This systematic review and meta-analysis identified 13 studies that compared RAL with CL in terms of perioperative outcomes of endometriosis surgery. The quantitative synthesis of our findings confirmed that RAL had no numerical advantages over CL in the aspects studied. Moreover, we found it to be inferior in terms of operating room and operative times. Subgroup analysis based on the pattern of endometriosis was not feasible with the available data.

In addition to previous studies, we examined operating room time as a new outcome [27,28,29]. However, we obtained similar results for all other perioperative outcomes. Our results did not show the expected benefit of RAL over CL in terms of intra- and postoperative complications, estimated blood loss, number of rehospitalizations, or days spent in hospital, and we observed even longer operative times by approximately 30 min. The latter can be attributed to an average docking time of approximately 18.2 min [30]. Operating time was found to be the most significant factor associated with postoperative complications and length of postoperative hospital stay. Magrina et al. found that for every additional 60 min of operating time, the odds of postoperative complications and prolonged hospital stay increased the chances by 57% and 103%, respectively [24]. This is partly explained by the disproportionate distribution of patients. In some articles, more radical procedures (e.g., endometriosis surgery with concomitant hysterectomy) were performed and patients with more advanced endometriosis were operated on with RAL according to the revised American Society for Reproductive Medicine (rASRM) staging. This might indicate that the surgeon favored a robotic approach and might have added bias to the results, contributing to differences in operating times [24, 31]. However, we did not find significant differences in rASM classification between the two approaches. (Supplementary Figures S17-S20) On the other hand, the experience of Nezhat suggests that procedures for the treatment of severe disease require multiple camera and instrument exchanges, making CL easier to perform [23].

It should be noted that the only RCT conducted by Soto 2017 found the mean operative time and blood loss within the range of time and volumes previously reported by other non-randomized studies. This suggests that their findings are unlikely to be related to patient selection and the experience of the surgeons or the team with various platforms [32]. Not surprisingly, in the articles describing more severe cases (e.g., bowel (deep infiltrating endometriosis (DIE), rASRM stage III/IV.), an even longer operative times could be observed compared to the mean difference we reported. Similar to the operative time, we obtained a significant difference of approximately 50 min in the operating room time. This difference could also be attributed to the necessary preparation procedures of the robot in addition to factors described influencing the length of the surgery time.

Intraoperative complications were crucial in determining intra- and postoperative outcomes, such as operative time (thus, operating room time), expected blood loss, the likelihood of conversion to open surgery, days spent in hospital, and postoperative complications. Most studies showed relatively low numbers of intraoperative and postoperative complications, indicating that both methods were safe and neither seemed to be superior in terms of complication rates. It should be noted that Carpentier et al. only operated on bladder DIE, and the relative frequency of postoperative complications in the RAL group was 60% versus 36% in the CL group [32]. Conversion to open surgery depended on several factors, including the previous abdominal surgeries of the patient and unexpected technical events, among other factors. However, the level of experience of the surgeon was a factor that needs to be highlighted.

Other meta-analyses had also been conducted on this topic; however, due to the low number of cases and methodological problems, we considered it necessary to conduct another meta-analysis. In the meta‐analysis by Chen (2016), RAL was compared with CL for endometriosis surgery; no difference was found in most aspects, except for operating time [27]. In 2020, Restainato et al. and in 2018, Balla et al. performed a meta-analysis and found no difference in the operating time or complication rates between robot-assisted and conventional laparoscopy. However, in the latter study, in which all patients underwent colorectal resection due to endometriosis, only a small fraction (1.7%) of the procedures were performed with RAL. Moreover, complications were not evaluated separately for RAL and CL [28, 29].

Our study showed that RAL did not offer a quantifiable advantage in the day-to-day surgical management of patients with endometriosis. However, the reality is more nuanced; an important finding is that longer operative time has been correlated with increased overall costs strongly associated with the robotic platform [23]. As for costs, no data were found for endometriosis surgeries; however, such data were reported in closely related areas. A database study of 36,188 patients showed that robotic hysterectomy was more expensive than laparoscopic hysterectomy ($9,640 vs. $6,973, P < 0.01). In gynecological oncology, for endometrial or cervical cancer, the extra cost of using RAL was €1,456 per intervention [33].

Le Gac et al. mentioned earlier that the learning curve of robotic surgery in general could have influenced docking and operative times, as well as the complication rate [22]. The articles by Lee et al. and Terzi et al. demonstrated that the learning curves of RAL and CL differed significantly. For RAL surgeries, the operation time for hysterectomy could be reduced after 23 surgeries of the same type, whereas for CL, 75 surgeries were required [34, 35].

However, experts of both RAL and CL have experienced the convenience of using RAL, as it provides comfort and increased precision in the operating technique. The RAL offers better visualization of the surgical site using 3D technology and 15 times magnification. Also, from an ergonomic point of view, instruments that mimic the movement of human hands, wrists, and fingers allow an extensive range of motion that is more precise than natural hand and wrist movements. Owing to the robotic arms, the sustained maintenance of positions demanding substantial force does not precipitate deleterious consequences or substantial fatigue. CL also has indisputable advantages, such as haptic tissue feedback, which, for example, is particularly beneficial for establishing the pathological-healthy border during the excision of a DIE nodule. Also, the esthetic effect is highlighted mainly from the patients' point of view. For CL, both the location of the ports on the trunk and the smaller diameter of the trocars are options preferred by patients. Robot-assisted surgery appears to have fewer negative cognitive and musculoskeletal impacts on surgeons than CL [36]. In 2021, Sers et al. found that performing laparoscopic surgery on patients, especially with high BMIs, increased the prevalence of non-neutral postures and could have further increased the risk of musculoskeletal disorders in surgeons [37]. However, to date, no studies have investigated the more serious, long-term, irreversible effects of CL on health, such as the potential development of knee and hip joint impairment. Current recommendations for the use of RAL in the surgical treatment of endometriosis vary depending on several factors, including the individual circumstances of the patients, the expertise of the surgeon, and the availability of resources and equipment. In 2013, the American Association of Gynecologic Laparoscopists (AAGL) recommended that RAL should not replace CL or vaginal procedures for women who could otherwise undergo CL or vaginal surgery for benign gynecologic diseases [38]. On the basis of the guidelines of the Danish Health Authority, RAL hysterectomy should only be preferred over CL hysterectomy after careful consideration because its beneficial effect is uncertain due to longer operating time [39]. Especially, with regard to advanced stage endometriosis, RAL is a possible first-line approach for the surgical treatment of bowel DIE [23, 40]. Furthermore, Lee et al. conclude that robot-assisted cystectomy in bilateral ovarian endometrioma is better than the laparoscopic approach for preserving ovarian function [20]. The decision to use robot-assisted laparoscopy for the treatment of endometriosis should be made on a case-by-case basis, taking into account the specific needs and circumstances of the patient as well as the experience and skill of the surgeon. Patients are advised to consult their healthcare providers to determine the most appropriate treatment approach for their individual situations. It is essential to highlight that this was only a snapshot. As time passes, expert surgeons who have spent most of their lives with laparoscopy will spend more time with the robot-assisted technique and could produce entirely new results.

We followed our rigorous protocol, which had been registered in advance. Our investigation covered a long study period, with a high number of cases. Although there had been previous meta-analyses on the topic, we were able to include more articles than the latest one from 2020. Compared with previous meta-analyses, our review examined operating room time as a new outcome.

As for the limitations of our analysis, most articles were retrospective studies, and only one RCT was included. In most of the included articles, patient selection was based on the availability of a robotic room. Also, some articles performed only certain organ-specific interventions and operated only on a specific severity of endometriosis, thus not representing the full range. Furthermore, it is imperative to underscore that the same author has contributed to some of the selected articles.