Abstract
Background/Aim: This study aimed to evaluate the learning curve and perioperative outcomes of robot-assisted hysterectomy (RAH). Patients and Methods: We retrospectively analyzed data from 45 patients who underwent RAH using the da Vinci Xi surgical system. The learning curve was evaluated using the cumulative summation method. Demographic data and various perioperative parameters, including total operative time, docking time, and console time, were obtained from the medical records. Results: Cumulative summation analysis indicated that proficiency regarding hysterectomy time was reached after 33 cases. There were two unique phases of the learning curve for console time: the introduction phase identified by the bottom point in the curve, and the proficient phase, identified by an upward line after the bottom point in the curve. There were no significant differences between the two phases in terms of patient age and body mass index. Total operative time, docking time, and console time were significantly decreased in the proficient phase compared with those in the introduction phase. There was a significant reduction in blood loss during operation in the proficient phase. The perioperative complication rates were 12.1% in the introduction phase and 0% in the proficient phase (p=0.5606). No blood transfusion or conversion to laparotomy was required in either phase. Conclusion: The introduction and proficient phases identified by cumulative summation analysis demonstrated progressive improvement of surgical performance in surgeons carrying out RAH.
- Robot-assisted hysterectomy
- cumulative summation analysis
- learning curve
- endometrial cancer
- benign gynecological tumor
Minimally invasive surgery is a rapidly developing and expanding field in gynecology. Robot-assisted surgery was approved in the United States in 2005, approximately 15 years after the first use of total laparoscopic hysterectomy (TLH) in 1988 (1, 2). Robot-assisted hysterectomy (RAH) for early-stage endometrial cancer and benign tumors was first covered by insurance in Japan in April 2018, in line with a revision of the medical payment system. A proctor system (surgical instructor) was developed by relevant academic societies to introduce and guide robot-assisted operations; however, facilities remain in the trial-and-error stage, with safety being the priority.
Robot-assisted surgery has several advantages, including magnified vision with a robotic camera, greater freedom of forceps movements, motion scaling, and a vessel-sealing system, and these devices have been associated with many successful procedures (3). Several randomized clinical trials have compared robot-assisted with laparoscopic surgery, with varying perioperative outcomes depending on the facility and surgeon (2-6). Robot-assisted surgery has a steep learning curve; however, team training improves the quality of the operations, because surgical assistants can perform trocar placement, docking, replacement forceps, hemostasis support, and conversion to open surgery (7-11).
The da Vinci Xi Surgical System® (Intuitive Surgical, Inc., Sunnyvale, CA, USA) was introduced in our department in October 2018 and robotic surgery was actively introduced in the same month. In this study, we reviewed the smooth introduction of robot-assisted surgery and compared the surgical outcomes of robotic-assisted total hysterectomy (RAH) to treat malignant and benign uterine tumors, before and after the attainment of surgical proficiency.
Patients and Methods
This was a retrospective study of 45 patients who underwent RAH between October 2018 and May 2021 in Fukuoka university hospital. Surgery was performed to treat uterine fibroids and adenomyosis (benign uterine tumors), and stage IA endometrial cancer with endometrioid adenocarcinoma grade 1/2 (G1/G2) (malignant uterine tumors) (International Federation of Gynecology and Obstetrics, FIGO 2008). The study was approved by the ethics committee of Fukuoka university hospital (approval number: U22-02-013), and written informed consent for the procedures and publication of this study was obtained from all patients.
RAH was carried out using the da Vinci Xi Surgical System® (Intuitive Surgical, Inc.) by four Intuitive Surgical-certified surgeons who conducted or assisted with the procedure. We invited an expert to provide guidance during the introduction of the technique. Patients were placed in a supine position with their legs open and head-down tilt (20-23 degrees). Surgery was performed using side docking from the patient’s right caudal side. Because of the changes in ocular pressure caused by the low head position, all patients were examined by an ophthalmologist before surgery to exclude the presence of glaucoma. Patients with a history of cerebral aneurysms were not eligible for the study. The average pneumoperitoneum pressure was 8-l0 mmHg with a maximum flow of 20 ml/min. For RAH, four 8-mm-diameter robotic working ports were placed near the navel on the abdominal wall and docked from the first to the fourth arms. The distance between the ports was standardized at 7.0-8.0 cm but adjusted for body shape to avoid the robot arms colliding. We also installed a 12-mm-diameter assist port to the left of the first arm to help during surgery and for placement of the surgical thread, needle, and gauze. Fenestrated bipolar forceps were used on the first arm, a 0-degree camera on the second arm, monopolar curved scissors, or vessel sealer on the third arm, and Cadiere forceps on the fourth arm.
The first step involved inserting a uterine manipulator (RUMI II Koh colpotomiser system®; UMH650; Ken Medical). If insertion of the uterine manipulator was difficult, e.g., due to cervical stenosis, we carried out vaginal incision using a Vagi-pipe® (AP-158-D; Hakko Medical). We sampled ascites in the abdominal cavity or peritoneal wash from saline lavage for peritoneal cytopathology before inserting the uterine manipulator in patients with endometrial cancer, after cauterizing both fallopian tubes. The round ligament was then coagulated and dissected, and the ureter and uterine artery were identified by exposing the retroperitoneal space. The infundibulopelvic ligament (or ovarian ligament in case of ovary preservation) was coagulated and dissected after isolation of the uterine artery for coagulation and dissection. The uterosacral ligaments were then coagulated and dissected, after dissection of the rectouterine pouch, and the vesicouterine pouch was then dissected, the bladder was sufficiently peeled to the caudal side of the vaginal dissection line, and the cardinal ligaments were ligated for dissection. We processed as close to the pelvic wall as possible up to the vaginal amputation line. Finally, we made an incision around the vaginal wall using the Koh Cup™ or Vagi-pipe® as a guide and extracted the uterus transvaginally. The vaginal stump was closed using an absorbable suture.
Patient age and body mass index (BMI) were recorded. The learning curve for RAH was analyzed using the cumulative summation (CUSUM) method, with the lowest point of the learning curve separating the introduction and proficiency stages. We investigated total operation time from skin incision to skin closure, docking time from placement of the trocar to the moment at which all the robotic arms were successfully connected to the trocar, console time from round ligament dissection to vaginal cuff closure, intraoperative blood loss, blood transfusion requirements, and complications for up to 6 months following surgery for each period. CUSUM analysis was used to quantitatively assess the learning curve for console time, as shown in previous studies (12-14). Non-parametric variables were analyzed by the Mann–Whitney U-test, and differences between proportions were compared using Fisher’s tests. A p-value <0.05 was regarded as significant. Statistical analysis was carried out using Prism 6® (GraphPad Software Inc., San Diego, CA, USA).
Results
The patient demographics and operative characteristics are shown in Table I. All cases were completed robotically, without blood transfusion or conversion to conventional laparoscopy or laparotomy.
Overall patient characteristics (n=45).
The raw console time was plotted against increasing case number (Figure 1A) and the learning curve was drawn by plotting the cumulative sequential differences between each console time point and the process average over time. The fitting model formula of the learning curve was: CUSUM=0.2472×2–10.791×–54.942 min (where x represents the case order). The goodness-of-fit was R2=0.935. This learning curve was divided into two distinct phases by CUSUM=33 (Figure 1B): the negative slope for the initial 33 cases, with an average less than the process average, was defined as the introduction phase (Figure 2A; linear regression y=–3.3041×–91.083; R2=0.2561), and the positive slope of the next 12 cases, with an average above the average, was defined as the proficient phase (Figure 2B; linear regression y=13.828×–631.27; R2=0.7441). Increased surgical competence was therefore demonstrated after the first 33 cases.
Learning curve of robot-assisted hysterectomy analyzed by cumulative sum (CUSUM). (A) Raw console time plotted against chronological case number. (B) Console time plotted against each case number (solid line). CUSUM curve of best modelled fit for the plot (dashed line). The CUSUM value of 33 divided the learning curve for console time into introduction and proficient phases.
Best-fit lines for introduction and proficient phases. (A) The introduction phase represents the initial learning curve and (B) the proficient phase represents increasing surgeon’s competence after the initial 33 cases. CUSUM: Cumulative sum.
We compared the patient characteristics and perioperative outcomes between the two phases identified by CUSUM analysis (Table II). There was no significant difference in terms of patient age (p=0.8147), BMI (p=0.3774), or numbers of patients with early-stage endometrial cancer with G1/G2, uterine fibroids, and adenomyosis (p>0.95). Total operative time decreased significantly during the proficient phase compared with the introduction phase (255 vs. 330 min, respectively; p=0.002). The docking time was significantly shorter in the proficient phase compared with the introduction phase (38 vs. 48 min, respectively; p=0.0009) and the median console time was also significantly shorter in the proficient phase than in the introduction phase (120 vs. 147 min, respectively; p=0.0346). Blood loss was significantly lower during the proficient phase compared with the introduction phase (20 vs. 70 ml, respectively; p=0.0473). Intraoperative and postoperative complications occurred in four patients (12.1%) in the introduction phase, but none in patients in the proficient phase. There was only one intraoperative complication, involving an approximately 2 cm laceration of the bladder, which was repaired immediately during surgery. Two patients with paralytic ileus (Clavien–Dindo grade I) required short-term restriction of oral food and fluids and intravenous electrolyte administration. One patient had a urinary tract infection (Clavien–Dindo grade II), which was treated with intravenous antibiotics.
Interphase comparisons of patient characteristics and perioperative outcomes.
Discussion
The learning curve for robot-assisted surgery is steeper than that for laparoscopic surgery, which stabilizes after 20-30 cases in terms of operation time, docking, and time to place ports (9). These parameters stabilized after carrying out 33 procedures in our department. However, it has been suggested that 50-100 operations are required to become fully proficient in robot-assisted operations, with serious complications occurring in approximately 30 cases (15-17). Utmost care and attention are therefore required when performing robot-assisted surgery.
The average time for laparoscopic surgery is 75±21 min, which is significantly less than that for robot-assisted surgery (106±29 min), according to two surgeons with experience of 500 laparoscopic surgeries and more than 30 robot-assisted operations (18). Five other surgeons who had performed 75-400 laparoscopic surgeries and more than 20 robot-assisted operations reported an average duration of laparoscopic surgery of 103±64 min (19), which was much less than the average time for robot-assisted operations (173±89 min). Laparoscopic surgery is thus quicker when performed by surgeons experienced in laparoscopic surgery. However, one study reported median times of 104 and 76 min for laparoscopic and robot-assisted surgery, respectively, although the surgeons’ experience levels were not reported (20). In addition, previous studies found no significant differences between RAH and TLH in terms of operative duration, bleeding volume, and hospitalization time (21-23). However, a recent study found that RAH outperformed TLH in terms of perioperative outcomes (24-26), particularly in obese patients with a BMI ≥30 kg/m2 (27, 28). The current study only included four patients with a BMI ≥30 kg/m2, and further research is therefore needed to determine the utility of RAH in obese patients in our hospital.
A previous study reported an increase in complications during the introductory period of RAH (29); however, RAH was introduced safely at our hospital for the treatment of early-stage endometrial cancer and benign gynecologic tumors for several reasons: (i) we adhered strictly to the Japan Society of Obstetrics and Gynecology’s “guideline for robot-assisted operations for gynecological diseases” (30), (ii) we assigned the same surgeons and assistants to standardize the operation, and the RAH team was thus well-trained, with co-operation from anesthesiologists, nurses, and clinical engineers from the surgical department, and (iii) routine simulations were performed to increase safety awareness and manage potential intraoperative issues.
Although robot-assisted surgery has several advantages, such as improved forceps operability, a clear field of view, and educational tools, the equipment and consumables are expensive. However, since the procedure has been licensed, we anticipate that domestic and international companies will participate to decrease the costs and accumulate evidence to improve the insurance score.
This study had the limitation of being a retrospective study with a small number of cases. Given that robot-assisted surgery is likely to replace minimally invasive surgeries for gynecologic tumors in the future, further studies of the oncological prognosis are needed, including large numbers of patients, and considering the surgeons’ and assistants’ education and experience, to ensure the safety and effectiveness of the procedures.
The current study indicated that robot-assisted surgery was introduced successfully in our department; however, surgeons’ and assistants’ education and experience must be taken into consideration, as well as continuing to increase the number of cases to ensure its safety and effectiveness.
Acknowledgements
The Authors would like to thank all the hospital physicians for their invaluable help. The Authors also thank Susan Furness, PhD, from Edanz (https://jp.edanz.com/ac) for editing a draft of this manuscript. This work was supported in part by JSPS KAKENHI Grant Number 18K09242 and a Grant-in-Aid from the Kakihara Science and Technology Foundation (Fukuoka, Japan) to Shingo Miyamoto.
Footnotes
↵* These Authors contributed equally to this work.
Authors’ Contributions
Conception and design of the study: Fusanori Yotsumoto, Ayako Sanui and Shingo Miyamoto; Acquisition of data: Fusanori Yotsumoto, Ayako Sanui, Tomohiro Ito, Daisuke Miyahara, Kenichi Yoshikawa, Koichiro Shigekawa and Yukiko Noguchi; Analysis and interpretation of data: Fusanori Yotsumoto and Ayako Sanui; Drafting of the article: Fusanori Yotsumoto, Ayako Sanui and Shingo Miyamoto; Critical revision of the article for important intellectual content: Shin’ichiro Yasunaga and Shingo Miyamoto. All Authors have read and approved the final version of the manuscript.
Conflicts of Interest
All Authors have no competing interests to declare that are relevant to the content of this article.
- Received May 20, 2022.
- Revision received May 29, 2022.
- Accepted May 30, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.