This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright

European Journal of Obstetrics & Gynecology and Reproductive Biology 150 (2010) 92 96 Contents lists available at ScienceDirect European Journal of Obstetrics & Gynecology and Reproductive Biology journal homepage: www.elsevier.com/locate/ejogrb Robotic hysterectomy versus conventional laparoscopic hysterectomy: Outcome and cost analyses of a matched case control study Dimitri Sarlos *, LaVonne Kots, Nebojsa Stevanovic, Gabriel Schaer Department of Obstetrics and Gynaecology, Kantonsspital Aarau, Tellstrasse, 5001 Aarau, Switzerland ARTICLE INFO ABSTRACT Article history: Received 5 August 2009 Received in revised form 8 December 2009 Accepted 2 February 2010 Keywords: Robotic hysterectomy Hysterectomy Robot-assisted Hysterectomy Laparoscopic Hysterectomy Robotic versus laparoscopic Laparoscopy Cost analysis Objective: Robotic surgery, with its technical advances, promises to open a new window to minimally invasive surgery in gynaecology. Feasibility and safety of this surgical innovation have been demonstrated in several studies, and now a critical analysis of these new developments regarding outcome and costs is in place. So far only a few studies compare robotic with conventional laparoscopic surgery in gynaecology. Our objective was to evaluate our initial experience performing total robotassisted hysterectomy with the da Vinci 1 surgical system and compare peri-operative outcome and costs with total laparoscopic hysterectomy. Study design: For this prospective matched case control study at our institution, peri-operative data from our first 40 consecutive total robot-assisted hysterectomies for benign indications were recorded and matched 1:1 with total laparoscopic hysterectomies according to age, BMI and uterus weight. Surgical costs were calculated for both procedures. Surgeons subjective impressions of robotics were evaluated with a self-developed questionnaire. Results: No conversions to laparotomy or severe peri-operative complications occurred. Mean operating time was 109 (113; 50 170) min for the robotic group and 83 (80; 55 165) min for the conventional laparoscopic group. Mean postoperative hospitalisation for robotic surgery was 3.3 (3; 2 6) days versus 3.9 (4; 2 7) days for the conventional laparoscopic group. Average surgical cost of a robot-assisted laparoscopic hysterectomy was s4067 compared to s2151 for the conventional laparoscopic procedure at our institution. For the robotic group wider range of motion of the instruments and better ergonomics were considered to be an advantage, and lack of direct access to the patient was stated as a disadvantage. Conclusion: Robot-assited hysterectomy is a feasible and interesting new technique with comparable outcome to total laparoscopic hysterectomy. Operating times of total laparoscopic hysterectomy seem to be achieved quickly especially for experienced laparoscopic surgeons. However, costs of robotic surgery are still higher than for conventional laparoscopy. Randomised clinical trials need to be conducted to further evaluate benefits of this new technology for patients and surgeons and analyse its costeffectiveness in gynaecology. ß 2010 Elsevier Ireland Ltd. All rights reserved. 1. Introduction Since its establishment in gynaecological surgery in the 1960s and 1970s, major improvements in visualisation technology and instrumentation in the mid-1980s have clearly endorsed laparoscopic techniques in gynaecology [1]. Procedures such as hysterectomy, staging of endometrial and cervical cancer, prolapse surgery or endometriosis resection meanwhile are standard procedures in gynaecological surgery. Despite promising results compared to laparotomy regarding equal outcomes and lower peri-operative morbidity, improvement of quality of life, shorter hospital stay and faster return to activity * Corresponding author. Tel.: +41 062 838 5083; fax: +41 062 838 5091. E-mail address: Dimitri.Sarlos@ksa.ch (D. Sarlos). [2,3], the percentage of laparoscopic hysterectomies is still very low. Abdominal hysterectomy remains the most common approach [4]. An unfavourable learning curve [5,6] and extensive training of surgeons and the whole surgical team are often cited as reasons. In recent years robotic surgery has been introduced in urological surgery [7], general surgery [8,9], cardiac surgery [10] and gynaecological surgery [11]. Several publications describe safety and feasibility of this new approach [11 14]. After the first excitement over the innovative and sophisticated technology has settled there also should be some consideration for a critical assessment of technique and costs, ideally with recommendations for further improvements by experienced laparoscopic surgeons. The da Vinci 1 surgical system has some advantages compared to conventional laparoscopic surgery, such as three-dimensional vision, better ergonomics, higher degree of freedom of the robotic instruments and reduction of tremor interference. The hope is that 0301-2115/$ see front matter ß 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejogrb.2010.02.012

D. Sarlos et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 150 (2010) 92 96 93 these advantages will one day lead to a more widespread use of minimally invasive techniques in gynaecological surgery. There is a need for more prospective data evaluating the da Vinci 1 system in gynaecological surgery as there are few studies on this topic with a small number of cases and most of them are retrospective [12,13,15,16]. With this matched case control study we report on our initial experience using the da Vinci 1 system for robot-assited hysterectomy. We evaluated peri-operative outcome and analysed costs in comparison with results of total laparoscopic hysterectomies in our institution. Subjective impressions of experienced endoscopic surgeons were recorded prospectively to evaluate possible advantages or disadvantages of robotic surgery. 2. Material and methods 2.1. Population From June 2007 to May 2009 we prospectively recorded perioperative data of the first 40 consecutive robot-assisted hysterectomies using the da Vinci 1 surgical system. All patients were recruited from our clinic and participated after informed consent. The local ethics committee waived committee approval for this study because laparoscopic interventions are standard procedures in our clinic. Indications for robot-assisted hysterectomy were benign lesions if vaginal hysterectomy was expected to be difficult because of large myomas or nulliparity. All patients with these indications were included in the study after informed consent and if the uterus weight was estimated to be <500 g. If patients did not consent for the study, or if the uterus weight was estimated to be over 500 g by ultrasound, a conventional laparoscopic hysterectomy was performed. In cases with an estimated uterus weight over 1200 g a laparotomy was performed. Indications for conventional total laparoscopic hysterectomy were the same as for the robotic group. 2.2. Surgical technique The three-armed da Vinci 1 standard surgical robot (Intuitive Surgical Inc., Sunnyvale, CA, USA) has been used in our public teaching hospital, Kantonsspital Aarau, Switzerland by the Urology Department since 2005. We started our first robot-assisted laparoscopic procedures in 2007. All operations were performed by two senior gynaecological surgeons experienced in laparoscopic surgery. They have been performing an average of 50 laparoscopic hysterectomies per year for almost 10 years now. All operations took place under general anaesthesia in a lithotomy position with a Foley catheter in the bladder. All patients received peri-operative antibiotic prophylaxes of 2 g intravenous cephazolin. A Clermont- Ferrand manipulator (Storz) was placed in the uterus. The surgical team consisted of a console surgeon, a bedside assistant and a surgical nurse standing on the left side of the patient. A second assistant was sitting between the legs of the patient and once the robot was in place underneath the arms of the robot for manipulation of the uterus. After the assistants and nurse were in position and the uterus manipulator was inserted vaginally, the surgeon performed the skin incisions for the trocars. These are the connections to the robot. Then the robot was rolled to the table and trocars were attached to the robot. The optic port was 12 mm in diameter and was placed 3 5 cm above the umbilicus. Two robotic working ports consisting of 8 mm trocars were placed lateral to the rectus abdominis muscle about 3 cm inferior to the level of the umbilicus. An additional 10 mm trocar was placed para-umbilically on the left side for assistance and transport of suturing. The following robotic instruments were used: a fenestrated EndoWrist 1 bipolar forceps on the left robotic arm and a monopolar EndoWrist 1 curved scissors on the right side. All surgeries were performed following our defined standard operating procedure for conventional laparoscopic hysterectomy. We started the hysterectomy by transection of the round ligament and then dissection of the broad ligament anteriorly and posteriorly. In premenopausal women we did not perform extra-adnexal hysterectomy. The bladder was dissected from the proximal vagina. The ascending branches of the uterine vessels were coagulated with bipolar current and transected by scissors. After transection of the cardinal ligaments, colpotomy was performed using the monopolar cutting current and the uterus was extracted vaginally. Large size uteri were cut into extractable parts with a knife and removed vaginally. In these cases the robot had to be undocked and the vagina was closed vaginally. The vaginal cuff was closed by robotic suturing using fiveinterrupted sutures, withintracorporal knotting technique used for all other cases. For conventional laparoscopic hysterectomy, endoscopic morcellation of large size uteri was performed and laparoscopic suturing was applied to close the vagina. 2.3. Outcome measures Time for robot docking and total operating time were recorded. Intra-operative uterine weight was recorded with a digital scale. Duration of surgery or total operating time was defined as the time from skin incision to the last skin closure suture. Robot docking time was defined as the time from when the robot is brought to the operating table until the surgeon starts the operation at the console, including attachment of the robot to trocars. The robot was brought to the table after trocars were inserted. According to protocol we recorded all intra- and postoperative complications (bowel and bladder lesions, urinary tract infections, blood loss, fever), dosage of postoperative analgesics and the total postoperative hospital stay. Data from this prospective evaluation of robotic hysterectomies were compared with retrospective data from a pool of 150 conventional laparoscopic hysterectomies performed by the two senior surgeons in 2007 and 2008. This pool was meticulously searched for comparable laparoscopic hysterectomies by an independent reviewer by handwork. These data were matched 1:1 according to surgeon, uterus weight, BMI and age. 2.4. Cost analysis Data were recorded to assess costs generated in the operating room. Personnel costs for surgeons, anaesthetist and nurses were calculated as costs per minute with different personnel factors based on salaries at our publicly funded hospital for each speciality. Material used during each procedure, such as reusable, disposable and special robotic instruments and medications for the robotic procedure, were recorded by medical staff, and costs for purchase of all devices and sterilisation only for reusable instruments were calculated. 2.5. Surgeons questionnaire Surgeons subjective assessments about the advantages of the robotic procedure, as well as ergonomics and different surgical aspects compared to conventional laparoscopic surgery or laparotomy, were recorded at the end of the procedure using a special designed questionnaire (Table 3). 2.6. Biometrics For statistical analysis, continuous variables expressed as mean, median and range (minimum/maximum) were given (median; range); Mann Whitney test was used to compare both groups.

94 D. Sarlos et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 150 (2010) 92 96 3. Results 3.1. Robotic group From June 2007 to May 2009, 40 robot-assisted total hysterectomies were performed at our institution according to the study protocol. All 40 consecutive cases were included for final analysis. Thirty-four patients received total hysterectomy without adnexectomy; in three cases an additional unilateral adnexectomy, and in another three cases an additional bilateral adnexectomy was performed. Mean intra-operative uterine weight was 217 g (160; 35 520 g). Mean duration of total surgery was 109 min (113; 50 180 min). Average docking time was 21 min (20; 14 60 min). Mean blood loss was 81 ml (80; 20 200 ml). There were no intraoperative complications, especially no severe bleeding or organ lesions. No conversions to conventional laparoscopy or laparotomy were needed and no technical problems related to the robot occurred. In 16 cases the vagina was closed by vaginal sutures after undocking the robot for vaginal morcellation of the uterus. Postoperatively one patient had a urinary tract infection and four patients had postoperative fever higher than 37.58. Postoperatively antibiotics were used for patients with urinary tract infection. Comparing operating time of the first 10 consecutive cases of robot-assisted hysterectomies with the last 10 cases there was no significant difference in operating times: 100 min versus 122 min (p > 0.05), but mean uterine weights were significantly higher in the last 10 cases, with 335 g compared to the first 10 procedures with 215 g (p < 0.05). For robot set-up time a decrease was noted comparing the first 10 consecutive cases (29.2 min) with the last 10 cases (10.8 min; p < 0.05). 3.2. Conventional laparoscopic group Operating time in the conventional laparoscopic group was 82.9 min (80; 55 165 min) and significantly shorter than in the robotic group (p < 0.05). Mean uterine weight was 195 g (217; 50 551 g). There were no significant differences (p < 0.05) between the two groups regarding complications, conversions to laparotomy, intra-operative bleeding and hospital stay. There was one patient with wound infection. The comparative peri-operative outcome is summarised in Table 1. 3.3. Costs Total personnel median costs for surgery were s1771 for total robot-assisted hysterectomy and s1329 for conventional total laparoscopic hysterectomy according to time recorded for surgery, i.e. time spent for preparation and surgery in the operating room. The material used was standard procedure material. Because there were no complications such as excess bleeding, no extra material was recorded and material costs were the same for each procedure of the respective groups. Total material costs came to s2217 for the robot-assisted procedure and s822 for conventional total laparoscopic hysterectomy. Taking together personnel and material costs, total surgical costs then amounted to s4067 for the robotic group and s2151 for the conventional laparoscopic group at our institution. Details are shown in Table 2. This does not include acquisition and amortisation as the robot was funded by the cantonal government of Aargau and is mostly used by other surgical specialities. Results of the surgeons questionnaire are shown in Table 3. 4. Discussion With this matched case control study of our first 40 cases we could show that total robot-assisted hysterectomy is feasible and safe and that the learning curve is quite fast if the surgeon is experienced in conventional laparoscopic surgery. The postoperative outcome seems to be similar to conventional total laparoscopic hysterectomy but operating times were significantly longer in our robotic group. Analysis of the surgeons questionnaire indicates that wider range of motion of robotic instruments and better ergonomics were the most important advantages of the da Vinci 1 surgical system for the surgeon. Lack of direct access to the patient and especially the uterine manipulator was considered to be the most significant disadvantage compared with conventional laparoscopic hysterectomy. We could also demonstrate that operating room costs for robotassisted hysterectomy were around s2000 higher than for conventional total laparoscopic hysterectomy and most of these higher costs are attributed to material costs. Unfortunately, studies analysing cost-effectiveness are too few and too small to allow valid evaluation. So far most studies comparing costs of robotic surgery with other approaches demonstrate higher costs for the robotassisted procedure [14]. Randomised controlled clinical trials are imperative to evaluate cost-effectiveness of robotic surgery. To our knowledge, this is the first matched case control study systematically analysing peri-operative data and costs of robot-assisted hysterectomy in comparison to conventional total laparoscopic hysterectomy. It is also the first to include a subjective evaluation of the da Vinci 1 surgical system by experienced laparoscopic surgeons with a defined questionnaire. Table 1 Demographic and peri-operative outcome. Robotic group Conventional laparoscopic group p-values Demographic data Age (years) 47 (45.5; 34 68) 43.6 (44; 33 58) 0.112 BMI 26 (24.5; 19 46) 26 (23.9; 19 38) 0.288 Surgery parameters Robot docking time 21 (20; 14 60) Operating time (min) 108.9 (113; 50 180) 82.9 (80; 55 165) <0.001 Severe intra-operative complications Blood loss (ml) 81 (80; 20 200) <50 Uterus weight 217 (160; 35 520) 195 (217; 50 551) 0.380 Postoperative data Postop. fever 4 Urinary tract infection 1 Wound infection 1 Hospital stay (days) 3.3 (3; 2 6) 3.9 (3; 2 7) 0.924 Data presented as mean (median and range).

D. Sarlos et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 150 (2010) 92 96 95 Table 2 Material and personnel costs. Robotic group Conventional laparoscopy p-values Material costs (s) 2296.08 821.68 Personnel costs (s) 1770.84 (1194 2288) a 1329.08 (1160 1707) a >0.05 Total costs surgery (s) 4066.84 2150.76 <0.05 a Data presented as mean (range). Table 3 Results of surgeon s questionnaire. Questions Yes No Was the use of the robot an advantage compared with conventional laparoscopic surgery? 37 3 Was the three-dimensional view a significant benefit? 40 Was the wider range of instrument motion an advantage? 32 8 Was the absence of haptic feedback a disadvantage? 16 24 Was the lack of direct access to the patient a disadvantage? 31 9 Was the use of the robot ergonomically better? 33 7 For better interpretation of these results it has to be mentioned that we compared the first 40 robotic procedures with a wellestablished procedure. Conventional total laparoscopic hysterectomy has been performed more than 500 times at our institution during the last 7 years. We assume that with more experience operating times will probably improve. Mean operating time of our first robot-assisted hysterectomies in this study was 109 min and significantly shorter than for the first cases reported by Reynolds and Advincula [12] and by Fiorentino et al. [15] with operating times of 242 and 192 min respectively, even though in our study mean uterine weight was significantly higher with 217 g compared with 131 g in the study of Reynolds and Advincula [12]. With 109 min as the mean skin-to-skin time for the robotic group, our study showed similar results as reported in the study by Payne and Dauterive [17] with 119 min, who also reported similar mean uterine weights of 266 g in the robotic cohort. In Payne s study [17] a significant decrease in operating time for the robotic group could be demonstrated comparing the last 25 with the first 25 procedures. In our study we did not see a significant decrease in operating time comparing the first 10 with the last 10 cases. However, the mean uterine weights were significantly higher in the last cases, which could explain these operating times in our study. It also has to be mentioned that in our institution we are still at the beginning of the learning curve for robotic surgery, however, in our opinion experience in conventional laparoscopic surgery is of advantage to help overcome entry to robotic surgery. This would also explain the shorter operating times in our and other studies where previous laparoscopic experience is stated [17] compared with other studies [12,15]. Therefore, the experience in conventional laparoscopic surgery of a surgical team should always be stated in publications analysing the outcome of robotic surgery; unfortunately, this is not mentioned in most studies. Set-up time of the robot is often reported as a disadvantage of the da Vinci 1 surgical system. In our study it was significantly shorter in the last 10 cases compared with the first 10 cases. This shows that set-up times can be improved, as demonstrated by other authors [17]. Complication rates and hospital stay in our study are similar for robotic and conventional laparoscopic surgery. A randomised controlled trial with sufficient cases including quality of life analysis is necessary to provide a well-grounded answer to this question. We are aware that the surgeons questionnaire gives only a subjective impression of possible advantages and disadvantages of robot-assisted surgery, but to date no validated questionnaire is available. By analysing the different surgical steps we found that the real benefit of the robot seems to be better ergonomics and wider range of motion of the robotic instruments. Especially in cases with larger uteri (>300 g), opening of the vagina was easier because of better angularity of the robotic instruments. Threedimensional vision does not seem to give a relevant benefit for the laparoscopic surgeon experienced in two-dimensional planes. Absence of haptic feedback is not a significant disadvantage of the da Vinci 1 surgical system as it seems to be compensated by optical control after a short learning curve. Lack of direct patient access, especially for manoeuvring of the uterine manipulator, was stated as one of the most important disadvantages of robotassisted surgery. It has to be mentioned that we used a threearmed robot and therefore guiding the uterine manipulator is more important. This might be less significant with a four-armed robot. Robotic surgery is of enormous interest for the future and in our opinion will significantly influence surgical procedures. Robotic surgery still is in its infancy and we believe that further improvements in technology and costs are needed. With rising demand to cut down expenses in health care systems worldwide, the high costs for acquisition, instrumentation and maintenance of the da Vinci 1 robot are a major drawback. Therefore, material costs should be reduced in order to guarantee a more widespread use of robotics. With further technical advances such as reducing bulkiness, better suturing techniques and implementation of learning software/simulators and teaching consoles, robotic surgery might help shorten the learning curve for endoscopic surgery and lead to a more widespread use of minimally invasive procedures. Conflict of interests There was no conflict of interest. Ethics approval The local ethics committee was informed about this study and waived committee approval because endoscopic interventions are standard procedures in our clinic. Acknowledgement The study was government funded by the Kanton (district government) Aargau and by our institution, the Kantonsspital Aarau.

96 D. Sarlos et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 150 (2010) 92 96 References [1] Spaner SJ, Warnock GL. A brief history of endoscopy, laparoscopy, and laparoscopic surgery. J Laparoendosc Adv Surg Tech A 1997 Dec;7(6):369 73. [2] Kluivers KB, Johnson NP, Chien P, Vierhout ME, Bongers M, Mol BW. Comparison of laparoscopic and abdominal hysterectomy in terms of quality of life: a systematic review. Eur J Obstet Gynecol Reprod Biol 2008 Jan;136(1):3 8. [3] Walsh CA, Walsh SR, Tang TY, Slack M. Total abdominal hysterectomy versus total laparoscopic hysterectomy for benign disease: a meta-analysis. Eur J Obstet Gynecol Reprod Biol 2009 May;144(1):3 7. [4] Wu JM, Wechter ME, Geller EJ, Nguyen TV, Visco AG. Hysterectomy rates in the United States, 2003. Obstet Gynecol 2007;110(November (5)):1091 5. [5] Boggess JF. Robotic surgery in gynecologic oncology: evolution of a new surgical paradigm. J Robotic Surg 2007;1(March (1)):69 74. [6] Lenihan Jr JP, Kovanda C, Cammarano C. Comparison of laparoscopic-assisted vaginal hysterectomy with traditional hysterectomy for cost-effectiveness to employers. Am J Obstet Gynecol 2004;190(June (6)):1714 20 [discussion 1720 2]. [7] Ficarra V, Cavalleri S, Novara G, Aragona M, Artibani W. Evidence from robotassisted laparoscopic radical prostatectomy: a systematic review. Eur Urol 2007;51(January (1)):45 55 [discussion 56]. [8] Hartmann J, Menenakos C, Ordemann J, Nocon M, Raue W, Braumann C. Longterm results of quality of life after standard laparoscopic vs. robot-assisted laparoscopic fundoplications for gastro-oesophageal reflux disease. A comparative clinical trial. Int J Med Robot 2009;5(March (1)):32 7. [9] Draaisma WA, Ruurda JP, Scheffer RC, et al. Randomized clinical trial of standard laparoscopic versus robot-assisted laparoscopic Nissen fundoplication for gastro-oesophageal reflux disease. Br J Surg 2006;93(November (11)):1351 9. [10] Gao C, Yang M, Wu Y, et al. Hybrid coronary revascularization by endoscopic robotic coronary artery bypass grafting on beating heart and stent placement. Ann Thorac Surg 2009;87(March (3)):737 41. [11] Advincula AP, Song A. The role of robotic surgery in gynaecology. Curr Opin Obstet Gynecol 2007;19(August (4)):331 6. [12] Reynolds RK, Advincula AP. Robot-assisted laparoscopic hysterectomy: technique and initial experience. Am J Surg 2006;191(April (4)):555 60. [13] Boggess JF, Gehrig PA, Cantrell L, et al. Perioperative outcomes of robotically assisted hysterectomy for benign cases with complex pathology. Obstet Gynecol 2009;114(September (3)):585 93. [14] Schreuder HW, Verheijen RH. Robotic surgery. BJOG 2009;116(January (2)): 198 213. [15] Fiorentino RP, Zepeda MA, Goldstein BH, John CR, Rettenmaier MA. Pilot study assessing robotic laparoscopic hysterectomy and patient outcomes. J Minim Invasive Gynecol 2006;13(January February (1)):60 3. [16] Diaz-Arrastia C, Jurnalov C, Gomez G, Townsend Jr C. Laparoscopic hysterectomy using a computer-enhanced surgical robot. Surg Endosc 2002;16(September (9)):1271 3. [17] Payne TN, Dauterive FR. A comparison of total laparoscopic hysterectomy to robotically assisted hysterectomy: surgical outcomes in a community practice. J Minim Invasive Gynecol 2008;15(May June (3)):286 9.