A Critical Analysis of the Learning Curve and Postlearning Curve Outcomes of Two Experience- and Volume-Matched Surgeons for Laparoscopic and Robot-Assisted Radical Prostatectomy

Abstract

Background and Purpose:

There remains equipoise with regard to whether laparoscopic radical prostatectomy (LRP) or robot-assisted radical prostatectomy (RARP) has any benefit over the other. Despite this, there is a trend for the increasing adoption of RARP at great cost to health services across the world. The aim was to critically analyze the learning curve and outcomes for LRP and RARP for two experience- and volume-matched surgeons who have completed the learning curve for LRP and RARP.

Patients and Methods:

Two experience- and volume-matched LRP and RARP surgeons who have completed the learning curve were compared with respect to their learning curve and outcomes for RARP and LRP. There were 531 RARP and 550 LRPs analyzed from April 2003 until January 2012 at two relatively high-volume United Kingdom centers. Outcome measures included operative time, blood loss, complication rate (Clavien-Dindo grade III), positive surgical margin (PSM) rate, and early continence rate.

Results:

Learning curves for blood loss, operative times, and complication rate were similar between groups. The overall PSM rate and pT₂ PSM rate learning curves were longer for RARP compared with LRP but shorter for early continence. Apical PSM showed no learning curve for RARP; however, a long learning curve for LRP and the rate was lower for RARP than for LRP (P=<0.001).

Conclusions:

This study of RARP and LRP identified that both modalities had long learning curves. Despite the long learning curve for RARP, significant benefits in lower PSM rates and better early continence in comparison with LRP exist. There are benefits to patients with RARP over LRP, especially those linked to better apical dissection (apical PSM and early continence).

Introduction

Prostate cancer is an important disease and represents the most common nondermatologic cancer in men.¹ Radical prostatectomy (RP) is a well-recognized curative treatment for patients with the disease and is increasingly being performed using minimally invasive approaches (laparoscopic radical prostatectomy [LRP] and robot-assisted radical prostatectomy [RARP]) rather than the traditional open (ORP) approach.² This change of approach was pioneered in Europe, and minimally invasive radical prostatectomy (MIRP) has been shown to provide fewer perioperative complications and improve recovery in comparison with ORP.^3,4

RARP is becoming the most widely practiced approach for MIRP in many parts of the world, including Europe and North America.⁵ LRP remains technically demanding and is considered ergonomically more challenging than RARP, which benefits from three-dimensional (3D) vision, increased degrees of freedom at instrument tips, and a comfortable position for the surgeon.^6,7

Perhaps as a consequence of these advantages over laparoscopy, RARP has been reported to have a shorter learning curve^8

–11 in comparison with the long learning curve reported for LRP.^12,13 Aggressive marketing of these advantages by the manufacturer and healthcare institutions has undoubtedly contributed to the dramatic rise in the popularity of RARP among surgeons and their patients.^14,15

Despite many reports in the literature, there is still controversy as to whether RARP truly enables a shorter learning curve compared with LRP.¹⁶ While early reports demonstrated a shorter learning curve for RARP,^8

–11 more recent studies^17,18 have questioned these findings.

The aim of this study was to critically analyze the learning curves and outcomes for LRP and RARP for two experience- and volume-matched surgeons who could be shown to have completed the learning curve for LRP and RARP.

Patients and Methods

Two surgeons prospectively collected outcome data on patients under their care undergoing RP at their institutions. Between April 2003 and October 2009, Surgeon 1 (Sx 1, DC), who had performed or assisted in 19 ORP as a trainee, performed 289 LRPs independently (Montsouris technique).¹⁹ This followed a clinical fellowship of 7 months, which involved dry and cadaveric laboratory training and performing five LRPs by the end of training. In August 2007, Sx 1 then transferred to RARP (transperitoneal antegrade technique), after completing an accredited intensive robotic workshop and a wet-lab course, and had performed 531 RARPs up to September 2013. He received initial mentoring by departmental colleagues experienced in robotic surgery for the first 20 cases until deemed competent for independent practice. During the period of transition to RARP, Sx 1 performed 53 LRPs and 66 RARPs before finally transitioning fully to RARP. The average number of RARPs performed per year was 90.

Surgeon 2 (Sx 2, SAM), who had considerable laparoscopic experience in upper tract surgery (160 laparoscopic nephrectomies performed until February 2006), performed 550 LRPs between February 2006 and January 2012. His training for LRP included approximately 20 ORPs as a trainee including a British Urology Foundation sponsored visit to Miami in 1999; a British Association of Urological Surgeons (BAUS)-sponsored preceptorship at the Cleveland Clinic in 2003; and a fellowship in endoscopic extraperitoneal radical prostatectomy in 2005.^20,21 The fellowship, at the University of Leipzig, Germany, involved animal laboratory training and a modular mentored approach to LRP. The fellowship lasted 4 weeks during which the various modules were performed in 25 LRPs. After this, he performed five further cases under supervision of a mentor before he was deemed competent for independent practice. The average number of LRPs performed per year was 92.

All 1370 patients undergoing surgery were extensively counseled regarding their treatment options and in consultation with a regional multidisciplinary team chose their preferred treatment. All anonymized data were collected in two separate prospective research databases, and both surgeons collected these data. The Declaration of Helsinki principles were adhered to.

To critically analyze the LRP and RARP learning curves for two experience- and aptitude-matched surgeons, we first analyzed both surgeons' outcome measures for LRP (Results section: LRP learning curve comparison for Sx 1 and Sx 2; aptitude for LRP analysis). Then we analyzed Sx 1's outcomes for RARP and Sx 2's outcomes for LRP (Results section: RARP and LRP postlearning curve comparison for Sx 1 and Sx 2).

Outcome measures

Data from both surgeons' prospectively maintained databases were analyzed for patient demographics, preoperative investigations, intraoperative data; pathologic and postoperative details including complications, early continence (EC), and positive surgical margins (PSMs).

The analysis of this study started with a comparison of LRP for Sx 1 and Sx 2 to determine if these showed similar aptitude with regard to learning curve and overall outcomes. After this comparison of initial learning curves for LRP, which we believe allows us to take into account some of the factors that usually confound comparisons between surgeons and technologies, a complete analysis of the postlearning curve outcomes was performed comparing RARP with LRP between Sx 1 and Sx 2. We also demonstrate for transparency the RARP learning curves.

Perioperative outcomes

Operative time was calculated from first incision to application of dressings in the theater and was standard between both centers. Estimated blood loss was calculated in millilitres. Complications were collected if occurring within 30 days of operation and classified using the Clavien-Dindo (CD) grading system with the grade III (CD-III) or above (IV and V) complications being used for comparison between groups.

Positive surgical margins (PSM)

All specimens in both institutions were processed according to International Society of Urological Pathology guidelines with whole mount step sectioning. A PSM was defined as tumor involving the inked margin of the cut surface. PSM location was also noted and used in the analysis; however, margin length was not collected routinely. There was no central pathology review of all the cases between the two institutions.

EC

This was assessed in clinic by asking patients how many pads they were using, with dry being defined as 0 pads/24 hours at 3-month follow-up. All patients were continent preoperatively. Sx 1 and Sx 2 did not routinely use a Rocco stitch (posterior reconstruction) in LRP; however, after October 2012, Sx 1 did routinely use a Rocco stitch in RARP.

Learning curve analysis

The different groups studied (Sx 1 LRP, Sx 1 RARP and Sx 2 LRP) were divided into consecutive groups of 50 patients. Outcomes within these sequential groups were then assessed and plotted on scatterplots for the various outcome measures. Locally weighted scatterplot smoothing was used to help determine when “plateau” of the learning curve was achieved. Plateau was defined when for increasing experience there was no further improvement in outcome measure.²²

Statistical analysis

Patients were divided into consecutive groups of 50 patients. Outcomes were then analyzed for each group of 50 patients for both LRP and RARP. All statistical analyses were performed using PASW statistics 19.0 (Chicago, IL), and statistical significance was set at P<0.05. Continuous data were analyzed using the Mann Whitney U test and categorical variables using the Pearson chi-square test.

Results

In total, we identified and included 1370 patients (289 in Sx 1 LRP group, 550 in Sx 2 LRP group and 531 in Sx 1 RARP group) in our study. Table 1 shows baseline demographics of these groups. The groups are of similar age, risk groupings, and prostate weight, although the RARP group were slightly younger (1 year), had lower mean prostate-specific antigen levels (1 μg/L), higher mean prostate weight (4 g), had more D'Amico high-risk patients (10%) than the LRP group. While these differences reached statistical significance, their clinical significance is questionable because these small differences are unlikely to have any impact on the outcomes we used in our analysis.

Table 1.

Demographic and Clinical Characteristics by Surgeon and Radical Prostatectomy Modality

Characteristic	Sx 1 LRP	Sx 2 LRP	P value	Sx 1 RARP	Sx 2 LRP	P value
Number of patients	289	289		531	550
Age at surgery (yr)
Mean	60	62	0.001	61	62	0.01
Range	36–76	40–76		42–75	40–76
Mean PSA (μg/L)	8	7.5	0.01	9	8	0.01
Range	1–35	2–32		1–85	2–32
Mean prostate weight (g)	58	49	0.001	56	52	0.01
Range	21–174	17–251		22–192	17–251
Biopsy Gleason score, (%)
<7	149 (51)	173 (60)	0.05	156 (29)	174 (32)	0.2
7	118 (41)	104 (36)	0.09	324 (61)	354 (64)	0.1
>7	22 (8)	12 (4)	0.003	51 (10)	22 (4)	0.001
D'Amico risk category, (%)
Low	121 (42)	143 (49)	0.01	158 (30)	219 (40)	<0.001
Intermediate	123 (43)	129 (45)	0.5	259 (49)	271 (49)	0.9
High	46 (16)	17 (6)	0.001	114 (21)	60 (11)	<0.001

LRP=laparoscopic radical prostatectomy; RARP=robot-assisted radical prostatectomy; PSA=prostate-specific antigen.

Table 2 shows the perioperative data for both of the LRP groups and also between the Sx 1 RARP and Sx 2 LRP groups for the entire series (including the learning curve). This table shows similar rates of stage pT₂/pT₃ disease and also similar rates of nerve-sparing procedures between the RARP and LRP groups. Operative time and length of stay were lower in the RARP group, while estimated blood loss was greater. The overall PSM rate, pT₃ PSM rate, and the complication rate were similar between LRP and RARP, but the pT₂ PSM rate was significantly lower in the RARP group.

Table 2.

Perioperative Data

Characteristic	Sx 1 LRP	Sx 2 LRP	P value	Sx 1 RARP	Sx 2 LRP	P value
Pathologic stage, (%)
pT₂	186 (64)	196 (68)	0.2	354 (67)	388 (71)	0.09
pT₃/₄	103 (36)	93 (32)	0.2	177 (33)	162 (29)	0.15
Gleason score, (%)
<7	112 (39)	173 (60)	0.001	71 (13)	273 (50)	<0.001
7	155 (54)	102 (35)	0.001	404 (76)	237 (43)	<0.001
>7	22 (8)	14 (5)	0.03	56 (11)	40 (7)	0.01
Operative time (min)
Median	150	150	0.3	120	134	<0.001
Quartiles	130–180	120–170		100–150	120–170
Blood loss (mL)
Median	200	200	0.9	200	200	<0.001
Quartiles	150–300	100–250		105–350	100–250
Length of hospital stay (days)
Median	2	3	0.01	1	3	<0.001
Quartiles	1–2	2–3		1–2	2–3
Nerve sparing (%)	161 (56%)	176 (61)	0.02	303 (57%)	335 (61)	0.07
Complication Clavien-Dindo grade III, (%)	18 (6)	6 (2)	0.001	8 (2)	9 (2)	0.5
PSM (%)	79 (27)	82 (28)	0.7	125 (23)	129 (23)	0.8
PSM, pT₂ (%)	31 (16)	40 (20)	0.1	44 (12)	63 (16)	0.01
PSM, pT₃/₄ (%)	48 (47)	42 (45)	0.3	78 (44)	66 (41)	0.4

PSM=positive surgical margin.

LRP learning curve comparison for Sx 1 and Sx 2 (aptitude for LRP analysis)

A comparison of the LRP learning curves for Sx 1 and Sx 2 is shown in Figures 1 and 2. Figure 1 shows similar rates of progression for the pT₂ PSM rate as experience increased with no significant differences (P=0.1). Similarly, Figure 2 shows the learning curve for estimated blood loss and operative time; again, both are closely matched with no significant differences between surgeons as their experience increased (P=0.3, 0.9, respectively). We believe that these figures demonstrate Sx 1 and Sx 2 have very similar aptitude for MIRP.

FIG. 1.

Positive surgical margin (PSM) rate for laparoscopic radical prostatectomy by surgeon (1 vs 2) for pT₂ disease.

FIG. 2.

Estimated blood loss (mL) and operative time (min) for laparoscopic radical prostatectomy by surgeon (1 vs 2).

RARP and LRP postlearning curve comparison for Sx 1 and Sx 2

Table 3 shows the mean outcomes for RARP and LRP for various outcomes calculated after the plateau point was reached and does not include procedures performed during the learning curve. The mean estimated blood loss was lower in the LRP vs RARP group. The overall PSM rate, pT₂ PSM rate, and EC rates were all significantly better in the RARP vs LRP group. The complication rate and mean operative time were similar in both groups.

Table 3.

Comparison of Mean Outcomes After Learning Curve Plateaued by Modality

Outcome measure	RARP	LRP	RARP learning curve plateau	LRP learning curve plateau	P value
Mean blood loss (mL)	272 (207–290)	202 (166–250)	250 cases	250 cases	0.002
Mean operative time (min)	124 (108–133)	131 (127–135)	250	250	0.151
Overall PSM rate (%)	14 (12–16)	19 (18–22)	300	200	0.001
pT₂ PSM rate (%)	6 (3–9)	12 (11–15)	300	250	0.003
pT₃ PSM rate (%)	31 (20–40)	37 (33–41)	250^*	200	0.113
Complication rate (C-D III grade, %)	0.3 (0–2)	0.8 (0–2)	250	250	0.439
Early continence (%)	77 (72–81)	35 (26–42)	100	350	<0.001

No plateau identified (continues to improve after more than 500 cases; however, for analysis, point taken from when crossed LRP learning curve line.

C-D=Clavien-Dindo.

Analysis of the RARP learning curve shows the length for estimated blood loss (250 cases), operative time (250 cases), and complication rate (250 cases). The learning curves for the RARP was also long for overall PSM rate (300 cases), pT₂ PSM rate (300 cases), but much shorter for EC (100 cases) (Fig. 3). The pT₃ PSM learning curve for RARP continues to improve even after 500 cases.

FIG. 3.

Early urinary continence (%) by modality (Sx 1 robot-assisted radical prostatectomy [RARP] vs Sx 2 laparoscopic radical prostatectomy [LRP]).

Analysis of the location and rate of PSM between LRP (Sx 2) and RARP revealed a significantly lower apical PSM rate in the RARP group (P=<0.001). Subanalysis of the learning curve for apical PSM showed no learning curve for RARP, but a 200-case learning curve for LRP (Fig. 4). There was a statistically significant difference in overall apical PSM between LRP and RARP (P=<0.001), which reached significance after only 50 cases. This remained significant when combining LRP apical PSM rates for both surgeons' LRP data and comparing this with the RARP apical PSM (P=<0.001).

FIG. 4.

Learning curve for apical PSM by modality (Sx 1 RARP vs Sx 2 LRP).

The EC rate is higher for RARP than LRP within the first group of 50 patients and remains significantly higher throughout the series, similar to apical PSM, which is consistently lower than LRPs throughout the entire series. Complication rates are similar between LRP and RARP.

Missing data

Data for EC outcomes were missing in both the LRP and RARP groups. In the last two groups (last 100 patients) in both the LRP and RARP groups, there were incomplete data in >50% of patients regarding EC, and therefore these groups were excluded. For the other groups that were included: In total, data were missing in 10/450 (2%) patients in the LRP group and 58/450 (12%) in the RARP group; these patients were excluded.

Discussion

This study investigated the learning curve for RARP and compared the postlearning curve outcomes between RARP and LRP for two experience- and volume-matched surgeons from the United Kingdom (UK). This study adds to the evidence base because it analyzes (with adjustment for confounders) outcomes and the learning curve for perioperative (estimated blood loss, operative time, complications), oncologic (PSM rate), and functional (EC rate) between the two minimally invasive approaches for RP using more than 1000 patients. The study reveals a significant learning curve for RARP, similar to our published report¹² for LRP groups in all three outcome areas assessed. The study, however, shows clear superiority in outcomes, after the learning curve plateaued, in PSM and EC for the RARP group.

The strengths of the study are the in-depth comparison of two experience- and volume-matched surgeons in the same region (UK) of the learning curve and outcomes for two minimally invasive RP approaches. The experience (more than 500 cases) for each surgeon enabled “plateauing” of the learning curve to be demonstrated for all parameters except the pT₃ margin rate. The analysis of the two surgeons' LRP learning curves clearly demonstrates similarity between both Sx 1 and Sx 2, which strengthens the resultant comparison of learning curves for RARP and LRP.

We are aware that the study has some weaknesses, which include the lack of validated questionnaires and quality of life data as previously reported by Thompson and associates.¹⁷ Standardized reporting of outcome variables were used in both groups, however. Unfortunately we did not have the data to compare long-term outcomes such as biochemical recurrence (BCR), urinary continence, and potency.

This study is also limited by the inclusion of only two surgeons' outcomes, despite being matched for experience, aptitude, and volume. Randomized controlled trials to investigate learning curves seem unlikely to be performed, and while including more surgeons in this type of study would be ideal, this is also limited by the difficulty of collection and maintaining large prospective databases with adequate follow-up. Another limitation is the fact that during his transition to RARP, Surgeon 1 also performed 53 LRPs, which is likely to have affected his learning curve of RARP when viewed in isolation. A further limitation is the change in approach for Sx 1; in his LRP experience, this was extraperitoneal; however, for his RARP experience, this changed to the transperitoneal approach, which may confound the results to a certain extent.

The change from LRP to RARP is an ever-increasing trend both in the UK and Europe, however, and demonstrates aptly the considerable learning curve required. The major outcome measures assessed would not be affected by this change in approach (PSM rate, EC rate, complication rate (CD-III and higher). Similarly, the use of a Rocco suture by Sx 1 after October 2012 (last 11 months of this study) in his RARP may confound the EC rates; however, the lack of a demonstrable change in EC rates for Sx 1's RARP experience suggests this was not the case (Fig. 3).

Furthermore, the potential for other confounders (such as body mass index, previous surgery, etc) to be important is possible, which is the risk of observational studies such as this. Our findings, however, corroborate those of two recent studies (Thompson and colleagues¹⁷ and Sooriakumaran and coworkers²³), which demonstrated a significant learning curve for RARP with the end result of improved outcomes.

An interesting finding in our study is the reduced apical PSM rate and better EC rate for RARP when compared with LRP, which reflects anecdotal experience of surgeons. This is likely the result of 3D vision, enhanced magnification, greater degrees of freedom, downscaled movements, controlled gentle traction provided with the robotic system that allows for improved apical dissection. It seems probable that this is related to the technological platform rather than an individual surgeon-related factor because better apical PSM and EC rates were apparent after only 50 cases, which became statistically significant within the first 50 cases and continued to be significantly better than LRP throughout the series.

This finding is particularly notable because we demonstrate that there are significant learning curves with RARP for most other parameters (overall, pT₂ and pT₃ PSMs; complications; operative time; estimated blood loss). This finding supports other reports stating an improved dissection with RARP in areas such as the prostatic apex where space is very limited.⁷

The learning curve for LRP is known to be long with multicenter and pentafecta studies on this indicating a learning curve of approximately 250 cases.^12,13,24 There have been many initial reports of RARP having a “short” learning curve^8,10,11; however, these were early reports and did not have sufficient numbers to identify the plateau of the learning curve. Two recent studies^17,23 have demonstrated a much longer learning curve than the 12 or 18 cases shown earlier. Our study demonstrates a long learning curve needed to plateau that is similar in length to that of LRP. Our study also demonstrates, however, that RARP enables the surgeon to improve PSM and EC rates beyond that of pure LRP once the learning curve has been achieved.

The learning curve for pT₂ PSM rates found in our study is also similar to recent publications, which demonstrated a learning curve of more than 300 cases.¹⁷ Furthermore, the recent multicenter study of PSM rates by Sooriakumaran and colleagues² demonstrated that >1000-case experience was needed before the pT₃ PSM rate plateaued, an observation consistent with our findings that there was continued improvement in the pT₃ PSM rate even after 500 cases.

The demonstration of continuing improvement in the pT₃ PSM rate is an interesting finding because pT₃ disease has a higher risk of BCR than pT₂ disease and the trend for superior outcomes in this regard may well translate into a lower rate of BCR in RARP than LRP. It is clear, however, that meaningful comparison of BCR in this group will only be possible by collecting data from surgeons who have completed more than 1000 cases and their pT₃ PSM rate has plateaued. Only by comparing BCR outcomes from such highly experienced surgeons will any true differences become apparent.

More data are needed on outcomes from LRP and RARP to further strengthen the learning curve evidence. Nationwide reporting of surgeon outcomes for RP such as by BAUS may well provide useful insight into the true learning curve for the different RP approaches.²⁵ These large databases will provide greater insight if data on surgeon experience, previous ORP, LRP, fellowship training, and volume are collected. The difficulties that surgeons face when trying to maintain a prospective database without administrative support may inhibit such data collection.

In some quarters, there remains equipoise with regard to whether one modality (LRP/RARP) has any benefits over the other. Systematic reviews and meta-analyses can provide some insight; however, they cannot be adequately adjusted for learning curves, experience, or aptitude. Despite this equipoise, there is a trend for the increasing adoption of RARP at great cost to health services across the world, in part driven by industry but also by patients demanding the best care. Our study helps to demonstrate that the robotic surgical system does not shorten the learning curve for RP in comparison with LRP. After adjusting for the learning curve, however, it appears that RARP has clear benefits in improved apical dissection of the prostate with the resultant benefits in pT₂, pT₃ PSM rates and EC rates.

Importantly, this study adds further evidence that MIRP (LRP or RARP) should only be practiced in high-volume institutions by surgeons with a high volume of cases to ensure best outcomes and lessen exposure of patients to the learning curve of individual surgeons. It also provides evidence of a considerable learning curve when transferring from LRP to RARP, something that is an ever-increasing trend in the UK and Europe.

The introduction of robotic simulators, both virtual reality and mechanical simulators, where trainees are able to practice specific tasks or steps in a procedure is a promising development.^26

–29 It should be noted that while mechanical simulators that have high face, content, and construct validity³⁰ are useful in helping trainees to improve their manipulative skills and in learning steps of a procedure, they are unable currently to add the subtle anatomic detail of prostatectomy.

Novel training curricula such as that from the European Association of Urology,³¹ which is a 12-week structured program, is a further positive development because this combines e-learning, simulator, and supervised modular training for RARP. Whether these novel training curricula and better simulators prove to be more effective in improving outcomes and shortening the learning curves for RARP in comparison with the traditional fellowship is still to be studied.

We have recently obtained funding for a prospective multisurgeon, two-center study comparing LRP vs RARP for outcomes and cost that will overcome many of the weaknesses of this retrospective study and are recruiting in January 2015.

Conclusion

This retrospective cohort comparison study of RARP and LRP with two experience-, aptitude-, and volume-matched surgeons identified that both approaches had similar, considerable learning curves. Despite the long learning curve for RARP, however, this study identified significant benefits in lower PSM rates and better EC in comparison with LRP. We think that this supports the concept of RP being centralized in a small number of high-volume centers and that there are significant benefits to patients with RARP over LRP, especially the outcomes linked to better apical dissection (apical PSM and early continence).

Footnotes

Disclosure Statement

No competing financial interests exist.

Abbreviations Used

References

Center

, Jemal

, Lortet-Tieulent

, et al. International variation in prostate cancer incidence and mortality rates. Eur Urol, 2012; 61:1079–1092.

Sooriakumaran

, Srivastava

, Shariat

, et al. A multinational, multi-institutional study comparing positive surgical margin rates among 22393 open, laparoscopic, and robot-assisted radical prostatectomy patients. Eur Urol, 2014; 66:450–456.

Rassweiler

, Stolzenburg

, Sulser

, et al. Laparoscopic radical prostatectomy—the experience of the German Laparoscopic Working Group. Eur Urol, 2006; 49:113–119.

Ficarra

, Novara

, Artibani

, et al. Retropubic, laparoscopic, and robot-assisted radical prostatectomy: A systematic review and cumulative analysis of comparative studies. Eur Urol, 2009; 55:1037–1063.

Lowrance

, Eastham

, Savage

, et al. Contemporary open and robotic radical prostatectomy practice patterns among urologists in the United States. J Urol, 2012; 187:2087–2092.

Ploussard

, de la Taille

, Moulin

, et al. Comparisons of the perioperative, functional, and oncologic outcomes after robot-assisted versus pure extraperitoneal laparoscopic radical prostatectomy. Eur Urol, 2014; 65:610–619.

Stolzenburg

, Qazi

, Holze

, et al. Evaluating the learning curve of experienced laparoscopic surgeons in robot-assisted radical prostatectomy. J Endourol, 2013; 27:80–85.

Tewari

, Srivasatava

, Menon

, Members of the VIP Team. A prospective comparison of radical retropubic and robot-assisted prostatectomy: Experience in one institution. BJU Int, 2003; 92:205–210.

Menon

, Tewari

, Baize

, et al. Prospective comparison of radical retropubic prostatectomy and robot-assisted anatomic prostatectomy: The Vattikuti Urology Institute experience. Urology, 2002; 60:864–868.

10.

Patel

, Tully

, Holmes

, Lindsay

. Robotic radical prostatectomy in the community setting—the learning curve and beyond: Initial 200 cases. J Urol, 2005; 174:269–272.

11.

Ahlering

, Skarecky

, Lee

, Clayman

. Successful transfer of open surgical skills to a laparoscopic environment using a robotic interface: Initial experience with laparoscopic radical prostatectomy. J Urol, 2003; 170:1738–41.

12.

Good

, Stewart

, Stolzenburg

, McNeill

. Analysis of the pentafecta learning curve for laparoscopic radical prostatectomy. World J Urol, 2014; 32:1225–1233.

13.

Vickers

, Savage

, Hruza

, et al. The surgical learning curve for laparoscopic radical prostatectomy: A retrospective cohort study. Lancet Oncol, 2009; 10:475–480.

14.

Wirth

, Hakenberg

. Surgery and marketing: Comparing different methods of radical prostatectomy. Eur Urol, 2009; 55:1031–3.

15.

Touijer

. Marketing versus science: A fight between necessary evil and stern good over the adoption of new technology in medicine. Eur Urol, 2010; 58:522–524.

16.

Graefen

, Beyer

, Schlomm

. Outcome of radical prostatectomy: Is it the approach or the surgical expertise?. Eur Urol, 2014; 66:457–458.

17.

Thompson

, Egger

, Böhm

, et al. Superior quality of life and improved surgical margins are achievable with robotic radical prostatectomy after a long learning curve: A prospective single-surgeon study of 1552 consecutive cases. Eur Urol, 2014; 65:521–531.

18.

Sharma

, Papadopoulos

, Lee

, et al. First 500 cases of robotic-assisted laparoscopic radical prostatectomy from a single UK centre: Learning curves of two surgeons. BJU Int, 2011; 108:739–747.

19.

Guillonneau

, Vallancien

. Laparoscopic radical prostatectomy: The Montsouris technique. J Urol, 2000; 163:1643–1649.

20.

Stolzenburg

, Kallidonis

, Minh

, et al. Endoscopic extraperitoneal radical prostatectomy: Evolution of the technique and experience with 2400 cases. J Endourol, 2009; 23:1467–1472.

21.

Stolzenburg

, Rabenalt

, Do

, et al. Intrafascial nerve-sparing endoscopic extraperitoneal radical prostatectomy. Eur Urol, 2008; 53:931–940.

22.

Khan

, Abboudi

, Khan

, et al. Measuring the surgical ‘learning curve’: Methods, variables and competency. BJU Int, 2014; 113:504–508.

23.

Sooriakumaran

, John

, Wiklund

, et al. Learning curve for robotic assisted laparoscopic prostatectomy: A multi-institutional study of 3794 patients. Minerva Urol Nefrol, 2011; 63:191–198.

24.

Secin

, Savage

, Abbou

, et al. The learning curve for laparoscopic radical prostatectomy: An international multicenter study. J Urol, 2010; 184:2291–2296.

25.

Laird

, Fowler

, Good

, et al. Contemporary practice and technique-related outcomes for radical prostatectomy in the UK: A report of national outcomes. BJU Int, 2014. Epub ahead of print.

26.

Abboudi

, Khan

, Aboumarzouk

, et al. Current status of validation for robotic surgery simulators—a systematic review. BJU Int, 2013; 111:194–205.

27.

Hung

, Zehnder

, Patil

, et al. Face, content and construct validity of a novel robotic surgery simulator. J Urol, 2011; 186:1019–1024.

28.

Liss

, Abdelshehid

, Quach

, et al. Validation, correlation, and comparison of the da Vinci trainer(™) and the daVinci surgical skills simulator(™) using the Mimic(™) software for urologic robotic surgical education. J Endourol, 2012; 26:1629–1634.

29.

Lallas

, Davis

. Robotic surgery training with commercially available simulation systems in 2011: A current review and practice pattern survey from the Society of Urologic Robotic Surgeons. J Endourol, 2012; 26:283–293.

30.

Alemozaffar

, Narayanan

, Percy

, et al. Validation of a novel, tissue-based simulator for robot-assisted radical prostatectomy. J Endourol, 2014; 28:995–1000.

31.

Volpe

, Ahmed

, Dasgupta

, et al. Pilot validation study of the European Association of Urology robotic training curriculum. Eur Urol, 2014. Epub ahead of print.