First Report on Joint Use of a Da Vinci® Surgical System with Transfer of Surgical Know-How between Two Public Hospitals

Robot-assisted radical prostatectomy (RARP) is widely used to treat localized prostate cancer. Its perioperative outcomes and complications as well as oncological and particularly functional outcomes are equal or slightly superior to those of open and laparoscopic prostatectomy, even in centers with limited caseloads [1, 2, 3, 4, 5, 6]. Unfortunately, the costs of robotic devices, their service contracts, material and depreciation are extremely high. These costs are a significant obstacle to small and mid-size institutions purchasing a robotic surgery device. A large caseload is therefore a major prerequisite for a hospital to offer this type of surgery [7]. By the same token, independent of costs, a large caseload is essential for the training of surgical teams to maintain and improve oncological and functional outcomes, especially for difficult procedures such as radical prostatectomy (RP) [8, 9, 10, 11].

In an effort to optimize surgical outcomes, the Department of Urology of the Kantonsspital in Lucerne (26 urology beds, 875 total beds) purchased a Da Vinci® surgical system in 2008. Over time it became clear that the caseload was insufficient to meet costs. In the search for a solution, the hospital learned that the Department of Urology of the Kantonsspital in St. Gallen (40 urology beds, 863 total beds) was looking for a way to offer robotic surgery to its patients without incurring the costs of purchasing its own Da Vinci® system. A joint use agreement was made.

The objective of this prospective study was (1) to assess the technical and logistic feasibility of sharing a standard four-arm Da Vinci® surgical device between the urological departments of two mid-sized public hospitals in Switzerland, (2) to evaluate the effectiveness of transferring surgical know-how during the first year of joint use, as reflected in the early oncological and functional results of RARP, and (3) to enumerate the technical problems and errors of the Da Vinci® system during one year of joint use and to determine whether and to what extent they were caused by transportation of the device between the two hospitals.

To our knowledge this is the first systematic prospective documentation of the joint use of a Da Vinci® surgical device owned by one public hospital and rented by another with the transfer of surgical know-how from the owning to the renting hospital.

The Department of Urology of the Kantonsspital in Lucerne and the Department of Urology of the Kantonsspital in St. Gallen entered into the current joint use arrangement on January 1, 2011. The objective was to share a standard four-arm Da Vinci® surgical device (Intuitive Surgical, Sunnyvale, Calif., USA), with transfer of surgical know-how from Lucerne to St. Gallen. The distance between the two hospitals is 140 kilometers. During the study period, the Da Vinci® device was transported between hospitals by a special truck with pneumatic shock absorbers and remained in each institute for 1 month per cycle. The Da Vinci® system was tested applying an in-house technical inspection protocol (table 1) before and after transportation, and all Da Vinci® system errors were noted in a log book.

The Da Vinci® surgical device is owned by the Lucerne hospital and rented by the St. Gallen hospital. The rent includes a fee for the service contract in St. Gallen, transportation fees and special insurance costs. The manufacturer of the robotic device adjusted the price of the service contract to allow for potential major device failures due to transportation.

The Lucerne hospital began its robotic surgery program in June 2008. By the end of 2010 it had performed a total of 171 robot-assisted urological operations, of which 127 were RARPs and 44 pyeloplasties. The oncological and functional outcomes of the RARPs and the functional outcomes of the pyeloplasties were comparable to those of leading institutions performing the same robot-assisted procedures [5, 12]. Robot-assisted surgical procedures performed at the Lucerne hospital during the present study were RARP, pyeloplasty, partial nephrectomy, nephropexy and retroperitoneal lymph node dissection; however, only the RARP procedures were prospectively documented.

Both hospitals assigned a single surgeon to perform all RARP procedures (A.M. in Lucerne, D.E. in St. Gallen). The St. Gallen surgeon (D.E., 15 years experience in urology, trained at the hospital of St. Gallen by H.P. Schmid) was mentored by the Lucerne surgeon (A.M., 15 years experience in urology and 6 years experience in laparoscopic and robot-assisted surgery, trained at the university of Berne, Switzerland by U.E. Studer and at the Clinic Saint Augustin in Bordeaux, France by R. Gaston). Training was also given to the operating room nursing staff. The two institutions applied the same surgical technique: monoblock technique for (extended) pelvic lymph node dissection (PLND) and tension- and energy-free technique for RP [13, 14]. During the first year of joint use, preoperative patient characteristics, perioperative data and complications according to the modified Clavien classification [15] as well as oncological and functional outcomes at 3-month follow-up were documented for all patients undergoing RARP. Early continence was tested by the cough stress test on postoperative day 5 after catheter removal [16]. Preoperative and 3-month postoperative continence were evaluated by pad use and the validated International Consultation on Incontinence Questionnaire-Short Form (ICIQ-SF), potency was measured by the validated Expanded Prostate Cancer Index Composite (EPIC) questionnaire [17].

All technical failures of the Da Vinci® surgical device were monitored and all maintenance steps covered by the service contract with the manufacturer (Intuitive Surgical) were documented.

Statistical analysis was performed by external professionals using Wilcoxon, exact Poisson and χ² tests.

A total of 103 patients underwent robot-assisted laparoscopic surgery during the first year of joint use, 75 in Lucerne (61 RARPs, 12 pyeloplasties, 1 partial nephrectomy, 1 retroperitoneal lymph node dissection) and 28 in St. Gallen (19 RARP, 8 pyeloplasties, 1 nephropexy).

The RARP patients of both hospitals exhibited no significant differences in the preoperative characteristics median age, body mass index, prostate-specific antigen (PSA), tumor stage, risk group, urinary continence and erectile function (tables 2, 3, 4). The median prostate volume of the St. Gallen patients was significantly larger than that of the Lucerne patients (40 cm³ [interquartile range 33-57 cm³] vs. 33 cm³ [interquartile range 24-40 cm³]; p = 0.02). The biopsy Gleason scores and postoperative histology stages of the Lucerne patients were significantly higher than those of the St. Gallen patients (p = 0.04 and p = 0.03, respectively) (tables 2, 3). The mean operation time of RARP including PLND was significantly longer in the St. Gallen hospital compared to the Lucerne hospital (mean 351 ± 50 vs. 252 ± 49 min; p = 0.0001). The proportion of attempted PLNDs (limited PLND for low-risk patients and extended PLND for intermediate- and high-risk patients according to the D'Amico classification [18]) and nerve-sparing procedures were comparable for the two hospitals (table 5).

Water tightness of the vesicourethral anastomosis as checked by cystography on the 5th postoperative day was similar for the two hospitals (98% for Lucerne patients and 95% for St. Gallen patients; p = 1.00). Patients were discharged after a median of 8 days (range 7-14 days) from the Lucerne hospital and 9 days (range 8-18 days) from the St. Gallen hospital (p < 0.01) (table 5). Positive surgical margin rates and postoperative complications were comparable for the two hospitals (tables 3, 6). The number of lymph nodes removed during lymph node dissection was higher in Lucerne (median 16 nodes, range 13-21) than in St. Gallen (median 15 nodes, range 8-16) (p = 0.02).

At 3-month follow-up, 98% of Lucerne patients and 100% of St. Gallen patients had a PSA value <0.2 ng/ml. Early continence on the day of catheter removal was 79% for Lucerne patients and 68% for St. Gallen patients (p = 0.54). Continence at 3-month follow-up was 95% (0-1 pad per day for safety only) or 92.3% (ICIQ-SF questionnaire) for Lucerne patients and 95% (0-1 pad per day) or 100% (ICIQ-SF questionnaire) for St. Gallen patients (p = 0.77) (table 4).

Erectile function (with or without phosphodiesterase type 5 inhibitors) in preoperatively potent patients undergoing unilateral or bilateral nerve-sparing surgery was 65% at 3-month follow-up for Lucerne patients and 40% for St. Gallen patients (table 4).

Technical failures of the Da Vinci® surgical device are shown in table 7. To our knowledge, none of the technical problems were unequivocally due to transportation of the device. The Da Vinci® device was transported from one hospital to the other a total of 11 times. After five of the eleven shipments the Da Vinci® robot worked error- and technical problem-free during the month-long stay at the receiving hospital. During the 12 months of the study a total of eight error signals were noted. The median time between transportation and occurrence of an error signal was 15 days (range 4-30 days). A comparison of all technical failures of our standard Da Vinci® device with those of a similar standard four-arm device of comparable age owned by an independent large public hospital in Switzerland performing almost 200 robotic procedures per year and used exclusively in a single operating room showed similar numbers and types of failure (table 8).

Joint use of the Da Vinci® surgical device resulted in a 40% reduction in costs (acquisition and maintenance, including transportation and customized service package) for each hospital.

The main goal of any surgical treatment for prostate cancer is to provide patients with optimal oncological and functional outcomes at affordable costs. Many nomograms have been made to predict RP outcomes related to patient characteristics [19, 20], but the surgical approach and technique remain a fundamental step to achieve good results. Optimal outcomes can be reliably achieved with a robotic surgery system. The high costs associated with these systems, however, constitute a major obstacle to their acquisition by low-volume centers. This is especially relevant for countries like Switzerland, which has numerous small to mid-sized centers offering RP. For such centers, material costs for the relatively small number of robotic procedures they perform are disproportionately high, RARP costing significantly more than open RP [7, 21, 22].

The health care system in Switzerland, as in many other countries, is under pressure to stem the rising costs of health care. To meet this challenge, we have searched for innovative solutions that would allow us to keep pace with modern advances in minimally invasive urological surgery, especially RARP. Moreover, robotic surgery is an important marketing item for recruitment of patients [23, 24, 25]. The learning curve for new technologies like RARP [26, 27, 28] can take years in small centers working entirely independently of one another. An internal survey of members of the Swiss Urological Society in 2010 found that the majority of urologists performed a maximum of 20 RPs per year and that only a few institutes exceeded 100 RPs per year (unpublished data). According to the studies of Vickers et al. [8] and Budäus et al. [9, 10], it takes years to reach good outcomes, few complications and brief hospital stays. During the start-up phase of a robotic system for RARP, the quality of surgical treatment must be comparable to the results of open RP pertaining at the same and neighboring institutions. This can only be achieved by a steep learning curve, for which a large caseload is imperative. Our joint use agreement for the Da Vinci® surgical device, therefore, stipulated the transfer of surgical know-how during mentoring of the surgeon (D.E.) and staff at St. Gallen by an experienced surgeon (A.M.) from Lucerne during the year of the agreement. The effectiveness of the mentoring is reflected in the similarity of the early oncological and functional outcomes at the two institutes. The limited number of patients (n = 80) treated by RARP at the two hospitals began with comparable preoperative characteristics and went on to share similar perioperative data such as nerve-sparing procedures, attempted lymph node dissections, complications, water tightness of the vesicourethral anastomosis, lengths of hospital stay, positive surgical margins, PSA relapse, continence 1 day after catheter removal and 3 months postoperatively, as well as potency 3 months postoperatively. Although the early potency rate is probably inconclusive due to the limited number of nerve-sparing procedures performed on preoperatively potent men, the rates appear to be close for the two hospitals. The results of our first year of sharing a Da Vinci® surgical device are similar to those of other centers with similar or even greater years of RARP experience and caseloads per year [1, 2, 3, 4].

Our data show that rigorous standardization of the procedure and continuous mentoring over 1 year and almost 20 RARP procedures can neutralize the learning curve for RARP with regard to oncological and functional outcomes, as already demonstrated in other centers [11]. The only significant difference between the two hospitals was, as expected, the longer operation time of the mentored staff in St. Gallen. Somewhat surprisingly, this did not lead to significantly more operation time-related complications in St. Gallen. Of great interest will be the results regarding outcomes in St. Gallen after the mentoring phase, planned only for the first year, has been completed. The present results show that joint use of a robotic system by two urology institutes, including monthly transportation from one institute to the other, is clearly feasible and may offer financial advantages to small institutes with limited caseloads without causing relevant transportation-associated technical problems. The sharing not only of material but also of technical know-how by the owning institute facilitated start-up of the robotic procedures in the renting institute. As quality in surgery is always also a matter of caseload, it appears clear that the performance of only about 20 procedures per year will never be sufficient to install a solid routine among the surgical team. It has to be pointed out that at the hospital of St. Gallen the number of RPs performed yearly varies between 60 and 70. Nevertheless, during the first year of sharing and introduction of the robotic technology, only one of the three senior surgeons started operating with the robot. A further and larger use for the robot was planned after introduction in the start team had been accomplished. Details of the service costs or the joint use agreement cannot be given here because they are covered by a confidentiality agreement between the Lucerne and St. Gallen hospitals as well as between the Lucerne hospital and the company supplying the Da Vinci® surgical system. The number and kinds of technical failures of our standard Da Vinci® surgical system were similar to those of a similar system of the same age but used exclusively in a single operating room. We therefore assume that the technical failures of our device were probably due more to its routine use and age than to its transportation between hospitals. This assumption is supported by the fact that five of the eleven shipments were followed by no technical problems or error signals. Moreover, the eight total technical problems or error signals noted during the 1-year study period did not appear until a median of 15 days (range 4-30 days) after transportation and after several surgical procedures had been performed at the receiving hospital.

Joint use of a Da Vinci® surgical device by two (or more) institutes can reduce costs without compromising the device's functionality due to its transportation between institutes if stringent safety precautions are applied. In addition, the continuous transfer of surgical know-how during the first year of joint use shortens the learning curve for RARP in institutes naïve to robotic surgery, thus securing the quality of early oncological and functional results.

The authors declare that they have no conflict of interest.

A. Mattei and M. Thoms contributed equally to the realization of the study and the draft of the manuscript.