Transition from Resident Robotic Training Program to Clinical Practice: Robotic-Assisted Radical Prostatectomy Benchmark for Perioperative Safety

Introduction

As robotic utilization in urology increases, the robotic training background of urologists is becoming more varied and diverse. There remains no standardized credentialing system to evaluate surgeon competency and safety with regard to robotic surgery. ¹ Many of the early robotic-assisted radical prostatectomy (RARP) series describing the learning curve of RARP are from experienced open surgeons who incorporated RARP into practice. ² Literature from fellowship-trained laparoscopic surgeons incorporating RARP into practice also outlined the learning curve and benchmarked the expected perioperative morbidity of early RARP series. ³ Fellowship-trained robotic surgeons have also provided benchmark data pertaining to initial experience with RARP. ⁴ The urologic literature is lacking benchmark outcomes data for new surgeons entering practice directly out of residency. Herein we evaluate the perioperative safety and outcomes of a newly trained resident entering practice performing RARP. This may serve as benchmark data for new surgeons entering practice directly out of residency with regard to RARP safety.

Subjects and Methods

Following Institutional Review Board approval, retrospective examination was performed of the first 100 RARPs performed by one surgeon (T.E.B.) immediately following residency training. The resident in training was involved in a total of 80 robotic cases during the residency and sat at the console for 51 cases during resident years 3–5 for various portions of the RARP. Seventeen cases were performed as a chief resident as console surgeon under the guidance of a fellowship-trained robotic surgeon. Of those 17 cases, all seven steps (bladder takedown, endopelvic fascia dissection, bladder neck dissection, seminal vesicle/vas deferens mobilization, nerve sparing, apex dissection, and anastomosis) of the operations were recorded with regard to resident participation. The resident performed bladder take-down and incised the endopelvic fascia 17 times. He dissected the bladder neck 11 times and elevated the seminal vesicles/vas deferens in 6 cases. The resident completed five anastomoses and never performed the apical dissection or nerve sparing under the guidance of the fellowship-trained surgeon.

Results

The first 100 RARPs were completed over a 17-month period. Preoperative data, intraoperative data, pathologic findings, and 30-day postoperative course were available for all 100 patients. No patients required re-operation within the 30-day postoperative period. Eighty-eight patients (88%) had a hospital stay of 2 days or less without postoperative complications and had urethral catheter removal within 14 days of surgery and without re-admission.

Table 1 depicts patient characteristics and operative data. Pelvic lymph node dissection was completed on 60 patients without complication. No patient had positive pelvic lymph nodes. Fourteen patients were noted to be high-risk preoperatively as defined by prostate-specific antigen (PSA) >10 ng/mL, primary Gleason score of 4 or higher, or palpable disease. All 14 of these patients had pelvic lymph node dissections completed. There was one intraoperative rectal injury that was noted intraoperatively and repaired. This is also the only patient who was converted to an open operation.

Table 2 notes surgical pathology results and postoperative data. The overall positive margin rate was 21%. The margin rate of pathologic T2 patients was 19.4% (18 of 93). For patients with pathologic T2 disease and primary Gleason score of 2 or 3, the positive margin rate was 17.2% (16 of 93). The positive margin rate was 10.5% (4 of 38) for patients with Gleason score 4–6, T2 disease. Two of the clinical T2c Gleason score 4–6 patients developed PSA recurrence over a 6-month follow-up period. Two patients with positive bladder neck margins had seminal vesicle invasion on final pathologic examination. Ten patients had positive apical margins on final pathologic examination. Eight of these patients had an undetectable PSA at 1-year follow-up. Two of the 8 patients with a positive posterior margin had PSA recurrences within 1 year.

Table 3 lists the surgical complications using the Clavien classification system of surgical complications. Five patients (5%) required blood transfusions in the perioperative period. Nine patients (9%) stayed in the hospital longer than 2 days, and 4 patients remained in the hospital longer than 3 days. Three patients stayed over 3 days for ileus/delayed bowel function. One patient stayed over 3 days to receive a blood transfusion for delayed pelvic bleeding. One patient had a postoperative myocardial infarction that was managed medically. One patient had an intraoperative rectal injury that was repaired and was kept nothing peroral until bowel function.

Six-month PSA follow-up was available on 91 patients. Seven of the 91 patients had a detectable PSA within the 6-month period. Four of these patients were Gleason score 7 or higher on pathologic examination, and 4 had positive surgical margins.

Discussion

Current RARP outcomes literature originates from high-volume surgeons from various training backgrounds.^2,3,7,8 The “learning curve” for RARP has been proposed to be anywhere from 20 to 250 cases. ⁹ There is a paucity of benchmark data in the literature pertaining to newly trained surgeons performing RARP. Hu ¹⁰ proposed that this lack of benchmark data leads to inexperienced surgeons claiming outcomes data of the experienced surgeons in the literature as their own. This can lead to an unrealistic portrayal of perioperative safety and outcomes for the patient. This article addresses a couple of issues pertaining to RARP. The first issue is to provide benchmark data to the literature for new surgeons out of residency entering practice performing RARP; the second issue is the current RARP training of graduating residents.

Currently there is no standardized robotic training required for the modern-day urologic resident. The early dilemma for residents in training was to teach RARP when the surgeons themselves had not perfected the technology. Approximately 29% of graduating residents believe they are adequately trained in robotic surgery. ¹¹ Multiple avenues exist to attain postgraduate robotic training, including urologic oncology fellowships, laparoscopic/endourology fellowships, and robotic fellowships. ¹² These fellowships vary in their length (1–3 years of duration) and accreditation. It is not known if one form of training is superior to another. What appears to be consistently successful in robotic training is a three-phase approach to learning surgical robotics and the guidance of an experienced mentor. The first phase involves learning the robotic technology including port placement, arm clutching, and overall familiarization with the robot. The second phase is assisting laparoscopically at the bedside. The third phase is completing individual steps of the robotic-assisted laparoscopic prostatectomy while sitting at the console.^13,14

Most medical malpractice claims surrounding robotic surgeries are secondary to systems malfunctions, and 75% of those malfunctions arise intraoperatively. ¹⁵ The most common causes of these intraoperative systems malfunctions are due to inexperience or lack of technical competence with the instrumentation/surgical device.^1,15 This fact points to the need for robotic simulation. Current robotic simulation is limited to simulating basic robotic tasks. Actual steps of robotic surgery (i.e., dissection) cannot currently be simulated or adequately replicated. However, having trainees learn basic robotic skills in a simulated environment (camera movement, clutching, hand movements, foot pedal coordination, and suturing) can decrease surgical risk and prepare the learner for more advanced tasks (prostate apical dissection, nerve sparing, etc.). What is still unknown is how much simulated learning is needed to improve safety and outcome and which simulated program provides the best pathway to the desired results. ¹⁶ Perhaps the addition of simulation to robotic training programs will increase the number of trainees who feel adequately trained in robotics once training is complete.

The safety profile of this series is comparable to those in the literature of experience robotic surgeons. The overall complication rate was 16%. The most common reported complication in mature RARP series is urinary leakage.^7,8 Five patients in this series required a catheter longer than 14 days secondary to urinary leakage. Urinary leakage is usually secondary to pelvic hematoma or a non-watertight anastomosis. Surgeons learning RARP note that the vesicourethral anastomosis is one of the most time-consuming and difficult steps to master.^17,18

There was one intraoperative rectal injury noted in the series. This rectal injury occurred in a patient with clinical T3 disease that was adherent to the rectum. A review of 6650 RARPs from six institutions noted the rectal injury rate to be 0.17%, with 72.7% recognized intraoperatively. ¹⁹ Most patients did well with primary closure. Analysis did not find conclusive association of rectal injury with surgeon experience. Three patients in this series had prolonged hospitalization secondary to postoperative ileus.

An interesting finding is that there were 60 pelvic lymph node dissections performed in the series. There was no patient with palpable disease or high-grade disease that failed to get a pelvic lymph node dissection. A previous analysis noted that fellowship-trained robotic surgeons were more likely to perform pelvic lymph node dissections in high-risk patients than open surgeons incorporating robotics into practice. ²⁰ It is proposed that the fellowship-trained surgeons felt more comfortable doing the lymph node dissection secondary to training. Exposure to pelvic lymph node dissection in residency may have the same benefit and allow more dissections to be completed.

A recent analysis of the first 100 RARPs performed by a fellowship-trained robotic surgeon revealed that 82% of patients had a hospital stay of 2 days or less without any postoperative complications and urethral catheter removal within 14 days of surgery and did not require re-admission to the hospital. ⁴ The positive margin rate was 16.7% (7 of 42) for patients with Gleason score 6, T2 disease. Clavien grade 1–4 complication rates were 4%, 10%, 1%, and 1%, respectively. There were no deaths, re-operations, or bladder neck contractures, and 1 patient (1%) required a blood transfusion. One mentionable difference between this series and the one represented in the current article is that a resident in training performed major portions of the operation in a majority of cases. Residents entering community practice directly from training will be able to perform all portions of the RARP compared with those entering into academic practice, where resident performance of the procedure is expected. Having trainees perform portions of RARP may affect outcomes and safety. Another mentionable difference in the studies is the inclusion of prospective quality of life data, continence data, and potency data in the analysis. This study focuses on perioperative safety and short-term outcomes. It lacks validated long-term follow-up instruments to evaluate long-term functional outcomes such as continence and potency.

A few items deserve special mention. The first is that the resident in training participated in a mentorship model type of residency. The resident worked one-on-one with an attending surgeon throughout training. This may provide benefits not available in larger residency training programs. The resident was also permitted on the console beginning in year 3 out of 5 of his training program. Earlier console exposure to residents has been proposed to increase their active involvement in learning robotic surgery. ²¹ In fact, many programs have increased their use of various bedside assistants such as surgical technicians and physician assistants to free up residents for more robotic console learning. ²¹ One limitation of the study is that the outcomes of only one surgeon are reported. Therefore this evaluation cannot be expected to speak for all surgeons out of residency training.

Conclusions

This study provides benchmark RARP data for residents entering practice directly from training. A mentorship residency model and early resident console exposure may aid in achieving acceptable perioperative safety and outcomes for residents entering directly into practice.