Incorporating Robot-Assisted Prostatectomy at a Veterans Affairs Hospital: Outcomes

Abstract

Purpose:

To evaluate a unique method of extended mentorship in robot-assisted laparoscopic prostatectomy (RALP) at VA San Diego Healthcare System (VASDHS). As novel robotic technologies diffuse into surgical practice, developing safe apprenticeships remains a challenge.

Patients and Methods:

Between October 2008 and November 2010, 90 RALPs were prospectively divided into three phases: Proctored, Independent, and Instructor. During the first 30 Proctored cases, an experienced robotic surgeon from the affiliated university-based hospital mentored a robotic novice attending surgeon with previous open retropubic and laparoscopic experience. The novice surgeon gained proficiency during the next 30 Independent cases, then introduced increasing resident participation during the last Instructor 30 cases. Patient demographics, tumor characteristics, operative measures, and length of hospital stay were compared. Functional outcomes were assessed using the Sexual Health Inventory for Men and an incontinence questionnaire. We used independent t test, analysis of variance, Mann-Whitney U test, Fisher exact test, Kruskal-Wallis, and Pearson chi-square tests for comparison in these patient populations.

Results:

All groups were similar in age, clinical T-stage, and D'Amico Risk Group. Preoperative prostate-specific antigen levels were significantly higher (P<0.001) and prostates were larger (P=0.044) in the middle Instructor Phase. The early Proctored Phase had the lowest Gleason scores and the lowest body mass indexes. Despite these differences favoring the Proctored Phase, immediate operative outcomes were similar with respect to safety, oncologic, and functional parameters.

Conclusions:

In the VASDHS cohort, RALPs were performed safely under the supervision of a newly proctored attending surgeon. Although longer follow-up could reveal subtle differences between groups, overall follow-up was similar to most existing studies. Extended mentorship by an experienced surgeon is a viable model for achieving proficiency in RALP in a setting such as a VA hospital affiliated with an academic hospital and increasing access to care for the veteran patient population.

Introduction

Robot-assisted laparoscopic prostatectomy (RALP) is an increasingly prevalent modality in the management of clinically localized prostate cancer. There is a need for reliable methods to safely introduce and incorporate RALP into the practice of surgeons unfamiliar with this platform. The introduction of RALP into practice generally follows one of two patterns: Either trainees with robotic experience in residency or fellowship launch the new program, or an existing surgeon or group of surgeons obtains postgraduate training by attending industry sponsored training and/or continuing medical education courses, often followed by a mentored clinical experience before credentialing. The length of mentoring can range from the extremes of none up to years or hundreds of cases. Numerous studies describe specific training methods with heterogeneous outcomes measures. Some studies report surgeon confidence,^1

–4 while others report operative outcomes, and yet others report operative, pathologic, and functional outcomes.^8

–11

The Veterans Affairs San Diego Healthcare System (VASDHS) is a large, tertiary care hospital providing care for more than 200,000 veterans with more than 5000 surgeries performed every year.¹² In 2008, the VASDHS was the first VA hospital in Southern California to purchase a single console da Vinci® S robot (Intuitive Surgical, Sunnyvale, CA). As in many other VA hospitals, VASDHS was staffed with a primary urologist, multiple university-based urologists, and rotating residents.

This study evaluates the safety of a unique prolonged proctorship for the primary VA attending urologist by analyzing the perioperative safety parameters, pathologic, and functional outcomes. We aim to safely train a VA attending physician in RALP so that the attending physician gains proficiency and proceeds to training residents. Many published reports fail to provide the operative, pathologic, and functional outcomes experience during the learning curve, and we thought that this was important to include. This approach is a viable model to introduce RALP in other VA hospitals with affiliated university-based urologists with expertise in RALP, or any similar clinical setting.

Patients and Methods

Between October 2008 and November 2010, all of the 90 RALPs performed at the VASDHS were identified and prospectively divided into three chronologic groups. The Proctored Phase consisted of the first 30 cases in which an experienced robotic surgeon from the academic hospital affiliated with the VASDHS provided extended mentorship for a novice attending robotic surgeon. The experienced robotic surgeon at that time had independently performed more than 1500 RALPs. The novice attending had experience in open retropubic prostatectomy, as well as laparoscopic prostatectomy and nephrectomy for a combined total of 123 laparoscopic cases before mentorship. The initial 30 Proctored cases allowed for the entire operating room team to become familiar with RALP during this time as well.

The subsequent Independent and Instructor Phases each had 30 cases, in which the mentored novice surgeon initially worked to increase proficiency and then proceeded to work with the urology trainees. During the Instructor Phase, one rotating senior resident was always present, and these senior residents had minimal but accruing experience in laparoscopy and robotic laparoscopy when they started at the affiliated academic hospital. Their level of involvement typically began with assisting the primary surgeon and incrementally increased to more participation on the console.

The following parameters were compared for all three phases: Patient demographics, tumor characteristics, operative measures, length of hospital stay, and functional outcomes. Because the length of the operation affects the safety of the surgery, “prolonged surgeries” were identified in each group and defined as lasting more than 5 hours. The Sexual Health Inventory for Men (SHIM) and a questionnaire developed by this group (Table 1) were mailed to all patients to assess functional outcome (incontinence and erectile function). Independent t test, analysis of variance, Mann-Whitney U test, Fisher exact test, Kruskal-Wallis, and Pearson chi-square tests compared the demographics and operative outcomes in these patient populations.

Table 1.

Robot-Assisted Laparoscopic Prostatectomy

Veterans Affairs Robot-Assisted Laparoscopic Prostatectomy Survey
1) How many pads a day do you wear for urinary incontinence? ____________________
2) On a scale from 1 to 5 where 1 is entirely dry, 2 are drips, 3 is moist, 4 is wet, and 5 is drenched, how would you rate your average pad when you change it? ___________________
3) If you had to live the rest of your life with your incontinence the way it is right now would you be 1) delighted 2) happy 3) satisfied 4) unhappy or 5) miserable? _________________
4) Is your incontinence continuing to improve (Y/N)?
5) Have you attempted sexual intercourse in the last month (Y/N)?
6) On a scale of 1 to 5 where 1 is always and 5 is never, how would you rate your confidence in obtaining an erection satisfactory of sexual activity? ___________________
7) On a scale of 1 to 5 where 1 is always and 5 is never, how often do you require medication to obtain an erection? ___________________
8) If you had to live the rest of your life with your erectile function the way it is right now would you be 1) delighted 2) happy 3) satisfied 4) unhappy or 5) miserable? _________________

Results

The three groups were similar in mean age, clinical T-stage, and D'Amico Risk Group. There was a statistically significant difference in the median preoperative prostate-specific antigen (PSA) levels of these populations (P<0.001), with the Instructor Phase having the highest median preoperative PSA level. These patients also had the largest prostates (P=0.044). The Proctored Phase had the lowest Gleason scores and the lowest body mass indexes (BMIs) (Table 2). Despite these preoperative differences favoring the Proctored phase, immediate operative outcomes were statistically similar between the groups (Table 3).

Table 2.

Demographic and Clinical Characteristics

	Proctored Phase (n=30)	Independent Phase (n=30)	Instructor Phase (n=30)	P value
Mean age±SD, years	61±6	62±4	61±7	0.867
Mean BMI±SD, kg/m²	26.5±3.8	30.1±3.9	29.2±5.5	0.007^*
Median preoperative PSA (IQR), ng/mL	4.6 (2.9–6.1)	6 (4–8.6)	8 (5–12.8)	< 0.001^*
Median prostate size (IQR), from US, cc	39.5 (33–50.3)	36.5 (30.8–47.3)	47.8 (35.8–62.9)	0.044^*
Prostate nodule	15 (50%)	17 (56.7%)	10 (33.3%)	0.175
Clinical T stage				0.247
T_1a-c	14 (46.7%)	15 (50%)	20 (66.7%)
T_2a-c	16 (53.3%)	15 (50%)	9 (30%)
T_3a-b	0	4 (2.9%)	1 (3.3%)
Biopsy Gleason score				0.050^*
≤6	15 (50%)	5 (16.7%)	7 (23.3%)
7	11 (36.7%)	17 (56.7%)	14 (46.7%)
≥8	4 (13.3%)	8 (26.7%)	9 (30%)
D'Amico Risk Group				0.076
Low risk	11 (36.7%)	3 (10%)	6 (20%)
Intermediate risk	15 (50%)	16 (53.3%)	14 (46.7%)
High risk	4 (13.3%)	11 (36.7%)	10 (33.3%)

Significant two-tailed.

P value by: Pearson chi-square test, analysis of variance, Kruskal-Wallis.

SD=standard deviation; BMI=body mass index; PSA=prostate-specific antigen; IQR=interquartile range.

Table 3.

Operative and Pathologic Details and Outcomes

	Proctored Phase (n=30)	Independent Phase (n=30)	Instructor Phase (n=30)	P value
Median operative time (IQR), min	263 (239–332)	281 (242–353)	273 (230–347)	0.755
Median estimated blood loss (IQR), mL	135 (50–250)	138 (98–200)	100 (69–150)	0.568
Blood transfusions	2 (6.7%)	0	0	0.129
Median hospital stay (IQR), days	1 (1–2)	1 (1–2)	2 (1–2)	0.075
Pathologic Gleason score				0.272
≤6	8 (27.6%)	3 (10%)	5 (16.7%)
7	17 (58.6%)	21 (70%)	16 (53.3%)
≥8	4 (13.8%)	6 (20%)	9 (30%)
Pathologic T stage				0.357
T_2a-c	25 (83.3%)	21 (70%)	22 (73.3%)
T_3a-b	4 (13.3%)	9 (30%)	8 (26.7%)
T₄	1 (3.3%)	0	0
Positive margins	6 (20%)	7 (23.3%)	5 (16.7%)	0.812
Patients with complications	8 (26.7%)	4 (13.3%)	8 (26.7%)	0.358

Significant two-tailed.

P value by: Pearson chi-square test, Kruskal-Wallis.

IQR=interquartile range.

Importantly, the safety of the initial patients in each group appeared to be no different from the safety of the latter patients. There was no statistically significant difference in the median operative times (263 minutes [range 239–332 min], 281 minutes [242–353 min], and 273 minutes [230–347 min], P=0.775), the number of prolonged surgeries (9/30, 30.0%; 11/30, 36.7%; and 13/30, 43.3%; P=0.563), the estimated blood loss (135 mL, 50–250; 138 mL, 98–200; and 100 mL, 69–150; P=0.568), or the median hospital stay (1 day, 1–2; 1 day, 1–2; and 2 days, 1–2; P=0.075), in the Proctored, Independent, and Instructor Phases, respectively. Complications were reported within 30 days of surgery (Table 4).

Table 4.

Complications

	Proctored Phase (n=30)	Independent Phase (n=30)	Instructor Phase (n=30)	Total	Clavien grade
Neuropathy^a	1		1	2	I
Prolonged urinary leak		1	2	3	I
Other^b	4	1	2	7	I
Total Clavien I				12
Delayed extubation^c		1		1	II
Wound infection	1			1	II
Urinary tract infection	2			2	II
Pneumonia			1	1	II
Transfusion	1			1	II
Total Clavien II				6
Conversion to open^d		1		1	IIIb
Bowel injury^e		1	2	3	IIIb
Total Clavien III				4

Neuropathy—one femoral, one sciatic.

Other—symptomatic orthostatic hypotension not necessitating transfusion, hyperkalemia, skin blisters, facial edema.

Delayed extubation was during a prolonged case because of laryngeal edema.

Conversion to open was because of hypoxia early in the case.

Bowel injuries all recognized and repaired intraoperatively.

Note that one patient in the Proctored Phase had both a wound infection and a urinary tract infection and was listed twice.

The rates of complications were similar between the groups (8, 26.7%; 5, 16.7%; and 8, 26.7%; P=0.358). These included five (16.7%) Clavien I and three (10%) Clavien II complications during the Proctored Phase (One patient had two Clavien I complications. Both are listed in Table 4, but this patient was counted once in the total number of complications.). There were two (6.6%) Clavien I, one (3.3%) Clavien II, and two (6.6%) Clavien IIIb complications during the Independent Phase. The two Clavien IIIb complications consisted of a conversion to open surgery because of early hypoxia, likely unrelated to surgical technique, and a bowel serosal tear that was recognized and repaired intraoperatively with no additional sequelae.

During the Instructor Phase, there were five (16.7%) Clavien I, one (3.3%) Clavien II, and two (6.6%) Clavien IIIb complications (one rectal and one sigmoid tear, also recognized and repaired intraoperatively). One also had no additional sequelae, and the other had a prolonged hospital stay of 3 days, which was because of limited mobility, preexisting comorbidity, and residing in another state. He had no delayed return in bowel function. Of note, the Clavien IIIb complications were not statistically significant between the groups (0, 2, and 2, P=0.223, data not shown). Blood transfusions were rare and were seen only in the Proctored Phase (P=0.129), suggesting no increased risk in the latter patients. Finally, the positive margin rate was not significantly different between groups with 6 (20%), 7 (23.3%), and 5 (16.7%), P=0.812, respectively, as above.

Functional outcomes were assessed using the SHIM and a questionnaire developed by this group (Table 1). The questionnaire contains four questions about stress urinary incontinence (SUI) (minimum score 2=no SUI) and four questions about erectile dysunction (ED) (minimum score 2=no ED). These were mailed on two separate occasions to all patients from the Proctored and Independent Phase, and nine of the patients from the Instructor Phase. There were no significant differences in number of SHIM and questionnaires returned, average SHIM scores, total questionnaire scores, or scores on the SUI subscore (questions 1–4) or ED subscore (questions 5–8) of our questionnaire (Table 5).

Table 5.

Functional Outcomes (Incontinence and Erectile Function)

	Proctored Phase (n=30)	Independent Phase (n=30)	Instructor Phase (n=9)	P value
First round
Respondents	17 (56.6%)	16 (53.3%)	7 (77.8%)	0.0593
Avg. SHIM	7.0 (range 1–21)	6.8 (1–24)	3.0 (0–5)	0.4037
Questionnaire
Total	11.57	10.13	15.00	0.5995
SUI score (Q1–4)	4.70	4.61	5.13	0.9841
ED score (Q5–8)	6.87	5.52	9.90	0.5081
Second round
Respondents	15 (50.0%)	18 (60.0%)	4 (44.4%)	0.2989
Avg. SHIM	4.13	3.61	5.00	0.8903
Questionnaire
Total	2.08	2.11	1.00	0.7936
SUI score (Q1–4)	3.83	3.89	4.50	0.9668
ED score (Q5–8)	5.29	9.06	4.00	0.3229
Average time from surgery to survey
Survey 1	18.5	9.8	6.0	0.0276^*
Survey 2	23.3	14.4	9.8	0.0511

P value by: Pearson chi-square test, analysis of variance; Kruskal-Wallis.

Significant two-tailed.

Avg=average; SHIM=Sexual Health Inventory for Men; SUI=stress urinary incontinence; ED=erectile dysfunction.

Questionnaire (see Table1).

The surveys were performed at the same time for all three groups and only after 9 of the Instructor Phase surgeries were complete. The time from surgery to survey was therefore less for surgeries performed later in the study period. The mean time from surgery to survey 1 was 18.5, 9.8, and 6.0 months in the Proctored, Independent, and Instructor Phases, respectively (P=0.027), and to survey 2 was 23.3, 14.4, and 9.8 months, respectively (P=0.0511), with an overall follow-up (of respondents) ranging from 5 months to 28 months. Overall, there were no significant differences in functional outcome, even given the lead time bias that would favor the Proctored group. There was also improvement across all groups between the first and second survey.

Discussion

The first innovators of RALP were Vallancien and Guillonneau, laparoscopic surgeons who collaborated with Menon and associates¹³ at the Vattikuti Institute to help develop the RALP and structure a training program with early limited outcomes data (50 RALPs).¹³ Menon with Badani and colleagues¹⁴ describe a training sequence for a technique they called the Vattikuti Institute Prostatectomy (VIP) procedure. This included familiarization of the surgeons with the robot via syllabus and videos, progressing to practice on a pelvic trainer, observing 10 to 20 live cases, participating as the left-sided assistant and then the right-sided assistant, followed by gradual stepwise work on the console for about 20 to 30 cases. Once all the steps were mastered, they suggested mentoring of about 10 cases by an expert. These initial experiences were supported by limited outcomes data. The logistics of this training schema, however, may be difficult in some practices that want to adopt a robotics program in which the novice surgeon does not have the luxury of continued training allowing observation and gradual acclimation to the procedure.

Since the initial experience of the Vattikuti Institute, different institutions have struggled to outline the requirements before performing RALPs to ensure patient safety. There currently is no evidence-based standard for training novice surgeons in the use of the da Vinci® robot. Contemporary training programs vary from observerships to simulators to animal models to hands-on proctored cases of various lengths.

Hazards posed by the varied diffusion of minimally invasive technology are not new. Wide variation in patient outcome and complications occurred with the diffusion of laparoscopic cholecystectomy that resulted in large part from nonstandardized, relatively short-term training of surgeons adopting this new technology.^15,16 Despite these early difficulties and lack of standardized training, laparoscopic cholecystectomy is now considered standard of care. Our data suggest that a mentorship program of approximately 30 cases is one of the many possible ways to safely proctor a novice attending surgeon.

There are several limitations to this study. It is a single case study, and applicability may be limited to training other surgeons. Also, small differences may not be apparent because of the relatively small number of patients in each group. For example, the difference in the number of prolonged surgeries was not statistically significant between groups; however, this may be because of a lack of power and may reflect a real difference. Furthermore, serious adverse events in RALP are rare. Previous studies^10,17,18 claim that the learning curve needed to achieve acceptable perioperative outcomes is 12 to 40 patients. Even with small groups of patients, these previous studies were able to capture statistically significant changes in outcomes. These previous study designs informed the design of our protocol group size.

Another limitation is the smaller number and shorter follow-up of functional outcomes for patients who underwent surgery during the last Instructor Phase. One may anticipate that this lead time bias would yield worse functional outcomes for this latter group with shorter follow-up. On the contrary, however, despite this lead time bias, there were no significant differences in functional outcomes across the three groups. Longer follow-up may reveal more subtle differences between the group over time, but the overall follow-up (of respondents) was 5 months to 28 months, which is comparable to most existing studies.

Important features of the VASDHS mentorship program include a relatively high volume center (single surgeon performing 1 to 2 cases per week) with a dedicated robotics team of nurses providing consistent ancillary support throughout the process. The program allowed graduated transition of surgical responsibility from proctor to novice attending surgeon. The proctored novice attending surgeon completed the initial cases and subsequently transitioned to proctoring other trainees in the late cases. Despite the limitations of the study, the similarities in the outcomes of the patients, even in the later patients with larger prostates and higher BMIs, demonstrate that the gradual mentoring process is safe.

Conclusions

A graduated, hands-on mentorship program appears to be safe with similar functional outcomes in this VA-derived patient population. Longer follow-up may reveal more subtle differences between the groups over time, but the overall follow-up (of respondents) was 5 months to 28 months, which is comparable to most existing studies. Important features of our program included extended hands-on mentorship at a high volume center with closely monitored outcomes. Further study is needed to determine the optimal time frame of mentorship and those factors most critical to optimizing the novice attending physician learning curve. Despite this, we believe that extended mentorship by an experienced surgeon is a viable model for achieving proficiency in RALP in a setting such as the VA hospitals that are affiliated with academic hospitals or other similar settings. This is a viable method for incorporating RALP and increasing access to care for the veteran patient population.

Footnotes

Acknowledgment

The authors thank Eileen Silva for her editorial assistance.

Disclosure Statement

Dr. Kane is a consultant for and has received honoraria from Intuitive Surgical, Inc. For the remaining authors, no competing financial interests exist.

Abbreviations Used

References

Mirheydar

, Jones

, Koeneman

, Sweet

. Robotic surgical education: A collaborative approach to training postgraduate urologists and endourology fellows. JSLS, 2009; 13:287–292.

McDougall

, Corica

, Chou

, et al. Short-term impact of a robot-assisted laparoscopic prostatectomy ‘mini-residency’ experience on postgraduate urologists' practice patterns. Int J Med Robot, 2006; 2:70–74.

Gamboa

, Santos

, Sargent

, et al. Long-term impact of a robot assisted laparoscopic prostatectomy mini fellowship training program on postgraduate urological practice patterns. J Urol, 2009; 181:778–782.

Sharma

, Papadopoulos

, Lee

, et al. First 500 cases of robotic-assisted laparoscopic radical prostatectomy from a single UK centre: Learning curves of two surgeons. BJU Int, 2011; 108:739–747.

Sammon

, Perry

, Beaule

, et al. Robot-assisted radical prostatectomy: Learning rate analysis as an objective measure of the acquisition of surgical skill. BJU Int, 2010; 106:855–860.

Dev

, Sharma

, Dawson

, et al. Detailed analysis of operating time learning curves in robotic prostatectomy by a novice surgeon. BJU Int, 2012; 109:1074–1080.

O'Malley

, Van Appledorn

, Bouchier-Hayes

, et al. Robotic radical prostatectomy in Australia: Initial experience. World J Urol, 2006; 24:165–170.

Wolanski

, Chabert

, Jones

, et al. Preliminary results of robot-assisted laparoscopic radical prostatectomy (RALP) after fellowship training and experience in laparoscopic radical prostatectomy (LRP). BJU Int, 2012; 110(suppl 4):64–70.

Di Pierro

, Baumeister

, Stucki

, et al. A prospective trial comparing consecutive series of open retropubic and robot-assisted laparoscopic radical prostatectomy in a centre with a limited caseload. Eur Urol, 2011; 59:1–6.

10.

Ahlering

, Skarecky

, Lee

, Clayman

. Successful transfer of open surgical skills to a laparoscopic environment using a robotic interface: Initial experience with laparoscopic radical prostatectomy. J Urol, 2003; 170:1738–1741.

11.

Chin

, Luke

, Pautler

. Initial experience with robotic-assisted laparoscopic radical prostatectomy in the Canadian health care system. Can Urol Assoc J, 2007; 1:97–101.

12.

Smith

. VA San Diego Healthcare System Annual Report. Available at: www.sandiego.va.gov/docs/publicaffairs/annualreport2011.pdf. Accessed: May 30, 2014 .

13.

Menon

, Shrivastava

, Tewari

, et al. Laparoscopic and robot assisted radical prostatectomy: Establishment of a structured program and preliminary analysis of outcomes. J Urol, 2002; 168:945–949.

14.

Badani

, Hemal

, Peabody

, Menon

. Robotic radical prostatectomy: The Vattikuti Urology Institute training experience. World J Urol, 2006; 24:148–151.

15.

Escarce

, Bloom

, Hillman

, et al. Diffusion of laparoscopic cholecystectomy among general surgeons in the United States. Med Care, 1995; 33:256–271.

16.

Ferreira

, Bennett

, Gilman

, et al. Diffusion of laparoscopic cholecystectomy in the Veterans Affairs health care system, 1991–1995. Eff Clin Pract, 1999; 2:49–55.

17.

Leroy

, Thiel

, Duchene

, et al. Safety and peri-operative outcomes during learning curve of robot-assisted laparoscopic prostatectomy: A multi-institutional study of fellowship-trained robotic surgeons versus experienced open radical prostatectomy surgeons incorporating robot-assisted laparoscopic prostatectomy. J Endourol, 2010; 24:1665–1669.

18.

Gumus

, Boylu

, Turan

, Onol

. The learning curve of robot-assisted radical prostatectomy. J Endourol, 2011; 25:1633–1637.