native English speakers and feedback was obtained from six urologists; two trainees and four consultant surgeons.

The questionnaire was designed using LimeSurvey (www.limesurvey.com) and hosted on their website. A link to the survey website was embedded in the covering letter and responses were captured using an automated process. The questionnaire was dynamic with the number and nature of the responses tailored towards the individual based on their previous answers. The questionnaire was open for one month, with a single reminder sent out after initial dissemination.

A survey request was considered valid if the potential participant opened the email containing the link to the survey. A valid respondent was defined as an individual who had answered at least the first five questions of the questionnaire. The response rate was calculated as the ratio between valid respondents to valid requests [17]. Incomplete but valid responses were not excluded.

3

2.2 Survey Analysis

Statistical analysis was performed in GraphPad Prism (GraphPad Software, La Jolla California USA, www.graphpad.com).

The analysis of Likert scales was performed by scoring data from 1-7 (strongly agree to strongly disagree). This allowed tests of statistical significance to be performed on the data and mean ranks to be generated. The Mann-Whitney U test was used to establish significance when comparing unpaired data. Wilcoxon’s rank-sum test was used to compare paired data.

Analysis of ordinally-ranked data was performed by scoring responses according to their ranking; a ranking of seven resulted in a score of one, a ranking of six resulted in a score of two and so forth. This allowed the generation of a ranking score to be used for statistical analysis (rank sum/number of respondents within the specific rank). Group comparison of consultants versus trainees was performed using the Kruskal-Wallis test and comparisons of the individual responses of trainees and consultants were performed using the student t-test.

Logistic regression was used to model for the effect of various training modalities on perceived adequacy of training. The Hosmer-Lemeshow test was used to assess the goodness of fit; with a higher p value equating to increased model accuracy.

For all statistical tests a p-value of <0.05 was taken to confer statistical significance.

When computing fractions or percentages the denominator was the number of relevant responses (i.e. all respondents; all respondents to have assisted in or performed robotic procedures; robotic surgeons or trainees) to a given question. This number varied as not all respondents answered all questions.

3. Results

3.1 Demographics

The survey was sent out to a total of 1,423 recipients. Of these, 828 (58.2%) opened the invitation email, with 239 individuals completing the survey (179 complete and 60 incomplete responses), giving a response rate of 28.9%. In all 71 trainees and 168 surgeons who had completed their training responded, and of the 168 who had completed training 117 performed robotic surgery. The majority of the respondents were currently working in robotic centres (194/239, 81.5%, figure 1), and had both trained (210/239, 87.9%) and worked in Europe (215/239, 90.0%). In all, 211 of the 239 (88.7%) respondents had undertaken or assisted in robotic urological procedures.

4

68.6% (77/112) of consultant robotic surgeons had a trainee working for them with 82.2% (71/77) offering robotic training to their trainee. 73.2% (82/112) had mentored a consultant colleague in a robotic procedure.

The majority of surgeons perform more than 50 cases per year (55.6%, 65/117) (figure 2) with robotic prostatectomy most commonly undertaken (figure 3). 3.2 Exposure and barriers to training amongst robotic surgeons and trainees

The majority of respondents had received ‘on-patient training’ either as the scrubbed assistant (67%) or on the console (70%) (Table 1). Interestingly, when robotic surgeons were analysed as a subgroup, 29.9% (35/117) were performing robotic surgery having received their training solely on patients.

When examining access to ‘off-patient’ training the largest group had received simulation training (57.3%, i.e. 121 of the 211 respondents who had received training in, performed or assisted in robotic surgery) with 40% (84/211) having had regular simulator access of one form or the other. This simulation had been undertaken as robotic skills training on the console (36.5%, 77/211) and/or, virtual reality simulation (42.7%, 90/211). Specific to units offering robotic surgical training, no significant difference in regular access to simulation training facilities was seen between robotic surgeons and trainees (47% (55/117) versus 38% (16/42), p=0.31).

Both robotic surgeons and trainees ranked lack of access to the robotic platform as the greatest barrier to training, with an average ranking score of 5.23 (out of a maximum of 6). The second most highly ranked barriers to training were lack of access to simulation training and a trainer’s operative learning curve. Rankings of barriers to training were found to be statistically significantly different within robotic surgeon and trainee groups (p<0.01 and p<0.01 respectively). However, no statistically significant difference was seen between trainee and robotic surgeon mean ranking scores for individual barriers to training (figure 5).

3.3 Delivery of training and its correlation to satisfaction

Overall 66.3% (134/212) of respondents agreed that their robotic training needs had been met, although satisfaction was lower for trainees than robotic surgeons (51.6% (28/55) versus 71.6% (106/157), p=0.01). Similarly trainees were less likely to agree that there training could not have been improved (figure 4). Both robotic surgeons and trainees were less satisfied with their robotic training compared with their general urological surgical training (robotic surgeons - 3.63 versus 4.03, p<0.01) (trainees - 3.07 versus 3.87, p<0.01). Significantly more trainees with regular access to a simulator (either skills-based console time or VR simulator) were satisfied with their training 87.5% (2/16) versus 36.8% (14/38), p<0.01).

The only independently significant predictor for training needs being met amongst trainee surgeons was access to a virtual reality simulator (OR 16.2,

5

p=0.03), with fellowship training and regular access to simulation (either VR or real world) tending towards significance (OR 12.7 and 12.7, p=0.07 and 0.08 respectively) (Table 2). Amongst robotic surgeons, console time was the only independently significant predictor (OR 4.1, p<0.01), with fellowship training also demonstrating a trend (OR=2.5, p=0.09). Goodness of fit was excellent for both the trainee and robotic surgeon models (Hosmer-Lemeshow values of p=0.997 and 0.607 respectively).

4. Discussion

This study, for the first time, has provided an overview of robotic urological training in Europe. The key finding is the apparent inability of robotic surgical training, as it currently stands, to meet the needs of a significant proportion of surgeons. This discontent was most apparent amongst trainees, with only 51.6% feeling their training needs had been met, but was also an issue amongst post-training surgeons who rated their robotic surgical training as significantly worse than their general urological training.

The reason for the dissatisfaction with robotic training is almost certainly multifactorial, and will differ depending an individual’s stage of training. Trainees, particularly those who are at an earlier stage of training, are more likely to put emphasis on skills based training and as such, perhaps unsurprisingly, access to simulation was shown to be a better predictor of satisfaction in training than console time (respective OR of 12.7 and 3.5). This is in contrast to post-training surgeons for whom the only significant predictor of satisfaction in training was console time.

The distance between the trainer and trainee may explain, in part, the differences in satisfaction between robotic and general urological training. During conventional laparoscopic and open surgery the trainee and mentor operate side-by-side with the role of primary surgeon in constant flux. This is in stark contrast to robotic surgery where the trainee will often be the scrubbed assistant or observing, resulting in the loss of this close training proximity and a subsequent reduction in experience as the primary operator. This decrease in training opportunities associated with the introduction of a surgical robot into a training environment was investigated by Robinson et al [5] who found that 68% of trainees felt that the introduction of a da Vinci robotic platform had been detrimental to their training. A potential solution to this increasing distance between surgeon and trainee is through the use of a dual console. This solution currently remains underutilised with only 16.6% of surgeons currently having access to the facility; this may be, in part, due to the expense of the technology.

A group of surgeons left unanalysed in previous, consensus gathering, questionnaires were those surgeons who had been trained after the completion of their formal urological training. Interestingly the post-training group in our survey had significantly higher satisfaction than the trainee group, 71.6 versus 51.6%. This may be due to a number of factors, with perhaps the most significant being a greater focus on procedure specific

6

rather than skills based training combined with increased access to on-console training.

Indeed a significant proportion of independently practising robotic surgeons (29.9%) had received the entirety of their training ‘on-patient’. This ‘on-patient’ apprenticeship, although a crucial part of a surgeon’s training, should represent a component rather than the entirety of training [19,20]. Palter and colleagues [20] examined the use of a formal curriculum of training for laparoscopic cholecystectomy and demonstrated that a surgeon’s learning curve could be largely removed from the operating theatre.

Although similar studies examining comprehensive robotics curricula have not been performed, literature exists supporting the notion, both for conventional laparoscopic [20–23] and robotic simulation [10,11,15], that proficiency at simulated surgical tasks is associated with improved surgical performance [24]. In addition to this precedent a validated, fundamentals of robotic surgery curriculum, has been proposed [16] and was shown to improve the skills of novice robotic surgeons. Palter et al’s [20] findings would suggest that if this curriculum was expanded to include procedure-specific training, it could remove elements of the robotic learning curve from the operating room improving quality and safety for the patient. Specific examples include the potential to improve oncological outcomes [25] and reduce the financial burden of a surgeon’s learning curve, estimated at $217,000 for robot assisted radical cystectomy [26].

The data presented from our study adds to a growing body of research [4,5,27,28] examining the effect of the introduction of robotic surgery on training. Previous survey-based studies [4,5,27] have also reported that robotic surgical training, as it stands, is insufficient to meet the needs of a significant proportion of trainees. A recent survey of robotic general surgery fellows in the USA found that only 33% of fellows were satisfied with their robotic surgical training [4], which equates favourably with the figure of 51.6% amongst urological trainees from our study. This low level of satisfaction amongst both urology trainees, and to a greater extent general surgical robotic fellows, may be, in part, due to robotic surgeons not having reached the plateau phase of their learning curve. This learning phase translates to ‘trainers’ being potentially less likely, or indeed comfortable, to relinquish the role of primary surgeon to their trainee.

Despite the additional insight this questionnaire has provided to robotic practice and training, it is not without its limitations. The most notable is the self-reported nature of the questionnaire meaning the data presented can only be seen to give a surrogate and subjective measure of training quality. However, until an objective scoring system for robotic training has been developed and validated this form of questionnaire obtained measure remains practical. A further limitation results from the way in which the data was grouped for analysis, in particular within the trainee cohort. Within this group no distinction was made between differing levels of training because the sample size was not big enough to stratify in this way; ideally trainees would

7

have been analysed separately according to postgraduate years of experience.

5. Conclusions

The study reinforces the perceived need for improved robotic training amongst robotic surgeons and trainees. Future training in robotic surgery should be based on structured, validated, curricula tailored to the individuals for whom it is directed. Increased access to simulation, as part of these curricula, has the potential to improve satisfaction with training and would allow for a proportion of the surgeon’s learning cure to be removed from the operating room. These improvements to robotic training have the potential to improve the quality and safety of robotic surgical practice and provide cost savings to healthcare systems.

References

[1] Hofer MD, Meeks JJ, Cashy J, Kundu S, Zhao LC. Impact of increasing prevalence of minimally invasive prostatectomy on open prostatectomy observed in the national inpatient sample and national surgical quality improvement program. J Endourol 2013;27:102–7.

[2] Ben-Or S, Nifong LW, Chitwood WR. Robotic surgical training. Cancer J 2013;19:120–3.

[3] Zorn KC, Gautam G, Shalhav AL, et al. Training, credentialing, proctoring and medicolegal risks of robotic urological surgery: recommendations of the society of urologic robotic surgeons. J Urol 2009;182:1126–32.

[4] Shaligram A, Meyer A, Simorov A, Pallati P, Oleynikov D. Survey of minimally invasive general surgery fellows training in robotic surgery. J Robot Surg 2013;7:131–6.

[5] Robinson M, Macneily A, Goldenberg L, Black P. Status of robotic-assisted surgery among Canadian urology residents. Can J Urol 2012;6:160–7.

[6] Fried GM, Feldman LS, Vassiliou MC, et al. Proving the Value of Simulation in Laparoscopic Surgery. Ann Surg 2004;240:518–28.

[7] Sweet RM, Beach R, Sainfort F, et al. Introduction and validation of the American Urological Association Basic Laparoscopic Urologic Surgery skills curriculum. J Endourol 2012;26:190–6.

[8] Ikonen TS, Antikainen T, Silvennoinen M, et al. Virtual reality simulator training of laparoscopic cholecystectomies - a systematic review. Scand J Surg 2012;101:5–12.

8

[9] Lee JY, Mucksavage P, Kerbl DC, et al. Validation study of a virtual reality robotic simulator--role as an assessment tool? J Urol 2012;187:998–1002.

[10] Finnegan KT, Meraney AM, Staff I, Shichman SJ. da Vinci Skills Simulator construct validation study: correlation of prior robotic experience with overall score and time score simulator performance. Urology 2012;80:330–5.

[11] Hung AJ, Zehnder P, Patil MB, et al. Face, content and construct validity of a novel robotic surgery simulator. J Urol 2011;186:1019–24.

[12] Hung AJ, Jayaratna IS, Teruya K, et al. Comparative assessment of three standardized robotic surgery training methods. BJU Int 2013:1–9.

[13] Dulan G, Rege R V, Hogg DC, et al. Developing a comprehensive, proficiency-based training program for robotic surgery. Surgery 2012;152:477–88.

[14] Hung AJ, Ng CK, Patil MB, et al. Validation of a novel robotic-assisted partial nephrectomy surgical training model. BJU Int 2012;110:870–4.

[15] Stegemann AP, Ahmed K, Syed JR, et al. Fundamental Skills of Robotic Surgery: A Multi-institutional Randomized Controlled Trial for Validation of a Simulation-based Curriculum. Urology 2013.

[16] Suh I, Mukherjee M, Oleynikov D, Siu K-C. Training program for fundamental surgical skill in robotic laparoscopic surgery. Int J Med Robotics Comput Assist Surg 2011;7:327–33.

[17] Eysenbach G. Improving the quality of Web surveys: the Checklist for Reporting Results of Internet E-Surveys (CHERRIES). J Med Internet Res 2004;6:e34.

[18] Edwards PJ, Roberts I, Clarke M, et al. Methods to increase response to postal and electronic questionnaires ( Review ). The Cochrane Collaboration 2010.

[19] Aggarwal R, Darzi A. Innovation in surgical education - A driver for change. Surgeon 2011;9:S30–1.

[20] Palter VN, Orzech N, Reznick RK, Grantcharov TP. Validation of a Structured Training and Assessment Curriculum for Technical Skill Acquisition in Minimally Invasive Surgery: A Randomized Controlled Trial. Ann Surg 2012;00:1–7.

[21] Kroeze SGC, Mayer EK, Chopra S, et al. Assessment of laparoscopic suturing skills of urology residents: a pan-European study. Eur Urol 2009;56:865–72.

9

[22] Banks EH, Chudnoff S, Karmin I, Wang C, Pardanani S. Does a surgical simulator improve resident operative performance of laparoscopic tubal ligation? Am J Obstet Gynecol 2007;197:541.e1–5.

[23] Kirby TO, Numnum TM, Kilgore LC, Straughn JM. A prospective evaluation of a simulator-based laparoscopic training program for gynecology residents. J Am Coll Surg 2008;206:343–8.

[24] Aggarwal R, Darzi A. From scalpel to simulator: a surgical journey. Surgery 2009;145:1–4.

[25] Atug F, Castle EP, Srivastav SK, et al. Positive surgical margins in robotic-assisted radical prostatectomy: impact of learning curve on oncologic outcomes. Eur Urol 2006;49:866–71; discussion 871–2.

[26] Yu H, Hevelone ND, Lipsitz SR, et al. Comparative analysis of outcomes and costs following open radical cystectomy versus robot-assisted laparoscopic radical cystectomy: results from the US Nationwide Inpatient Sample. Eur Urol 2012;61:1239–44.

[27] Sfakianos GP, Frederick PJ, Kendrick JE, et al. Robotic surgery in gynecologic oncology fellowship programs in the USA : a survey of fellows and fellowship directors. Int J Med Robotics Comput Assist Surg 2010;6:405–12.

[28] Duchene D a, Moinzadeh A, Gill IS, Clayman R V, Winfield HN. Survey of residency training in laparoscopic and robotic surgery. The Journal of Urology 2006;176:2158–66; discussion 2167.

10

Appendix 1

11

Question Asked Question TypeDemographics

In which country do you usually practice? Single best answer

In which country have you undertaken the majority of your training? Single best answer

At what stage of training are you? Multiple choiceHave you ever assisted in, performed or received training in robotic surgery?

Yes or No

Are urological robotic procedures performed where you currently work?

Yes or No

Robotic ExperienceWhich robotic urological procedures do you perform? Multiple Choice*How many robotic procedures have you been table-side assistant for?

Single best answer

How many robotic procedures have you performed part of? Single best answer

How many robotic procedures have you been primary surgeon for? Single best answer

How many robotic urological procedures do you perform per year? (as the primary surgeon)

Single best answer

Robotic TrainingWhich robotic training modalities have you had access to? Multiple ChoiceDo you have regular access to a robotic simulator and/or skills based console-training time?

Yes or No

Do you feel this training met your robotic training needs? Yes or NoMy general urology surgical training to date could not have been improved.

Likert Scale

My robotic surgical training to date could not have been improved. Likert ScaleDo you currently have a trainee(s) working for you? Yes or NoHave you ever trained a consultant colleague to perform a robotic procedure?

Yes or No

Which modalities do you utilise to train your trainee(s)? Multiple ChoiceWhich modalities do you utilise to train your consultant colleague? Multiple ChoiceDo you have access to a dual console setup to train your trainee? Yes or NoDo you have access to a dual console setup to train your consultant colleagues?

Yes or No

What do you think is the greatest barrier to training in robotic surgery?

Ranking

How do you think training in robotic surgery could be improved? Free textQuestions in italics were only asked to post training robotic surgeons*Also asked to provide the number of each procedure performed

Overall (n=211)

Trainees (n=55)

Surgeons post training (n=156)

p-value

On-Console 70.1% 58.2% 74.4% 0.03Scrubbed Assistant 66.8% 61.8% 68.6% 0.41Classroom 36% 27.3% 39.1% 0.14VR 42.7% 45.5% 41.7% 0.64Box trainer 36.5% 29.1% 39.1% 0.19Formal robotic curriculum 19.4% 14.5% 21.2% 0.33Access to dual console 16.6% 16.4% 16.7% 1.00Fellowship 23.7% 12.7% 27.6% 0.03Table 1 – The way in which robotic training is currently being delivered

Training modality Robotic Surgeons (n=148)

Trainees (n=54)

OR p-value 95% CI OR p-value 95% CIConsole Time 4.1 <0.01 1.7 – 10.1 3.5 0.14 0.1 – 18.9Table-Side Assistance 1.6 0.31 0.6 – 3.9 0.8 0.77 0.1 - 4.7Dual Console 2.7 0.17 0.7 – 11.0 0.7 0.75 0.1 – 8.3Virtual Reality Simulator 1.3 0.58 0.5 – 3.7 16.2 0.03 1.2 – 211.1Box Trainer 0.5 0.19 0.2 – 1.4 0.4 0.56 0.0 – 7.6Regular Simulator Access 1.5 0.5 0.5 – 4.6 12.7 0.08 0.8 – 208.6Formal Curriculum 1.4 0.53 0.5 – 3.9 0.8 0.89 0.1 – 13.4Classroom Teaching 0.6 0.20 0.2 – 1.4 4.2 0.31 0.0 – 3.7Fellowship 2.5 0.09 0.9 – 7.5 12.7 0.07 0.8 – 206.7

12

Table 2 – Odds of reporting training needs being met according to training modalities.

Number of Surgeons

Figure 1

Geographical distribution of respondents working in European robotic centres

<4

8-11

12-14

23-41

5-7

15-22

13

Figure 2

Figure 3

14

0 20 40 60204060

Figure 4

Figure 5

15