What credentials are required for robotic‐assisted surgery in reconstructive and functional urology?

Abstract Introduction The increasing popularity of robotic assisted surgery (RAS) as it is implemented in to sub specialities poses many challenges to ensuring standards in quality and safety. The area of Reconstructive and Functional Urology (RFU) has a wide range and largely complex heterogeneous procedures. In recent years RFU has started to incorporate RAS as the primary method to undertake these procedures due to improved vision, dexterity, and access to deep cavities. To ensure patient safety majority of institutions maintain minimal requirements to operate using RAS however across specialities and institutions these greatly vary. Methods A narrative review of all the relevant papers known to the author was conducted. Results Specific challenges facing RFU is the inability to rely on case numbers as a surrogate means to measure competency as well the ongoing consideration of how to differentiate between surgeons with robotic training and those with the clinical experience specific to RFU. Conclusion This review explores current models of training and credentialling and assess how it can be adapted to suggest a standardised guideline for RFU to ensure the highest standards of patient care.

process is required. This area of urology has unique challenges for the credentialling as there is considerable heterogeneity of procedures within this umbrella term.
Broad guidelines have been released by associations such as the British Association of Urological Surgeons (BAUS) on principles for training and credentialling of surgeons. These policies focus on urooncological procedures such as RARP and RAPN (robot-assisted partial nephrectomy). The Society of European Robotic Gynaecological Surgery (SERGS) has recently addressed the important aspects of credentialling of robotic gynaecology procedures. 3 The Society published a pilot curriculum in the form of a fellowship-programme using a standardised modular and stepwise educational programme and validated tests as proof of efficacy similar to BAUS guidelines for RARP and RAPN. These frameworks can be adapted as templates for a proposed curriculum for robot-assisted surgery in urology and urogynaecology.
In addition to heterogeneity, RFU is challenging to credential because many of the procedures are complex and technically challenging and the case volume for each procedure to be credentialed is low compared to RARP or RAPN. Further the conditions themselves require specialist training to ensure surgical judgement regarding appropriate choice of procedure.
This review explores the training and credentialling process that is described for RAS in other subspecialities to determine aspects that can be adapted to creation of a training and credentialling guideline for reconstructive and functional urology to ensure safe patient care and good outcomes are achieved with 'onboarding'. What is wrong with the SERGS draft that we are not just adopting it?  4 Robot-assisted sacrocolpopexy has been extensively reported and it is now considered the gold standard for apical prolapse. Large meta-analysis demonstrated an objective cure rate of 84-100%, a reoperation rate of 3.3% and a recurrence rate of 6.4%. 5 The RAS approach was found to be associated with increased operating time, but reduced blood loss and length of stay. 5 It has considerably gained in popularity after the first FDA warning about transvaginal mesh for pelvic organ prolapse in 2008. 6 Colposuspension has seen a revival in popularity following the transvaginal mesh controversy. There is very limited literature on the RAS Burch Colposuspension. A small pilot study comparing robotic hysterectomy and robotic colposuspension with open hysterectomy and colposuspension demonstrated no difference in continence rates, however there were advantages when comparing blood loss and hospitalisation time. 7 Tan et al reported on 28 cases of robotic Burch Colposuspension with promising improvements in pad rates and quality of life scores. 8 As mesh free options become more popular there has been success in several small series using autologous fascia robotic sacral colpopexy. Bock et at showed similar short-term anatomic outcomes to mesh for women with apical pelvic organ prolapse 9 and Scott et al reported no intra-or postoperative complications, and the median Patient Global Impression of Improvement (PGI-I) response was 2 (range, 1-3, very much to a little better). 10

| Bladder neck reconstruction and artificial urinary sphincter
Bladder neck contracture, though uncommon, is a challenging complication of prostate surgery and pelvic trauma. A robotic-assisted approach to V-Y-plasty of the bladder neck has recently been described by two small case series by Granieri  appear to favour the use of RAS. 14,15 The use of RAS innovates an established implantation technique with potential significant advantages. However female artificial sphincter is not a common operation and tends to be used in complicated cases rather than as first or second line for severe stress incontinence. In this small series was a high rate of bladder (18%) and vagina (18%) injury reported. 14 Learning curves play a major role in determining quality and safety, not just the robotic experience but indeed understanding of the anatomy itself and possibly prior experience with low volume procedure even by open surgery.

| Upper urinary tract procedures: Pyeloplasty and ureteric reimplantation and reconstruction
Multiple studies [16][17][18] have demonstrated comparable outcomes with a reduction in operating time for the robotic approach compared to open pyeloplasty surgery, the relative ease of suturing being the key advantage. 1 Data on RAS ureteric reimplant surgery demonstrates similar complication rates, re-stricture rates and operative times but with decreased blood loss and length of stay. 19,20 Compared to conventional open surgery, RAS surgery generally results in reduced wound related morbidity. 21 Ureteric reconstruction has also utilised RAS and buccal mucosa grafts in a select patient cohort of previously failed pyeloplasty. This small multi-centre study demonstrated recurrence-free rate of 95% at 2 years 22 comparable to both open and laparoscopic rates of conventional re-do pyeloplasty. 23 RAS is also being utilised in the paediatric population for re-do pyeloplasty for recurrent pelvi-ureteric junction obstruction (PUJO).

| Other RAS reconstructive urological procedures
There are several other RAS reconstructive urological procedures described by expert pioneers, generally with case series describing feasibility but in most series with relatively small numbers including: bladder reconstruction and urinary diversion, [24][25][26] repair of genitourinary fistula, 27 gender affirming neo-vaginoplasty, 27  They demonstrated that a stepwise approach of increasing difficulty guards against surgeons attempting steps that are too complex and promotes the acquisition of competent operating and patient safety. 32

| Non-urologic procedures
Rice et al describe the vastly improved learning curves with progression from non-structured toward structured training from robotic pancreatico-duodenectomy. From self-taught pioneers (1st generation) through mentorship alone (2nd generation) to mentorship and modular curriculum, they demonstrated significant reduction in operating times accompanied by a fall in complication rates and blood loss between 1st versus 2nd/3rd generations. 33 Strikingly, the operating times were lower from the first case for each successive generation: the operating time for 2nd generation was 251.8 min faster than the first generation. Viewed as case load, the starting point for the second generation was greater than 90 cases compared with first-generation surgeons. The starting point for third-generation surgeons was greater than 90 cases compared with second-generation surgeons and greater than 150 cases from first-generation surgeons. 33

| NON-ROBOTIC FUNCTIONAL UROLOGY TRAINING
Case selection requires advanced non-technical skills: history and physical examination relevant to functional urology problems, interpretation of appropriate urodynamic testing, medical imaging and knowledge of the surgical procedures that have been performed previously. For each procedure, adequate knowledge of the other options and their relative risks and benefits are required.

| Robotic training requirements: General
Specific knowledge is required of the risks of a RAS in any given individual due to the need for the extended head down position and the risks of operating in the abdominal cavity. The position of steep Trendelenburg at 25-45 degrees can lead to brain and upper airway oedema in addition to increase in intracranial pressure and cerebral blood flow. 34 Common risks with positioning have been reported at 0.13% to 3% for Corneal Abrasion, aspiration 6.3%-15%, pressure lesions 35% and airway oedema 0.7%-26%. 35 The training requirements are tailored to the surgeon's existing skill level, which can be categorised as follows: • Trainee (novice) surgeon (resident/registrar/fellow) This is in conjunction with observation of RAS procedures being performed by experts. Emphasis should be on the steps in conducting a safe and effective robotic procedure. This is the approach generally adopted for RARP training elsewhere.

| Simulation training
There are several simulation services available: the Robotic Surgical Simulator (RoSS), Mimic ™ Simulation Software and the Simsurgery Educational Platform (SEP), which have proven face, content and construct validity. The simulation aspect should be considered as a step with its own curriculum with scores of >80% in three consecutive tests to be achieved, that is, a competency-based evaluation. This is juxtaposed with the number of hours on the console, which is less relevant.
To ensure simulation practice is effective it should be supervised by an expert surgeon or educationalist. Virtual reality (VR) may also become an increasingly prevalent tool to enhance simulation. A study from

| Credentialing
This requires the establishment of standardised criteria to credential or judge a surgeon's competency. The development and subsequent validation of The Global Evaluative Assessment of Robotic Skills (GEARS) tool was created by deconstructing the fundamental elements of robotic surgical procedures. The six domains evaluated include depth perception, bimanual dexterity, efficiency, autonomy, force-sensitivity and robotic control. GEARS is the first and currently only such tool, its domains being applicable to any speciality. The tool is applied to a step of the operation that reflects a pivotal point/complex task, which in turn can be applied to various operations-index or complex to assess competency of a surgeon. 40 Recently, a study was published applying GEARS to Simulation Model for Robotic Sacrocolpopexy concluding construct and face validity in addition to high interrater reliability. 41

| Modular training
The final aspect of training is applying the theoretical knowledge and skillset to procedures. Modular training breaks down a procedure into key steps/components and entails advancing through the steps of increasing difficulty. 32 This precedes advancement to operating supervised in theatre with a mentor, eventually with the capacity to independently performing a procedure to a proficient standard. 42 It is generally expected that each module should be completed prior to progressing to the next step. 3 Once all modules are completed under supervision, the surgeon or trainee is ready to perform the procedure independently under supervision. Following completion of performing independently for the entirety of the case this will need to be repeated under proctorship-the final step of the credentialing pathway. One of the most well-known forms of modular training in Urology is the EAU Robotic Urology Section (ERUS) structured curriculum that focuses on RARP. 43 Recently, there has been publications that works to address establishing performance metrics for the RARP procedure. A study by The learning curve is still substantial for an experienced laparoscopic surgeon to transition to robotic surgery. A CUSUM learning curve study of sacrocolpopexy found a learning curve of 78 cases for two surgeons who had each already completed 300 laparoscopic sacrocolpopexies, based on intraoperative bladder/bowel complications. 46 They had an experienced robotic surgeon supervising for their first two robotic cases. Accordingly, BAUS guidelines 2014 suggest that 'Mentoring and training must occur in high volume centres by surgeons who have surpassed their "learning curves"'. 37  Generally, the numbers required to be deemed competent are arbitrary. If a given trainee has competency in RARP with >20 supervised cases, it would follow that specific RFU credentialling may require fewer cases than where such prior credentialing exists.

| ONGOING CREDENTIALLING REQUIREMENTS
Once credentialling is achieved, the next issues faced by institutions is how are they maintained. Generally, it is expected that a surgeon will continue to operate using RAS.
Though training has shifted towards competency-based models, one of the shortcomings remains low case numbers/low volume centres. It is generally accepted that high case numbers do not necessarily equate to competency. A high volume of cases does however provide the opportunity to develop and apply skills and in turn demonstrate competency. Most credentialling requires completing a certain number of procedures, making it difficult to undertake at low volume centres. Given the uncommon nature of some operations, considerations need to be given to grouping procedures with similar skillsets to allow this to be achievable without significant impact on patient outcome. Proposed methods to maintain standards are annual simulation and regular undertaking of operative assistance to remain familiar with the docking and undocking of the robots and its equipment. Another means to assess surgical skill is surgeons should include audit of operative times, blood loss and complications based on Clavien-Dindo Classification. 49 Lastly patient-reported outcome measures could be considered to fill a vital gap in our perceived knowledge and expertise on outcomes that matter most to patients.

| ADDRESSING CHALLENGES IN ROBOTIC FUNCTIONAL AND RECONSTRUCTIVE UROLOGY
The variety and heterogeneity of cases, combined with low numbers of some types of cases presents a challenge that differs significantly from existing robotic uro-oncologic and other specialty procedures. Our group would like to identify a RFU curriculum with: • modular curriculum with 'components', which could encompass skills across several different procedures to address volume constraints • emphasis on index procedures to establish transferable skills, where possible, such as Burch Colposuspension or sacrocolpopexy or Burch takedown?
• develop the use of mixed reality/simulation models to support the opportunities to perform component skills in a safe environment, especially for complex procedures.
F I G U R E 1 Proposed framework for robotic assisted surgery (RAS) training in functional and reconstructive urology.

| CONCLUSION
Our group would like to use the above experiences to develop a guideline on training and credentialing of robotics in reconstruction and functional urology through the use of a modified Delphi study engaging specific societies and international expertise. The challenge in developing such a framework (Figure 1) in RFU is the relatively low case numbers and the low numbers of subspecialised surgeons. The use of reporting critical operative variables and complications provides a more standardised approach and would be more reflective on a surgeon's capabilities. Specifically, for RFU we would propose the addition of specific variables such as bowel, bladder, ureteric, vascular injury, fistula rates and chronic pain to the commonly reported operative time, blood loss and Clavien-Dindo complications. To overcome the low numbers, it will likely require a coordinated effort between sub-speciality experts to develop the simulation and technology to ensure the appropriate proctorship is undertaken before credentialling.

AUTHOR CONTRIBUTIONS
EF, HY, HH and HO created the initial concept of the work. FH wrote the initial manuscript. All authors refined the final manuscript, and agree to be accountable for all aspects of the work.