ORIGINAL ARTICLE ANZJSurg.com

Supervisor volume affects oncological outcomes of trainees performing open radical prostatectomy Dermot O’Kane,*† Nathan Papa,*‡§ Nathan Lawrentschuk,*¶**†† Rodney Syme,†† Graham Giles‡§ and Damien Bolton*† *Department of Surgery, The University of Melbourne, Melbourne, Victoria, Australia †Urology Department, Austin Hospital, Melbourne, Victoria, Australia ‡Cancer Council Victoria, Cancer Epidemiology Centre, Melbourne, Victoria, Australia §Centre for Epidemiology and Biostatistics, Melbourne School of Population and Global Health, The University of Melbourne, Melbourne, Victoria, Australia ¶Urology Department, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia **Olivia Newton-John Cancer Research Institute, Melbourne, Victoria, Australia and ††Urology Unit, Epworth Freemasons Hospital, Melbourne, Victoria, Australia

Key words biochemical failure, learning curve, radical prostatectomy, surgical training. Correspondence Dr Dermot O’Kane, Urology Department, Austin Hospital, 145 Studley Road, Heidelberg, Melbourne, Vic. 3084, Australia. Email: [email protected] D. O’Kane MB, BCh; N. Papa MBBS, MEpi; N. Lawrentschuk PhD, FRACS; R. Syme MBBS, FRACS; G. Giles MSc, PhD; D. Bolton MD, FRACS. Accepted for publication 8 March 2015. doi: 10.1111/ans.13112

Abstract Background: The relationship between biochemical failure (BF) rate and surgeon experience following open radical prostatectomy (ORP) has been well established, but BF when ORP is performed by urology trainees who are supervised by urologists of differing volume has not. We aimed to compare the oncological outcomes from ORP when a urology trainee as primary operator and is supervised by a high- or low-volume consultant urologist. Methods: Using a centralized whole of population dataset, created through the Victorian Radical Prostatectomy Registry, patients were classified as either those where a consultant was the primary operator, a urology trainee was the primary operator and supervised by a high-volume consultant or those where a urology trainee was supervised by a low-volume consultant. BF- and prostate cancer (PCa)-specific mortality was compared between these latter two groups and the consultant-only group. Results: We found BF- and PCa-specific mortality rate to be poorer when ORP was performed by a urology trainee supervised by a low-volume consultant compared with consultant-led surgery (hazard ratio (HR) = 1.33, P = 0.022; subhazard ratio (SHR) = 2.31, P = 0.010, respectively). When a urology trainee, as primary operator, was supervised by a high-volume consultant, there was no statistical difference in BF- or PCa-specific mortality rate following ORP compared with consultant-led surgery (HR = 1.19, P = 0.234; SHR = 1.53, P = 0.346, respectively). There was a trend evident with decreasing supervisor volume leading to worse oncological and mortality outcomes for trainee-led cases. Conclusion: This study demonstrates the value of high-volume and fellowshiptrained urologists in performing and teaching ORP. As outcomes are increasingly scrutinized with audits, the best strategy for clinicians to maintain standards and optimal patient outcomes is to understand these elements and direct trainees to appropriate centres for training and fellowships.

Introduction Despite the widespread adoption of laparoscopic approaches to radical prostatectomy in recent years, the open retropubic technique remains the most common approach in many regions of the world and exposure to laparoscopic techniques remains variable for © 2015 Royal Australasian College of Surgeons

urology trainees.1 Learning the art of open radical prostatectomy (ORP) is a daunting task for the urology trainee. Studies have shown that the number of cases required to develop proficiency in this procedure is of the order of 250–1000.2 During this learning curve, positive surgical margins (PSMs) and resultant oncological outcomes have been shown to be inferior.3,4 ANZ J Surg •• (2015) ••–••

2

Following the learning curve with ORP, a well-established relationship has been shown to exist between surgeon volume and patient outcomes, including perioperative morbidity, quality of life outcomes and oncological outcomes.5 The effect of urology trainees on patient outcomes following ORP has also been shown, but to date these outcomes have only included post-operative complications and length of hospital stay.6 There has been no published literature addressing the effect of urology trainee involvement in ORP and long-term oncological outcomes when the trainee is supervised by a high-volume or low-volume surgeon.

Methods Patients Patients were identified from the Victorian Radical Prostatectomy Register (VRPR). The VRPR is a prospective whole of population series of all men who underwent ORP for the treatment of localized prostate adenocarcinoma between 1995 and 2000 in Victoria, Australia. The VRPR was established within the Victorian Cancer Registry (VCR), in which all cancer cases in the state of Victoria, other than non-melanoma skin cancer, are recorded. The VCR is managed by the Cancer Council of Victoria and the VRPR was approved by the Cancer Council of Victoria’s Human Research Ethics Committee and established within the VCR.7 Specific details of patient registration and data collection have previously been described.8

Data obtained Data pertaining to the operating surgeon for each of these cases were collated. For every case, the de-identified code for the operating surgeon was collected. It was noted whether the primary operator was a consultant urologist or a urology trainee. This information was determined by the recorded primary and secondary operators as per the operation report. The operator who performed the majority of the operation, particularly the latter steps, was deemed the primary operator. Trainees were stratified according to whether they were operating with a high- or low-volume consultant. We defined highvolume consultant as having performed >80 ORP (as primary operator or with a trainee as primary operator), and low-volume consultant as having performed 0.2 ng/mL) and prostate cancer (PCa)-specific mortality.

O’Kane et al.

ables and the non-parametric Kruskal–Wallis test for PSA level due to significant deviations from normality. BF-free survival was modelled with multivariable Cox proportional hazard regressions with all variables entered simultaneously. Time from surgery was set as the time axis. Kaplan–Meier survival curves were plotted with differences in these assessed with the log-rank test. The proportional hazard assumption was tested by analysis of Schoenfeld residuals and was found not to be violated. PCa-specific mortality was analysed by a competing hazard regression using the method of Fine and Grey. All tests were two sided and significance set at

Supervisor volume affects oncological outcomes of trainees performing open radical prostatectomy.

The relationship between biochemical failure (BF) rate and surgeon experience following open radical prostatectomy (ORP) has been well established, bu...
512KB Sizes 0 Downloads 6 Views