Abstract
Objective
To assess the learning curve for radical prostatectomy (RP) with the Versius system (CMR Surgical, Cambridge, UK) and evaluate the transfer of skills from previous robotic and laparoscopic practice.
Subjects, Patients and Methods
This prospective cohort study included patients with localised prostate cancer undergoing Versius‐assisted RP. Four surgeons participated: two with extensive prior robotic experience (>500 Da Vinci Surgical System [Intuitive Surgical Inc., Sunnyvale, CA, USA] RPs) and two with laparoscopic but no robotic background, operating under structured proctoring. Outcomes included objective measures such as console time or novel hand motion volumetry. Proficiency was defined by Learning Curve Cumulative Sum (LC‐CUSUM) analysis with thresholds of console time <165/180 min (without/with lymphadenectomy) and absence of significant positive surgical margins.
Results
A total of 151 consecutive patients were included. Pre‐ and postoperative characteristics were comparable across surgeons. There were clinically and statistically significant improvements in operative time and hand motion metrics with increasing experience. Proficiency was achieved after 23 cases by experienced surgeons and after 37 by novel surgeons. The overall positive margin rate was 27.8%, with only two cases exceeding the ≥3 mm threshold in nerve‐sparing procedures. Major complications (Clavien–Dindo Grade ≥III) only occurred in four (2.6%) patients.
Conclusion
Surgical proficiency in Versius‐assisted RP was achieved in <40 cases, particularly among surgeons with prior robotic experience. Structured proctoring shortened the learning curve, and hand motion volumetry emerged as an objective marker of skill, supporting its use in training and competency assessment.
Keywords: robotics, radical prostatectomy, prostate cancer, Versius, learning curve, skill transference
Abbreviations
- BSU
bedside unit
- IQR
interquartile range
- LC‐CUSUM
Learning Curve Cumulative Sum
- NVB
neurovascular bundle
- PSM
positive surgical margin
- RP
radical prostatectomy
- SD
standard deviation
Introduction
Robotic surgery has been progressively implemented across multiple surgical specialties. In Urology, radical prostatectomy (RP) is one of the primary indications due to its inherent complexity, a robotic approach offers multiple advantages over open and laparoscopic surgery for prostate cancer [1]. In addition to enabling greater technical precision, robotic surgery may facilitate proctoring when remote procedures are performed thanks to the development in telecommunications networks [2].
In recent years, new robotic systems beyond the traditional Da Vinci platform (Intuitive Surgical Inc., Sunnyvale, CA, USA) have emerged, such as the modular Hugo robot‐assisted surgery (RAS) system (Medtronic, Minneapolis, MN, USA) and Versius system (CMR Surgical, Cambridge, UK), together with the democratisation of robotic surgery, means that surgeons will likely need to operate across different robotic platforms throughout their careers [3, 4].
The Versius robot is a modular system with an open console with joystick‐style controllers without any pedals, its design may require specific adaptation by the surgeon (Supporting Information).
At our centre, both the DaVinci and Versius platforms have been available since 2023, making us pioneers in the clinical use of the latter in Spain. These two systems present significant differences in aspects such as docking, joystick design, and the use (or not) of foot pedals, which poses additional challenges in mastering both platforms.
Two key aspects arise in this context: first, the learning curve of surgeons by adopting a new robotic platform; and second, the potential transfer of skills from other surgical techniques (laparoscopy) or platforms, as studies suggest improved performance in robotic surgery if one has prior experience in robotic or laparoscopic procedures [5, 6, 7].
The aim of this study was to assess the implementation and learning progression of the Versius system in surgeons with prior minimally invasive surgery experience, as well as to analyse the learning curve of surgeons with or without prior robotic surgery experience who were proctored by experts.
Subjects, Patients and Methods
We conducted an analysis of our prospective cohort of patients diagnosed with prostate cancer who underwent Versius‐assisted RP at our centre between March 2023 and June 2025, we collected various clinical and oncological variables. All consecutive patients undergoing the procedure during the study period were included to minimise selection bias. The study question, variables collected and statistical analyses were defined prior to data extraction.
The initial procedures were performed by two expert surgeons (Senior 1 and 2), with >500 previous robot‐assisted RPs using the Da Vinci system, these surgeons were proctorised for their first three cases by an external proctor, after completing 20 cases (these cases also included other surgeries like radical nephrectomies, which are not included in the present study) they were appointed as proctors according to the company's platform policy. These surgeons (Senior 1 and 2) subsequently acted as proctors for two other surgeons (Junior 1 and 2), who had no prior robotic experience. Both had extensive laparoscopic experience, each having performed >100 laparoscopic RPs, and one (Junior 1) was a regular video game player—a factor associated in the literature with enhanced psychomotor skills transferable to robotic surgery [8]. The novice surgeons began their first operations in May 2024 under direct supervision.
Furthermore, the surgery data, which were captured automatically by the software and accessed through the Versius Connect® app, included total surgery time, docking/setup time, console time, number of robotic arm collisions (clashes), and hand motion volumetry for both hands.
We defined a significant positive surgical margin (PSM) as ≥3 mm, as described by John et al. [9] in their meta‐analysis, some studies suggest that margins <3 mm may have a risk of biochemical recurrence similar to negative margins. Moreover, many of these may be capsular denudations, particularly in cases with neurovascular bundle (NVB) preservation. Hence, surgical margin status alone may underestimate the overall quality of surgical resection after RP, as not all capsular violations result in a PSM [10]. Patient selection reflects standard clinical practice, and NVB preservation was determined intraoperatively based on oncological safety and anatomical considerations by the surgeon.
The hand motion volumetry parameter is recorded by the Versius system software. It represents the average movement of the controls in centimetres and reflects the fluidity and efficiency of motion during surgery through clutch usage. Lower volumetry is associated with shorter, more controlled movements, indicating better ergonomics and potentially correlating with higher technical skills (Supporting Information).
Statistical Analysis
A descriptive and univariate analysis was performed to compare patient characteristics and performance variables by surgeon. Chi‐square tests were used to compare the distribution of categorical variables, while one‐way ANOVA F‐tests were applied to compare continuous variables. Continuous variables are reported as mean (SD) or median (interquartile range [IQR]) as appropriate; categorical variables as frequency and percentage.
Linear mixed‐effects models were fitted to evaluate the association between surgical variables. A log‐transformed case order was included to model the learning curve. To facilitate interpretation, we reported the expected change in outcome associated with a 10% increase in case order, that was defined as ‘surgeon's experience’. Prostate volume >60 g, performance of lymphadenectomy, and NVB preservation were included as covariates in the model, as these factors were considered likely to increase operative difficulty and consequently prolong console time. A random intercept for each surgeon was included to account for intra‐surgeon correlation and baseline performance heterogeneity.
Predicted values from the fitted model for console time were used to estimate the minimum number of procedures required to achieve competence. The model was built using data from Senior 1, the most experienced surgeon in the series, whose performance was used as the reference standard. Specifically, we plotted the estimated console time across the full range of experience (case order) while holding other variables constant. The point of initial deceleration and the stabilisation threshold were then identified by examining the marginal reduction in estimated console time per additional procedure, defining these points as the cases at which this reduction fell below 1.0 and 0.5 min per procedure, respectively. This threshold represents the estimated case volume required to achieve a level of efficiency comparable to that of experienced surgeons [11].
The Learning Curve Cumulative Sum (LC‐CUSUM) method was applied to determine when a trainee achieves a pre‐defined level of performance. The LC‐CUSUM monitors the cumulative sum of a score assigned to each case: a positive score is added for a success and a negative score for a failure, with the curve signalling competence when it crosses a pre‐specified control limit (further methodological detail is provided in the Supporting Information). To define the success criteria for the LC‐CUSUM analysis, the surgical metrics of Senior 1 were used as a reference, with consideration of whether surgery included or not lymphadenectomy due to its impact on surgical duration as demonstrated in the multivariable analysis. Senior 2 was excluded from the LC‐CUSUM analysis due to an insufficient number of cases to allow reliable estimation of the learning curve. Console time was chosen to avoid confounding factors such as anaesthetic duration and strictly assess the technical performance of the procedure, using the median + 2 SDs of console time of Senior 1. Finally, we defined the following success criteria: console time <165/180 min (without/with lymphadenectomy) and no PSM ≥3 mm.
Hand motion volumetry was analysed descriptively and compared across surgeons as an exploratory parameter. Given its novelty and the lack of established benchmarks, it was not incorporated into the LC‐CUSUM success criteria.
All tests were two‐sided and a P < 0.05 was considered statistically significant. Univariate analyses, statistical models, and learning curves were implemented using Stata release 17 (Stata Corp., College Station, TX, USA) and R software (version 4.4.1, R Foundation for Statistical Computing, Vienna, Austria).
Results
Baseline Characteristics and Comparison by Surgeon
A total of 151 patients were included in the analysis. Senior 1, the surgeon with the most procedures, performed 57 cases, Senior 2 completed 19 procedures and Junior 1 and Junior 2 did 38 and 36 procedures, respectively. The mean (SD) age of the patients was 63.8 (6.7) years, the median (IQR) PSA level was 6.3 (5–9.3) ng/mL, and the median (IQR) prostate volume was 40 (30–52.3) mL. The majority of patients (122 cases [80.8%]) were in the intermediate‐risk group according to the D'Amico classification [12]. Among all procedures, 56 lymphadenectomies (37.1%) were performed, and NVB preservation was achieved in 38 cases (25.2%). The overall mean (SD) setup time was 8.2 (3.5) min, the mean (SD) surgical time was 145.3 (42.7) min, and the mean (SD) console time was 129.9 (38) min. A detailed breakdown of these values is provided in the Supporting Information.
Tables 1 and 2 present the preoperative, postoperative, and surgical data categorised by surgeon. Statistically significant differences between groups are highlighted. Significant differences were observed between surgeons in mean age of patients, NVB preservation and all surgery‐related time variables. Senior 1 recorded the lowest mean console time and consistently shorter times across all measured intervals. In contrast, Junior 1 demonstrated the lowest hand motion values across all metrics and had the lowest number of clashes. Two surgeries had to be switched to laparoscopic approach for technical problems with the console. The PSM rate was higher in Junior 1 and Junior 2 (39.5% and 36.1%, respectively) compared to Senior 1 and Senior 2 (17.5% and 21.2%, respectively), although this difference did not reach statistical significance (P = 0.066). There were no statistically significant differences when considering only clinically relevant PSM ≥3 mm rates across surgeons (P = 0.547) (Table 1).
Table 1.
Demographics, preoperative and postoperative data categorised by surgeons.
| Senior 1 (n = 57) | Junior 1 (n = 38) | Junior 2 (n = 36) | Senior 2 (n = 19) | P | |
|---|---|---|---|---|---|
| Preoperative data | |||||
| Age, years, mean (SD) | 64.9 (6.7) | 60.9 (7.7) | 64.6 (5.4) | 65.1 (5.8) | 0.018 |
| PSA level, ng/mL, median (IQR) | 6.4 (4.9–10) | 6.2 (5.2–9.6) | 6.1 (5.0–9.1) | 6.5 (5.5–9) | 0.822 |
| DRE, n (%) | |||||
| T1 | 40 (70.2) | 30 (78.9) | 32 (88.9) | 16 (84.2) | 0.169 |
| T2 | 17 (29.8) | 8 (21.1) | 4 (11.1) | 3 (15.8) | |
| Biopsy Gleason Group, n (%) | |||||
| 1 | 4 (7) | 2 (5.3) | 3 (8.3) | 2 (10.5) | |
| 2 | 23 (40.4) | 20 (52.6) | 20 (55.6) | 5 (26.3) | |
| 3 | 20 (35.1) | 14 (36.8) | 12 (33.3) | 12 (63.2) | 0.138 |
| 4 | 8 (14) | 2 (5.3) | 0 | 0 | |
| 5 | 2 (3.5) | 0 | 1 (2.8) | 0 | |
| Risk group (D'Amico), n (%) | |||||
| Low | 4 (7) | 2 (5.3) | 2 (5.6) | 2 (10.5) | |
| Intermediate | 42 (73.7) | 31 (81.6) | 32 (88.9) | 16 (84.2) | 0.492 |
| High | 11 (19.3) | 5 (13.2) | 2 (5.6) | 1 (5.3) | |
| Prostate volume, mL, median (IQR) | 38 (31.8–52) | 42.5 (27.3–50) | 40 (27–56) | 46 (32–58) | 0.723 |
| Postoperative data | |||||
| Specimen Gleason Group, n (%) | |||||
| 1 | 0 | 2 (5.3) | 3 (8.3) | 2 (10.5) | |
| 2 | 14 (25.9) | 12 (31.6) | 12 (33.3) | 4 (21.1) | |
| 3 | 26 (48.1) | 17 (44.7) | 16 (44.4) | 11 (57.9) | 0.609 |
| 4 | 10 (18.5) | 6 (15.8) | 3 (8.3) | 4 (21.1) | |
| 5 | 4 (7.4) | 1 (2.6) | 2 (5.6) | 0 | |
| Pathological T stage, n (%) | |||||
| T0 | 1 (1.8) | 0 | 0 | 0 | |
| T2 | 37 (64.9) | 35 (92.1) | 27 (75) | 14 (73.7) | |
| T3a | 12 (21.1) | 1 (2.6) | 6 (16.7) | 4 (21.1) | 0.124 |
| T3b | 7 (12.3) | 2 (5.3) | 3 (8.3) | 1 (5.3) | |
| NVB preservation, n (%) | |||||
| No | 47 (82.5) | 20 (52.6) | 31 (86.1) | 14 (73.7) | |
| Unilateral | 6 (10.5) | 4 (10.5) | 4 (11.1) | 2 (10.5) | 0.002 |
| Bilateral | 4 (7) | 14 (36.8) | 1 (2.8) | 3 (15.8) | |
| PSMs (all), n (%) | 10 (17.5) | 15 (39.5) | 13 (36.1) | 4 (21.1) | 0.066 |
| PSMs ≥3 mm, n (%) | 3 (5.3) | 4 (10.5) | 2 (5.6) | 2 (10.5) | 0.547 |
| Lymphadenectomy, n (%) | 24 (42.1) | 15 (39.5) | 8 (22.2) | 8 (42.1) | 0.228 |
| Node number, median (IQR) | 10 (7–13.8) | 10 (8–14) | 12 (10.3–16.5) | 8.5 (7.2–12.2) | 0.310 |
| Positive nodes, n (%) | 1 (4.2) | 1 (6.7) | 1 (11.1) | 0 | 0.851 |
| Complications, n (%) | 10 (17.5) | 6 (15.8) | 9 (25) | 2 (10.5) | 0.563 |
| Clavien–Dindo Grade, n (%) | |||||
| I | 3 (30) | 2 (33.3) | 1 (11.1) | 0 | |
| II | 6 (60) | 4 (66.7) | 7 (77.8) | 0 | |
| IIIa | 1 (10) | 0 | 0 | 1 (50) | 0.413 |
| IIIb | 0 | 0 | 1 (11.1) | 1 (50) | |
Table 2.
Surgery data categorised by surgeon.
| Surgery data | Senior 1 (n = 57) | Junior 1 (n = 38) | Junior 2 (n = 36) | Senior 2 (n = 19) |
|---|---|---|---|---|
| Setup time, min, mean (SD) | 7.2 (2.8) | 7.6 (2.8) | 9.7 (3.9) | 9.5 (4.8) |
| Surgery time, min, mean (SD) | 123.3 (33) | 143.7 (39.9) | 164.9 (36.2) | 179.3 (48.2) |
| Console time, min, mean (SD) | 111.2 (32.2) | 124.7 (27.5) | 149 (34.3) | 161.1 (44.6) |
| Left hand motion, cm, mean (SD) | 30.2 (2.6) | 27.4 (2.8) | 34.6 (2.5) | 27.6 (2) |
| Right hand motion, cm, mean (SD) | 26.7 (2.7) | 22.2 (2.4) | 29.7 (2.8) | 26.3 (1.5) |
| Total hand motion, cm, mean (SD) | 36 (3.1) | 31.6 (3.1) | 40.8 (2.9) | 34 (2) |
| Clashes, n, median (IQR) | 29.5 (16.2–47.5) | 21 (11.5–28) | 59 (32.5–117) | 47 (25–65) |
Intraoperative Performance Metrics
A comparison of surgical performance among Senior 1 and both junior surgeons is illustrated in Figs 1 and 2, setup time for Senior 1 showed a notable improvement after the initial cases and remained stable thereafter. In comparison, setup times for both junior surgeons remained relatively consistent throughout their procedures. Console time decreased progressively for Senior 1 and Junior 1, whereas Junior 2 exhibited stable console times throughout. Procedures involving lymphadenectomy were associated with longer console times across all groups. Patterns in hand motion mirrored these trends, with consistently higher values observed for the left hand among all surgeons. Regarding the number of instrument collisions, it is important to consider the previously discussed limitations of this metric (Fig. S5). Nonetheless, Senior 1 and Junior 1 followed similar patterns to those seen in setup time, whereas Junior 2 demonstrated greater variability in this parameter across cases.
Fig. 1.
Graphical comparison of setup time and console time data between the three surgeons (Senior 1, Junior 1 and Junior 2). Blue, surgery WITH lymphadenectomy; Red, surgery WITHOUT lymphadenectomy.
Fig. 2.
Graphical comparison of volumetry by hands, left hand path (A) and right hand path (B) between the three surgeons (Senior 1, Junior 1 and Junior 2). Blue, surgery WITH lymphadenectomy; Red, surgery WITHOUT lymphadenectomy.
Predictors of Console Time: Multivariable Analysis
Multivariable analysis revealed that lymphadenectomy was associated with a statistically significant increase in all surgery‐related time variables. Regarding console time, the estimated difference was 18.0 min (95% CI 6.3–29.6 min). NVB preservation did not significantly affect any time metrics, except for setup time in cases of unilateral preservation. Prostate volume was significantly associated with right‐hand and total hand motion only, with no statistically significant effect on any time‐related variable. Surgeon's experience demonstrated a significant effect on all time‐related variables and on right‐hand motion. A full description of the variables and the multivariable mixed model is available in the Supporting Information. The predictive model estimated a substantial reduction in console time with increasing surgical experience. Based on modelled values, console time decreased by 33.2 min over the first 14 procedures (20.5% without lymphadenectomy and 18.4% with lymphadenectomy), with a mean estimated reduction of 2.6 min per case, defining the point of initial deceleration (marginal reduction <1.0 min/case). Between cases 14 and 26 the rate of improvement slowed to an estimated 0.65 min per case, reaching modelled stabilisation at case 26 (<0.5 min/case), beyond which only a further 10.5 min of reduction was projected over the remaining 34 procedures (Fig. S6).
Learning Curve Assessment and Proficiency Thresholds
The LC‐CUSUM analysis demonstrated that Senior 1 met the pre‐defined proficiency criteria after 23 procedures, following an initial phase of performance deviation. Junior 1 demonstrated performance deviations during the intermediate phase of the learning curve and achieved proficiency after 37 procedures. Junior 2 had not yet met the proficiency threshold at the end of the study period (Fig. 3).
Fig. 3.
The LC‐CUSUM analysis for performance monitoring of the procedure comparing three surgeons (Senior 1 and Junior 1 and Junior 2). The left panel (blue graph) shows the cumulative number of failures used to assess attainment of competency for each surgeon. The right panel displays the corresponding LC‐CUSUM score, with the calculated control limit for competency indicated by a dashed red line. The learning curve is reflected by the progressive evolution of the score, with a penalty increment applied after each failure.
Discussion
This article presents the longest reported series to date of RP performed using the Versius platform. The findings demonstrate a transfer of skills from prior laparoscopic and robotic surgery experience, enabling a rapid adaptation to the new system. Furthermore, structured proctoring appears to contribute to reduced setup times and facilitates the achievement of procedural expertise in <40 cases, with an even shorter learning curve observed in surgeons with previous experience using the Da Vinci system. Although there is significant variability depending on the outcome measure used to define competency, previous literature has reported that up to 250 cases may be required to achieve expertise—particularly when surgeons have no prior robotic experience [13].
In a recent meta‐analysis comparing surgery times in RP across different robotic platforms, mean surgery times with the Da Vinci system ranged from 71 to 268 min, whereas those with other robotic platforms ranged from 126 to 267 min [4]. Only one of these studies included the Versius system, Dibitetto et al. [14] reported a median (IQR) surgery time of 170 (158–202) min in 51 cases. Surgery time is defined inconsistently across studies and may include not only the actual surgical procedure and this variability may not necessarily reflect surgical competency and can introduce bias. In our study, console and surgery times were lower than those reported for other alternative robotic platforms and were therefore comparable to those achieved with the Da Vinci system. This supports the conclusion that, with sufficient experience, procedures performed using this novel robotic platform can achieve outcomes equivalent to those obtained with Da Vinci system.
It is important to note that setup time can be reduced through adequate training and by optimising the configuration of the bedside units in modular platforms [15]. In our series, setup time improved after the initial cases for the experienced surgeons. In contrast, it remained relatively consistent for both novice robotic surgeons, this finding is consistent with other published series [4, 14, 16, 17, 18, 19, 20].
The use of PSM is not always considered when evaluating learning curves [13]. Intraoperatively, a capsular denudation may occur (e.g., during NVB preservation) that is subsequently repaired after the capsule has already been incised; such cases will still be reported as PSMs [10]. Unfortunately, our pathology reports did not include information on benign capsular violations.
Similarly, not all patients with PSM develop biochemical recurrence, there is evidence that suggests that the length of the PSM may act as an independent prognostic factor for biochemical recurrence, with a threshold of ≥3 mm proposed as a potential cut‐off to identify patients at higher risk [9]. In another study, Kir et al. [21] reported that PSMs measuring 0–2 mm carried a risk equivalent to that of negative surgical margins.
In our series, the overall PSM rate was 27.8%, which is comparable to our institutional rate over the previous 5 years (25.9%). Among patients undergoing NVB preservation, there were 10 PSMs, but only two were ≥3 mm in length. All patients, except one with negative surgical margins, had a negative postoperative PSA level at 2 months.
Although the differences in PSM rates between surgeons did not reach statistical significance, those with prior robotic experience achieved lower PSM rates compared to junior surgeons. This trend supports previous findings that PSM rates tend to decrease with increasing surgical experience [22]. Additionally, variations in the proportion of NVB preservation between surgeons may have contributed to the differences observed, potentially increasing the risk of capsular denudation as described above.
The observed rate of NVB preservation may appear relatively low at first glance. However, it is important to note that the decision to preserve the NVB was made intraoperatively by the operating surgeon, taking into account individual patient anatomy, tumour characteristics, and oncological safety considerations. Although a high proportion of patients in the cohort presented with lower‐risk disease (Gleason Grade Groups 1–2, 52.3%), the study reflects routine real‐world clinical practice rather than a protocolised trial setting. Consequently, nerve‐sparing decisions were individualised and sometimes limited by intraoperative findings. For transparency, we have reported the NVB preservation rates as observed, as they provide important context for interpreting the results of the present study.
Hand motion volumetry was included as an exploratory parameter of robotic performance. While lower motion has been associated with greater experience, it cannot always be interpreted as a direct measure of surgical skill, as reduced movements could result from omission of steps or lower procedural quality. Ideally, correlation with independent video‐based assessment would be required. In the present study, technical progression was primarily evaluated using console time as a surrogate for efficiency and PSMs as a marker of procedural quality. We acknowledge that RP is influenced by multiple factors, and motion metrics are presented here for exploratory purposes only.
The available evidence suggests that shorter and more controlled movement trajectories correlate with greater surgical experience with different metrics described—conceptually analogous to volumetry. Although most of these studies have been conducted in simulated environments, one study stands out for comparing surgeons performing RP with the Da Vinci system, it was based on several performance metrics derived from recorded surgical procedures rather than automated software [23]. Overall, all of these studies demonstrate that objective motion‐based metrics are valid and effective tools for assessing technical skill in robotic surgery [24, 25, 26]. The application of automatically measured volumetry during live clinical procedures using the Versius system represents a novel contribution to the field, offering valuable insights into tracking and improving surgical skills over time, in contrast to existing literature that has primarily focused on simulation‐based assessments. Further research comparing standardised kinematic parameters in multicentre studies could enhance its validity and clinical applicability.
In the present study, while a reduction over time was observed in right‐hand motion, the left‐hand metric consistently showed higher overall values across all surgeons. This is likely due to the fact that all participants were right‐handed, and because traction movements using the grasper (typically broader movements) were performed with the left hand. It is also notable that Junior 1 exhibited a significant reduction in both hand motion metrics, even surpassing the decrease observed in Senior 1, despite the latter's extensive experience in robotic surgery.
There is existing evidence suggesting that surgeons with video gaming experience tend to perform better in simulation‐based and robotic surgical tasks compared to non‐gaming counterparts [8, 27, 28]. These findings support the interpretation that Junior 1's substantial reduction in hand motion metrics may be related to his video gaming background and suggests ergonomic familiarity and dexterity transfer from gaming to robotic console control.
As limitations, the retrospective design and absence of a control group should be acknowledged when interpreting the findings, there are several factors that could influence surgical time, such as patients’ prior abdominal surgeries or body mass index, were not included in the analysis, and there are certain variables, such as preoperative MRI staging and surgical blood loss that were not consistently available for all patients or surgeons during the study period; therefore, these data were not included in the analysis. Instrument clashes were not precisely measured, and they could not be included in the learning curve analysis; further software development would be required to capture these data reliably. The study also did not assess short‐term outcomes, such as functional or continence results, as it was designed to focus on intraoperative technical performance and perioperative outcomes; however, this remains an important area for future research and potential correlation with motion metrics. It is also important to note that objective metrics have some limitations as mentioned before, so they cannot be included in the LC‐CUSUM analysis. Finally, the number of surgeries, when categorised by four individual surgeons, may be insufficient to draw definitive conclusions about learning curves or inter‐surgeon variability, potentially limiting the statistical power of subgroup analyses.
This study presents the largest series of robot‐assisted RPs performed using the Versius system to date, demonstrating that surgical expertise can be achieved in <40 cases, especially with prior robotic experience and structured proctoring. This prior robotic experience provides foundational skills, but individual performance on a novel system may vary, and conclusions regarding its effect must be interpreted cautiously. This study also highlights hand motion volumetry as a novel metric of technical skill, potentially serving as an objective quantitative marker of surgical performance.
Disclosure of Interests
Guillermo Fernández‐Conejo has received honoraria as proctor of the Versius robotic platform from CMR surgical. The authors declare no other conflicts of interest.
Funding
No funds have been received for the purpose of this research output.
Supporting information
Fig. S1. Bedside unit (BSU) showing all the junctions inside the arm.
Fig. S2. Joystick‐style Versius controller.
Fig. S3. Open console of the Versius robot.
Fig. S4. Example of the image obtained after a surgery where the volumetry of each hand is shown.
Fig. S5. Graphical comparison of total hand volumetry and clashes between the three surgeons (Senior 1, Junior 1 and Junior 2). Red: surgery WITHOUT lymphadenectomy. Blue: surgery WITH lymphadenectomy.
Fig. S6. Predictive model to estimate the number of cases to achieve a significant reduction in console time with the Senior 1's time data. Blue, surgeries without lymphadenectomy; Orange, surgeries with lymphadenectomy; x‐axis, number of surgeries; y‐axis: console time (min).
Table S1. Patients characteristics and surgery.
Table S2. Description variables of the surgery data and its relationship with lymphadenectomy, NVB preservation and prostate volume.
Table S3. Multivariable analysis with the surgery data variables. Statistically significant differences are highlighted (P < 0.05).
Table S4. Descriptive surgery data of Senior 1 categorised in cases with or without lymphadenectomy.
Table S5. Predictive model to estimate the number of cases to achieve a significant reduction in console time with the Senior 1's time data.
Data Availability Statement
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
References
- 1. Kim DK, Moon YJ, Chung DY et al. Comparison of robot‐assisted, laparoscopic, and open radical prostatectomy outcomes: a systematic review and network meta‐analysis from KSER update series. Medicina 2025; 61: 61 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Moustris G, Tzafestas C, Konstantinidis K. A long distance telesurgical demonstration on robotic surgery phantoms over 5G. Int J Comput Assist Radiol Surg 2023; 18: 1577–1587 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Ficarra V, Rossanese M, Giannarini G et al. Evaluation of clinical research on novel multiport robotic platforms for urological surgery according to the IDEAL framework: a systematic review of the literature. Eur Urol Open Sci 2024; 67: 7–25 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Reitano G, Tumminello A, Ghaffar U et al. Perioperative, oncological, and functional outcomes of new multiport robotic platforms in urology: a systematic review and meta‐analysis. Eur Urol Open Sci 2025; 74: 44–70 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Sighinolfi MC, Terzoni S, Scanferla E et al. Prior robotic console expertise may improve basic skills at the new Hugo RAS simulator: results from a cohort trial and implications for skill transference across platforms. Eur Urol Open Sci 2023; 53: 83–89 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Quinn KM, Chen X, Griffiths C, Chen G, Osayi S, Husain S. Skill transference and learning curves in novice learners: a randomized comparison of robotic and laparoscopic platforms. Surg Endosc 2023; 37: 8483–8488 [DOI] [PubMed] [Google Scholar]
- 7. Prontera PP, Prusciano FR, Marco L et al. Impact of laparoscopic experience on learning curves in robotic‐assisted radical prostatectomy (RaRP): a comparative analysis of oncological and functional outcomes. Arch Ital Urol Androl 2025; 97: 13640 [DOI] [PubMed] [Google Scholar]
- 8. Fernández‐Conejo G. The transfer of video game skills to robotic and laparoscopic surgery: feasibility in contemporary practice. Actas Urol Esp (Engl Ed) 2026; 50(5):501813. 10.1016/j.acuroe.2025.501813 [DOI] [PubMed] [Google Scholar]
- 9. John A, Lim A, Catterwell R, Selth L, O'Callaghan M. Length of positive surgical margins after radical prostatectomy: does size matter? – a systematic review and meta‐analysis. Prostate Cancer Prostatic Dis 2023; 26: 673–680 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Williams SB, Sutherland DE, Frazier HA 2nd, Schwartz A, Engel JD. Pathologic analysis of capsular and incisional denudation and positive margin status in the development of a robot‐assisted laparoscopic prostatectomy program. J Robot Surg 2009; 3: 137 [DOI] [PubMed] [Google Scholar]
- 11. Biau DJ, Porcher R. A method for monitoring a process from an out of control to an in control state: application to the learning curve. Stat Med 2010; 29: 1900–1909 [DOI] [PubMed] [Google Scholar]
- 12. European Association of Urology . EAU guidelines. 2025. Presented at: EAU Annual Congress, Madrid, Spain. ISBN:978‐94‐92671‐29‐5
- 13. Chahal B, Aydin A, Amin MS et al. The learning curves of major laparoscopic and robotic procedures in urology: a systematic review. Int J Surg 2023; 109: 2037–2057 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Dibitetto F, Castellucci R, Russo P et al. Extraperitoneal robot‐assisted radical prostatectomy by the da Vinci and Versius system: first comparative analysis. Minerva Urol Nephrol 2024; 76: 570–577 [DOI] [PubMed] [Google Scholar]
- 15. Fernández‐Conejo G, Subiela Henríquez J, Tagalos Muñoz A et al. Standardizing radical prostatectomy with the Versius robotic system: technical setup and early institutional experience. J Robot Surg 2025; 19: 470 [DOI] [PubMed] [Google Scholar]
- 16. Abdelhakim MA, Abdelwahab M. Robot‐assisted radical prostatectomy with the Versius surgical platform: an objective criticism and a guide for an optimal surgical setup. Arab J Urol 2024; 23: 152–159 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Marino F, Moretto S, Rossi F et al. Robot‐assisted radical prostatectomy with the Hugo RAS and da Vinci surgical robotic systems: a systematic review and meta‐analysis of comparative studies. Eur Urol Focus 2024; 2: 195 [DOI] [PubMed] [Google Scholar]
- 18. Scarcia M, Filomena GB, Moretto S et al. Extraperitoneal robot‐assisted radical prostatectomy with the Hugo RAS system: initial experience at a high‐volume robotic centre. J Clin Med 2024; 13: 5916 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Ragusa A, Prata F, Iannuzzi A et al. Safety and feasibility of “three arms settings” robot‐assisted radical prostatectomy using the Hugo RAS system: surgical set‐up in a double‐center large case series. World J Urol 2024; 42: 517 [DOI] [PubMed] [Google Scholar]
- 20. Antonelli A, Veccia A, Malandra S et al. Intraoperative performance of DaVinci versus Hugo RAS during radical prostatectomy: focus on timing, malfunctioning, complications, and user satisfaction in 100 consecutive cases (the COMPAR‐P trial). Eur Urol Open Sci 2024; 63: 104–112 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Kir G, Arikan EA, Seneldir H et al. Determining the cut‐off values of tumor diameter, degree of extraprostatic extension, and extent of surgical margin positivity with regard to biochemical recurrence of prostate cancer after radical prostatectomy. Ann Diagn Pathol 2020; 44: 151431 [DOI] [PubMed] [Google Scholar]
- 22. Bravi CA, Dell'Oglio P, Piazza P et al. Positive surgical margins after anterior robot‐assisted radical prostatectomy: assessing the learning curve in a multi‐institutional collaboration. Eur Urol Oncol 2024; 7: 821–828 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Hung AJ, Chen J, Jarc A, Hatcher D, Djaladat H, Gill IS. Development and validation of objective performance metrics for robot‐assisted radical prostatectomy: a pilot study. J Urol 2018; 199: 296–304 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Bitner DP, Choksi S, Carsky K et al. Kinematic metrics and surgeon experience in robotic cholecystectomies: a pilot study on breaking down technical performance. Surg Endosc 2024; 38: 913–921 [DOI] [PubMed] [Google Scholar]
- 25. Liang K, Xing Y, Li J, Wang S, Li A, Li J. Motion control skill assessment based on kinematic analysis of robotic end‐effector movements. Int J Med Robot 2018; 14: e1845 [DOI] [PubMed] [Google Scholar]
- 26. Olsen RG, Svendsen MBS, Tolsgaard MG, Konge L, Røder A, Bjerrum F. Automated performance metrics and surgical gestures: two methods for assessment of technical skills in robotic surgery. J Robot Surg 2024; 18: 297 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Hvolbek AP, Nilsson PM, Sanguedolce F, Lund L. A prospective study of the effect of video games on robotic surgery skills using the high‐fidelity virtual reality RobotiX simulator. Adv Med Educ Pract 2019; 10: 627–634 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Gupta A, Lawendy B, Goldenberg MG, Grober E, Lee JY, Perlis N. Can video games enhance surgical skills acquisition for medical students? A systematic review. Surgery 2021; 169: 821–829 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Fig. S1. Bedside unit (BSU) showing all the junctions inside the arm.
Fig. S2. Joystick‐style Versius controller.
Fig. S3. Open console of the Versius robot.
Fig. S4. Example of the image obtained after a surgery where the volumetry of each hand is shown.
Fig. S5. Graphical comparison of total hand volumetry and clashes between the three surgeons (Senior 1, Junior 1 and Junior 2). Red: surgery WITHOUT lymphadenectomy. Blue: surgery WITH lymphadenectomy.
Fig. S6. Predictive model to estimate the number of cases to achieve a significant reduction in console time with the Senior 1's time data. Blue, surgeries without lymphadenectomy; Orange, surgeries with lymphadenectomy; x‐axis, number of surgeries; y‐axis: console time (min).
Table S1. Patients characteristics and surgery.
Table S2. Description variables of the surgery data and its relationship with lymphadenectomy, NVB preservation and prostate volume.
Table S3. Multivariable analysis with the surgery data variables. Statistically significant differences are highlighted (P < 0.05).
Table S4. Descriptive surgery data of Senior 1 categorised in cases with or without lymphadenectomy.
Table S5. Predictive model to estimate the number of cases to achieve a significant reduction in console time with the Senior 1's time data.
Data Availability Statement
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.