*3.3. Risk of Bias*

The Cochrane RoB 2.0 and ROBINS-I tools were used to assess the quality of the included papers.

Additionally, regarding the four randomized controlled trials, only one study [33] reported a low risk of bias. The other three studies [27,32,34] showed a high risk of bias, with the major bias due to deviations from the intended interventions, i.e., conversions to an open approach. Two conversions from laparoscopy were described by Draaisma et al. [27], and two conversions from the robotic approach were reported by Morino et al. and Nakadi et al. [32,34].

Due to the nature of the surgical interventions, blinding was impossible, but the results are unlikely to be affected by the lack of blinding.

All non-randomized studies reported a risk of bias due to baseline confounding. Only three [22,26,31] authors with a consequent moderate risk of bias performed propensity score matching, while the other nine [23–25,28–30,35–37] had a severe risk of bias due to insufficient adjustment for confounding domains.

The results of the RoB 2.0 andROBINS-I quality assessments arereported inFigures 2a,b and 3a,b, which were created with Robvis (Risk-of-Bias VISualization), a web app that facilitates rapid production of publication-quality risk-of-bias assessment figures. *J. Pers. Med.* **2023**, *13*, x FOR PEER REVIEW 8 of 16

### *3.4. Primary Outcomes*

The functional outcomes after laparoscopic (LF) and robotic fundoplication (RF) were analysed during a mean follow up period of 1-93.6 months, as described in Table 3.

(**a**)

(**b**)

**Figure 2.** Summary of risk of bias for RCT studies [27,32–34]. **Figure 2.** Summary of risk of bias for RCT studies [27,32–34].


(**a**)

**Figure 3.** Summary of risk of bias for non-RCT studies [22–26,28–31,35–37] **Figure 3.** Summary of risk of bias for non-RCT studies [22–26,28–31,35–37].

*3.4. Primary Outcomes*

*J. Pers. Med.* **2023**, *13*, 231

**Table 3.** Functional outcomes.


#### 3.4.1. 30-Days Readmission Rates

Only seven authors [22,23,25,26,33,35,37] reported 30-day readmission rates including 434 patients (207 RF and 227 LF) with no significant differences between the two groups [*<sup>p</sup>* = 0.73, RD = 0.00, 95% CI (−0.02, 0.03)]. No heterogeneity among the studies [Tau<sup>2</sup> = 0.00; Chi<sup>2</sup> = 0.85; df = 6 (*p* = 0.99); I<sup>2</sup> = 0%] was reported (Figure 4). *J. Pers. Med.* **2023**, *13*, x FOR PEER REVIEW 11 of 16

*J. Pers. Med.* **2023**, *13*, x FOR PEER REVIEW 11 of 16

#### **Figure 4.** Forest plot of laparoscopic vs. robotic 30day readmission rates [22,23,25,26,33,35,37] 3.4.2. Persistence of Symptoms Eleven studies [22,24,25,27,29,31,33–37] investigated the persistence of symptomatol-

3.4.2. Persistence of Symptoms Eleven studies [22,24,25,27,29,31,33–37] investigated the persistence of symptomatology after almost 1 month of follow-up. In addition, apart from two studies [25,37], which reported no ongoing symptoms, in the other nine [22,24,27,29,31,33–36], a total of 164 of the 641 patients referred reported lasting symptomatology without statistically significant differences between robotic and laparoscopic procedures (*p* = 0.60). Neither heterogeneity among the studies [Tau<sup>2</sup> = 0.00; Chi<sup>2</sup> = 6.19; df = 10 (*p* = 0.80); I<sup>2</sup> = 0%] was described (Figure Eleven studies [22,24,25,27,29,31,33–37] investigated the persistence of symptomatology after almost 1 month of follow-up. In addition, apart from two studies [25,37], which reported no ongoing symptoms, in the other nine [22,24,27,29,31,33–36], a total of 164 of the 641 patients referred reported lasting symptomatology without statistically significant differences between robotic and laparoscopic procedures (*p* = 0.60). Neither heterogeneity among the studies [Tau<sup>2</sup> = 0.00; Chi<sup>2</sup> = 6.19; df = 10 (*p* = 0.80); I<sup>2</sup> = 0%] was described (Figure 5). ogy after almost 1 month of follow-up. In addition, apart from two studies [25,37], which reported no ongoing symptoms, in the other nine [22,24,27,29,31,33–36], a total of 164 of the 641 patients referred reported lasting symptomatology without statistically significant differences between robotic and laparoscopic procedures (*p* = 0.60). Neither heterogeneity among the studies [Tau<sup>2</sup> = 0.00; Chi<sup>2</sup> = 6.19; df = 10 (*p* = 0.80); I<sup>2</sup> = 0%] was described (Figure 5).

**Figure 5.** Forest plot of laparoscopic vs. robotic persistence of symptomatology **Figure 5.** Forest plot of laparoscopic vs. robotic persistence of symptomatology [22,24,25,27,29,31,33–37].

**Figure 5.** Forest plot of laparoscopic vs. robotic persistence of symptomatology [22,24,25,27,29,31,33–37] [22,24,25,27,29,31,33–37] In particular, thirteen studies [22–24,26,27,30–37] reported the presence of postoperative dysphagia without significant differences between the two groups [39/387 RF and In particular, thirteen studies [22–24,26,27,30–37] reported the presence of postoperative dysphagia without significant differences between the two groups [39/387 RF and 47/529 LF, *p* = 0.77, RD = 0.00, 95% CI (−0.03, 0.02)]. Supplementary Materials: Figure S1.

In particular, thirteen studies [22–24,26,27,30–37] reported the presence of postoperative dysphagia without significant differences between the two groups [39/387 RF and 47/529 LF, *p* = 0.77, RD = 0.00, 95% CI (−0.03, 0.02)]. Supplementary Materials: Figure S1. 47/529 LF, *p* = 0.77, RD = 0.00, 95% CI (−0.03, 0.02)]. Supplementary Materials: Figure S1. Seven authors [22,23,25,28,30,36,37] reported data regarding delayed gastric emptying: a total of 6 patients in each group had gastric paresis at follow-up, with no statistically Seven authors [22,23,25,28,30,36,37] reported data regarding delayed gastric emptying: a total of 6 patients in each group had gastric paresis at follow-up, with no statistically significant differences between the two groups [6/215 RF vs. 6/242 LF; *p* = 0.99, RD = 0.00, 95% CI (−0.02, 0.02)]. Supplementary Materials: Figure S2.

Seven authors [22,23,25,28,30,36,37] reported data regarding delayed gastric emptying: a total of 6 patients in each group had gastric paresis at follow-up, with no statistically significant differences between the two groups [6/215 RF vs. 6/242 LF; *p* = 0.99, RD = 0.00, 95% CI (−0.02, 0.02)]. Supplementary Materials: Figure S2. significant differences between the two groups [6/215 RF vs. 6/242 LF; *p* = 0.99, RD = 0.00, 95% CI (−0.02, 0.02)]. Supplementary Materials: Figure S2. Only five articles [31,32,34–36] described the presence of postoperative pyrosis in 11 of 117 patients for the robotic group and 15 of 122 patients for the laparoscopic group, with no statistically significant differences between the two groups [*p* = 0.58, RD = −0.02, Only five articles [31,32,34–36] described the presence of postoperative pyrosis in 11 of 117 patients for the robotic group and 15 of 122 patients for the laparoscopic group, with no statistically significant differences between the two groups [*p* = 0.58, RD = −0.02, 95% CI (−0.09, 0.05)]. Supplementary Materials: Figure S3.

Only five articles [31,32,34–36] described the presence of postoperative pyrosis in 11 of 117 patients for the robotic group and 15 of 122 patients for the laparoscopic group,

95% CI (−0.09, 0.05)]. Supplementary Materials: Figure S3.

3.4.3. Recurrence

3.4.3. Recurrence

#### 3.4.3. Recurrence scribed in 22 patients (5 RF and 17 LF), with no significant differences between the two

In nine studies [22,24,25,27,30,31,33,34,37], recurrence of symptoms of reflux was described in 22 patients (5 RF and 17 LF), with no significant differences between the two groups [*p* = 0.36, RD = −0.02, 95% CI (−0.07, 0.03)]. The moderate heterogeneity among the studies [Tau<sup>2</sup> = 0.00; Chi<sup>2</sup> = 13.42; df = 8 (*p* = 0.10); I<sup>2</sup> = 40%] was reported (Figure 6). In nine studies [22,24,25,27,30,31,33,34,37], recurrence of symptoms of reflux was described in 22 patients (5 RF and 17 LF), with no significant differences between the two groups [*p* = 0.36, RD = −0.02, 95% CI (−0.07, 0.03)]. The moderate heterogeneity among the studies [Tau<sup>2</sup> = 0.00; Chi<sup>2</sup> = 13.42; df = 8 (*p* = 0.10); I<sup>2</sup> = 40%] was reported (Figure 6). groups [*p* = 0.36, RD = −0.02, 95% CI (−0.07, 0.03)]. The moderate heterogeneity among the studies [Tau<sup>2</sup> = 0.00; Chi<sup>2</sup> = 13.42; df = 8 (*p* = 0.10); I<sup>2</sup> = 40%] was reported (Figure 6).

In nine studies [22,24,25,27,30,31,33,34,37], recurrence of symptoms of reflux was de-

*J. Pers. Med.* **2023**, *13*, x FOR PEER REVIEW 12 of 16

**Figure 6.** Forest plot of laparoscopic vs. robotic recurrence of reflux symptoms **Figure 6.** Forest plot of laparoscopic vs. robotic recurrence of reflux symptoms [22,24,25,27,30,31,33,34,37]. 3.4.4. Reoperation

#### [22,24,25,27,30,31,33,34,37] 3.4.4. Reoperation A total of nine studies [22,25,27,30,31,33,34,36,37] reported reintervention during fol-

3.4.4. Reoperation A total of nine studies [22,25,27,30,31,33,34,36,37] reported reintervention during follow-up involving 22 patients with no significant differences between the two groups [*p* = 0.81, RD = 0.00, 95% CI (−0.04, 0.03)]. In addition, ten patients underwent reintervention after an initially successful robotic fundoplication: five experienced troublesome dysphagia [27,30,33,36], three had persistent GERD symptoms [36], one had an incisional hernia at the umbilicus [27], and another patient with a gastric torsion underwent a laparoscopic procedure with reduction of the torsion and fixation of the anterior gastric wall to the abdominal wall [34]. In the laparoscopic group, twelve patients were subject to reoperation during follow-up for persistent symptoms: six presented recurrent GERD symptoms [30,31,36] and six had severe dysphagia [27,30]. No heterogeneity among the studies [Tau<sup>2</sup> A total of nine studies [22,25,27,30,31,33,34,36,37] reported reintervention during follow-up involving 22 patients with no significant differences between the two groups [*p* = 0.81, RD = 0.00, 95% CI (−0.04, 0.03)]. In addition, ten patients underwent reintervention after an initially successful robotic fundoplication: five experienced troublesome dysphagia [27,30,33,36], three had persistent GERD symptoms [36], one had an incisional hernia at the umbilicus [27], and another patient with a gastric torsion underwent a laparoscopic procedure with reduction of the torsion and fixation of the anterior gastric wall to the abdominal wall [34]. In the laparoscopic group, twelve patients were subject to reoperation during follow-up for persistent symptoms: six presented recurrent GERD symptoms [30,31,36] and six had severe dysphagia [27,30]. No heterogeneity among the studies [Tau<sup>2</sup> = 0.00; Chi<sup>2</sup> = 2.58; df = 8 (*p* = 0.96); I<sup>2</sup> = 0%] was reported (Figure 7). low-up involving 22 patients with no significant differences between the two groups [*<sup>p</sup>* =0.81, RD = 0.00, 95% CI (−0.04, 0.03)]. In addition, ten patients underwent reintervention after an initially successful robotic fundoplication: five experienced troublesome dysphagia [27,30,33,36], three had persistent GERD symptoms [36], one had an incisional hernia at the umbilicus [27], and another patient with a gastric torsion underwent a laparoscopic procedure with reduction of the torsion and fixation of the anterior gastric wall to the abdominal wall [34]. In the laparoscopic group, twelve patients were subject to reoperation during follow-up for persistent symptoms: six presented recurrent GERD symptoms [30,31,36] and six had severe dysphagia [27,30]. No heterogeneity among the studies [Tau<sup>2</sup>= 0.00; Chi<sup>2</sup> = 2.58; df = 8 (*p* = 0.96); I<sup>2</sup> = 0%] was reported (Figure 7).


### *3.5. Publication Bias*

**Figure 7.** Forest plot of laparoscopic vs. robotic reoperation rates [22,25,27,30,33,34,36,37] *3.5. Publication Bias* It is recognized that publication bias can affect the results of meta-analyses; thus, we attempted to assess this potential bias using funnel plot analysis performed with Comprehensive Meta-analysis Software (CMA v.2). In evaluating all the analysed outcomes, we *3.5. Publication Bias* It is recognized that publication bias can affect the results of meta-analyses; thus, we attempted to assess this potential bias using funnel plot analysis performed with Comprehensive Meta-analysis Software (CMA v.2). In evaluating all the analysed outcomes, we observed a symmetrical distribution of the studies without any publication bias by the It is recognized that publication bias can affect the results of meta-analyses; thus, we attempted to assess this potential bias using funnel plot analysis performed with Comprehensive Meta-analysis Software (CMA v.2). In evaluating all the analysed outcomes, we observed a symmetrical distribution of the studies without any publication bias by the Egger's linear regression method (30-day readmission *p* = 0.84; recurrence *p =* 0.23; reoperation *p* = 0.60; persistence of symptomatology *p* = 0.12) (Supplementary Materials: Figures S4–S7).

observed a symmetrical distribution of the studies without any publication bias by the

#### **4. Discussion**

The functional disease of the esophago-gastric junction (EGJ) is a common health problem that often causes a serious deterioration of quality of life. Sometimes GERD patients do not achieve complete control of the symptoms with medical treatment with a proton pump inhibitor (PPI), needing surgical management.

Up until now, laparoscopic fundoplication has been considered the gold standard for the surgical treatment of functional disease of the EGJ [3,4].

Additionally, after the first robotic-assisted Nissen fundoplication (RALF) [5], it has been debated if the robotic approach could improve surgical outcomes compared with the conventional laparoscopic fundoplication (CLF) [6], considering the documented safety and feasibility of the robot-assisted approach in this setting [7–10].

It could be rational to hypothesize advantages of robotic surgery to improve functional results; limitations of laparoscopic procedures due to lack of dexterity, lack of tactile sense, magnification of natural tremors, and two-dimensional visualization could be overcome.

However, the robotic technique presents an important limitation that is related to the high functional costs, as shown by Hartmann et al. [29], Morino et al. [32], and Albassam et al. [22], due to the instrumentation and reusable materials, the nursing costs, the investment costs, and the maintenance costs [34].

According to current literature, there is no clear evidence as to which minimally invasive surgical approach is superior for the treatment of functional diseases of the EGJ. In addition, to the best of our knowledge, this is the first meta-analysis reported on the comparative efficacy of available interventions in the management of functional diseases of the EGJ.

Several limitations must be considered in our study: first, because of the novelty of this topic, few studies are present in the literature. In fact, only 16 studies [22–37] published between 2002 and 2021 could be included in this meta-analysis, with a narrow sample size of 1064 patients suffering from GERD, hiatal hernia, or paraesophageal hernia.

Then, only four studies were RCTs [27,32–34] and two were prospective trials [31,37]. All the other ten included studies were retrospective [22–26,28–30,35,36]; the observed results in each study could be affected by many factors, such as standards in patients' selection, the surgeon's experience, or technical details.

It is important to highlight that no one study had functional results as its primary outcome. According to our results, both robotic and laparoscopic fundoplication are effective as well, reporting no significant differences between the two groups in terms of 30-day readmission rates (*p* = 0.73), lasting of symptomatology at almost 1 month of follow-up (*p* = 0.60), recurrence of symptoms of reflux (*p* = 0.36), and needing for reintervention during follow-up (*p* = 0.81). Moreover, two conversions from laparoscopy were described by Draaisma et al. [27], and two conversions from the robotic approach were reported by Morino et al. and Nakadi et al. [32,34]. Although nowadays the risk of conversion to open surgery has decreased due to higher surgeon expertise, it is important to underline that the conversion rate from robotic surgery is lower than that from laparoscopy, according to current literature [38–40].

Furthermore, regarding the persistence of symptomatology after 1 month from the intervention, no differences were found in postoperative dysphagia (*p* = 0.77), gastric paresis (*p* = 0.99) and postoperative pyrosis (*p* = 0.58). It is fair to specify that both persistence of symptomatology and recurrence could appear even after years with a treatment failure rate of 40%, as shown by Spechler SJ [41]. However, there is a lack of data concerning a follow-up longer than five years for both the medical and surgical approaches.

Even if, on the basis of our results, we can state that the robotic approach was effective and feasible for the surgical treatment of the functional disease of EGJ, we cannot declare any advantage on the basis of the functional analysis results. Both laparoscopic and robotic approaches could be selected to perform a Nissen fundoplication. On the other hand, the current literature presents a lack of ad hoc papers evaluating some important features, such as:


The rationale is that robotic surgery should improve functional outcomes due to the magnified view and endowrist technology. However, it required making a call for future well-designed multicentre high-quality randomized controlled studies to evaluate the functional outcomes after robotic surgery for the treatment of functional disease of the EGJ, indications and parameters for GERD-surgery, and the long-term follow-up longer than 5 years.

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

**Supplementary Materials:** The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/jpm13020231/s1. Figure S1: Forest plot of laparoscopic vs. robotic persistence of postoperative dysphagia [22–24,26,27,30–37]; Figure S2: Forest plot of laparoscopic vs. robotic presence of gastric paresis [22,23,25,28,30,36,37]; Figure S3: Forest plot of laparoscopic vs. robotic presence of postoperative pyrosis [31,32,34–36]. Figure S4: Funnel plot of 30-day readmission rates; Figure S5: Funnel plot of recurrence of reflux symptoms; Figure S6: Funnel plot of reoperation; Figure S7: Funnel plot of persistence of symptomatology.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki. This study involved a systematic review and meta-analysis; as such, ethical review and approval were waived.

**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.
