**1. Introduction**

Robotic surgery was introduced in the early 2000s to overcome some technical limitations of conventional laparoscopic surgery. However, even if some benefits of the robotic approach over laparoscopy have been described [1–5], it is currently considered the gold standard treatment only for radical prostatectomy [6].

Specific interventions that could benefit from the robotic approach are yet to be identified. It is worth mentioning that one of the most extensively reported advantages of robotic surgery is the lower rate of unplanned open conversions [7–14]. Conversion to open surgery can be multifactorial, and when all causes of conversion were examined in ROLARR Randomized Controlled Clinical Trial (RCT), no difference was found between robotic and laparoscopic techniques during rectal cancer surgery [15].

Intra-abdominal adhesions due to prior abdominal surgery are a common and wellrecognised risk factor for conversion [16–18], and it is not known whether the robotic approach could allow a lower conversion rate than laparoscopy in patients with adhesions. The rationale lies in the potential technical advantages of robotic surgery—magnified 3D

**Citation:** Milone, M.; Manigrasso, M.; Anoldo, P.; D'Amore, A.; Elmore, U.; Giglio, M.C.; Rompianesi, G.; Vertaldi, S.; Troisi, R.I.; Francis, N.K.; et al. The Role of Robotic Visceral Surgery in Patients with Adhesions: A Systematic Review and Meta-Analysis. *J. Pers. Med.* **2022**, *12*, 307. https://doi.org/10.3390/ jpm12020307

Academic Editor: Volker J. Schmidt

Received: 2 December 2021 Accepted: 14 February 2022 Published: 18 February 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

vision with a more stable operative field, preservation of natural eye-hand-instrument alignment, precisely controlled EndoWrist instruments with better ergonomics and reduced physiologic tremor—heightened in case of distortion of the normal abdominal anatomy related to adhesions, which makes the visualisation more difficult and increases the difficulty of surgical procedure.

Since the indications for the robotic technique outside prostatectomy are far from being established by high levels of evidence, a meta-analysis of the available literature addressing pertinent questions related to the possible benefits of the robotic approach over laparoscopy is required to guide the expansion of the application of the robotic techniques. The aim of this study was to systematically review the literature and pool the evidence in order to evaluate and compare the adhesion-related conversions to open surgery are in patients undergoing robotic and laparoscopic surgery across all specialities.

#### **2. Materials and Methods**

#### *2.1. Literature Search and Study Selection*

To identify all available studies, an electronic search of Cochrane Library (including the Cochrane Central Register of Controlled Trials), EMBASE, PubMed, SCOPUS and Web of Science was conducted according to PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) guidelines [19]. This systematic review was performed following the meta-analysis of observational studies epidemiology (MOOSE) guidelines [20].

The search terms "laparoscopic" and "robotic" were used. The search was limited to studies regarding humans and published in English between June 1993 and March 2020.

Inclusion criteria were as follows: (1) comparative studies evaluating patients undergoing laparoscopic and robotic surgery; (2) studies reporting data on conversion to open surgery for each group due to adhesions; and (3) studies including at least 5 patients in each group, to minimise the imprecision associated with very small populations. Indexed abstracts of posters and podium presentations at international meetings were not included. Systematic reviews and meta-analyses were consulted to find additional studies of interest. Reference lists of the selected studies were screened to find additional studies of interest. If the same author or institution published overlapping series in different articles, only the most recent study was included. Two reviewers (Mi.Ma. and S.V.) independently assessed the reports for eligibility at the title and abstract level. In case of discrepancies, a third author (M.M.) was consulted, and an agreement was reached by consensus.

#### *2.2. Data Extraction and Quality Assessment of Included Studies*

The following data were extracted from each included study: first author, year of publication, study design, propensity score analysis, surgical field, diagnosis, type of intervention, total number of patients, number of patients undergoing laparoscopic and robotic surgery, and number of conversions related to intraoperative adhesions. Although widely reported by surgical studies, the definition of conversion within the literature varies [21]; therefore, we searched for this information in all the included studies. Surgeons' expertise (classified as novice vs. expert) has been described in many of the included studies, even if only a few studies reported the number of procedures performed by the surgeons. None of the studies provided an exact definition of the various steps of the surgical procedure. Thus, the criteria to define expertise remains heterogeneous.

Furthermore, attempts to examine the quality assurance of surgical techniques of the studies according to Foster JD et al. [22] was performed for the assessment of surgeondependent performance bias.

The following patients' characteristics were extracted and registered: gender, mean age, mean BMI, American Society of Anesthesiologists (ASA) score and previous abdominal surgery.

Study quality assessment for non-randomised clinical trial was performed using the Newcastle Ottawa Scale (NOS) [23]. This scoring system encompasses three major domains (selection, comparability and exposure), with a resulting score that varies between 0 (low quality) and 9 (high quality). In the case of randomised controlled trial (RCTs), the risk of bias was evaluated according to the Cochrane Collaboration Tool for assessing the risk of bias [24]. According to this scoring system, seven domains were evaluated as "Low risk of bias" or "High risk of bias" or "Unclear" according to reporting on sequence generation, allocation concealment, blinding of participants, blinding of outcome assessment, incomplete outcome data, selective outcome reporting and other potential threats to validity. The results of the quality assessment are reported in Table 1.


**Table 1.** NOS quality assessment of the included non-randomised trials.


#### **Table 1.** *Cont.*


#### **Table 1.** *Cont.*


#### **Table 1.** *Cont.*

#### *2.3. Statistical Analysis*

Statistical analysis was performed using RevMan (Version 5.4, Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2020).

The primary outcome of this study was the open conversion rate to open surgery due to adhesions. The odds ratio (OR) along with 95% confidence interval (CI) was used as an effect estimate for dichotomous outcomes, with OR values < 1 indicating fewer events in the robotic group. In the case of zero events, a 0.5 correction was added to incorporate all available data in the meta-analysis and to maintain analytic consistency [94]. When studies provided only means for continuous variables and sample size of the trial, a standard deviation was imputed, according to Furukawa et al. [95]. The summary estimate was computed according to the random effect model described by DerSimonian and Laird [96]. A conservative random effect model was chosen a priori in consideration of foreseen heterogeneity among the studies, which were from different surgical fields. The heterogeneity among studies was tested by Q statistic and quantified by I2 statistic, with I2 values < 25%, between 25 and 50%, and >50% indicating respectively low, moderate, and high heterogeneity [97].

With the aim to assess if that differences among included studies may be affected by demographic (gender, age and BMI) and clinical variables (ASA Score and previous abdominal surgery), we planned to perform meta-regression analyses in case of the significance of the meta-analysis after implementing a regression model with incidence of the main outcome as dependent variable (y) and the above-mentioned covariates as independent variables (x). Meta-regression analyses were performed with Comprehensive Meta-analysis (Version 2.2, Biostat Inc., Englewood, NJ, USA, 2005), provided by Biostat Inc. [98].

The presence of publication bias was investigated through a funnel plot where the summary estimate of each study (OR) was plotted against a measure of study precision (standard error). In addition to visual inspection, funnel plot symmetry was tested using Egger's linear regression method [99]. *p* values < 0.05 were considered statistically significant.

Furthermore, different subgroups analyses, including studies about each surgical field (colorectal, oesophagogastric, hepatobiliary, pancreatic, endocrine, urologic and gynaecologic surgery) and the surgeons' expertise (novice and expert) were performed. Furthermore, in case of a statistically significant difference in any of the above-mentioned surgical fields, further analyses were performed to understand if the significance was present in the case of benign and or malignant disease.

#### **3. Results**

#### *3.1. Study Selection*

The electronic search provided a total of 49,891 results. After the removal of duplicates, 10,489 studies underwent screening on the basis of title. Of the 4050 full-text articles assessed for eligibility, 3978 studies were excluded for several reasons: 431 were not published in the English language, 444 were case reports, 2179 were reviews, 535 were off-topic after scanning abstract, and for 391, data were not available. At the end of the selection process, 70 studies were included in the meta-analysis [13,25–93].

#### *3.2. Study Characteristics*

The selected studies included a total of 14,329 patients, of whom 6472 underwent robotic surgery and 7857 laparoscopic surgery. Fifty-one studies were retrospective [13,25–27,32,33,37, 39–49,51–53,56–58,63–69,71,73–90,92,93], eighteen were prospective [28–31,34–36,38,50,54,55, 59–62,65,72,91], and there was only one randomised controlled trial [70]. Studies were from different fields of surgery, including colorectal (*n* = 19), oesophagogastric (*n* = 10), hepatobiliary (*n* = 5), pancreatic (*n* = 6), gynaecologic (*n* = 19), urologic (*n* = 5), endocrine (*n* = 3) vascular (*n* = 1), abdominal wall (*n* = 1) and splenic surgery (*n* = 1). In six studies, robotic surgery was performed by early surgeons, and by expert surgeons in other 47 studies. The other 17 studies did not provide these data. The characteristics of the included studies are reported in Table 2.


**Table 2.** Characteristics of the included studies.


#### **Table 2.** *Cont.*


#### **Table 2.** *Cont.*


#### **Table 2.** *Cont.*

GERD—gastroesophageal reflux disease; pNET—pancreatic neuroendocrine tumour; IS—in situ; IBD—intestinal bowel disease; TME—total mesorectal excision; LAR—low anterior resection; APR—abdominoperineal resection; AGB—adjustable gastric banding; VSG—vertical sleeve gastrectomy; RYGB—Roux-en-Y gastric bypass; VBG—vertical banded gastroplasty; MAL—median arcuate ligament; PD—pancreaticoduodenectomy; CME—complete mesocolic excision; NR—not reported.

#### *3.3. Quality Assessment of Studies and Performance*

All studies had NOS quality scores greater than 6, indicating that all these studies had a high methodological quality. Specifically, twenty-one studies had NOS quality score = 9; thirty studies had NOS quality score = 8; eighteen studies had NOS quality score = 7. The NOS quality score is shown in Table 1. The only randomised controlled trial (RCT) showed a low risk of bias.

Among the expert surgeons, none of the included studies reported on the quality assurance of surgical technique as described by Foster et al. [22]. Thus, it was not possible to perform further analyses on the quality of surgical performance among expert or early surgeons.

#### *3.4. Conversion to Open Surgery Due to Adhesions*

Seventy studies provided data about the conversion to open surgery due to anastomotic adhesions [13,25–93], even if only nine of them [13,35,41,43,45,46,50,78,84] reported the definition of conversion. The robotic approach was associated with a reduced risk of conversion (OR 1.53, 95% CI 1.12–2.10, *p* = 0.007, Figure 1), with consistent results across all the 70 studies since no heterogeneity was observed (I<sup>2</sup> = 0%, *p* = 0.95).

Regarding surgeons' expertise, 47 studies classified surgeons as "expert" [25,26,28,31–33, 35–37,40,41,43–49,51,53,55–57,59,60,62–64,66–70,73,74,78,79,82,84–86,88,89,91–93] and 6 studies as "novice" [13,30,34,80,81,83]. The analysis of the procedures performed by expert surgeons involved 11,172 procedures, of which 6283 laparoscopic and 4889 robotic and showed a statistically significant difference in favour of robotic surgery (OR 1.48, 95% CI 1.03–2.12, *p* = 0.03), with no heterogeneity among the studies (I<sup>2</sup> = 0%, *p* = 0.71). The analysis of the procedures performed by "novice" surgeons involved 622 procedures, of which 307 laparoscopic and 315 robotic and showed no significant difference between the two groups (OR 1.53, 95% CI 0.44–5.28, *p* = 0.50), without any heterogeneity among the studies (I<sup>2</sup> = 0%, *p* = 0.91). Data on surgeons' expertise are shown in Figure 2.

Our meta-regression analysis showed that no demographic or clinical outcomes significantly impacted conversion, as shown in Table 3.


**Figure 1. Figure 1.**  Meta-analysis of the included studies on conversion due to adhesions. Meta-analysis of the included studies on conversion due to adhesions.


**Figure 2.** Conversions due to adhesions according to surgeons' expertise: (**a**) in procedures performed by expert surgeons; (**b**) in procedures performed by "early" surgeons. **Figure 2.** Conversions due to adhesions according to surgeons' expertise: (**a**) in procedures performed by expert surgeons; (**b**) in procedures performed by "early" surgeons.

Our meta-regression analysis showed that no demographic or clinical outcomes

significantly impacted conversion, as shown in Table 3.


**Table 3.** *p*-values of the meta-regression analysis.

BMI—body mass index.

### *3.5. Subgroup Analysis*

#### 3.5.1. Colorectal Surgery

The results of the studies about colorectal surgery are shown in Figure 3. Nineteen studies [30,33,37,41,43,45,46,52,57,61,65,71,72,78,80,81,84,91,92] included in the final analysis were including colorectal surgery cases and involved 2969 procedures, of which 1548 laparoscopic and 1421 robotic. Of the included studies, eleven were on colorectal cancer [30,41,45,46,52,57,61,65,80,84,92], three on rectocele or rectal prolapse [71,72,91] and three on diverticular disease patients [33,37,43]. Ozben et al. [78] described surgical procedures related to both benign and malign diseases. Rencuzogullari et al. [81] was the only one to report surgical proctectomy performed for IBD, so it was excluded from the subgroup analysis.

In the overall colorectal surgery analysis, a significant difference in terms of conversion rate related to adhesions was observed between the two groups in favour of robotics (OR 2.22, 95% CI 1.18–4.19, *p* = 0.01), with no heterogeneity among the included studies (I<sup>2</sup> = 0%, *p* = 0.93).

Meta-regression analysis showed that none of the demographic and clinical parameters (gender, age, BMI, ASA and tumoural stage) significantly impacted the conversion rate due to adhesions, with the exception of "previous abdominal surgery" (*p* = 0.03).

In a further analysis about colorectal cancer the significance was confirmed (OR 2.62, 95% CI 1.20–5.72, *p* = 0.02), with no heterogeneity among the included studies (I<sup>2</sup> = 0%, *p* = 0.89). Even including only studies about rectal cancer [30,41,45,46,52,57,61,65,80,84,92], the significance was confirmed (OR 2.54, 95% CI 1.10–5.88, *p* = 0.03), with no heterogeneity among the included studies (I<sup>2</sup> = 0%, *p* = 0.79).

Meta-regression analysis on colorectal cancer showed that none of the demographic or clinical parameters significantly impacted the analysed outcome.

No statistically significant differences in terms of conversion rate due to adhesions were observed between robotics and laparoscopy in the studies about rectocele/rectal prolapse [72,73,91] and diverticular disease [33,37,43] (OR 1.72, 95% CI 0.27–11.16, *p* = 0.57 and OR 1.36, 95% CI 0.10–18.02, *p* = 0.81, respectively), with no significant heterogeneity among the studies (I<sup>2</sup> = 0%, *p* = 1.00 and I<sup>2</sup> = 53%, *p* = 0.12, respectively).

Within the colorectal surgery studies, surgeons were classified as "expert" in eleven studies [33,37,41,43,45,46,57,78,84,91,92] and as "novice" in other three studies [30,80,81]. Five studies did not provide these data [52,61,65,71,72]. The analysis about expertise in colorectal surgeries showed that a significant difference in terms of conversion rate related to adhesions was found in colorectal surgery performed by expert surgeons in favour of robotic approach (OR 2.34, 95% CI 1.07–5.11, *p* = 0.03), while no statistically significant differences were observed among colorectal (OR 1.35, 95% CI 0.25–7.40, *p* = 0.73) "novice" surgeons.

**Figure 3.** Results of the studies about colorectal surgery: (**a**) conversion due to adhesions in colorectal surgery; (**b**) conversion due to adhesions in colorectal cancer surgery; (**c**) conversion due to adhesions in rectal cancer surgery; (**d**) conversion due to adhesions in colorectal surgery for rectal prolapse/rectocele; (**e**) conversion due to adhesions in colorectal surgery for diverticular disease; (**f**) conversion due to adhesions in colorectal surgery performed by expert surgeons; (**g**) conversion due to adhesions in colorectal surgery performed by "early" surgeons.

#### 3.5.2. Oesophagogastric Surgery 1.45, 95% CI 0.58–3.64, *p* = 0.43), with no heterogeneity among the included studies (I2 =

3.5.2. Oesophagogastric Surgery

The results of the studies about oesophagogastric surgery are shown in Figure 4. Ten studies addressed oesophagogastric surgery [25,27,29,31,35,49,53,62,63,79], involving 3504 procedures, 2254 of which were laparoscopic and 1250 robotic. Of the included studies, five were about gastric cancer [27,49,62,63,79], four about morbid obesity [29,31,35,53] (two about Roux-en-Y gastric bypass [31,35], one about sleeve gastrectomy [29] and one about different surgical procedures for bariatric revisional surgery [53]) and one on Nissen fundoplication for reflux disease [25]. No statistically significant differences were found between the two groups in terms of conversion rate related to adhesions (OR 1.45, 95% CI 0.58–3.64, *p* = 0.43), with no heterogeneity among the included studies (I<sup>2</sup> = 0%, *p* = 0.47). 0%, *p* = 0.47). Surgeons' expertise was reported by eight studies [25,31,35,49,53,62,63,79], that classified surgeons as experts. The other two studies did not report on these data [27,29]. The meta-analysis about surgeons' expertise showed that no statistically significant differences were found among the expert surgeon between robotic and laparoscopic conversion rate (OR 1.12, 95% CI 0.41–3.10, *p* = 0.82), with no heterogeneity among the studies (I2 = 0%, *p* = 0.44). The analysis about "novice" surgeons was not possible because none of the included studies reported these data.

**Figure 3.** Results of the studies about colorectal surgery: (**a**) conversion due to adhesions in colorectal surgery; (**b**) conversion due to adhesions in colorectal cancer surgery; (**c**) conversion due to adhesions in rectal cancer surgery; (**d**) conversion due to adhesions in colorectal surgery for rectal prolapse/rectocele; (**e**) conversion due to adhesions in colorectal surgery for diverticular disease; (**f**) conversion due to adhesions in colorectal surgery performed by expert surgeons; (**g**) conversion due

The results of the studies about oesophagogastric surgery are shown in Figure 4. Ten studies addressed oesophagogastric surgery [25,27,29,31,35,49,53,62,63,79], involving 3504 procedures, 2254 of which were laparoscopic and 1250 robotic. Of the included studies, five were about gastric cancer [27,49,62,63,79], four about morbid obesity [29,31,35,53] (two about Roux-en-Y gastric bypass [31,35], one about sleeve gastrectomy [29] and one about different surgical procedures for bariatric revisional surgery [53]) and one on Nissen fundoplication for reflux disease [25]. No statistically significant differences were found between the two groups in terms of conversion rate related to adhesions (OR

*J. Pers. Med.* **2022**, *12*, x FOR PEER REVIEW 15 of 28

to adhesions in colorectal surgery performed by "early" surgeons.

**Figure 4.** Results of the studies about oesophagogastric surgery: (**a**) conversion due to adhesions in oesophago-gastric surgery; (**b**) conversion due to adhesions in oesophagogastric surgery performed by expert surgeons.

Surgeons' expertise was reported by eight studies [25,31,35,49,53,62,63,79], that classified surgeons as experts. The other two studies did not report on these data [27,29]. The meta-analysis about surgeons' expertise showed that no statistically significant differences were found among the expert surgeon between robotic and laparoscopic conversion rate (OR 1.12, 95% CI 0.41–3.10, *p* = 0.82), with no heterogeneity among the studies (I<sup>2</sup> = 0%, *p* = 0.44). The analysis about "novice" surgeons was not possible because none of the included studies reported these data.

#### 3.5.3. Gynaecologic Surgery

The results of the studies about gynaecologic surgery are shown in Figure 5. Nineteen studies about gynaecologic surgery were included in the meta-analysis [34,38–40,44,47,50, 51,55,56,58,60,67,68,70,76,82,83,86]. Of the included studies, 18 reported data on hysterectomies performed for benign [50,60] or malignant conditions [34,39,40,44,47,51,55,56,58,67, 68,70,82,83,86] or the combination of malignancy and benign diseases [38] and one about salpingo-oophorectomy due to early ovarian cancer [76]. The included studies involved 3124 procedures, of which 1772 were laparoscopic and 1352 robotic, with no statistically significant difference between the two groups (OR 1.36, 95% CI 0.82–2.25, *p* = 0.24) in terms of conversion rate related to adhesions and no heterogeneity among the included studies (I<sup>2</sup> = 0%, *p* = 0.73).

Fourteen studies classified the surgeons as "experts" [40,44,47,50,51,55,56,58,60,67,68, 70,82,86], two as "novice" [34,83] and four did not report on these data [38,39,50,76]. No significant difference in terms of conversion rate related to adhesions was found between the two groups in the procedures performed by both expert or novice surgeons (OR 1.52, 95% CI 0.87–2.65, *p* = 0.14 and OR 1.18, 95% CI 0.12–11.43, *p* = 0.89), with no heterogeneity among the studies (I<sup>2</sup> = 0%, *p* = 0.77 and I<sup>2</sup> = 0%, *p* = 0.37).

#### 3.5.4. Hepatobiliary Surgery

The results of the five hepatobiliary surgery studies are shown in Figure 6 [42,48,73,85,89]. Of the included studies, three [73,85,89] included 511 liver resections, 343 laparoscopic and 168 robotic, and all the studies classified surgeons as experts.

Cuendis-Velazquez A. et al. [41] and Gangemi et al. [48] reported hepaticojejunostomy performed for bile duct injury and cholecystectomy, respectively, so they were excluded from our analysis.

No differences were found in terms of conversion due to adhesions between the two groups (OR 1.41, 95% CI 0.15–13.30, *p* = 0.76), without a significant heterogeneity among the included studies (I<sup>2</sup> = 41%, *p* = 0.76).

#### 3.5.5. Pancreatic Surgery

The results of the studies about pancreatic surgery are shown in Figure 7. Six studies [13,26,36,66,69,75] reporting data about conversion due to adhesions in pancreatic surgery were included in the meta-analysis, involving 558 procedures, 333 laparoscopic and 225 robotic. Of the included studies, four [13,26,36,66] reported data on distal pancreatectomies for pancreatic tumours [13,36,66] or neuroendocrine tumours (pNETs) [26], one about pancreaticoduodenectomies for periampullary neoplasms [69] and one about distal pancreatectomies or pancreatic enucleations for benign and borderline tumours [75]. The analysis showed no statistically significant difference in terms of conversion rate related to adhesions between the two groups (OR 1.03, 95% CI 0.40–2.68, *p* = 0.95), with no heterogeneity among the studies (I<sup>2</sup> = 0%, *p* = 0.53).

Surgeons were classified as "expert" in four studies [26,36,66,69] and as "novice" in one study [13], while one study [75] did not report on these data.

In the case of surgery performed by expert surgeons, the analysis showed no significant differences in terms of conversion due to adhesions between the two groups (OR 0.74, 95% CI 0.25–2.15, *p* = 0.58), with no heterogeneity among the studies (I<sup>2</sup> = 0%, *p* = 0.54).

#### 3.5.6. Urologic Surgery

The results of the studies about urologic surgery are shown in Figure 8. Five studies [28,32,54,88,93] included partial nephrectomies [28,32,54,88] for renal cancer or simple enucleation with single layer renorrhaphy for localized renal tumours [93] were included in the analysis, involving 1087 procedures, 557 laparoscopic and 530 robotic.



**Figure 5.** Results of the studies about gynaecologic surgery: (**a**) conversion due to adhesions in gynaecologic surgery; (**b**) conversion due to adhesions in gynaecologic surgery performed by expert **Figure 5.** Results of the studies about gynaecologic surgery: (**a**) conversion due to adhesions in gynaecologic surgery; (**b**) conversion due to adhesions in gynaecologic surgery performed by expert surgeons; (**c**) conversion due to adhesions in gynaecologic surgery performed by "early" surgeons.

surgeons; (**c**) conversion due to adhesions in gynaecologic surgery performed by "early" surgeons.

laparoscopic and 168 robotic, and all the studies classified surgeons as experts.

The results of the five hepatobiliary surgery studies are shown in Figure 6 [42,48,73,85,89]. Of the included studies, three [73,85,89] included 511 liver resections, 343

3.5.4. Hepatobiliary Surgery

they were excluded from our analysis.

the included studies (I2 = 41%, *p* = 0.76).

Cuendis-Velazquez A. et al. [41] and Gangemi et al. [48] reported hepaticojejunostomy performed for bile duct injury and cholecystectomy, respectively, so

No differences were found in terms of conversion due to adhesions between the two groups (OR 1.41, 95% CI 0.15–13.30, *p* = 0.76), without a significant heterogeneity among

**Figure 6.** Conversion due to adhesions in hepatobiliary surgery. **Figure 6.** Conversion due to adhesions in hepatobiliary surgery.

**Figure 7.** Results of the studies about pancreatic surgery: (**a**) conversion due to adhesions in pancreatic surgery; (**b**) conversion due to adhesions in pancreatic surgery performed by expert **Figure 7.** Results of the studies about pancreatic surgery: (**a**) conversion due to adhesions in pancreatic surgery; (**b**) conversion due to adhesions in pancreatic surgery performed by expert surgeons. *J. Pers. Med.* **2022**, *12*, x FOR PEER REVIEW 20 of 28


surgeons.

**Figure 8.** Results of the studies about urologic surgery: (**a**) conversion due to adhesions in urologic surgery; (**b**) conversion due to adhesions in urologic surgery performed by expert surgeons. **Figure 8.** Results of the studies about urologic surgery: (**a**) conversion due to adhesions in urologic surgery; (**b**) conversion due to adhesions in urologic surgery performed by expert surgeons.

No significant difference was found between the two groups in terms of conversion due to adhesions (OR 1.52, 95% CI 0.24–9.49, *p* = 0.65), with no heterogeneity among the

Surgeons' expertise was reported by two studies [59,74], classifying surgeons as

Analysis of the studies about expert surgeons showed no significant differences between the two groups (OR 1.00, 95% CI 0.10–9.80, *p* = 1.00), with no heterogeneity

experts. One study [77] did not report on these data.

analysis, involving 286 procedures, 155 laparoscopic and 131 robotic.

3.5.7. Endocrine Surgery

studies (I2 = 0%, *p* = 0.52).

among the studies (I2 = 0%, *p* = 0.34).

No statistical difference was found in the two groups in terms of conversion due to adhesions (OR 0.74, 95% CI 0.17–3.10, *p* = 0.68), with no heterogeneity among the studies (I<sup>2</sup> = 0%, *p* = 0.77).

Four studies [28,32,88,93] classified surgeons as experts, while one study [54] did not report on these data.

The analysis of the studies about expert surgeons showed no significant differences between the two groups (OR 0.92, 95% CI 0.18–4.58, *p* = 0.92), with no heterogeneity among the studies (I<sup>2</sup> = 0%, *p* = 0.83).

#### 3.5.7. Endocrine Surgery

The results of the endocrine surgery studies are shown in Figure 9. Three studies [59,74,77] that addressed adrenalectomies for adrenal cancer were included in the analysis, involving 286 procedures, 155 laparoscopic and 131 robotic. *J. Pers. Med.* **2022**, *12*, x FOR PEER REVIEW 21 of 28

**Figure 9.** Results of the studies about endocrine surgery**:** (**a**) conversion due to adhesions in endocrine surgery; (**b**) conversion due to adhesions in endocrine surgery performed by expert **Figure 9.** Results of the studies about endocrine surgery: (**a**) conversion due to adhesions in endocrine surgery; (**b**) conversion due to adhesions in endocrine surgery performed by expert surgeons.

surgeons. 3.5.8. Other Surgical Fields No significant difference was found between the two groups in terms of conversion due to adhesions (OR 1.52, 95% CI 0.24–9.49, *p* = 0.65), with no heterogeneity among the studies (I<sup>2</sup> = 0%, *p* = 0.52).

Khrucharoen et al. [64] described the median arcuate ligament (MAL) release for median arcuate ligament syndrome. Vasilescu et al. [87] reported splenectomy for Surgeons' expertise was reported by two studies [59,74], classifying surgeons as experts. One study [77] did not report on these data.

hereditary spherocytosis. Warren et al. [90] described ventral hernia repair. These were individual studies for each respective surgical field, so it was not possible to perform a meta-analysis. Analysis of the studies about expert surgeons showed no significant differences between the two groups (OR 1.00, 95% CI 0.10–9.80, *p* = 1.00), with no heterogeneity among the studies (I<sup>2</sup> = 0%, *p* = 0.34).

#### *3.6. Publication Bias*  3.5.8. Other Surgical Fields

radical prostatectomy [6].

laparoscopic approach [3,4,105–110].

Visual inspection of the funnel plot (Figure S1) showed symmetry, which was confirmed by Egger's linear regression test (*p* = 0.12), indicating no publication bias. In the subgroup analyses, a symmetrical distribution of the studies was observed in all surgical fields except from pancreatic surgery, in which the visual inspection of the funnel plot Khrucharoen et al. [64] described the median arcuate ligament (MAL) release for median arcuate ligament syndrome. Vasilescu et al. [87] reported splenectomy for hereditary spherocytosis. Warren et al. [90] described ventral hernia repair. These were individual studies for each respective surgical field, so it was not possible to perform a meta-analysis.

#### suggested an asymmetric distribution of studies around the mean and the Egger's test *3.6. Publication Bias*

curve.

confirmed a significant publication bias (*p* = 0.0029). **4. Discussion**  Since its introduction in the early 1990s, laparoscopic surgery has become the gold standard treatment of many benign and malignant conditions [100–103]. Visual inspection of the funnel plot (Figure S1) showed symmetry, which was confirmed by Egger's linear regression test (*p* = 0.12), indicating no publication bias. In the subgroup analyses, a symmetrical distribution of the studies was observed in all surgical fields except from pancreatic surgery, in which the visual inspection of the funnel plot

The advantages of a minimally invasive approach over an open approach are well proven [6,104], but laparoscopic surgery is technically challenging with a long learning

Robotic surgery was introduced in the early 2000s to overcome these challenges of laparoscopic surgery, but to date, it is considered the gold standard treatment only for

The efficacy and the feasibility of the robotic technique have been shown in various procedures across many surgical fields and demonstrate some benefits over the

One of the reported benefits of robotic surgery is the lower rate of unplanned conversions to open surgery compared to laparoscopy [8–14,110]. This was, however, not supported by the results of an RCT on rectal cancer surgery and comparing conversions for all causes in robotic and laparoscopic procedures [15]. A cause–effect analysis is suggested an asymmetric distribution of studies around the mean and the Egger's test confirmed a significant publication bias (*p* = 0.0029).

#### **4. Discussion**

Since its introduction in the early 1990s, laparoscopic surgery has become the gold standard treatment of many benign and malignant conditions [100–103].

The advantages of a minimally invasive approach over an open approach are well proven [6,104], but laparoscopic surgery is technically challenging with a long learning curve.

Robotic surgery was introduced in the early 2000s to overcome these challenges of laparoscopic surgery, but to date, it is considered the gold standard treatment only for radical prostatectomy [6].

The efficacy and the feasibility of the robotic technique have been shown in various procedures across many surgical fields and demonstrate some benefits over the laparoscopic approach [3,4,105–110].

One of the reported benefits of robotic surgery is the lower rate of unplanned conversions to open surgery compared to laparoscopy [8–14,110]. This was, however, not supported by the results of an RCT on rectal cancer surgery and comparing conversions for all causes in robotic and laparoscopic procedures [15]. A cause–effect analysis is required to specifically target conversions related to adhesions and appraise the true impact of the robotic technique in comparison to laparoscopy in order to support the adoption of the robotic technique across all surgical fields.

By pooling together 14,329 patients, 6472 of whom were undergoing robotic surgery and 7857 laparoscopic surgery, we were able to observe that the robotic approach seems to be associated with a lower number of conversions due to abdominal adhesions compared to laparoscopic surgery, with an overall OR of 1.5.

However, to reduce the heterogeneity in the included studies, we performed subgroups analyses to assess if the statistical significance was confirmed in each surgical field.

Our subgroups analysis performed on colorectal patients confirmed the reduced conversion rate due to adhesions in the robotic surgery population, as obtained in the overall analysis, with an OR of 2.22 (95% CI 1.18–4.19, *p* = 0.01). Furthermore, the analysis on different colorectal procedures showed that this significance was present only in colorectal procedures performed in cancer patients (OR 2.62, 95% CI 1.20–5.72, *p* = 0.02), while the colorectal procedures for other diseases did not significantly impact the results (OR 1.72, 95% CI 0.27–11.16, *p* = 0.57 for rectal prolapse and OR 1.36, 95% CI 0.10–18.02, *p* = 0.81 for diverticular disease).

One potential explanation of these findings is that surgery for colorectal cancer often requires access to various quadrants of the abdomen: frequently both the supra- and the infra-mesocolic spaces. Thus, the presence of adhesions in those cases could significantly affect this type of surgical procedure, more than other speciality procedures that are confined to one compartment in the abdomen or the pelvis.

Evaluating the role of surgeons' experience was of paramount importance, being a potential confounding factor considering the study's primary endpoint (conversion to open). We performed this subgroup analysis to ensure that the results of the two techniques were comparable and not affected by different experience levels.

Our results showed that the robotic approach significantly reduced the conversion rate in the case of expert surgeons (OR 1.48, 95% CI 1.03–2.12, *p* = 0.03), while no significant difference was found in the case of procedures performed by "novice" surgeons (OR 1.53, 95% CI 0.44–5.28, *p* = 0.50). This finding was also observed in the overall conversion analysis and in the colorectal surgery subgroup.

In the analysis on the colorectal surgery subgroup performed by expert surgeons, a statistically significant difference favouring robotic surgery was observed (OR 2.34, 95% CI 1.07–5.11, *p* = 0.03), while no statistically significant difference was observed among colorectal (OR 1.35, 95% CI 0.25–7.40, *p* = 0.73) "novice" surgeons. One possible

explanation of these results is that the benefits of the robotic approach in colorectal surgery are maximised and become evident only after completing the learning curve.

However, the criteria to define the expertise remains heterogeneous. In fact, only five studies reported the number of procedures [45,46,61,65,84] performed by the surgeons, and none of the studies provided an exact definition of the various steps of the surgical procedure. We attempted to apply rigorous criteria to evaluate the quality of the techniques, but none of the studies reported on surgeons' credentialing, standardisation of techniques and objective evaluation and monitoring of surgeons' skills. Nevertheless, the pooled data in this study highlight the importance of optimal training in robotic surgery in order to achieve the maximum benefits for the patients.

Our study has several strengths. To date, this is the first meta-analysis on the risk of conversion due to intraabdominal adhesions comparing robotic and laparoscopic surgery. In this setting, clinical decisions of adopting one technique over the other could be supported by our meta-analysis, which comprises a large number of studies and cases and therefore enhances the external validity and generalizability.

Based on these results, we could encourage the use of robotic surgery in patients with known or suspected abdominal adhesions and due to undergo a colorectal cancer resection.

However, several limitations should also be acknowledged. By only including studies published in English with full text, a language bias could not be excluded. Results from retrospective studies inevitably contained potential selection bias, confounding bias and missing data bias.

We could not fully adjust for confounding factors, including the causes and the extent and severity of the adhesions. Additionally, in the included studies, the definition of expertise is heterogeneous, with an increased risk of surgeons-dependent performance bias.

Further efforts are required to implement a quality assurance framework when reporting on advanced surgical skills [21].

No ad hoc studies were currently available specifically addressing the role of robotic versus laparoscopic surgery determining conversion related to adhesions.

Additionally, the definition of conversion to open was not adequately standardised; in fact, only nine studies provide this information. [13,35,41,43,45,46,50,78,84] and an optimal information prevalence of conversions for adhesions cannot be obtained.

#### **5. Conclusions**

Limitations notwithstanding, this state-of-the-art review provides a lens through which to scrutinise and appraise the currently available evidence on abdominal robotic and laparoscopic surgery with a focus on conversion rates due to intraabdominal adhesions.

Our study should not be interpreted as an arbitrary conclusion that any planned colorectal intervention with certain or presumed adhesions should be treated by a robotic approach. Instead, our findings should support surgeons in the process of selecting the optimal technique and highlight the potential advantages of the robotic approach when performing surgery with a high risk of necessitating complex adhesiolysis.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jpm12020307/s1. Figure S1: Forest plot analysis of the included studies.

**Author Contributions:** M.M. (Marco Milone), conceptualisation of the study and supervision; M.M. (Michele Manigrasso), P.A., A.D. and S.V., data curation and investigation; M.M. (Michele Manigrasso) and M.C.G., formal analysis; M.M. (Michele Manigrasso), S.V., G.R., M.C.G., R.I.T. and G.D.D.P., drafting of the article; U.E., M.M. (Marco Milone), R.I.T., G.D.D.P. and N.K.F., review and editing of the article; M.M. (Marco Milone), M.M. (Michele Manigrasso), P.A., A.D., U.E., M.C.G., G.R., S.V., R.I.T., G.D.D.P. and N.K.F., final approval. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki. Ethical review and approval were not required for a systematic review and meta-analysis.

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**

