Real-time assessment of learning curve for robot-assisted laparoscopic prostatectomy

Hospital learning curves for robot-assisted surgeries: a population-based analysis

BACKGROUND

Robot-assisted surgery has been rapidly adopted. It is important to define the learning curve to inform credentialling requirements, training programs, identify fast and slow learners, and protect patients. This study aimed to characterize the hospital learning curve for common robot-assisted procedures.

STUDY DESIGN

This cohort study, using administrative health data for Ontario, Canada, included adult patients who underwent a robot-assisted radical prostatectomy (RARP), total robotic hysterectomy (TRH), robot-assisted partial nephrectomy (RAPN), or robotic portal lobectomy using four arms (RPL-4) between 2010 and 2021. The association between cumulative hospital volume of a robot-assisted procedure and major complications was evaluated using multivariable logistic models adjusted for patient characteristics and clustering at the hospital level.

RESULTS

A total of 6814 patients were included, with 5230, 543, 465, and 576 patients in the RARP, TRH, RAPN, and RPL-4 cohorts, respectively. There was no association between cumulative hospital volume and major complications. Visual inspection of learning curves demonstrated a transient worsening of outcomes followed by subsequent improvements with experience. Operative time decreased for all procedures with increasing volume and reached plateaus after approximately 300 RARPs, 75 TRHs, and 150 RPL-4s. The odds of a prolonged length of stay decreased with increasing volume for patients undergoing a RARP (OR 0.87; 95% CI 0.82-0.92) or RPL-4 (OR 0.77; 95% CI 0.68-0.87).

CONCLUSION

Hospitals may adopt robot-assisted surgery without significantly increasing the risk of major complications for patients early in the learning curve and with an expectation of increasing efficiency.

"Single-Surgeon" versus "Dual-Surgeon" Robot-Assisted Radical Prostatectomy and Pelvic Lymph-nodes Dissection: Comparative Analysis of Perioperative Outcomes

PURPOSE

To compare the perioperative outcomes of robot-assisted radical prostatectomy (RARP) with pelvic lymph-nodes dissection (PLND) when the same surgeon performs RARP and PLND versus one surgeon performs RARP and another surgeon performs PLND.

MATERIALS AND METHODS

From January 2022 to March 2023, data of consecutive patients who underwent RARP with PLND were prospectively collected. The surgeries were performed by two "young" surgeons with detailed profile. Specifically for the study purpose, one surgeon performed RARP, and the other surgeon performed PLND. A set of surgeries performed according to the standard setup (i.e., the same surgeon performing both RARP and PLND) was retrieved from the institutional database and used as comparator arm. To test the study hypothesis, patients were divided into two groups: "dual-surgeon" versus "single-surgeon".

RESULTS

Fifty patients underwent RARP and PLND performed according to dual-surgeon setup and were compared to the last 50 procedures performed according to the standard single-surgeon setup. Patients in the groups had comparable baseline characteristics. Dual-surgeon interventions had significantly shorter median total operative (194 [IQR 178-215] versus 174 [IQR 146-195] minutes, p<0.001) and console time (173 [IQR 158-194] versus 154 [IQR 129-170] minutes, p<0.001). No significant differences were found in terms of blood loss, intraoperative complications, postoperative outcomes, and final pathology results.

CONCLUSIONS

The present analysis found that when RARP and PLND are split onto two surgeons, the operative time is shorter by 20 minutes compared to when a single surgeon performs RARP and PLND. This is an interesting finding that could sponsor further studies.

CUSUM learning curves: what they can and can't tell us

INTRODUCTION

There has been increased interest in assessing the surgeon learning curve for new skill acquisition. While there is no consensus around the best methodology, one of the most frequently used learning curve assessments in the surgical literature is the cumulative sum curve (CUSUM) of operative time. To demonstrate the limitations of this methodology, we assessed the CUSUM of console time across cohorts of surgeons with differing case acquisition rates while varying the total number of cases used to calculate the CUSUM.

METHODS

We compared the CUSUM curves of the average console times of surgeons who completed their first 20 robotic-assisted (RAS) cases in 13, 26, 39, and 52 weeks, respectively, for their first 50 and 100 cases, respectively. This analysis was performed for prostatectomy (1094 surgeons), malignant hysterectomy (737 surgeons), and inguinal hernia (1486 surgeons).

RESULTS

In all procedures, the CUSUM curve of the cohort of surgeons who completed their first 20 procedures in 13 weeks demonstrated a lower slope than cohorts of surgeons with slower case acquisition rates. The case number at which the peak of the CUSUM curve occurs uniformly increases when the total number of cases used in generation of the CUSUM chart changes from 50 to 100 cases.

CONCLUSION

The CUSUM analyses of these three procedures suggests that surgeons with fast initial case acquisition rates have less variability in their operative times over the course of their learning curve. The peak of the CUSUM curve, which is often used in surgical learning curve literature to denote "proficiency" is predictably influenced by the total number of procedures evaluated, suggesting that defining the peak as the point at which a surgeon has overcome the learning curve is subject to routine bias. The CUSUM peak, by itself, is an insufficient measure of "conquering the learning curve."

Robot-Assisted Laparoscopic Prostatectomy Experience and Pathological Quality: Are They Always Linked?

Objective: We investigated whether pathological outcomes improved with experience and surgeon generation after robot-assisted laparoscopic prostatectomy (RALP). Materials and Methods: The study included 1338 patients who underwent RALP between February 2010 and April 2020. We created learning curves for pelvic lymph node dissection (PLND), number of lymph nodes (LNs) removed, and positive surgical margin (PSM) after adjustment for confounders. We compared the outcomes between the first and second generation of surgeons in regression models. Results: The learning curve regarding PLND indications showed a significant increase with experience for the first generation, whereas the second generation had a learning curve that remained flat at a higher level (92.3%) and significantly better than the first generation (p < 0.001). Similarly, the number of LN removed showed a significant increase with experience in both generations, but the overall median number of LN removed was significantly higher in the second generation compared with the first generation (12 vs 10, p < 0.001). However, the learning curve for PSM remained flat at ∼20% after adjustment and did not show improvement with experience in both generations of surgeons (p = 0.794). Conclusions: Surgeons showed improvement with experience and education with RALP with respect to the indications for PLND and number of LNs removed. However, there was no improvement over time and generations for PSM. Experience based solely on the number of patients operated on is not an intrinsic factor in the pathological quality of RALP. Factors other than experience may also play a role in oncologic improvement.

Cumulative sum analysis of the learning curve of laparoendoscopic single-site robot-assisted radical prostatectomy

BACKGROUND

Radical prostatectomy has become the gold standard for treating localized prostate cancer. Improvement in the single-site technique and surgeon's skill reduces not only the hospital duration but also the number of wounds. Realizing the learning curve for a new procedure can prevent unnecessary mistakes.

OBJECTIVE

To analyze the learning curve of extraperitoneal laparoendoscopic single-site robot-assisted radical prostatectomy (LESS-RaRP).

METHODS

We retrospectively evaluated 160 patients diagnosed with prostate cancer during June 2016 to December 2020 who underwent extraperitoneal LESS-RaRP. Calculated cumulative sum analysis (CUSUM) was used to evaluate the learning curves for the extraperitoneal setting time, robotic console time, total operation time, and blood loss. The operative and functional outcomes were also assessed.

RESULTS

The learning curve of the total operation time was observed in 79 cases. For the extraperitoneal setting and robotic console times, the learning curve was observed in 87 and 76 cases, respectively. The learning curve for blood loss was observed in 36 cases. No in-hospital mortality or respiratory failure was observed.

CONCLUSION

Extraperitoneal LESS-RaRP using the da Vinci Si system is safe and feasible. Approximately 80 patients are required to achieve a stable and consistent operative time. A learning curve for blood loss was observed after 36 cases.

Robotic-assisted radical prostatectomy: learning curves and outcomes from an Australian perspective

Background

Robot-assisted radical prostatectomy (RARP) has been a treatment for men who suffer from intermediated to high-risk prostate cancer in Australia since 2003. The primary outcomes in relation to learning curves in robotic surgery have been extensively researched in overseas populations, but there is no study from a cohort of Australian surgeons performing RARP. This study aims to highlight the effect of RARP learning curves on primary surgical outcomes in a high-volume Australian centre.

Methods

A retrospective audit of all RARP performed at Epworth Healthcare from 2016 to 2021 was performed. The primary outcome data collected included operating time (OT), estimated blood loss (EBL), and positive surgical margins (PSM). Exclusion criteria were applied. Positive outcomes were set at OT 240 min, blood loss 310 mL, and negative surgical margins.

Results

A total of 3969 cases were analysed for a cohort of 53 surgeons. Of these surgeons, 24 surgeons have performed >50 operations to be able to undergo learning curve analysis. The median OT was 229 min, the median blood loss was 353 mL, and most cases had negative surgical margins (>1 mm, n = 3681, 92.7%). The mean learning curve transition point was 65 cases. There was a significant difference in the EBL and rate of PSM for the higher volume cohort (p = 0.002 and <0.0001, respectively).

Conclusion

We perform a retrospective study of all RARP performed at a high-volume Australian centre. Higher volume surgeons demonstrate that primary outcomes improve with a higher caseload (EBL, PSM). Learning curve transition points for RARP are comparable to international high-volume surgeons. Learning curve data could form the benchmark for RARP training and skills development.

Learning curves in laparoscopic and robot-assisted prostate surgery: a systematic search and review

PURPOSE

To perform a systematic search and review of the available literature on the learning curves (LCs) in laparoscopic and robot-assisted prostate surgery.

METHODS

Medline was systematically searched from 1946 to January 2021 to detect all studies in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement, reporting on the LC in laparoscopic radical prostatectomy (LRP), laparoscopic simple prostatectomy (LSP), robot-assisted radical prostatectomy (RARP) and robot-assisted simple prostatectomy (RSP).

RESULTS

In total, 47 studies were included for qualitative synthesis evaluating a single technique (LRP, RARP, LSP, RSP; 45 studies) or two techniques (LRP and RARP; 2 studies). All studies evaluated outcomes on real patients. RARP was the most widely investigated technique (30 studies), followed by LRP (17 studies), LSP (1 study), and RSP (1 study). In LRP, the reported LC based on operative time; estimated blood loss; length of hospital stay; positive surgical margin; biochemical recurrence; overall complication rate; and urinary continence rate ranged 40-250, 80-250, 58-200, 50-350, 110-350, 55-250, 70-350 cases, respectively. In RARP, the corresponding ranges were 16-300, 20-300, 25-200, 50-400, 40-100, 20-250, 30-200, while LC for potency rates was 80-90 cases.

CONCLUSIONS

The definition of LC for laparoscopic and robot-assisted prostate surgery is not well defined with various metrics used among studies. Nevertheless, LCs appear to be steep and continuous. Implementation of training programs/standardization of the techniques is necessary to improve outcomes.

A review of simulation training and new 3D computer-generated synthetic organs for robotic surgery education

We conducted a comprehensive review of surgical simulation models used in robotic surgery education. We present an assessment of the validity and cost-effectiveness of virtual and augmented reality simulation, animal, cadaver and synthetic organ models. Face, content, construct, concurrent and predictive validity criteria were applied to each simulation model. There are six major commercial simulation machines available for robot-assisted surgery. The validity of virtual reality (VR) simulation curricula for psychomotor assessment and skill acquisition for the early phase of robotic surgery training has been demonstrated. The widespread adoption of VR simulation has been limited by the high cost of these machines. Live animal and cadavers have been the accepted standard for robotic surgical simulation since it began in the early 2000s. Our review found that there is a lack of evidence in the literature to support the use of animal and cadaver for robotic surgery training. The effectiveness of these models as a training tool is limited by logistical, ethical, financial and infection control issues. The latest evolution in synthetic organ model training for robotic surgery has been driven by new 3D-printing technology. Validated and cost-effective high-fidelity procedural models exist for robotic surgery training in urology. The development of synthetic models for the other specialties is not as mature. Expansion into multiple surgical disciplines and the widespread adoption of synthetic organ models for robotic simulation training will require the ability to engineer scalability for mass production. This would enable a transition in robotic surgical education where digital and synthetic organ models could be used in place of live animals and cadaver training to achieve robotic surgery competency.

Prediction of a positive surgical margin and biochemical recurrence after robot-assisted radical prostatectomy

The positive surgical margin (PSM) and biochemical recurrence (BCR) are two main factors associated with poor oncotherapeutic outcomes after prostatectomy. This is an Asian population study based on a single-surgeon experience to deeply investigate the predictors for PSM and BCR. We retrospectively included 419 robot-assisted radical prostatectomy cases. The number of PSM cases was 126 (30.1%), stratified as 22 (12.2%) in stage T2 and 103 (43.6%) in stage T3. Preoperative prostate-specific antigen (PSA) > 10 ng/mL (p = 0.047; odds ratio [OR] 1.712), intraoperative blood loss > 200 mL (p = 0.006; OR 4.01), and postoperative pT3 stage (p < 0.001; OR 6.901) were three independent predictors for PSM while PSA > 10 ng/mL (p < 0.015; hazard ratio [HR] 1.8), pT3 stage (p = 0.012; HR 2.264), International Society of Urological Pathology (ISUP) grade > 3 (p = 0.02; HR 1.964), and PSM (p = 0.027; HR 1.725) were four significant predictors for BCR in multivariable analysis. PSMs occurred mostly in the posterolateral regions (73.8%) which were associated with nerve-sparing procedures (p = 0.012) while apical PSMs were correlated intraoperative bleeding (p < 0.001). A high ratio of pT3 stage after RARP in our Asian population-based might surpass the influence of PSM on BCR. PSM was less significant than PSA and ISUP grade for predicting PSA recurrence in pT3 disease. Among PSM cases, unifocal and multifocal positive margins had a similar ratio of the BCR rate (p = 0.172) but ISUP grade > 3 (p = 0.002; HR 2.689) was a significant BCR predictor. These results indicate that PSA and pathological status are key factors influencing PSM and BCR.