Several studies have shown the safety and feasibility of robot-assisted antireflux surgery but comparative data are lacking.
Several studies have shown the safety and feasibility of robot-assisted antireflux surgery but comparative data are lacking.
Fifty consecutive patients scheduled for laparoscopic antireflux surgery were randomized into two groups. Twenty-five patients underwent robot-assisted surgery and 25 standard laparoscopic fundoplication. All robot-assisted procedures were performed with the da Vinci Surgical System®.
There were no significant differences in age, sex, body mass or preoperative reflux pattern between the groups. Operating times were significantly longer for robot-assisted than standard laparoscopic operations (mean total operating time 131·3 versus 91·1 min, P < 0·001; skin-to-skin time 78·0 versus 63·5 min, P = 0·001). There was no conversion to open surgery. Conversion to standard laparoscopy was necessary in one of 25 robot-assisted procedures. The length of hospital stay was similar in both groups. Robot-assisted surgery was associated with significantly higher mean total costs (€3157 versus €1527; P < 0·001). There were no significant differences in clinical, endoscopic and functional outcomes between groups. There was no procedure-related mortality.
Robot-assisted laparoscopic fundoplication is comparable to the standard laparoscopic procedure in terms of feasibility and outcome, but costs are higher owing to longer operating times and the use of more expensive instruments. Copyright © 2006 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd.
Robotic technology has recently been introduced into laparoscopic clinical practice with the aim of improving surgical performance by eliminating tremors and fatigue. In 1997 Cadière et al.1 performed the first laparoscopic procedure using a robot; laparoscopic cholecystectomy was accomplished using the da Vinci Robotic Surgical System® (Intuitive Surgical, Mountain View, California, USA). This device has the benefit of three-dimensional vision2. A variety of clinical applications of the da Vinci® system have been described3, 4 including feasibility in laparoscopic antireflux surgery1, 4–18. Data comparing robotic fundoplication with the standard laparoscopic procedure are limited to two retrospective studies3, 19, one prospective non-randomized series20 and a randomized trial that contained only ten patients in the robotic group21. Postoperative functional results have never been reported.
After training comprising ten laparoscopic robot-assisted procedures (cholecystectomy, fundoplication and Heller's myotomy), a randomized clinical trial was initiated to compare the feasibility and clinical results of robot-assisted fundoplication using the da Vinci® system with those of traditional laparoscopic fundoplication in a consecutive series of patients.
Patients with gastro-oesophageal reflux disease (GORD) were randomized to undergo standard or robot-assisted laparoscopic Nissen fundoplication. Inclusion criteria were clinical GORD that necessitated surgery according to the criteria of Hinder et al.22 and an American Society of Anesthesiologists (ASA) score of I–II. Exclusion criteria were giant hiatal hernia (larger than 6 cm on preoperative barium meal), ASA score III–IV, previous upper abdominal surgery and contraindications to pneumoperitoneum. The study was approved by the institutional ethics committee and all patients gave written consent. Randomization was performed on the day before surgery by means of sealed opaque envelopes containing computer-generated random numbers.
All patients completed a standardized preoperative symptom assessment form using the Gastro-oesophageal Reflux Health-Related Quality of Life scale (GORD-HRQOL)23. Preoperative assessment included endoscopy, barium oesophagography, manometry and 24-h pH monitoring (De Meester score and total H+ area24). If duodenogastro-oesophageal reflux was suspected at pHmetry, patients underwent 24-h bilimetry and scintigraphy (six patients in the standard laparoscopy group and eight in the robot-assisted group).
In both groups, patients were placed in a leg-abducted, mild reverse Trendelenburg position. A standard five-cannula technique was used for conventional laparoscopy25.
The da Vinci® system was used for robot-assisted procedures. This consisted of a robotic manipulator with three arms (the central arm held the camera and the two outer arms the surgical instruments) and a remote console, where the operating surgeon sat during the procedure. Three surgeons, all proficient in laparoscopic procedures, were involved in all operations. The operating surgeon controlled the robot at the console; the first and second assistants placed the trocars once the pneumoperitoneum had been created, connected the arms and handled the laparoscopic instruments inserted in the accessory trocars. Instruments were semidisposable; the robot tracked the number of times they were used and would not operate an instrument after the tenth use. Most dissection was done at a 2 : 1 or 3 : 1 scale. A 30° angled scope was used in all cases.
After a 12-mmHg pneumoperitoneum had been created with a Veress needle, three robotic trocars (one 12-mm disposable for the camera, two 7-mm non-disposable) were inserted in the supraumbilical position, and in the left and right midclavicular line below the costal margin. Two additional trocars were placed for exposure. One (10 mm) was inserted in the right anterior axillary line below the costal margin and was used to introduce the liver retractor. The other (12 mm) was inserted in the left anterior axillary line and used for accessory instruments. Operative steps were similar to those of the laparoscopic technique25. In both groups, a ‘short and floppy’ 360° fundoplication26, 27 was performed and a crural repair was added whenever deemed necessary. The short gastric vessels were never divided. The wrap was not sutured to the oesophagus or to the diaphragm.
Standard protocols for anaesthesia, and preoperative and postoperative management were used. Analgesia comprised ketorolac and tramadol during the first 12 h after surgery and at the request of the patient thereafter. Care was taken to prevent vomiting or coughing in the early postoperative course, with appropriate pharmacological therapy in both groups.
Follow-up involved personal interview at 1, 3, 6 and 12 months after surgery. Patients were questioned specifically about heartburn, regurgitation, retrosternal pain or dysphagia. Outcomes were graded using the GORD-HRQOL. Patients underwent oesophageal manometry and 24-h pH monitoring 3 months after surgery and endoscopy at 6 months.
Variables analysed were patient age, sex, body mass index (BMI), preoperative manometric and 24-h pH data, postoperative complications, length of hospital stay, in-hospital costs and functional results at 3 and 6 months after surgery. Total operating time (including anaesthesia) and operative skin-to-skin time were noted, along with trocar positioning time, hiatal dissection time, hiatal pillar closure time and time taken to suture the wrap. The set-up time needed for wrapping the arms and positioning the robot cart was also recorded.
Cost evaluation included the use of the operating room (nursing and technical staff, surgical devices and maintenance), and the cost of surgical tools (disposable instruments, trocars and wires) and the hospital stay. The initial cost of the robot was not included in the analysis.
The primary endpoint of the study was in-hospital cost of the procedure. Secondary endpoints were skin-to-skin and total operating time. The required sample size was calculated based on the assumption of a difference of 30 per cent in costs between robot-assisted and standard laparoscopic fundoplication, and a difference of 20 per cent in skin-to-skin and total operating times.
Categorical variables were compared by χ2 test, with Yates' correction and Fisher's exact test (two tailed) as necessary. Continuous variables were compared by Student's t test or the Mann–Whitney U test, depending on distribution. All P values were two sided. P < 0·050 indicated a statistically significant difference. Data were analysed on an intention-to-treat basis. All calculations were done with SPSS® version 10.0 (SPSS, Chicago, Illinois, USA).
Between February 2002 and February 2004, 88 patients with GORD underwent laparoscopic fundoplication. Thirty-eight patients were excluded from the study for various reasons (Fig.1). Fifty consecutive patients were enrolled in the study, 25 (seven women, 18 men) who underwent traditional laparoscopic fundoplication and 25 (six women, 19 men) robot-assisted laparoscopic fundoplication. There were no significant differences between the groups in terms of sex, age, mean BMI, manometric or 24-h pH data (Table1).
|Sex ratio (M : F)||19 : 6||18 : 7||—|
|Body mass index (kg/m2)||25·5(2·9) (21·2–32·7)||26·1(2·3) (21·3–31·9)||0·372|
|Age (years)||43·0(12·8) (26–76)||46·3(11·3) (26–69)||0·336|
|Mean LOS pressure (mmHg)||8·1(2·2) (4–13)||7·9(2·5) (5–13)||0·797|
|De Meester score||40·6(23·0) (12·8–74·6)||48·1(22·8) (15·4–81·7)||0·309|
|No. of acid refluxes||71·7(38·7) (11–175)||82·1(41·2) (40–169)||0·353|
|No. of refluxes >5 min||4·2(3·3) (0–11)||5·7(4·2) (0–15)||0·164|
|Total H+ area||354(154·8) (52·1–594·8)||408·0(161·3) (50·7–651·7)||0·396|
There were no intraoperative complications in either group. All standard operations were completed laparoscopically. Conversion to standard laparoscopic surgery was required in one patient in the robot-assisted group because of difficulty in pursuing the dissection by robotic techniques with a prolonged operating time.
Operating times are shown in Table2. Total operating time and skin-to-skin time were significantly shorter for conventional laparoscopy. The total operating time and skin-to-skin time for the first ten robot-assisted procedures were no different from those for the final ten (mean 138·2 versus 132·3 min, and 75·2 versus 80·1 min). Crural repair was performed in 20 patients in the standard laparoscopy group and 19 in the robot-assisted group.
|Robot set up (min)||23·1(6·5) (12–39)||—|
|Trocar positioning (min)||16·2(6·5) (6–29)||11·6(3·5) (8–18)||0·003|
|Hiatal dissection (min)||24·8(12·7) (12–60)||29·2(8·7) (13–48)||0·165|
|Hiatal repair (min)||11·1(4·2) (5–23)||9·9(2·4) (6–14)||0·283|
|Suture of the wrap (min)||19·5(5·6) (10–30)||14·9(3·5) (9–21)||0·001|
|Skin-to-skin time (min)||78·0(17·5) (48–104)||63·5(13·3) (46–84)||0·001|
|Total operating time (min)||131·3 (18·3) (90–162)||91·1(10·6) (72–106)||<0·001|
There were no postoperative complications with either approach. Mean hospital stay was 2·9 (range 2–6) days in standard laparoscopy group and 3·0 (range 2–7) days in the robot-assisted group (P = 0·588).
Mean follow-up was 22·3 (range 6–32) months. At 1 month, three patients in each group had mild transient dysphagia. No clinical differences between the two groups were found using the GORD-HRQOL at 3, 6 and 12 months. None of the 50 patients had endoscopic oesophagitis at 6 months, although Barrett's metaplasia remained present in all patients in whom this was diagnosed before surgery. Lower oesophageal sphincter resting pressures were similar after both procedures, with a mean of 21·8 mmHg after robot-assisted surgery and 22·3 mmHg after standard laparoscopic fundoplication (P = 0·503). Postoperative 24-h pH values were normalized in all patients, and there were no significant differences between the two groups; mean De Meester scores were 5·8 (range 0–11) and 4·2 (range 0–12) respectively (P = 0·231), and total H+ area 22·1 (range 1·0–69·9) and 17·1 (range 0·7–68·2) (P = 0·423).
The cost of a robot-assisted procedure was significantly higher than that for standard laparoscopic fundoplication (Table3).
|Unit cost||Robot assisted||Laparoscopic|
|Time||Cost (€)||Time||Cost (€)|
|Total operating time||367 € per h||131·3 min||803||91·1 min||557|
|Hospital stay||300 € per day||3·0 days||900||2·9 days||870|
Several studies have shown the feasibility and safety of robot-assisted antireflux surgery5–17. Only two prospective studies have compared robot-assisted with standard laparoscopic fundoplication. In a non-randomized clinical trial that included 20 patients, Melvin et al.20 reported significantly longer operating times in the robotic group (141 versus 97 min; P < 0·001). Morbidity and postoperative hospital stay were similar. During 7 months of follow-up there was a significant difference in the number of patients needing regular antisecretory medication (0 per cent in robotic group versus 30 per cent in control group). In a randomized trial of ten robotic versus ten conventional laparoscopic fundoplications21, operating time was significantly longer for robotic procedures (76 versus 52 min; P < 0·010), but morbidity rates (10 versus 9 per cent), blood loss (10 ml in both groups) and hospital stay (both 1 day) were similar. There were no deaths or conversions in either group. Function and quality of life were not studied.
The present trial, although based on a limited number of patients, included an evaluation of clinical, endoscopic and functional benefits and drawbacks of robot-assisted laparoscopic antireflux surgery. In agreement with published data3–21, the results confirmed that robot-assisted laparoscopic Nissen fundoplication is comparable to traditional laparoscopy in terms of complications, mortality and length of hospital stay.
The main disadvantages of robot-assisted laparoscopic fundoplication were prolonged operating time and increased cost compared with standard laparoscopic fundoplication. Because operating time did not decrease with experience a learning curve effect seems unlikely. Prolonged operating time has been reported in all published series of robot-assisted abdominal surgery including cholecystectomy28, adrenalectomy29, colonic resection30 and fundoplication20, 21, other than one retrospective study of antireflux surgery3. In the present study the prolonged operating time was mainly due to the robot set-up time, the more difficult trocar positioning and increased time taken to suture the wrap.
The da Vinci® system has certain disadvantages that may reflect current technological limitations. The lack of tensile feedback is a significant limitation and seems to require an ‘intuitive method’ of hand–eye co-ordination to improve hand movements2. The range of robotic instruments available is limited, because the system was designed principally for use in cardiac surgery and switching instruments is laborious.
Another drawback is the higher cost of the robotic procedure. The initial cost of the robotic system was not taken into consideration, yet there were still increased expenses, mainly associated with the semidisposable robotic instruments. Although no economic benefit could be shown for the robot-assisted approach, multidisciplinary use of the system (for example in urology, cardiac, laparoscopic and thoracic surgery) would result in a faster realization of return on the initial investment.
In agreement with other studies29, 31, 32, current robotic systems are not of significant benefit to routine surgical practice. Developments in robotic technology such as new instruments, additional but smaller robotic arms and tactile feedback could, however, prompt re-evaluation of its use in antireflux surgery.