- Email: [email protected]
Transfer of Laparoscopic Radical Prostatectomy Skills From Bench Model to Animal Model: A Prospective, Single-Blind, Randomized, Controlled Study
Investigative Urology
Transfer of Laparoscopic Radical Prostatectomy Skills From Bench Model to Animal Model: A Prospective, Single-Blind, Randomized, Controlled Study Robert Sabbagh, Suman Chatterjee, Arun Chawla, Jen Hoogenes, Anil Kapoor and Edward D. Matsumoto* From the Division of Urology, Department of Surgery, Faculty of Medicine and Health Sciences, Université de Sherbrooke (RS), Sherbrooke, Québec and Division of Urology, Department of Surgery, Faculty of Health Sciences (SC, AC, AK, EDM) and Departments of Surgery and Clinical Epidemiology and Biostatistics (JH), McMaster University, Hamilton, Ontario, Canada
Purpose: Learning laparoscopic urethrovesical anastomosis is a crucial step in laparoscopic radical prostatectomy. Previously we noted that practice on a low fidelity urethrovesical model was more effective for trainees than basic suturing drills on a foam pad when learning laparoscopic urethrovesical anastomosis skills. We evaluated learner transfer of skills, specifically whether skills learned on the urethrovesical model would transfer to a high fidelity, live animal model. Materials and Methods: A total of 28 senior residents, fellows and staff surgeons in urology, general surgery and gynecology were randomized to 2 hours of laparoscopic urethrovesical anastomosis training on a urethrovesical model (group 1) or to basic laparoscopic suturing and knot tying on foam pads (group 2). All participants then performed timed laparoscopic urethrovesical anastomosis on anesthetized female pigs. A blinded urologist scored subject videotaped performance using checklist, global rating scale and end product rating scores. Results: Group 1 was significantly more adept than group 2 at the laparoscopic urethrovesical anastomosis pig task when measured by the checklist, global rating scale and end product rating (each p ⬍0.05). Time to completion was similar in the 2 groups. No statistically significant difference was noted in global rating scale and checklist scores for laparoscopic urethrovesical anastomosis performed on the urethrovesical model vs the pig. Conclusions: Training on a urethrovesical model is superior to training with basic laparoscopic suturing on a foam pad for performing laparoscopic urethrovesical anastomosis skills on an anesthetized female pig. Skills learned on a urethrovesical model transfer to a high fidelity, live animal model.
Abbreviations and Acronyms CL ⫽ task specific checklist GRS ⫽ global rating scale LRP ⫽ laparoscopic radical prostatectomy LUA ⫽ laparoscopic urethrovesical anastomosis UVM ⫽ urethrovesical model Submitted for publication June 16, 2011. Study received approval from Hamilton Health Sciences and St. Joseph’s Healthcare Hamilton research ethics boards. Supported by internal funding from the Department of Surgery, Division of Urology, McMaster University. * Correspondence: Division of Urology, McMaster University, St. Joseph’s Healthcare Hamilton, 50 Charlton Ave. East, Hamilton, Ontario L8N 4A6, Canada (telephone: 905-521-6186; FAX: 905-308-7205; e-mail: [email protected]).
Key Words: prostate; laparoscopy; surgical anastomosis; models, educational; clinical competence LAPAROSCOPIC radical prostatectomy is a minimally invasive surgical approach for localized prostate cancer.1,2 Many surgeons learning to perform LRP consider intracorporeal suturing and knot tying one of the more daunting parts of the surgery. Training and teaching laparoscopy is a costly, time-consuming
process associated with a steep learning curve. For complex laparoscopic surgery the literature suggests that practice outside the operating room in specially equipped training facilities, such as an animal or surgical skills laboratory, may improve this learning curve.3– 6
0022-5347/12/1875-1861/0 THE JOURNAL OF UROLOGY® © 2012 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION
Vol. 187, 1861-1866, May 2012 Printed in U.S.A. DOI:10.1016/j.juro.2011.12.050
AND
RESEARCH, INC.
www.jurology.com
1861
1862 TRANSFER OF LAPAROSCOPIC RADICAL PROSTATECTOMY SKILLS FROM BENCH TO ANIMAL MODEL
A question in surgical education remains unanswered. Does surgical skills training in a laboratory transfer to the real world operating room environment? We explored some factors attributable to this transfer from ex vivo to in vivo minimally invasive surgical environments. There are several bench models to teach a learner how to perform LUA in the setting of a surgical skills laboratory, including chicken skins and legs, and porcine bladders and urethras as models in pelvic trainers.7–11 However, some models lack the anatomical elements considered essential to simulate the procedure.12 High fidelity models provide surgical environments that are more realistic and allow a learner to practice procedures from beginning to end.13,14 Such models include animal models, cadavers and commercial simulators.15 Animal models, such as the pig model used to train in laparoscopic surgery, allow for realistic tissue handling and bleeding.14 However, animal models are costly, require veterinary assistance and raise social and ethical questions.16 Commercially available high fidelity simulators offer the advantages of reusability and realistic anatomy but they are costly and require maintenance.12 In a recent single-blind, randomized, controlled study we developed and tested a low fidelity latex UVM to practice LUA on a pelvic trainer (fig. 1).17 Our main objective was to evaluate the impact on the performance of intracorporeal urethrovesical anastomosis by comparing the outcome of task specific bench model training (LUA on the UVM) to that of a basic laparoscopic suturing and knot tying task done on a foam pad. According to CL and GRS scores subjects who trained on the UVM performed LUA more quickly and effectively than those who trained on the foam pad. Results suggested that task specific training on a UVM would be an effective way of learning how to perform LUA. Taking what we learned from these findings, we considered it necessary to next assess the transferability of skills from a low fidelity model to a higher fidelity model. We determined whether skills learned on the UVM would transfer to live anesthetized pigs.
We further assessed the face validity of the pig as a surrogate for human urethrovesical suturing.
MATERIALS AND METHODS The study was approved by the Hamilton Health Sciences and St. Joseph’s Healthcare Hamilton research ethics boards. Participants in this prospective, single-blind, randomized, controlled study were the same as those enrolled in the study reported previously.17 Participants were senior surgical residents (postgraduate year 4 –5), fellows or staff surgeons in urology, general surgery or gynecology. All participants provided written informed consent to be in the study and were provided a unique identification number to maintain confidentiality. Participants were recruited via surgical rounds and online message board advertisements. Study exclusion criteria included whether the participant had ever performed LRP. All participants completed a questionnaire on their laparoscopic experience. To show participants how the skills chosen for this study related to the actual procedure they were shown a 15-minute video of LUA during LRP. Participants were randomly assigned to groups using sealed envelopes prepared by our statistician. For the training portion of the study members in group 1 were allocated to practice on the latex UVM to perform running urethrovesical suture anastomoses on a body pelvic trainer, similar to the way that actual anastomoses are done in the operating room. The UVM had anatomical and tissue characteristics, ie thickness, similar to those of the human bladder and urethra (fig. 1). Subjects were given specific instructions and feedback during the practice session by an experienced laparoscopic urologist. Those allocated to group 2 were assigned to practice random suturing and knot tying on a foam pad (Limbs & Things, Bristol, United Kingdom). Subjects were instructed to place single sutures across the foam and tying. As in group 1, group 2 received feedback throughout the training session. After 2 hours of training the groups were asked to perform a post-training laparoscopic suturing test consisting of 2 tasks. The first task was to complete 5 figures-of-8 and 3 knot ties on the foam pad using a 2-zero silk suture on an SH needle (Ethicon, Johnson & Johnson, Markham, Ontario, Canada). The second task was to perform running LUA on the UVM. Training was previously described in detail.17
Figure 1. A, UVM. B, UVM positioned in pelvic trainer simulated orientation in male pelvis and beginning of posterior running suture anastomosis. C, resident performs laparoscopic anastomosis with UVM in pelvic trainer.
TRANSFER OF LAPAROSCOPIC RADICAL PROSTATECTOMY SKILLS FROM BENCH TO ANIMAL MODEL 1863
End-product rating Candidate Identification No. Please rate the candidate's performance on the following scale (maximum total score 25): Tissue integrity
Torn tissues
1
Suture placement
Disorganized suture alignment
1
Watertight
1
2
2
More than 3 sutures passed through the urethral catheter
4
Overall and anastomosis score
Poor
3
5 Most of sutures are well aligned
4
Minimum to moderate leak, more than 100 cc of methylene blue- normal saline return after aspiration
2
5 Intact tissues
3
2
Urethral catheter integrity
1
4
Almost half of the sutures are aligned
Huge leak. Most of the methylene bluenormal saline leak into the pelvis
1
3 Minimally torn
3
5 No leak
4
5
Less than 3 sutures passed through the urethral catheter Urethral catheter intact
2
3
4
Good
5 Excellent
Figure 2. End product rating form used to score LUA done by each participant in female pig
After the exercise in the dry laboratory all participants performed LUA on a high fidelity bench model, that is an anesthetized female pig. Subjects used 2, 15 cm 3-zero Monocryl® sutures (1 dyed and 1 not dyed) tied tail to tail with an RB needle at each end. Before starting this task the pig urethrovesical junction was transected laparoscopically. Participants performed LUA on each pig within a 2-hour time frame. Participant performance was video recorded and timed. Performance was evaluated by a blinded urologist using a CL and a GRS. All evaluated items were adapted from a previously validated CL and GRS.17–19 An end product rating was also used to evaluate pig LUA (fig. 2). Pig LUA watertightness was assessed by injecting 100 ml methylene blue tinted saline via a Foley catheter. All participants were given a questionnaire designed to preliminarily evaluate face validity to assess the UVM and the high fidelity model. Statistical analysis was performed using SigmaStat®, version 3.10.0. Based on an earlier study of bench model training in ureteroscopy20 we found a difference of 10.2 between the surgical training intervention arm and a didactic group. Using these data and a pooled SD of 5 we calculated a sample size of 6 participants per arm. To account for the possibility of study dropouts, and for differences in the complexity of the procedure and training intervention compared with those of the original ureteroscopy study we recruited an additional 8 participants per group. This allowed us to detect a difference of 15% between the 2 groups. Comparisons between groups 1 and 2 were made using the Mann-Whitney U rank test with p ⬍0.05 considered statistically significant.
RESULTS A total of 28 participants were enrolled in this study. At baseline no difference was detected in previous laparoscopic experience, surgical specialty distribution or training level between the groups.17 For the pig urethrovesical anastomosis task group 1 had a significantly higher mean CL score (p ⫽ 0.02), mean GRS score (p ⫽ 0.005) and mean end product rating (p ⫽ 0.04) than group 2 (see table). However, group 1 did not perform LUA more quickly than group 2 (p ⫽ 0.38, see table). The transfer of LUA performance skills from the UVM to a high fidelity model was assessed in the 2 groups (table). For group 1 vs 2 there was no statistically significant difference in the GRS score (p ⫽ 0.427 vs 0.9) or the CL score (p ⫽ 0.567 vs 0.646) LUA performance results in female pigs and skill transfer from bench to pig model
Female pig: Performance time (mins) CL score (max 12) GRS score (max 35) End product rating score (max 25) UVM:17,† CL score (max 12) GRS score (max 35)
Mean ⫾ 95% CI Group 1
Mean ⫾ 95% CI Group 2
50.1 ⫾ 14.5 11.0 ⫾ 1.2 29.9 ⫾ 1.7 20.5 ⫾ 2.4
58.3 ⫾ 13.3 8.9 ⫾ 2.0* 23.2 ⫾ 1.5* 15.9 ⫾ 2.0*
10.9 ⫾ 1.2 29.6 ⫾ 3.7
8.1 ⫾ 1.9 22.8 ⫾ 3.1
* p ⬍0.05 vs group 1. † Not significantly different vs female pig.
1864 TRANSFER OF LAPAROSCOPIC RADICAL PROSTATECTOMY SKILLS FROM BENCH TO ANIMAL MODEL
between LUA done on the UVM and on the pig (table). The end study questionnaire revealed that all participants (100%) thought that operating on the pelvis of the live anesthetized pig was an excellent representation of a human and virtually identical to the video of urethrovesical anastomosis performed on a patient. Participants also thought that training on the UVM significantly helped them acquire the skills needed to accomplish the circular anastomosis.
DISCUSSION Traditional surgical training, that is the “see one, do one, teach one” model, is becoming outdated. This model emphasizes intraoperative teaching, which relies on having time during a surgical case to teach. The staff surgeon has the challenge of balancing the needs of the learner with the need to ensure that the case is done in timely fashion. However, with the increasing financial constraints faced by many health care systems and hospitals, greater pressure is placed on the staff surgeon to teach less and operate more quickly. Also, with the emergence of new technologies such as laparoscopy and robotics it is common to see the staff surgeon balancing learning new skills while teaching them to others. This has necessitated an evolution in how surgeons are trained with defocusing on intraoperative practice and a shift toward training outside the operating room in a surgical skills laboratory. Several studies have shown that training on bench models or simulators and computer animated devices improves surgical skills.19,21–23 However, low fidelity bench models do not simulate the exact operating room setting. It is difficult to simulate a bleeding vessel or reproduce the feeling of tissue consistency. On the other hand, low fidelity bench models are economical, can be used repetitively and do not need the special care required by live animals. High fidelity models have a closer resemblance to actual anatomy and provide an operative field similar to real life operating room situations. They may be more important to teach more complex procedures. Sidhu et al performed a randomized, controlled trial of the effects of low and high bench model fidelity on vascular anastomosis.24 Results revealed better skill transfer from the bench model to live animals when practicing on high fidelity models. However, live animals are more expensive and may be a source of infection, and their use carries ethical concerns. Bench models offer the convenience of being available to a learner with little preparation and planning. Some surgical skills lab-
oratories are accessible 24 hours per day, offering flexibility for a busy surgical resident schedule. Thus, bench models will continue to have an important role in surgical skills training. Thus, it becomes more critical that the models are valid, reliable teaching tools. In our study LUA on the pigs was performed better by group 1 (UVM model) than group 2 (control) (table). This supports the notion of task specific surgical skills training, in that a model incorporating the critical steps of a procedure will ultimately be a better tool. Furthermore, global assessment and specific checklist scores for LUA on the pig were similar to those on the UVM. Thus, subjects in group 1 performed LUA as well as they did on the bench model. These results suggest that the skills learned with the UVM bench model transferred to live anesthetized female pigs, a higher fidelity model. Although there was no difference in time to task completion between the groups, notably quickness does not necessarily mean better surgical performance. This is why surgical skills assessment should also always include ratings of performance to capture the other important nuances of surgery, such as tissue handling, finesse, efficiency of motion, etc. This may explain why there was no difference in time to completion between the groups but there were better performance scores for the group trained using the task specific UVM. Task specific training is extremely important to acquire a new skill.24 Reviews of extensive experience with flight simulation revealed the importance of training designed to attain task specific performance goals.25–27 Thus, practicing the bench model surgical maneuvers, which are crucial in a complex procedure, should increase the chance of being able to perform the procedure successfully in the operating room. McDougall et al compared a silicone model and pelvic trainer to a virtual reality simulator to teach laparoscopic suturing and knot tying.15 When transferred to a task specific procedure, ie laparoscopic cystorrhaphy in a porcine model, there was no difference in performance. This study suggests that task specific training is more critical than the model for successful skills acquisition. For task specific training a great deal of thought and planning must be put into the design of the training program. Even a high technology virtual reality surgical simulator can be a short-lived teaching tool if trainees are not taught the steps specific and relevant to the actual procedure. To our knowledge we report the first study to use an anesthetized pig as a high fidelity in vivo model to validate the efficacy of LUA training on a bench model. The best way to validate our UVM would have been to perform LUA on a patient. However,
TRANSFER OF LAPAROSCOPIC RADICAL PROSTATECTOMY SKILLS FROM BENCH TO ANIMAL MODEL 1865
for ethical reasons this would be unacceptable due to participant lack of experience. Live anesthetized pigs represent a high fidelity model for urethrovesical anastomosis training. The urethrovesical junction and the pelvic dimensions of the pig have anatomical and tissue characteristics similar to those of humans. The bladder and urethra are in the pelvis at the same anatomical site as in humans. We used female pigs since it is easier to cut the urethrovesical junction and perform the anastomosis. The LUA performed on the pig mimics the urethrovesical anastomosis in humans except for the narrower pelvis in the pig. Working on a slightly more challenging model may possibly prepare the learner for the real procedure. Performing LUA on the anesthetized pig was an excellent surrogate for a human model, providing the much needed tissue handling characteristics, potential for bleeding and anatomy to enable an adequate, realistic training environment.3 Furthermore, study participants used equipment similar to that used at our hospital, making the simulation as realistic as possible.3 To incorporate a model into a surgical training program it is essential to establish the validity of the assessment tools. Validity broadly implies that the instrument appropriately measures what it is intended to measure. To that end various subjective and objective benchmarks have been developed to assess model validity. Subjective assessment comprises face validity and content validity, which describe the appropriateness of the test in the opinion of experts. Face validity establishes that the test seems reasonable and appropriate while content validity ensures that test contents cover the relevant areas of the subject being assessed.14 All participants confirmed that the pig urethrovesical anastomosis that they performed resembled the human urethrovesical anastomosis on the video that they watched before randomization. Also, all trainees agreed that performing the
running suture on the UVM helped them perform the pig LUA. Although the UVM had good face validity, further investigation is needed to establish the face, construct and criterion validity of our assessment tools. Future research may include validation of the porcine model by having expert laparoscopic surgeons perform the same tasks under the same conditions and performance appraisals. This study had limitations. Inclusion criteria were strictly reserved to participants at an advanced training level (staff, fellow or postgraduate year 4 –5) in a specialty in which laparoscopy is currently used (urology, general surgery and gynecology). Due to the limited number of urologists and urology residents in our area and program we could not rely exclusively on them to attain the needed sample size. However, the study population was homogenous since no participant had previously performed LRP.
CONCLUSIONS LUA is one of the most challenging steps during LRP. We developed and validated a task specific bench model that allows trainees to practice and improve LUA skills. The skills learned on the UVM transfer to a higher fidelity pig model. LUA in a pig appears to have good face validity and serves as a good surrogate for a patient. Training on models such as this UVM may shorten the learning curve for difficult techniques and allow learners to practice and prepare for surgery with less risk to patients. Further studies are required to evaluate the longterm maintenance and durability of skills acquired beyond the post-training interval and assess the efficacy of UVM for robotics training.
ACKNOWLEDGMENTS Dr. Tania Fayad assisted with the manuscript.
REFERENCES 1. Zorn KC and Lee BR: Laparoscopic radical prostatectomy: an established minimally invasive procedure with proven oncologic track record. J Endourol 2008; 22: 2053. 2. Bollens R, Sandhu S, Roumeguere T et al: Laparoscopic radical prostatectomy: the learning curve. Curr Opin Urol 2005; 15: 79. 3. Stelzer MK, Abdel MP, Sloan MP et al: Dry lab practice leads to improved laparoscopic performance in the operating room. J Surg Res 2009; 154: 163. 4. Scott DJ and Dunnington GL: The new ACS/APDS Skills Curriculum: moving the learning curve out
of the operating room. J Gastrointest Surg 2008; 12: 213. 5. Scott DJ, Bergen PC, Rege RV et al: Laparoscopic training on bench models: better and more cost effective than operating room experience? J Am Coll Surg 2000; 191: 272. 6. Fiolka A, Gillen S, Meining A et al: ELITE—the ex vivo training unit for NOTES: development and validation. Minim Invasive Ther Allied Technol 2010; 19: 281. 7. Ramachandran A, Kurien A, Patil P et al: A novel training model for laparoscopic pyeloplasty using chicken crop. J Endourol 2008; 22: 725.
8. Yang RM and Bellman GC: Laparoscopic urethrovesical anastomosis: a model to assess surgical competency. J Endourol 2006; 20: 679. 9. Nadu A, Olsson LE and Abbou CC: Simple model for training in the laparoscopic vesicourethral running anastomosis. J Endourol 2003; 17: 481. 10. Katz R, Nadu A, Olsson LE et al: A simplified 5-step model for training laparoscopic urethrovesical anastomosis. J Urol 2003; 169: 2041. 11. Teber D, Dekel Y, Frede T et al: The Heilbronn laparoscopic training program for laparoscopic suturing: concept and validation. J Endourol 2005; 19: 230.
1866 TRANSFER OF LAPAROSCOPIC RADICAL PROSTATECTOMY SKILLS FROM BENCH TO ANIMAL MODEL
12. Lucas S, Tuncel A, Bensalah K et al: Virtual reality training improves simulated laparoscopic surgery performance in laparoscopy naive medical students. J Endourol 2008; 22: 1047. 13. Kenney PA, Wszolek MF, Gould JJ et al: Face, content, and construct validity of dV-trainer, a novel virtual reality simulator for robotic surgery. Urology 2009; 73: 1288. 14. Wignall GR, Denstedt JD, Preminger GM et al: Surgical simulation: a urological perspective. J Urol 2008; 179: 1690. 15. McDougall EM, Kolla SB, Santos RT et al: Preliminary study of virtual reality and model simulation for learning laparoscopic suturing skills. J Urol 2009; 182: 1018. 16. Royston CM, Lansdown MR and Brough WA: Teaching laparoscopic surgery: the need for guidelines. BMJ 1994; 308: 1023. 17. Sabbagh R, Chatterjee S, Chawla A et al: Taskspecific bench model training versus basic laparoscopic skills training for laparoscopic radical
prostatectomy: a randomized controlled study. Can Urol Assoc J 2009; 3: 22. 18. Martin JA, Regehr G, Reznick R et al: Objective structured assessment of technical skill (OSATS) for surgical residents. Br J Surg 1997; 84: 273. 19. Reznick R, Regehr G, MacRae H et al: Testing technical skill via an innovative “bench station” examination. Am J Surg 1997; 173: 226. 20. Matsumoto ED, Hamstra SJ, Radomski SB et al: The effect of bench model fidelity on endourological skills: a randomized controlled study. J Urol 2002; 167: 1243.
23. Sweet RM and McDougall EM: Simulation and computer-animated devices: the new minimally invasive skills training paradigm. Urol Clin North Am 2008; 35: 519. 24. Sidhu RS, Park J, Brydges R et al: Laboratorybased vascular anastomosis training: a randomized controlled trial evaluating the effects of bench model fidelity and level of training on skill acquisition. J Vasc Surg 2007; 45: 343. 25. Ericsson KA: Deliberate practice and the acquisition and maintenance of expert performance in medicine and related domains. Acad Med 2004; 79: S70.
21. Anastakis DJ, Regehr G, Reznick RK et al: Assessment of technical skills transfer from the bench training model to the human model. Am J Surg 1999; 177: 167.
26. Hays R: Systems concepts for training systems development. IEEE Trans Syst Man Cybernet 1992; 22: 258.
22. Wanzel KR, Matsumoto ED, Hamstra SJ et al: Teaching technical skills: training on a simple, inexpensive, and portable model. Plast Reconstr Surg 2002; 109: 258.
27. Bell HH and Waag WL: Evaluating the effectiveness of flight simulators for training combat skills: a review. Int J Aviation Psychology 1998; 8: 223.
Transfer of Laparoscopic Radical Prostatectomy Skills From Bench Model to Animal Model: A Prospective, Single-Blind, Randomized, Controlled Study Robert Sabbagh, Suman Chatterjee, Arun Chawla, Jen Hoogenes, Anil Kapoor and Edward D. Matsumoto* From the Division of Urology, Department of Surgery, Faculty of Medicine and Health Sciences, Université de Sherbrooke (RS), Sherbrooke, Québec and Division of Urology, Department of Surgery, Faculty of Health Sciences (SC, AC, AK, EDM) and Departments of Surgery and Clinical Epidemiology and Biostatistics (JH), McMaster University, Hamilton, Ontario, Canada
Purpose: Learning laparoscopic urethrovesical anastomosis is a crucial step in laparoscopic radical prostatectomy. Previously we noted that practice on a low fidelity urethrovesical model was more effective for trainees than basic suturing drills on a foam pad when learning laparoscopic urethrovesical anastomosis skills. We evaluated learner transfer of skills, specifically whether skills learned on the urethrovesical model would transfer to a high fidelity, live animal model. Materials and Methods: A total of 28 senior residents, fellows and staff surgeons in urology, general surgery and gynecology were randomized to 2 hours of laparoscopic urethrovesical anastomosis training on a urethrovesical model (group 1) or to basic laparoscopic suturing and knot tying on foam pads (group 2). All participants then performed timed laparoscopic urethrovesical anastomosis on anesthetized female pigs. A blinded urologist scored subject videotaped performance using checklist, global rating scale and end product rating scores. Results: Group 1 was significantly more adept than group 2 at the laparoscopic urethrovesical anastomosis pig task when measured by the checklist, global rating scale and end product rating (each p ⬍0.05). Time to completion was similar in the 2 groups. No statistically significant difference was noted in global rating scale and checklist scores for laparoscopic urethrovesical anastomosis performed on the urethrovesical model vs the pig. Conclusions: Training on a urethrovesical model is superior to training with basic laparoscopic suturing on a foam pad for performing laparoscopic urethrovesical anastomosis skills on an anesthetized female pig. Skills learned on a urethrovesical model transfer to a high fidelity, live animal model.
Abbreviations and Acronyms CL ⫽ task specific checklist GRS ⫽ global rating scale LRP ⫽ laparoscopic radical prostatectomy LUA ⫽ laparoscopic urethrovesical anastomosis UVM ⫽ urethrovesical model Submitted for publication June 16, 2011. Study received approval from Hamilton Health Sciences and St. Joseph’s Healthcare Hamilton research ethics boards. Supported by internal funding from the Department of Surgery, Division of Urology, McMaster University. * Correspondence: Division of Urology, McMaster University, St. Joseph’s Healthcare Hamilton, 50 Charlton Ave. East, Hamilton, Ontario L8N 4A6, Canada (telephone: 905-521-6186; FAX: 905-308-7205; e-mail: [email protected]).
Key Words: prostate; laparoscopy; surgical anastomosis; models, educational; clinical competence LAPAROSCOPIC radical prostatectomy is a minimally invasive surgical approach for localized prostate cancer.1,2 Many surgeons learning to perform LRP consider intracorporeal suturing and knot tying one of the more daunting parts of the surgery. Training and teaching laparoscopy is a costly, time-consuming
process associated with a steep learning curve. For complex laparoscopic surgery the literature suggests that practice outside the operating room in specially equipped training facilities, such as an animal or surgical skills laboratory, may improve this learning curve.3– 6
0022-5347/12/1875-1861/0 THE JOURNAL OF UROLOGY® © 2012 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION
Vol. 187, 1861-1866, May 2012 Printed in U.S.A. DOI:10.1016/j.juro.2011.12.050
AND
RESEARCH, INC.
www.jurology.com
1861
1862 TRANSFER OF LAPAROSCOPIC RADICAL PROSTATECTOMY SKILLS FROM BENCH TO ANIMAL MODEL
A question in surgical education remains unanswered. Does surgical skills training in a laboratory transfer to the real world operating room environment? We explored some factors attributable to this transfer from ex vivo to in vivo minimally invasive surgical environments. There are several bench models to teach a learner how to perform LUA in the setting of a surgical skills laboratory, including chicken skins and legs, and porcine bladders and urethras as models in pelvic trainers.7–11 However, some models lack the anatomical elements considered essential to simulate the procedure.12 High fidelity models provide surgical environments that are more realistic and allow a learner to practice procedures from beginning to end.13,14 Such models include animal models, cadavers and commercial simulators.15 Animal models, such as the pig model used to train in laparoscopic surgery, allow for realistic tissue handling and bleeding.14 However, animal models are costly, require veterinary assistance and raise social and ethical questions.16 Commercially available high fidelity simulators offer the advantages of reusability and realistic anatomy but they are costly and require maintenance.12 In a recent single-blind, randomized, controlled study we developed and tested a low fidelity latex UVM to practice LUA on a pelvic trainer (fig. 1).17 Our main objective was to evaluate the impact on the performance of intracorporeal urethrovesical anastomosis by comparing the outcome of task specific bench model training (LUA on the UVM) to that of a basic laparoscopic suturing and knot tying task done on a foam pad. According to CL and GRS scores subjects who trained on the UVM performed LUA more quickly and effectively than those who trained on the foam pad. Results suggested that task specific training on a UVM would be an effective way of learning how to perform LUA. Taking what we learned from these findings, we considered it necessary to next assess the transferability of skills from a low fidelity model to a higher fidelity model. We determined whether skills learned on the UVM would transfer to live anesthetized pigs.
We further assessed the face validity of the pig as a surrogate for human urethrovesical suturing.
MATERIALS AND METHODS The study was approved by the Hamilton Health Sciences and St. Joseph’s Healthcare Hamilton research ethics boards. Participants in this prospective, single-blind, randomized, controlled study were the same as those enrolled in the study reported previously.17 Participants were senior surgical residents (postgraduate year 4 –5), fellows or staff surgeons in urology, general surgery or gynecology. All participants provided written informed consent to be in the study and were provided a unique identification number to maintain confidentiality. Participants were recruited via surgical rounds and online message board advertisements. Study exclusion criteria included whether the participant had ever performed LRP. All participants completed a questionnaire on their laparoscopic experience. To show participants how the skills chosen for this study related to the actual procedure they were shown a 15-minute video of LUA during LRP. Participants were randomly assigned to groups using sealed envelopes prepared by our statistician. For the training portion of the study members in group 1 were allocated to practice on the latex UVM to perform running urethrovesical suture anastomoses on a body pelvic trainer, similar to the way that actual anastomoses are done in the operating room. The UVM had anatomical and tissue characteristics, ie thickness, similar to those of the human bladder and urethra (fig. 1). Subjects were given specific instructions and feedback during the practice session by an experienced laparoscopic urologist. Those allocated to group 2 were assigned to practice random suturing and knot tying on a foam pad (Limbs & Things, Bristol, United Kingdom). Subjects were instructed to place single sutures across the foam and tying. As in group 1, group 2 received feedback throughout the training session. After 2 hours of training the groups were asked to perform a post-training laparoscopic suturing test consisting of 2 tasks. The first task was to complete 5 figures-of-8 and 3 knot ties on the foam pad using a 2-zero silk suture on an SH needle (Ethicon, Johnson & Johnson, Markham, Ontario, Canada). The second task was to perform running LUA on the UVM. Training was previously described in detail.17
Figure 1. A, UVM. B, UVM positioned in pelvic trainer simulated orientation in male pelvis and beginning of posterior running suture anastomosis. C, resident performs laparoscopic anastomosis with UVM in pelvic trainer.
TRANSFER OF LAPAROSCOPIC RADICAL PROSTATECTOMY SKILLS FROM BENCH TO ANIMAL MODEL 1863
End-product rating Candidate Identification No. Please rate the candidate's performance on the following scale (maximum total score 25): Tissue integrity
Torn tissues
1
Suture placement
Disorganized suture alignment
1
Watertight
1
2
2
More than 3 sutures passed through the urethral catheter
4
Overall and anastomosis score
Poor
3
5 Most of sutures are well aligned
4
Minimum to moderate leak, more than 100 cc of methylene blue- normal saline return after aspiration
2
5 Intact tissues
3
2
Urethral catheter integrity
1
4
Almost half of the sutures are aligned
Huge leak. Most of the methylene bluenormal saline leak into the pelvis
1
3 Minimally torn
3
5 No leak
4
5
Less than 3 sutures passed through the urethral catheter Urethral catheter intact
2
3
4
Good
5 Excellent
Figure 2. End product rating form used to score LUA done by each participant in female pig
After the exercise in the dry laboratory all participants performed LUA on a high fidelity bench model, that is an anesthetized female pig. Subjects used 2, 15 cm 3-zero Monocryl® sutures (1 dyed and 1 not dyed) tied tail to tail with an RB needle at each end. Before starting this task the pig urethrovesical junction was transected laparoscopically. Participants performed LUA on each pig within a 2-hour time frame. Participant performance was video recorded and timed. Performance was evaluated by a blinded urologist using a CL and a GRS. All evaluated items were adapted from a previously validated CL and GRS.17–19 An end product rating was also used to evaluate pig LUA (fig. 2). Pig LUA watertightness was assessed by injecting 100 ml methylene blue tinted saline via a Foley catheter. All participants were given a questionnaire designed to preliminarily evaluate face validity to assess the UVM and the high fidelity model. Statistical analysis was performed using SigmaStat®, version 3.10.0. Based on an earlier study of bench model training in ureteroscopy20 we found a difference of 10.2 between the surgical training intervention arm and a didactic group. Using these data and a pooled SD of 5 we calculated a sample size of 6 participants per arm. To account for the possibility of study dropouts, and for differences in the complexity of the procedure and training intervention compared with those of the original ureteroscopy study we recruited an additional 8 participants per group. This allowed us to detect a difference of 15% between the 2 groups. Comparisons between groups 1 and 2 were made using the Mann-Whitney U rank test with p ⬍0.05 considered statistically significant.
RESULTS A total of 28 participants were enrolled in this study. At baseline no difference was detected in previous laparoscopic experience, surgical specialty distribution or training level between the groups.17 For the pig urethrovesical anastomosis task group 1 had a significantly higher mean CL score (p ⫽ 0.02), mean GRS score (p ⫽ 0.005) and mean end product rating (p ⫽ 0.04) than group 2 (see table). However, group 1 did not perform LUA more quickly than group 2 (p ⫽ 0.38, see table). The transfer of LUA performance skills from the UVM to a high fidelity model was assessed in the 2 groups (table). For group 1 vs 2 there was no statistically significant difference in the GRS score (p ⫽ 0.427 vs 0.9) or the CL score (p ⫽ 0.567 vs 0.646) LUA performance results in female pigs and skill transfer from bench to pig model
Female pig: Performance time (mins) CL score (max 12) GRS score (max 35) End product rating score (max 25) UVM:17,† CL score (max 12) GRS score (max 35)
Mean ⫾ 95% CI Group 1
Mean ⫾ 95% CI Group 2
50.1 ⫾ 14.5 11.0 ⫾ 1.2 29.9 ⫾ 1.7 20.5 ⫾ 2.4
58.3 ⫾ 13.3 8.9 ⫾ 2.0* 23.2 ⫾ 1.5* 15.9 ⫾ 2.0*
10.9 ⫾ 1.2 29.6 ⫾ 3.7
8.1 ⫾ 1.9 22.8 ⫾ 3.1
* p ⬍0.05 vs group 1. † Not significantly different vs female pig.
1864 TRANSFER OF LAPAROSCOPIC RADICAL PROSTATECTOMY SKILLS FROM BENCH TO ANIMAL MODEL
between LUA done on the UVM and on the pig (table). The end study questionnaire revealed that all participants (100%) thought that operating on the pelvis of the live anesthetized pig was an excellent representation of a human and virtually identical to the video of urethrovesical anastomosis performed on a patient. Participants also thought that training on the UVM significantly helped them acquire the skills needed to accomplish the circular anastomosis.
DISCUSSION Traditional surgical training, that is the “see one, do one, teach one” model, is becoming outdated. This model emphasizes intraoperative teaching, which relies on having time during a surgical case to teach. The staff surgeon has the challenge of balancing the needs of the learner with the need to ensure that the case is done in timely fashion. However, with the increasing financial constraints faced by many health care systems and hospitals, greater pressure is placed on the staff surgeon to teach less and operate more quickly. Also, with the emergence of new technologies such as laparoscopy and robotics it is common to see the staff surgeon balancing learning new skills while teaching them to others. This has necessitated an evolution in how surgeons are trained with defocusing on intraoperative practice and a shift toward training outside the operating room in a surgical skills laboratory. Several studies have shown that training on bench models or simulators and computer animated devices improves surgical skills.19,21–23 However, low fidelity bench models do not simulate the exact operating room setting. It is difficult to simulate a bleeding vessel or reproduce the feeling of tissue consistency. On the other hand, low fidelity bench models are economical, can be used repetitively and do not need the special care required by live animals. High fidelity models have a closer resemblance to actual anatomy and provide an operative field similar to real life operating room situations. They may be more important to teach more complex procedures. Sidhu et al performed a randomized, controlled trial of the effects of low and high bench model fidelity on vascular anastomosis.24 Results revealed better skill transfer from the bench model to live animals when practicing on high fidelity models. However, live animals are more expensive and may be a source of infection, and their use carries ethical concerns. Bench models offer the convenience of being available to a learner with little preparation and planning. Some surgical skills lab-
oratories are accessible 24 hours per day, offering flexibility for a busy surgical resident schedule. Thus, bench models will continue to have an important role in surgical skills training. Thus, it becomes more critical that the models are valid, reliable teaching tools. In our study LUA on the pigs was performed better by group 1 (UVM model) than group 2 (control) (table). This supports the notion of task specific surgical skills training, in that a model incorporating the critical steps of a procedure will ultimately be a better tool. Furthermore, global assessment and specific checklist scores for LUA on the pig were similar to those on the UVM. Thus, subjects in group 1 performed LUA as well as they did on the bench model. These results suggest that the skills learned with the UVM bench model transferred to live anesthetized female pigs, a higher fidelity model. Although there was no difference in time to task completion between the groups, notably quickness does not necessarily mean better surgical performance. This is why surgical skills assessment should also always include ratings of performance to capture the other important nuances of surgery, such as tissue handling, finesse, efficiency of motion, etc. This may explain why there was no difference in time to completion between the groups but there were better performance scores for the group trained using the task specific UVM. Task specific training is extremely important to acquire a new skill.24 Reviews of extensive experience with flight simulation revealed the importance of training designed to attain task specific performance goals.25–27 Thus, practicing the bench model surgical maneuvers, which are crucial in a complex procedure, should increase the chance of being able to perform the procedure successfully in the operating room. McDougall et al compared a silicone model and pelvic trainer to a virtual reality simulator to teach laparoscopic suturing and knot tying.15 When transferred to a task specific procedure, ie laparoscopic cystorrhaphy in a porcine model, there was no difference in performance. This study suggests that task specific training is more critical than the model for successful skills acquisition. For task specific training a great deal of thought and planning must be put into the design of the training program. Even a high technology virtual reality surgical simulator can be a short-lived teaching tool if trainees are not taught the steps specific and relevant to the actual procedure. To our knowledge we report the first study to use an anesthetized pig as a high fidelity in vivo model to validate the efficacy of LUA training on a bench model. The best way to validate our UVM would have been to perform LUA on a patient. However,
TRANSFER OF LAPAROSCOPIC RADICAL PROSTATECTOMY SKILLS FROM BENCH TO ANIMAL MODEL 1865
for ethical reasons this would be unacceptable due to participant lack of experience. Live anesthetized pigs represent a high fidelity model for urethrovesical anastomosis training. The urethrovesical junction and the pelvic dimensions of the pig have anatomical and tissue characteristics similar to those of humans. The bladder and urethra are in the pelvis at the same anatomical site as in humans. We used female pigs since it is easier to cut the urethrovesical junction and perform the anastomosis. The LUA performed on the pig mimics the urethrovesical anastomosis in humans except for the narrower pelvis in the pig. Working on a slightly more challenging model may possibly prepare the learner for the real procedure. Performing LUA on the anesthetized pig was an excellent surrogate for a human model, providing the much needed tissue handling characteristics, potential for bleeding and anatomy to enable an adequate, realistic training environment.3 Furthermore, study participants used equipment similar to that used at our hospital, making the simulation as realistic as possible.3 To incorporate a model into a surgical training program it is essential to establish the validity of the assessment tools. Validity broadly implies that the instrument appropriately measures what it is intended to measure. To that end various subjective and objective benchmarks have been developed to assess model validity. Subjective assessment comprises face validity and content validity, which describe the appropriateness of the test in the opinion of experts. Face validity establishes that the test seems reasonable and appropriate while content validity ensures that test contents cover the relevant areas of the subject being assessed.14 All participants confirmed that the pig urethrovesical anastomosis that they performed resembled the human urethrovesical anastomosis on the video that they watched before randomization. Also, all trainees agreed that performing the
running suture on the UVM helped them perform the pig LUA. Although the UVM had good face validity, further investigation is needed to establish the face, construct and criterion validity of our assessment tools. Future research may include validation of the porcine model by having expert laparoscopic surgeons perform the same tasks under the same conditions and performance appraisals. This study had limitations. Inclusion criteria were strictly reserved to participants at an advanced training level (staff, fellow or postgraduate year 4 –5) in a specialty in which laparoscopy is currently used (urology, general surgery and gynecology). Due to the limited number of urologists and urology residents in our area and program we could not rely exclusively on them to attain the needed sample size. However, the study population was homogenous since no participant had previously performed LRP.
CONCLUSIONS LUA is one of the most challenging steps during LRP. We developed and validated a task specific bench model that allows trainees to practice and improve LUA skills. The skills learned on the UVM transfer to a higher fidelity pig model. LUA in a pig appears to have good face validity and serves as a good surrogate for a patient. Training on models such as this UVM may shorten the learning curve for difficult techniques and allow learners to practice and prepare for surgery with less risk to patients. Further studies are required to evaluate the longterm maintenance and durability of skills acquired beyond the post-training interval and assess the efficacy of UVM for robotics training.
ACKNOWLEDGMENTS Dr. Tania Fayad assisted with the manuscript.
REFERENCES 1. Zorn KC and Lee BR: Laparoscopic radical prostatectomy: an established minimally invasive procedure with proven oncologic track record. J Endourol 2008; 22: 2053. 2. Bollens R, Sandhu S, Roumeguere T et al: Laparoscopic radical prostatectomy: the learning curve. Curr Opin Urol 2005; 15: 79. 3. Stelzer MK, Abdel MP, Sloan MP et al: Dry lab practice leads to improved laparoscopic performance in the operating room. J Surg Res 2009; 154: 163. 4. Scott DJ and Dunnington GL: The new ACS/APDS Skills Curriculum: moving the learning curve out
of the operating room. J Gastrointest Surg 2008; 12: 213. 5. Scott DJ, Bergen PC, Rege RV et al: Laparoscopic training on bench models: better and more cost effective than operating room experience? J Am Coll Surg 2000; 191: 272. 6. Fiolka A, Gillen S, Meining A et al: ELITE—the ex vivo training unit for NOTES: development and validation. Minim Invasive Ther Allied Technol 2010; 19: 281. 7. Ramachandran A, Kurien A, Patil P et al: A novel training model for laparoscopic pyeloplasty using chicken crop. J Endourol 2008; 22: 725.
8. Yang RM and Bellman GC: Laparoscopic urethrovesical anastomosis: a model to assess surgical competency. J Endourol 2006; 20: 679. 9. Nadu A, Olsson LE and Abbou CC: Simple model for training in the laparoscopic vesicourethral running anastomosis. J Endourol 2003; 17: 481. 10. Katz R, Nadu A, Olsson LE et al: A simplified 5-step model for training laparoscopic urethrovesical anastomosis. J Urol 2003; 169: 2041. 11. Teber D, Dekel Y, Frede T et al: The Heilbronn laparoscopic training program for laparoscopic suturing: concept and validation. J Endourol 2005; 19: 230.
1866 TRANSFER OF LAPAROSCOPIC RADICAL PROSTATECTOMY SKILLS FROM BENCH TO ANIMAL MODEL
12. Lucas S, Tuncel A, Bensalah K et al: Virtual reality training improves simulated laparoscopic surgery performance in laparoscopy naive medical students. J Endourol 2008; 22: 1047. 13. Kenney PA, Wszolek MF, Gould JJ et al: Face, content, and construct validity of dV-trainer, a novel virtual reality simulator for robotic surgery. Urology 2009; 73: 1288. 14. Wignall GR, Denstedt JD, Preminger GM et al: Surgical simulation: a urological perspective. J Urol 2008; 179: 1690. 15. McDougall EM, Kolla SB, Santos RT et al: Preliminary study of virtual reality and model simulation for learning laparoscopic suturing skills. J Urol 2009; 182: 1018. 16. Royston CM, Lansdown MR and Brough WA: Teaching laparoscopic surgery: the need for guidelines. BMJ 1994; 308: 1023. 17. Sabbagh R, Chatterjee S, Chawla A et al: Taskspecific bench model training versus basic laparoscopic skills training for laparoscopic radical
prostatectomy: a randomized controlled study. Can Urol Assoc J 2009; 3: 22. 18. Martin JA, Regehr G, Reznick R et al: Objective structured assessment of technical skill (OSATS) for surgical residents. Br J Surg 1997; 84: 273. 19. Reznick R, Regehr G, MacRae H et al: Testing technical skill via an innovative “bench station” examination. Am J Surg 1997; 173: 226. 20. Matsumoto ED, Hamstra SJ, Radomski SB et al: The effect of bench model fidelity on endourological skills: a randomized controlled study. J Urol 2002; 167: 1243.
23. Sweet RM and McDougall EM: Simulation and computer-animated devices: the new minimally invasive skills training paradigm. Urol Clin North Am 2008; 35: 519. 24. Sidhu RS, Park J, Brydges R et al: Laboratorybased vascular anastomosis training: a randomized controlled trial evaluating the effects of bench model fidelity and level of training on skill acquisition. J Vasc Surg 2007; 45: 343. 25. Ericsson KA: Deliberate practice and the acquisition and maintenance of expert performance in medicine and related domains. Acad Med 2004; 79: S70.
21. Anastakis DJ, Regehr G, Reznick RK et al: Assessment of technical skills transfer from the bench training model to the human model. Am J Surg 1999; 177: 167.
26. Hays R: Systems concepts for training systems development. IEEE Trans Syst Man Cybernet 1992; 22: 258.
22. Wanzel KR, Matsumoto ED, Hamstra SJ et al: Teaching technical skills: training on a simple, inexpensive, and portable model. Plast Reconstr Surg 2002; 109: 258.
27. Bell HH and Waag WL: Evaluating the effectiveness of flight simulators for training combat skills: a review. Int J Aviation Psychology 1998; 8: 223.