Abstract
Objective:
This study examines perioperative outcomes of resident involvement during various steps of robot-assisted radical prostatectomy (RARP).
Methods:
The RARP procedure was divided into seven steps: bladder takedown (BTD), endopelvic fascia, bladder neck (BN), seminal vesicle/vas deferens, pedicle/nerve sparing, apex, and anastomosis. Three hundred seventy-two RARPs performed by a single surgeon were analyzed. Resident console time during each of the seven steps was recorded. Perioperative variables were compared to surgeon-only cases.
Results:
Residents performed on the console for 232 of 372 cases (62.4%). Estimated blood loss (p = 0.09), transfusion (p = 0.11), and complications (p = 0.33) were no different between surgeon-only and resident-involved cases. Mean operating room time (ORT) was less for the surgeon-only cases (190.4 vs 206.4 minutes, p = 0.003). There was no difference in positive margins (p = 0.79), length of stay (LOS) (p = 0.30), catheter days (p = 0.17), readmission (p = 0.33), or reoperation (p = 0.73) when comparing surgeon-only to resident-involved cases. Residents performing the BN step had no effect on BN margins (p = 0.73) or prolonged catheterization (p = 0.62). ORT was significantly prolonged if BTD was performed by a resident (233.0 vs 191.7 minutes, p < 0.0001). Residents performing anastomosis had no effect on prolonged catheter time (p = 0.62) or LOS (p = 0.20). Residents were more likely to be involved in at least one portion of RARP following the purchase of a Mimic simulator (Mimic Technologies, Inc., Seattle, WA) in January 2012.
Conclusions:
Supervised resident console involvement in RARP does not affect perioperative outcomes, although, it prolongs ORT, with the BTD step having the most effect on ORT.
Introduction
T
Materials and Methods
Three hundred seventy-two RARP performed by a single fellowship-trained surgeon from July 1, 2007 to February 28, 2015, at Mayo Clinic in Jacksonville, Florida, were analyzed retrospectively following Institutional Review Board approval. All RARP were performed with the da Vinci Surgical System (Intuitive Surgical Corporation, Sunnyvale, CA). RARP was divided into seven steps: bladder takedown (BTD), endopelvic fascia, bladder neck (BN), seminal vesicle/vas deferens, pedicle/nerve sparing, apex, and anastomosis. A variation of the step breakdown was previously described with additional steps, including ports and docking. 9 Resident console time during and involvement in each of the seven key steps was recorded on a per case basis with the resident needing to complete half of the step for the resident involvement in the step to be recorded. The attending surgeon determined who completed each step. Perioperative variables were compared between surgeon-only and resident-involved cases. Vital outcomes of resident performance on the key steps were compared using a two-sample t-test or Wilcoxon rank-sum test.
Training program
Mayo Clinic in Jacksonville, Florida, has an Accreditation Council for Graduate Medical Education (ACGME)-accredited 5-year urology residency program utilizing the mentorship model of training. Currently, there are two residents per postgraduate year (PGY). Resident trainees are introduced to robotic surgery in PGY-1 year, when possible, and starting in PGY-2 (Urology-1 year) to bedside assisting. The trainee is required to participate in a simulation-based bedside assistant course, as previously published. 6 The PGY-2 trainee then participates in RARP as a bedside assistant for numerous cases and is expected to utilize the Mimic Robotic Simulator (Mimic Technologies, Inc., Seattle, WA) to acquire the basic foot and hand skills necessary to operate on the console. The trainee is introduced to the console for various steps of the RARP at the supervised discretion of the operating surgeon. The order of the console steps introduced to the trainee is discussed further in the Discussion section. 7
Surgical technique
Our surgical technique was previously described in the literature. 10 RARP was performed with the da Vinci S (early in series) and Si (later in series) surgical system. Veress needle technique was used to gain abdominal access in the supraumbilical midline. Assistant ports were utilized on the patients' right-hand side and the fourth arm was utilized on the patients' left-hand side. RARP was performed transperitoneally with an anterior approach starting with the BTD. The dorsal vein was oversewn with 2-0 vicryl. An athermal nerve-sparing technique was utilized when clinically appropriate. Nonabsorbable polymer surgical clips were used on the pedicles for hemostatsis. The anastomosis was performed with two 3-0 monocryl sutures tied together running in a van Velthoven manner over an 18F 2-way catheter. 11 A 10-round Jackson-Pratt (JP) drain was left close to the anastomosis to monitor for leak. Patients were given a clear liquid diet on postoperative day (POD) 1, with discharge planned for POD 1 or 2. Early ambulation was encouraged. The JP drain output was assessed and, if appropriate, the JP drain was removed just before discharge. A Foley catheter remained in place at discharge. Foley catheter removal was planned for 7 to 14 days. Cystograms were performed more frequently in the earlier cases, but were eventually only utilized at the surgeons' discretion.
Pathologic analysis
Pathologic specimens were sent to the laboratory, weighed, and inked. The specimens were then sectioned and evaluated for margin status by the pathologist. A positive margin was defined as tumor present at the inked margin during microscopic analysis. Apex, posterior, lateral, and BN margins were examined.
Data acquisition
Demographic and preoperative data collected included age, body mass index, prostate-specific antigen (PSA), preoperative biopsy Gleason score (6–7, 8–10), American Society of Anesthesiologists (ASA) score, American Urological Association (AUA) score, and Sexual Health Inventory for Men score. Operative data included resident grade (PGY-2, -3, -4, or chief), operating room time (ORT), presence of median lobe, nerve sparing (no, bilateral, unilateral), pelvic lymph node dissection (right, left, bilateral), estimated blood loss (EBL), transfusion, and any complication. At the completion of the case, the attending surgeon recorded who had performed each step of the operation. If a resident completed more than half of a step, it was recorded that the resident was involved.
Perioperative outcomes
Outcomes in the perioperative period were obtained from final pathology as well as perioperative data. The perioperative data collected included specimen weight, total Gleason (final pathology, 6–7, 8–10), margins (positive or negative for apex, posterior, BN, and lateral), length of stay (LOS), catheter days, home with drain, 30-day readmission, 30-day reoperation, and 6-month PSA. LOS was considered prolonged if it was over 3 days. Catheter time was considered prolonged if it was over 14 days. Any primary Gleason score of ≥4 or total Gleason score of eight defined high-grade cancer. Outcomes were compared between surgeon-only and resident-involved cases.
Statistical analysis
Descriptive statistics for categorical variables are reported as frequency and percentage, while continuous variables are reported as mean (standard deviation) and median (range). Categorical variables were compared between surgeon-only cases and those with resident involvement using a chi-square test or Fisher exact test, and continuous variables were compared using a two-sample t-test or Wilcoxon rank-sum test, where appropriate. The trend of residents' involvement over the case series was tested using a Cochran–Armitage trend test. Univariate and multivariate logistic regression models were used to identify risk factors that are associated with having >14 catheter days. All statistical tests were two-sided with the alpha level set at 0.05 for statistical significance. SAS version 9.3 (SAS Institute, Inc., Cary, NC) was used for data analysis.
Results
A total of 372 RARP were performed by a single fellowship-trained surgeon from July 2007 to February 2015. Residents participated on the console in 62.4% of the cases (232/372). Table 1 lists the demographic and preoperative characteristics of the cohort. Variables, including body mass index (mean, 28.4), PSA (mean, 6.7), Gleason score (6–7 88.7%; 8–10 11.3%), ASA score (mean, 2.4), AUA score (mean, 8.9). and Sexual Health Inventory for Men score (mean, 15.9), were similar between surgeon-only and resident-involved cases. Patients were older in resident-involved cases (mean, 62.4 vs 60.7; p = 0.03).
ASA = American Society of Anesthesiologists; AUA = American Urological Association; BMI = body mass index; PSA = prostate-specific antigen; SHIM = sexual health in men.
Table 2 describes the perioperative and outcome data of the cohort. EBL (p = 0.09), transfusion (p = 0.11), and complications (p = 0.33) were no different between surgeon-only and resident-involved cases. Residents were more likely to operate the console with increased resident PGY status. The only perioperative variable that was associated with resident involvement was operative time (190.4 vs 206.4 minutes, p = 0.003).
OR = operating room; PLND = pelvic lymph node dissection.
Table 3 represents factors associated with no resident console time during a case. Second-year resident status (p < 0.0001) and preoperative PSA > 10 (p = 0.0093) were associated with no resident console time during a case.
When analyzing the BN portion of the surgery, there were 154 cases where the resident performed the BN resection compared to 217 surgeon-only cases. Residents performing BN resection were not associated with BN-positive margins (3.9% resident vs 3.2% surgeon, p = 0.73), prolonged catheter time (13.9% vs 15.8%, p = 0.62), or home with drain (2.6% vs 4.6%, p = 0.4121).
When the resident performed the anastomosis (154 resident involved compared to 217 surgeon only), they were more likely to be upper-level residents (fourth year and chief, p = 0.001). There were only two instances where a second-year resident performed the anastomosis and 22 instances where a third-year resident performed the anastomosis. Regardless, resident performance of the anastomosis was not associated with prolonged catheter time (13.9% vs 15.8%, p = 0.62), home with drain (2.6% vs 5.5%, p = 0.41), hospital LOS over 3 days (2.6% vs 5.5%, p = 0.20), or ORT (mean, 193.2 resident minutes vs 205.7 surgeon minutes; p = 0.08).
The residents were more likely to perform the anastomosis and BN steps following the installation of an Intuitive Si robotic system (residents performed the BN step in 30.6% of cases before Si system compared to 63.4% of cases following the instillation). Residents performed the anastomosis step in 15.3% of cases before the Si instillation compared to 48% of cases after the instillation. Whether or not a resident performed, the BN dissection step of the anastomosis did not appear to have a relationship to prostate size.
When the resident performed the BTD step (79 residents vs 292 surgeon), the ORT was significantly longer for the surgery (233 resident minutes vs 191.7 surgeon minutes; p = < 0.001). Residents performing the dissection of the vas deferens and seminal vesicles did not affect the overall operative time of the procedure (194.9 resident minutes vs 202.7 surgeon minutes; p = 0.8170).
Discussion
Sixty-two percent of urologists believe that residents should be proficient in robotic surgery when they enter the practice following residency training. 4 One dilemma is determining how to best train residents proficiently in robotic surgery, while providing safe oncologically appropriate patient care. Recently, simulation has been highly utilized as a training tool and shown to have good face, content, and construct validity in multiple studies. 12 –15 Over the last 5 years, virtual-reality robotic simulators increased in availability from 14% of American urology residency training programs to nearly 60%. 5 Educators also utilize the porcine model as a training platform. Although this provides good hands-on experience, it is expensive and may not provide realistic situational feedback.
There have been many publications describing various training protocols utilized to train residents in a safe manner without affecting patient outcomes. 1,2,16 Many of these describe protocols that utilize simulation followed by bedside assisting and then finally introduce residents to the robotic console. A true hands-on experience cannot be fully replicated with simulation, porcine models, or assisting. Early exposure to console time in residency could be a great adjunct to these other training techniques to produce safe, high-quality surgeons. Our study examines the effect of early introduction to hands-on console experience.
Schroeck et al. was the first to report that resident involvement early in the learning curve for robotic prostatectomy did not affect outcomes, including operative time, blood loss, or positive surgical margin rate. 3 This is in contrast to our study, which showed a significant difference in operative times if the resident did the BTD step of the case when compared to the surgeon performing that step. No other step was associated with a significant difference in operative time. Residents at our institution are now introduced to RARP in a step-wise manner during their PGY-2 (urology year 1). The first step a resident performs on the console is the endopelvic fascia/dorsal venous complex stitch. This is then followed by the anterior BN. Once residents have become more proficient in these steps, they then start performing the BTD. This is in contrast to most other institutions in which the very first step a resident performs on the console is typically the BTD. We made this change secondary to the initial difficulty of the BTD step as well as the time involved, while learning it as the first task on the robot.
Each step of RARP has its own inherent challenges that must be overcome for mastery of that step. It has been previously stated that the hardest step to master in RARP may be the anastomosis. 9 A previous simulation study also showed that residents felt the tubes 2 drill (made to replicate the vesicourethral anastomosis) was the most difficult to complete on the Mimic Robotic Simulator (Mimic Technologies, Inc.). 5 Because of this paradigm, mainly PGY-4 and PGY-5 residents performed the anastomosis majority of the time in our series. When the anastomosis was performed by the resident, there was no difference found in operative time, urethral catheter time, or whether or not the patient went home with a drain. Similarly, when the residents performed the BN dissection, there was no difference in the rate of positive BN margin, prolonged catheter time, or going home with a drain.
Early on in this series, the da Vinci S Surgical System (Intuitive Surgical Corporation) was utilized before the purchase of a da Vinci Si System. Also, a Mimic Robotic Simulator was purchased by our institution in January 2012. This simulator was available at all times in the simulation center on the hospital campus. Resident involvement on the console increased after the purchase of the Mimic Robotic Simulator as well as the da Vinci Si Surgical System. One can speculate that better visualization, wider working space, and easier instrument manipulation lead to more comfort in-resident utilization of the console. The simulator provided residents the means to become familiar with the robotic console, controls, foot pedals, and manipulation of the camera and cautery well before operating the console on a live patient. Now many institutions have installed dual consoles for the da Vinci Si and Xi Surgical Systems to allow better mentor-resident interaction, while performing robotic surgery. Morgan et al. recently reported that dual-console technology with RARP significantly reduced operative time, intraoperative complications, and postoperative complications. 8 With two-console availability, the resident can be operating, while the mentor is watching each move in three dimensions as opposed to the two-dimensional laparoscopic display that was utilized before dual-console technology.
Our study reassures resident educators that early supervised resident console involvement can be safely undertaken without affecting patient care outcomes. We believe the results are valid as it was a single surgeon series, so there was minimal bias as to which steps were performed by the resident as well as how the steps were recorded for this analysis. Our study is also not without limitations. Perioperative variables were analyzed, but functional outcomes pertinent to RARP (continence, erectile function) were not. In future studies, it would be beneficial to evaluate whether resident involvement has an effect on these functional outcomes. Also, it should be noted that the residents did not perform the apical dissection and rarely performed the nerve-sparing portion in potent males. This study did not monitor timing of each individual step performed, so it was not possible to note improved timing of the trainees as they progressed in their training.
Conclusions
Supervised resident console involvement in RARP does not negatively affect perioperative patient outcomes, although, it prolongs ORT compared to surgeon-only cases, with the BTD step adding the most operative time. The acquisition of a robotic simulator appears to have increased resident exposure to the robotic console in various steps of RARP.
Footnotes
Author Disclosure Statement
No competing financial interest exists.