Table 1. Studies Examining the Feasibility of Telemanipulators.

Study	Type	Size	Results			Comment
Le Bret et al. France 2002	Surgical closure of patent ductus arteriosus. Observational prospective comparative.	Pediatric patients. Group 1: 28 videothoracoscopic technique, mean age 33 months Group 2: 28 robotically assisted (ZEUS), mean age 20 months	Total operating room time: 83.52 min vs. 162 min, p<0.01 Surgical procedure time: 24.24 min vs. 49.9 min, p<0.01 Conversion for technical failure or surgical problems: 0 vs. 1 conversion to classical thoracoscopy due to poor exposition caused by insufficient lung retraction. Accidents of dissection: 0 vs. 0 Postoperative There was no difference between the 2 groups in terms of ICU and hospital length of stay. ICU stay was less than 6 hours, and postoperative ventilation time was less than 2 hours. Reversible laryngeal nerve injury noted on one patient from each group. A persisting shunt was observed in 3 patients (1 in group 1 and 2 in group 2). The 3 children were reoperated on the same day by the thoracoscopic approach. In all patients, the persisting shunt was related to an incomplete dissection of the ductus and misplacement of the clips. No residual shunt noted at discharge. No wound infection. No hemorrhage observed. No midterm complications including recurrence of ductal shunting.			Operating room time and surgical procedure time significantly greater in the robotic group. “Robotic approach did not prove to be either superior or inferior to the videothoracoscopic technique in terms of safety, quality of outcome and reduction of complication. It appears more complicated, demanding and time consuming and presently has no particular advantage over the regular technique”. Lack of detail of how patients were allotted to the treatment groups. Children ranged in age from 3 weeks to 15 years. Robotic PDA surgery may have an advantage in very young/small infants compared with older, larger children.
D’Attellis et al. France 2002	Robotic assisted cardiac surgery: CABG or valve surgery. Case series.	20 patients. Mean (±SD) age 53±5 years. da Vinci robot.	15 patients (75%) were extubated within 6 hours and discharged from the cardiac surgery ICU on postoperative day 1. 2 patients (10%) were reexplored in the immediate postoperative period (1 for postoperative bleeding and 1 for revision of the coronary artery anastomosis). 2 additional conversions to thoracotomy (both were valve patients). Conversions were due to video system malfunction (n=1) and patient robot conflict (n=1). One reoperation at 6 months (rerepair of mitral valve by conventional surgery for recurrent bacterial endocarditis that was also preoperative) and 1 late death at 6 months (75 year old woman with preoperative NYHA IV heart failure with persistent low cardiac output after mitral valve repair). At 1 year follow-up, good functional results in 18 cases observed.			Patient positioning is important in decreasing patient robot conflicts.
			Intraoperative difficulties: Video system dysfunction Mammary artery bleeding Tool manipulation Patient robot conflict Conversion to thoracotomy	n=1 n=1 n=2 n=3 n=2
Marescaux et al. France 2001	Telerobotic laparoscopic cholecystectomies. Case series.	25 patients. Median age=59 years (range 28-81). ZEUS robot.	Cholecystectomies successfully performed on 24/25 patients. Median time for dissection =25 min (range 14-109). Median total time for set-up and takedown of the robotic arms was18 min (range 13-27). One conversion to conventional laparoscopic procedure in a patient with acute cholecystitis. In 3 cases, minor technical adjustments were made: 2 cases resulting from a nonfunctioning grasper and 1 case from a malfunctioning robotic arm sensor. Mean postoperative hospital stay was 3 days. Follow-up at 1 week and 1 month showed 1 patient with symptoms of reflux disease that had been present before surgery. Responded to medical treatment. One patient reported upper abdominal wall pain at a site distant from the port insertion site. Responded to conservative treatment.			Feasible.
Detter et al. Munich Germany 2002	Robotically assisted coronary artery surgery with and without cardiopulmonary bypass. Case series.	41 patients. ZEUS robot introduced step by step: IMA harvest n=12 patients Coronary anastomoses on arrested heart n=13 patients Coronary anastomoses on beating heart after median sternotomy n=6 patients Endoscopic CABG on arrested heart n=2 patients Endoscopic CABG on beating heart n=8 patients	IMA harvest ranged from 48-110 minutes and completed in all cases. Robotic anastomosis time averaged 21 min on the arrested and 25 min on the beating heart respectively. Endoscopic anastomosis was 41 min on the arrested heart and 36.5 min on the beating heart with an overall duration of surgery between 4.0 and 8.0 hours. One endoscopic case was intraoperatively converted to a MIDCAB with manual anastomosis. Total patency rate of all graft anastomoses was 97%. One patient underwent a reoperation with an uneventful postoperative course.			Feasible. Time to perform anastomoses was shorter in the sternotomy group than in the endoscopic group.
Bucerius et al. Leipzig Germany 2002	To evaluate postoperative pain levels after endoscopic versus conventional internal thoracic artery (ITA) dissection for minimally invasive direct CABG (MIDCAB). Results compared with pain levels associated with conventional CABG via a median sternotomy. Observational, prospective, comparative.	24 patients robotic ITA takedown (robotic MIDCAB) using da Vinci robot. 73 patients direct vision MIDCAB. 93 patients conventional CABG via a median sternotomy (CABG). Standardized questionnaire used to assess pain.	Overall pain levels were significantly lower in the robotic MIDCAB group vs. direct vision MIDCAB and CABG groups respectively (p<0.001). No significant difference between pain levels in direct vision MIDCAB and CABG. The only statistically significant difference in pain medication between the groups was for ibuprofen: robotic MIDCAB required less ibuprofen postoperatively compared with direct vision MIDCAB and CABG (p=0.001 and p=0.018 respectively).			Robotic MIDCAB may lead to reduced postoperative pain levels possibly due to less rib retraction. Robotic takedown of the ITA can also be performed by conventional endoscopic techniques. There was no discussion regarding the use of endoscopic, nontelemanipulator assisted MIDCAB compared with telemanipulator assisted MIDCAB. The authors stated that 48 of the patients receiving MIDCAB had been randomly assigned to either robotic ITA takedown (n=24) or direct vision MIDCAB (n=24). In addition, another 142 patients received routine surgical treatment, either direct vision MIDCAB (n=49) or conventional CABG (n=93) “as indicated by medical demands”.32 These 142 patients were combined with the previous 48 prospectively allotted patients in order to obtain the three treatment groups of interest. Therefore, the study groups are “unclean” since the direct vision MIDCAB group contains a mixture of patients from different sources. It is not clear when the 142 patients were enrolled, if they were historical controls or how they were selected.
Bentas et al. Frankfurt Germany 2003	Laparoscopic radical prostatectomy Case series.	40 consecutive patients. da Vinci robot.	Procedure was completed laparoscopically in all but two patients. Mean procedure time was 8.3 hours. Patients recovered rapidly after surgery with early oncological and functional results that were similar to those obtained with standard radical prostatectomy technique. No intra or postoperative deaths. 4 (10%) intraoperative adverse events and 1 reoperation. Trocar injury to an epigastric artery necessitated open revision on postoperative day 1. An intraoperative partial injury of the obturator nerve. Two instances of hemostatic complications at the dorsal vein complex. Complications Pulmonary embolism n=2 (5%) Deep vein thrombosis n=1 (3%) Obturator nerve injury n=1 (3%) Trocar injury to epigastric artery n=1 (3%) Venous plexus bleeding n=2 (5%) Urinary tract infection n=2 (5%) Prolonged anastomotic leak n=4 (10%)			Feasible.
Dogan et al. Frankfurt Germany 2002	Robotically enhanced totally endoscopic CABG on the arrested heart. Case series.	45 patients. Mean (±SD) age 63±6 years old. Consecutive single (n=37) or double vessel (n=8) operations. da Vinci robot.	Morbidity and Complications Conversion to minithoracotomy n=7 Conversion to full sternotomy n=3 Bleeding from the anastomosis n=2 Prolonged crossclamp time n=4 ITA injury n=1 Port access failure n=3 Hypovolemic shock n=1 Myocardial infarction n=1 Hypoxic brain damage n=1 Moderate reperfusion injury n=1 Retrograde aortic dissection n=1 Operating time: single vessel 4.2±0.9 hours, double vessel 6.3±1.0 hours. ICU stay: single vessel 24±21 hours, double vessel 74±64 hours. Hospital stay: single vessel 8.6±2.7 days, double vessel 15.4±6.4 days.			Feasible.
Kappert et al. Dresden Germany 2001	Robotic enhanced CABG. Case series.	201 patients. Median age 64±10.5 years. Group A, n=156. robotic system used to harvest the left or right IMA or both. Manual anastomoses performed via chest incision (MIDCAB or by the “Dresden Technique” REDTCAB). Group B, n=37. Harvest of the IMA and anastomoses performed totally endoscopically (TECAB). Group C, n=8. Robotic enhanced CABG via median sternotomy already preoperatively planned (open CABG). da Vinci robot.	99.4% survival rate. One patient died due to pneumonia on postoperative day 16. (Which Group?) 9 patients had to undergo reexploration due to bleeding. (Which Groups?) In Group B, 3 patients had an explorative second look due to increased postoperative drainage. Delayed wound healing at the site of the chest incision found only in Group A: 10 patients. No patients in Group B revealed any signs of delayed wound healing. Of patients with intent to treat TECAB, 19 (33.9%) were actually converted to a MIDCAB procedure. This was due to: LAD identification not possible endoscopically n=5 Diffuse sclerosis of the LAD n=5 Difficulties with endoscopic stabilization n=3 Pleural adhesions n=2 Intramural LAD course n=2 Insufficient occlusion of the LAD n=2 No significant differences in ICU or hospital stay were noted. ICU stay: 25.6±18.8 and 24.9±6.4 hours, A and B respectively. Hospital stay: 7±1 and 6±1 days, A and B respectively.			Feasible.
Bodner et al. Innsbruck Germany 2002	Robotic assisted cholecystectomies. Case series.	25 patients. Median age 48 (range 22-78). da Vinci robot.	Successful in 23 patients. Two procedures were converted to conventional laparoscopy due to system bread downs. Median operating time was 100 min (range 60 to 171). Median robot setup and dismantle time was 60 min (range 49 to 82). Operating room occupied for median of 160 min. “Intraoperative events” included: Serosal lesion of the colon n=1 Bleeding n=1 Perforation with leakage of bile into the abdominal cavity n=2 Gall bladder could not be clutched by the robotic instruments until 20 mL of bile removed by an assistant n=1 Diffuse nonsurgical bleeding from gall bladder bed or leakage of bile during gall bladder dissection caused a drain to be placed in subhepatic space via incision of accessory port at end of operation n=8 Redo operation due to bleeding at a port site on the second postoperative day n=1 No peri or postoperative mortality. Median hospital stay 4 days (range 2-15).			Feasible.
Isgro et al. Ludwigshafen Germany 2003	Internal mammary artery takedown. Case series.	56 patients. Mean (±SD) age 64.9±8.6 years. ZEUS robot.	One patient ITA takedown was completed conventionally. All harvesting performed without complications. Mean setup time 24±12 min. Mean IMA takedown time 58±17 min. IMA was patent in all 56 patients.			Feasible.
Giulianotti et al. Italy 2003	Robotics in “general surgery”. Case series.	193 patients underwent a minimally invasive robotic procedure. 207 robotic surgical operations performed (abdominal, thoracic and vascular). Mean age 55.9 years. da Vinci robot.	179 were single procedures. 14 were double procedures (2 operations on the same patient). 4 conversions to open surgery and 3 to conventional laparoscopy (conversion rate 3.6%; 7 of 193 patients). The 4 procedures requiring conversion to open surgery included: 1 Nissen fundoplication due to traumatic hepatic lesion caused by the retractor. 1 pulmonary lobectomy due to pleural adhesions 1 pancreatoduodenectomy due to neoplastic infiltration of the portal vein 1 total gastrectomy due to neoplastic infiltration of the pancreas. The 3 procedures converted to conventional laparoscopy included: A cholecystectomy and a splenic aneurysmectomy due to robotic technical problems. A Nissen fundoplication due to peritoneal adhesions. Perioperative morbidity rate was 9.3% (18 of 193 patients). 2 patients had iatrogenic lesions (1 hepatic tear due to blind retraction, and 1 splenic lesion due to tear by traction on adhesions). 6 patients (3.1%) required a reoperation. Postoperative mortality rate was 1.5% (3 of 193 patients). Two patients died due to septic complications caused by anastomotic leakage and mediastinitis after total esophagectomy, and Boerhaave syndrome after pancreatoduodenectomy respectively. The third patient who underwent subtotal gastrectomy died owing to respiratory failure after a reoperation for hemoperitoneum.			Feasible. “Best indications still have to be defined”. “This report could serve as a future prospective randomized trial.”
Ruurda et al. Netherlands 2002	Robot assisted laparoscopic cholecystectomies Case series.	40 patients. Median age 45 years (range 22 to 72) da Vinci robot.	One conversion to open procedure due to surgeons’ inability to expose the gall bladder sufficiently due to severe cholecystitis. Robot related technical problems occurred in 3 cases. The replaceable hook of the electrocautery instrument detached during the procedure. Hook removed laparoscopically in 2 of the 3 cases, but resulted in laparotomy in 1 case because the hook could not be seen in an obese patient. Median total hospitalization time was 2 days (range 1 to 10). No postoperative morbidity or mortality at the time of patient release and during short term follow-up (length of follow-up not stated).			Feasible.
Horgan and Vanuno USA 2001	Robotic laparoscopic surgery. (gastric bypass, Heller myotomies, nephrectomies, gastrojejunostomies adrenlaectomy, Nissen fundoplication, Toupet fundoplication, cholecystectomy) Case series.	34 patients. Gastric bypass for morbid obesity n=7 Heller myotomies for achalasia n=9 Donor nephrectomies n=11 Gastrojejunosteomies n=2 Adrenalectomy n=1 Nissen fundoplication n=1 Toupet fundoplication n=1 Cholecystectomy n=1 Pyloroplasty n=1 da Vinci robot.	“No robot related complications”.			Feasible.
Nifong et al. USA 2003	Mitral valve repair. Case series	38 patients. da Vinci robot.	No intraoperative deaths, strokes or device related complications. All patients had successful valve repairs. One patient required valve replacement for hemolysis and one patient was reexplored for bleeding. No incisional conversions. Two adverse events – one resulted in death at 20 days. One patient reexplored through the same incision 6 hours post surgery for pacing wire site bleeding and was discharged from the hospital 3 days later. One patient developed a leak that was directed against a prosthetic chord causing hemolysis. Had mechanical valve replacement 19 days post operatively through a median sternotomy. The patient had a fatal stroke 1 day after the valve replacement while on warfarin therapy.			Feasible.
Talamini et al. USA 2002	Robotically assisted gastrointestinal surgical procedures. Anti reflux n=25 Bowel resection n=18 Cholecystectomy n=8 Heller myotomy n=5 Splenectomy n=5 Exploratory laparoscopy n=4 Case series	60 patients da Vinci robot.	The conversion rate (either to standard laparoscopy or to open procedures) for the following procedures were reported: Anti reflux 12% Bowel resection 11% Cholecystectomy 12% Heller myotomy 20% Splenectomy 60% Exploratory laparoscopy 25% No operative deaths. Postoperative complications attributable to the robot by the authors: Trocar slippage n=4 Arm positioning n=2 System positioning n=2 (conversion to standard laparoscopy in 1 case) System failure n=2 (conversion to standard laparoscopy in 1 case) Dropped cautery hook n=1 Postoperative complications attributable to the operation and not the robot by the authors: Misshapen Nissen n=1 Gastric Leak n=1			Feasible.
Tewari et al. USA 2003	Robot assisted prostatectomy compared with radical retropubic prostatectomy. Prospective, observational, single centre.	100 consecutive patients retropubic prostatectomy (RRP) (reference standard). Mean age 63.1 years (range 42 to 72). 200 consecutive patients robotic assisted prostatectomy (VIP). Mean age 59.9 (range 40 to 72) da Vinci robot. Inclusion criteria: Surgical candidates. Choice between RRP and VIP was offered to all patients who had a 10 year life expectancy and had prostate cancer of Gleason score ≥6. Patients assigned on basis of their personal preferences. Follow-up was for 556 days in the RRP group and 236 days in the VIP group. No explicit explanation was provided regarding this difference.	No deaths in either treatment arm. The percentage cancer, Gleason scores and pathological states were comparable between either treatment arm. Mean operative duration comparable and not significantly different. The number of catheterization days was significantly different between RRP and VIP groups. 15.5(7-28) vs. 7(1-18), p<0.05. Blood loss was 910 and 150 mL for RRP and VIP respectively and transfusion was greater after RRP (67% vs. none; both p<0.001). Four times as many complications after RRP (20% vs. 5%), p<0.05. Hospital stay longer for RRP 3.5 vs. 1.2 days, p<0.05. 93% of VIP and none of RRP patients were discharged within 24 hours, p<0.001. The 50% return of continence occurred in 160 days (RRP) vs. 44 days (VIP), p<0.05. The 50% return of erection occurred in 180 days (VIP) vs. 440 days (RRP), p<0.05.			One team performed VIP, however, 8 different surgeons in the same institution undertook the RRPs. Different lengths of follow-up between treatment arms.
			Complications Aborted Conversion Rectal injuries Postoperative ileus Wound dehiscence/hernia Postop. fever/pneumonia Lymphocele Obturator neuropathy DVT Postop. MI Postop.bleeding/reexploration Total	RRP 1 - 1 3 1 4 2 2 1 1 4 20	VIP 2 0 0 3 2 0 p<0.05 0 0 1 0 1 5 p<0.05
Perez et al. USA 2003	Robotically assisted laparoscopic cholecystectomy. Case series	20 patients. Mean age 47±4 years. da Vinci robot.	All patients had successful procedures without complications or need for conversion to conventional laparoscopic cholecystectomy. Mean procedure time was 152±8 min. The large proportion of operating time associated with the robotically assisted surgery was related to robotic positioning and adjustments rather than surgeon directed tissue manipulation.			Feasible.
Melvin et al. USA 2002	Antireflux surgery. Robotic enhanced fundoplication compared with standard laparoscopic control procedures. observational historical controls.	20 consecutive patients entered into each treatment group. da Vinci robot.	Operative times were significantly longer in the robot group (97 vs. 141 minutes. No complications and most patients went home the first postoperative day. Length of follow-up was 11.2 months for the laparoscopic group and 6.7 months for the robotic group (p<0.001). At follow-up, symptoms were similar in both groups; however, there was a significant difference in the number of patients taking antisecretory medication - none in the robotic group but 6 in the laparoscopic group reported regular usage (p<0.001). None of the 6 patients taking daily medication reported symptoms of heartburn while on medication.			At current level of development, robotic enhanced fundoplication appeared to offer no clear advantages in operative outcomes compared with standard laparoscopic approaches.
Damiano et al. USA 2001	Robotically assisted CABG Prospective, multicentre case series	32 consecutive patients. Mean age 63±9 years. ZEUS robot.	30/32 patients available for late follow-up. No intraoperative device related complications. 3 intraoperative graft revisions. 3 patients required a return to the operating room the evening of surgery for excessive mediastinal hemorrhage. 3 patients had postoperative atrial fibrillation. The average length of stay in the intensive care unit was 1.3±1.0 days. The average hospital stay was 5.5±2.7 days. 8-12 weeks after operation, 26/28 grafts were patient. At a mean follow-up of 16±4 months, 28/30 patients were doing well.			Feasible. “Further clinical trials are warranted to explore the potential of this new technology and establish its precise role in the treatment of patients with coronary artery disease”.