Modular surgical training for endoscopic extraperitoneal radical prostatectomy
Jens-Uwe Stolzenburg, Department of Urology, University of Leipzig, Liebigstrasse 20, 04103 Leipzig, Germany.
To develop a modular training scheme which enabled the use of individual steps of laparoscopic radical prostatectomy (RP) for teaching and training surgeons with varied experience, including residents with no experience in open RP, as in extending laparoscopic surgery to more complex operations like RP, the proper training of urologists is crucial.
SUBJECTS AND METHODS
The technique of endoscopic extraperitoneal RP (EERP) was divided into 12 individual steps of differing complexity. The levels of difficulty were called ‘modules’ and graded according to their requisite skills from module 1 (lowest level of difficulty) to module 5 (highest level). Based on this modular system we established a training programme whereby the trainee learns the procedure in a mentor-initiated schedule. During each training operation the trainee only performs the modules (steps) of the operation, which correspond with his or her actual skill level. The mentor performs all the other steps, with the trainee assisting. Four trainees with different surgical experience participated in the study.
After a phase of assisting and camera holding during EERP, the trainees entered the modular training programme and required 32–43 procedures until they were considered to be competent. An analysis of the first 25–50 procedures done independently by the trainee showed mean operative times of 176–193 min and a transfusion rate of 1.3%. Complications during and after EERP requiring re-intervention were one each of recto-urethral fistula, haemorrhage, symptomatic lymphocele and anastomotic leak. The positive margin rate for pT2 disease was 12.2% and for pT3 tumours 37%.
The modular concept for teaching EERP is an attractive concept, which overcomes many of the problems involved in complex laparoscopic procedures. Based on a highly standardized technique, this concept offers a short learning curve; it enables training on different sites in cooperation with a high-volume centre, and it makes it possible to start with this complex procedure as a beginner or with no experience in open RP.
(endoscopic extraperitoneal) radical prostatectomy.
There has been an exponential growth in minimally invasive surgical techniques in recent years. The obvious advantages for the patients have captured the imagination of the general public and have driven clinicians to achieve similar levels of excellence to those of standard open surgical techniques. Herein lies the major challenge of these innovations; that of learning the skills of minimally invasive surgery in an era where skilled teachers are few and modern attitudes to clinical governance inhibit free learning and opportunity.
Laparoscopic surgery is rapidly becoming an integral part of urological surgery. With the extension of laparoscopic surgery to more complex operations like radical prostatectomy (RP), cystectomy and partial nephrectomy, the proper training of urologists is crucial. This is even more crucial considering the high standard and relatively low complication rates of these procedures, when done in the standard open way. The learning curve for laparoscopic urological surgery is considered to be long and steep. Factors such as intracorporal suturing and the restricted space in the male pelvis make radical laparoscopic RP one of the most demanding procedures in laparoscopic surgery. Hence, the need for training has been estimated at 40–60 cases for skilled laparoscopic surgeons and 80–100 cases for laparoscopically naïve surgeons [1–3].
Economic problems, like operative time, time and cost needed for training, the lack of ‘easy’ urological laparoscopic ‘training procedures’ like cholecystectomy and hernia repairs, and the undeniable fact that open RP has reached a very high standard , demand faster training and a minimum of patient morbidity, to be able to establish this procedure.
‘Dry laboratory’ training models, using various simulators [5,6] are certainly a cornerstone of laparoscopic training, but of different quality and sometimes quite expensive. Even more expensive are models based on virtual reality computer software [7,8]. Both models cannot replace the challenge and atmosphere of live surgery.
Training with live animals is another, more realistic alternative. The possible ablative procedures (nephrectomy and prostatectomy) are really only useful training exercises for the beginner or intermediate-level laparoscopic trainee. However, there is a risk of trainees developing a falsely high impression of their ability in such ‘wet laboratory’ courses and this needs to be taken into account by the trainers .
To overcome these problems, we developed a modular training scheme, which enabled us to use individual steps of endoscopic (laparoscopic) RP for teaching and training. Endoscopic extraperitoneal RP (EERP) was developed to overcome the limitations of laparoscopic RP, using a strictly extraperitoneal access and combining the advantages of an open, retropubic RP with the undeniable advantages of minimally invasive surgery [10,11].
SUBJECTS AND METHODS
The technique of EERP was described previously in detail [12,13] and to date >800 procedures have been performed in the high-volume centre (Leipzig). We divided EERP into 12 individual steps of differing complexity (Table 1), to maximize standardization of the procedure. Each surgeon was required to perform the procedure in an identical stepwise fashion.
The 12 steps of EERP, with each step classified in one of five modules according to the surgical skills required
| 1||Placing trocar and dissecting the preperitoneal space||X|| || || || |
| 2||Pelvic lymphadenectomy|| ||X|| || || |
| 3||Incising the endopelvic fascia and dissecting the puboprostatic ligaments||X|| || || || |
| 4||Ligating Santorini's plexus|| || ||X|| || |
| 5||Bladder neck dissection|| || || || || |
|Anterior and lateral bladder neck|| ||X|| || || |
|Dorsal bladder neck|| || ||X|| || |
| 6||Identifying and dissecting the vasa deferentia|| || ||X|| || |
| 7||Dissecting the seminal vesicles|| || ||X|| || |
| 8||Incising posterior Denonvilliers’ fascia and mobilizing the dorsal surface of the prostate from the rectum|| || ||X|| || |
| 9||Dissecting the prostatic pedicles|| || ||X|| || |
|10||Nerve-sparing|| || || || ||X|
|11||Apical dissection|| || || ||X|| |
|12||Urethro-vesical anastomosis|| || || || || |
|Dorsal circumference (4, 5, 6, 7, 8 o’clock sutures)|| || || ||X|| |
|3 and 9 o’clock sutures|| ||X|| || || |
|Bladder neck closure and 11 and 1 o’clock sutures|| || ||X|| || |
The levels of difficulty were called ‘modules’ and graded according to their requisite skills, as 1–5, with 5 being the most difficult. For instance, placing the trocar and dissecting the preperitoneal space were classified as module 1, and nerve-sparing as module 5. The urethro-vesical anastomosis comprises three levels of difficulty (module 2, 3 and 4).
Four trainees with different surgical experience participated in the study. Trainees 1 and 2 were registrars from the department in Leipzig, whereas Trainees 3 and 4 were visiting surgeons from the UK. All subjects attended at least one dry-laboratory course before the start of the programme. Two had attended one wet-laboratory course each. Only two had performed open radical pelvic procedures before. Previous laparoscopic experience was variable, ranging from five varicocelectomies to 80 procedures performed as the main surgeon (Table 2).
The surgical experience of the four trainees before starting the EERP training programme, number of cases needed by each trainee before proceeding to the next level of difficulty, the data for the first 25–50 consecutive patients who had EERP by the trainees independently (with the results of the mentor for his cases 520–570)
|Dry laboratory, n courses attended||1||1||1||2|| |
|Wet laboratory, n courses attended||0||0||1||1|| |
|Open retropubic RP, n cases as assistant/surgeon||0||0||100/20||120/5|| |
|Previous laparoscopic experience, n cases as assistant/as surgeon||122/80*||0/5†||95/25‡||0/40¶|| |
|Previous experience with EERP, n cases as camera-holder/as assist||4/140||62/15||20/45||5/18|| |
|Number of operations on each module:|
|Module 1||2||5||5||6|| |
|Module 2||7||8||10||5|| |
|Module 3||15||14||12|| 11|| |
|Module 4||7||4||6||4|| |
|Module 5||12||7||9||6|| |
|Total operations by each trainee||43||38||42||32|| |
|No. of patients||50||50||25||25||50|
|Age, years||63.9 (57–74)||65.2 (53–76)||64.9 (52–74)||61.4 (53–72)||63.9 (49–76)|
|PSA, ng/mL||10.2 (4–24.3)||9.51 (3.3–25.3)||8.8 (2.7–40)||12.9 (2.2–80)||12.95 (4.2–50.7)|
|Previous pelvic surgery (procedure, n patients)||HR (5)||HR (7)||HR (2)||HR (3)||HR (9)|
|AE (9)||TEP (1)||AE (1)||AE (2)||TEP (2)|
|AE (5)|| || ||AE (11)|
|Mean (range) duration§ of EERP, min||176 (110–285)||173 (110–270)||196 (120–270)||193 (130–290)||120 (50–185)|
|Hernia repair; uni-, bilateral, n||Uni (2)||Uni (3)||0||Uni (1)||Uni (4)|
|Bi (1)|| || || |
|Pelvic lymphadenectomy, n||13||12||7|| 11||16|
|Transfusion rate, n (%)||2 (4)||0||0||0||0|
|Complications, n||Rectal injury (2)||None||None||None||None|
|Mean (range) prostate weight, g||52.7 (36–78)||54.6 (28–83)||51.2 (34–92)||49.9 (24–76)||73.3 (24–200)|
|Mean (range) catheterization, days||7.6 (4–20)||6.4 (4–20)||5.9 (5–10)||5.4 (5–7)||6.2 (4–14)|
|Complications within 1 month of EERP||1 RU, 1 H||1 SL||0||1 MAL||1 OP, 1 SL|
|Treatment||C, LF||PD|| ||LR||Co, LFe|
|Positive margins according to pT stage|
|pT2, n/N (%)||2/14 (14)||3/26 (12)||3/20 (15)||1/14 (7)||3/26 (12)|
|pT3,||14/36 (39)||7/24 (29)||3/5 (60)||5/11 (45)||8/24 (33)|
Initially, the trainees were advised to assist the mentor, first as camera-holder and then as first assistant (Table 2). In our department, EERP is performed as a three-surgeon procedure during the learning phase. As they progressed the trainees started to operate with the mentor as the assistant, but the trainees did not perform the whole procedure themselves. By contrast, they performed individual steps of the operation, according to the module system. For example, the trainee started with module 1 and performed steps 1 and 3 of the operation in a particular patient. The mentor then performed all the other steps of the operation, with the trainee assisting. This schedule was repeated during the following operations, until the mentor felt that the trainee was able to proceed to the next level, i.e. the trainee would gradually perform more modules of the operation.
A good example for this kind of tutoring is the urethro-vesical anastomosis that we carry out using interrupted sutures. As in most described surgical techniques for any kind of anastomosis, we started with the dorsal circumference. These sutures are the most difficult; the mounting of the needle has to be changed between forehand for the bladder neck and backhand for the urethra. Furthermore, the ureteric orifices can be very close to the margin at the bladder neck. Hence the mentor would place the dorsal 4, 5, 6, 7 and 8 o’clock position sutures, which are classified as module 4. Then the trainee would take over and place the sutures at the 3 and 9 o’clock positions, which are the easiest of these anastomotic sutures (module 2). The mentor would finish the anastomosis by closing the bladder neck and placing the 11 and 1 o’clock sutures, which are generally more difficult (module 3). This schedule would be repeated during every training operation, until the trainee had developed the requisite skills to progress to modules 3 and 4 in the anastomosis without the help of the mentor.
As an adjunct to the training, a custom made ‘pelvitrainer’ was used for suturing practice. This dry-laboratory station consists of a wooden box with two lateral ports for a needle holder and a forceps. To diminish costs, we used a mounted web-cam connected via Image Studio 7.3 (Logitech®, USA) to a Fujitsu-Siemens personal computer. The computer monitor was positioned in front of the pelvitrainer box so that it accurately simulated the picture obtained in a laparoscopic stack. Transected silicon indwelling catheters were used to simulate the transected urethra, creating a model for the urethro-vesical anastomosis. The sutures used were 2/0 polyglactin on a 5/8 needle. The training programme was used in consecutive patients scheduled for EERP in our department; there was no case selection for training purposes.
To assess whether this approach speeded learning, and to evaluate and put the progress of the trainees into perspective, we analysed the first 25–50 consecutive cases carried out by the trainees independently. Again, there was no case selection for training purposes. We consider 50 cases as reasonable for a proper analysis but because Trainees 3 and 4 were foreign Fellows with a limited period available we included their 25 first cases in the evaluation. We think that the performance of these trainees, who were not official members of the department, is particularly interesting in terms of assessing the value of our module programme. The results of the trainees were then compared with 50 consecutive cases performed by the mentor during the same period.
A cystogram was taken 5 days after EERP and the catheter removed if there was no anastomotic leak. Because longer follow-up data are currently not available for the patients operated by trainees 3 and 4, we evaluated the continence at 3 months after EERP, using the ICS questionnaire. Patients were considered to be continent if they required no more than one pad per day. When applicable, statistical significance was determined using the nonparametric Mann–Whitney U-test.
All four trainees completed the whole training programme and reached level 5 successfully, which enabled them to perform EERP with nerve-sparing independently. Depending on their aptitude and previous experience, the trainees performed different numbers of operations on a particular module. For instance, trainee 1 required only two operations on module 1 before proceeding to module 2. Trainee 4 performed six operations on module 1 until the mentor felt it appropriate for him to proceed to module 2 (Table 2). During this phase, two epigastric vessels were injured during EERP, caused by trocar placement, which were both managed successfully by laparoscopic coagulation or clipping. There were no other complications during EERP.
Six patients had six complications after EERP and during the modular training phase. One patient had a re-intervention (laparoscopic fenestration for a symptomatic lymphocele). Two patients with asymptomatic lymphoceles were monitored by watchful waiting, with no sequelae. One deep vein thrombosis, caused by a lymphocele, was managed conservatively. Two patients with urinary extravasation were treated by prolonged catheterization (>14 days) and recovered with no problems, including incontinence or stricture.
The data before EERP for the first consecutive 25–50 patients on whom the trainees performed EERP independently were, according to the consecutive recruitment of the patients, heterogeneous but similar for all trainees and the mentor (Table 2). Data during and after EERP for these patients were analysed and compared to 50 consecutive cases performed by the mentor, to assess the impact of the modular training system. (Table 2). The mean prostate weight was comparable among all five surgeons. There was no statistically significant difference in the number of additional intraoperative procedures, i.e. hernia repairs and pelvic lymphadenectomy, and no significant difference in operative duration among the four trainees, but there was between the trainees and the mentor (P < 0.005). None of the patients had a conversion to open surgery. The transfusion rate was comparable among all trainees (1.3% overall) and the mentor (none). There were two intraoperative rectal injuries caused by trainee 1, both repaired during EERP with a laparoscopic double-layer suture. There were no other complications during EERP in the whole series (Table 2).
Table 2 also shows the results of EERP; there was no statistically significant difference in catheter duration between the mentor and the trainees (P = 0.939). There were four complications caused by the trainees and two caused by the mentor. Of the 150 patients operated on by the trainees, four required re-interventions, compared to one operated on by the mentor (Table 2). All patients recovered after the re-interventions with no further problems. Most patients had pT3 (50%) and pT2c (26.5%) tumours. The positive margin rates were slightly higher for the trainees than the mentor for pT2 (12% vs 8%) and pT3 (38% vs 8%) disease (Table 2). Because of the variable number of cases a statistical analysis was not possible for this factor.
The overall continence rate 3 months after EERP, defined as the use of ≤ 1pad/day, was 67% for the trainees (64%, 72%, 64% and 68%) and 78% for the mentor; the differences were not statistically significant.
Laparoscopic RP has been considered to be one of the most difficult laparoscopic procedures; its principle advantages are that it is minimally invasive, provides superior visualization of the operative field through the magnification by the optical system, it gives an exact and watertight anastomosis, reduced blood loss, and a shorter hospital stay [14–16].
Training and accreditation emphasize the need for testing and scoring. Many attempts have been made to solve this problem [17–19], including sophisticated tools like continuous assessment using the multimode learning curve or the moving average method, or the cumulative sum method, which evaluates complication and conversion rates . However, as recently discussed by Guilloneau  in an excellent review, tests are not as sensitive as suggested, and many tests are unable to differentiate between beginners and surgeons with either little or moderate experience. He questions whether ‘hands-on’ sessions are an appropriate way to train numerous surgeons in complex techniques. One of the main concerns is the considerable time that must be given for such an undertaking. In another report Fabrizio et al. proposed the mentor-initiated approach, where an experienced surgeon serves as the mentor and performs a series of consecutive operations with the trainee acting as assistant. Once the trainee has gained enough confidence the subsequent procedures are performed by the trainee with the mentor acting as assistant. This is a very realistic way of teaching, which can be improved by using regular reviews of video documentation .
We think that a mentor-initiated approach, combined with intensive dry- and wet-laboratory training, is the way forward, certainly for laparoscopic RP. It is preferable to have an operation that fulfils the training requirements and that can be practised often. Although RP is indeed regularly performed it remains too complex for trainees who have varied and little previous laparoscopic experience to learn the procedure. The modular training system overcomes these shortcomings and can be adapted to train those with an intermediate level of laparoscopic skill, whether they have previous open pelvic surgical experience or not.
An important issue is whether this complex and difficult operation can serve as a training tool for urological surgeons and centres to acquire the laparoscopic skills without impeding patient safety and functional results. In educational settings, e.g. simulation models, tasks are standardized, which facilitates training and enables the trainer to gauge progress between trainees. We have shown that our modular teaching concept of EERP can overcome most of the problems encountered in laparoscopic training for complex pelvic operations. Rigid adherence to the modular teaching phases at the start of training confirmed that the surgeons develop the requisite skills to overcome intraoperative difficulties and progress rapidly from module to module.
Previous open or laparoscopic experience did not seem to dramatically affect the performance of the trainees in our programme. Interestingly, two of the trainees had never seen an open RP before. Furthermore, none of the trainees were required to convert the procedure to open surgery. This questions the widely held view that surgeons can only consider laparoscopic RP once they have a considerable experience in the anatomy and process of the open operation. Adherence to the modular teaching phases at the start of training for this operation confirmed that the surgeons develop the requisite skills to overcome intraoperative difficulties laparoscopically, thereby negating the need for open conversion. It therefore seems that the surgeon's previous experience does not influence the final outcome, and the progression from module to module can be variable in time and effort spent by trainee and mentor.
The question of how many procedures are deemed appropriate for learning a procedure has never been answered convincingly. A literature review focusing on laparoscopic procedures in general urology found large variations in the number of procedures required for proficiency, e.g. 8–200 for cholecystectomy and 20–60 for fundoplication . For laparoscopic RP, surgeons continued to improve in terms of operative time even after 300 cases . However, numbers alone are meaningless, because surgeons have different aptitudes, and therefore the rate at which they learn to master the technical challenges varies.
The many patients undergoing surgery in our department (370 per year) allows for more than one mentor to train the newer trainees. Indeed, more experienced trainees can mentor the newer trainees in the easier modules, thereby relieving the head of the department of the total responsibility of teaching each trainee. This is a fundamental advantage of the modular concept, as the traditional routine of trainer spending many hours patiently with the trainee is overcome.
The modular concept also allows for preliminary training in the less complicated modules to be done remotely from the high-volume centre. This creates a particularly attractive possibility for training surgeons in a setting where mentors are few, there are few patients for RP per urology unit, and consultant commitments and service obligations make it almost impossible to travel to other hospitals to teach. Provided that the steps of the procedure stay the same and the volunteer mentor is committed to adhere strictly to the standardized technique, there is opportunity for surgeons to learn this procedure partly in a local environment. The final steps can be learned on a substantially shortened fellowship to a high-volume centre.
Probably the most important issue in training and learning in medicine is the functional outcome and safety of the patients. The modular training scheme enables the mentor to ensure that the trainee has acquired the requisite skill of each level before progression to the next. This ensures patient safety and the completion of each module to the highest of standards. A further advantage is that this allows timely intervention by the mentor when the trainee is failing to progress, thereby keeping operative times within an acceptable range.
There were six complications caused by the trainees during their training, only one of which required re-intervention. Once the trainees operated independently there were only two major intraoperative complications, with trainee 1 (a complication rate of 1.3%), which compares favourably with other reports. In all, the overall complication rate, including problems after EERP requiring re-interventions, was 3.3% (rectal injuries, haemorrhage and lymphoceles). Guilloneau et al. reported a similar complication rate of 3.7% in 567 patients. Two other series [26,27] reported initial complication rates of 23% and 13.7%, respectively, which decreased to 3.2% and 6.4% with more patients. The mean operative duration for the trainees was 173–196 min, among the shortest reported to date , despite including the time for laparoscopic hernia repair and bilateral pelvic lymphadenectomy in a third of the patients. Only two patients operated by trainee 1 required a blood transfusion, amounting to an overall transfusion rate for the trainees of 1.3%, which again compares favourably to other reports . Most tumours were pT2c and pT3a, resulting in a positive margin rate in pT2 tumours of 12% for the trainees and 8% for the mentor; the respective rates for pT3 tumours were 38% and 33% (not significantly different). El-Feel et al. reported an 18% positive margin rate for pT2, 45% for pT3a and 50% for pT3b tumours, respectively, analysing the consecutive results of two experienced and two junior surgeons. Another group  described their experience of one surgeon with his first 100 laparoscopic RPs, reporting an overall positive margin rate of 12.8% in pT2 and 31.8% in pT3 tumours.
In conclusion, the proposed modular concept for teaching EERP is a feasible and attractive concept, which overcomes many of the problems that are encountered by trainees and departments who want to establish complex laparoscopic procedures. Based on a highly standardized technique, this concept offers a short learning curve, enables training on different sites in cooperation with a high-volume centre and it makes it possible to start with this complex procedure as a beginner or with no experience in open RP. Parts of the procedure can be taught independently from each other during consecutive operations. However, the main advantage of this modular training is that it provides training in highly complex laparoscopic procedures without putting patients at risk.
CONFLICT OF INTEREST