Since the advent of RAS, there has been growing interest in applying this technology to VVF repair, especially given the limitations of conventional laparoscopy, namely the restricted operative field, challenging intracorporeal suturing, and a steep learning curve [7, 8]. To our knowledge, this systematic review and quantitative synthesis represents the most comprehensive synthesis to date, comprising outcomes of 14 studies evaluating robotics-assisted repair of VVFs. More than a third (35.9%) of the cases included in this review were recurrent fistulas, reflecting prior failed attempts at repair using other surgical techniques. Additionally, a substantial proportion involved complex fistulas or patients with a history of prior pelvic surgeries, where extensive adhesions were anticipated. These factors likely influenced the choice of RAS in the studies included, underscoring its utility in challenging clinical scenarios where conventional transvaginal or laparoscopic approaches may be technically unfeasible or less effective.

Regarding LOS, only three studies reported longer LOS than 5 days. Bora et al. [13] and Chandna et al. [22] (median of 7.5 and mean of 9.5 days respectively) contextualized these results with patient-related and local health care system factors, such as the majority of patients being housewives from remote areas, which meant that they were not restricted by formal employment obligations such as sick leave and the affordability of care. Matei et al. [19] (who reported a mean time of 6.6 days) also offered “geographic reasons” as a justification for their longer hospitalization times. When discarding these three outliers, LOS was relatively short (under 5 days), both in the overall review and within each technique subgroup. This trend toward shorter postoperative hospitalization aligns with the reduced morbidity and faster recovery typically associated with minimally invasive approaches, particularly robotics-assisted techniques.

The wide variation in reported CTs across studies reflects the heterogeneity of post-surgical care standards in the collected literature, particularly with regard to removal of the urinary catheter in these patients: some authors used a cystogram or a voiding cystourethrography to confirm the patency of the bladder closure before removing the catheter [4, 19, 23, 24]; others removed it “based on operating surgeons’ preference” [26]; and others routinely removed it after a certain postoperative period, which varied among studies [20, 21].

As described above, our study’s findings support the robotics-assisted repair of VVF as a safe and effective technique, given its low recurrence rate coupled with minimal EBL, short hospital stays and a low complication rate. However, one of the most frequently stated concerns regarding RAS is its potential for increases in OT, owing to setup and docking requirements [9, 30]. It has been well established in the literature that TA minimally invasive approaches are associated with a longer OT than the TV route [31]. However, it is less clear if the robotics-assisted repair of VVF involves, in fact, a longer OT than the other TA alternatives (conventional laparoscopy and laparotomy). Evidence from Gupta et al. [9] demonstrates that OTs are comparable between RAS and open abdominal surgery. In their retrospective comparative study of recurrent VVF repair, the mean OT was 140 min for robotics-assisted repair and 148.5 min for open repair, with no statistically significant difference observed (p > 0.05). However, given the small sample size (N = 32) and retrospective nature of the study, the authors concluded that, regarding this specific point, future comparative prospective studies with larger population pools are needed to clarify further [9]. With respect to conventional laparoscopy, no comparative studies were conducted comparing the OT of RAS and this technique. The present review’s overall mean OT of 169.0 ± 69.5 min is longer than some of the largest previously published case series using only conventional laparoscopy, such as those by Miklos and Moore [32] (n = 44), Sharma et al. [33] (n = 22), and Tilala et al. [34] (n = 36), which reported mean OTs of 144.8, 140, and 128.5 min respectively. That said, some of the longer mean OTs reported in the present review have been attributed by authors to factors such as limited surgeon experience—as in the initial series of five patients by Sundaram et al. [24] (233 min)—or the performance of concomitant procedures, as described by Agrawal et al. [4] (214 min), where three patients underwent ureteral reimplantation and one required a sigmoid colectomy [4, 13]. Furthermore, some authors point to other variables that could significantly impact OT, such as non-uniformity in reporting methodology, as few doctors record the console time [26], the use or omission of tissue interposition, and the type of interpositional tissue employed [23]. Another important factor is the surgical technique adopted for VVF repair. An increasing trend has been observed toward the use of a modified O’Conor technique—the Mini O’Conor—which entails a limited posterior cystotomy rather than full-bladder bivalving. This approach has been associated with reduced OT, as reported by Hemal et al. [21]. In conclusion, in order to definitively state that robotics-assisted VVF repair entails longer OT than conventional laparoscopy, future comparative studies need to control for the aforementioned confounding factors.

Robotics-assisted surgical repair of VVF may serve as a particularly valuable approach in the management of recurrent cases, which are often surgically challenging owing to increased technical complexity. Two case series focused specifically on recurrent fistulas [9, 21]; in total, these included 19 patients with a total of 25 previous attempts at fistula repair (16 TA and 9 TV), given that some patients underwent multiple previous attempts. Recurrence rate was 0% in both studies; however, although Hemal et al. [21] reported a follow-up period of 3–12 months, Gupta et al. [9] did not report on their follow-up period. Both studies achieved this result with a low EBL (means of 90 and 88 ml respectively), a POC rate of 0%, short mean hospital stays (3 and 3.1 days respectively), and mean OT inferior to the review’s overall combined mean (141 and 140 min versus 169 min) [9, 21]. As previously mentioned, Gupta et al. [9] also compared outcomes of RAS with a matched control group of patients who underwent open repair for recurrent VVF in their retrospective study. Their findings demonstrated that RAS was associated with reduced blood loss (88 ml vs 170 ml; p < 0.05) and shorter hospital stays (3.1 days vs 5.6 days; p < 0.05), while achieving comparable success rates and OTs. These results suggest that RAS might be a safe and effective option for the repair of recurrent VVF and might offer advantages over open surgery in this context. However, to date, and to our knowledge, no comparative studies have evaluated RAS against conventional laparoscopy in this specific patient subgroup.

In established laparoscopic series, Miklos and Moore [32] reported a mean OT of 144.8 min and a mean EBL of 51 ml following laparoscopic EV VVF repair, with a short mean hospital stay of 1.2 days, no major POC, and a 98% success rate at a mean follow-up of 17.3 months. Similarly, Tilala et al. [34], in a comparative analysis of laparoscopic TV and EV repairs, observed OT of 117 and 140 min respectively, with greater blood loss (up to 210 ml) and longer hospitalization associated with the TV approach, while reporting success rates of 94.45% and 83.33% at mean follow-up durations of approximately 13–15 months. Differences between robotics-assisted and laparoscopic outcomes likely reflect the heterogeneity in case selection, including the proportion of recurrent and complex fistulas in robotics-assisted series, as well as variation in perioperative protocols and follow-up duration. Overall, available evidence suggests that robotics-assisted VVF repair achieves outcomes comparable with well-established laparoscopic techniques, although the predominance of retrospective studies and heterogeneous reporting limits definitive conclusions regarding comparative effectiveness. POC profiles were similarly favorable across approaches. Neither Miklos and Moore [32] nor Tilala et al. [34] reported any major POC (Clavien–Dindo III–IV) in their laparoscopic cohort. These findings mirror the low major complication rates reported in robotics-assisted series, supporting the safety of both minimally invasive approaches. Finally, regarding success outcomes, Miklos and Moore [32] reported a similar success rate (98%) to the one demonstrated in the present review (98%), whereas Tilala et al. [34] reported a lower success rate in the TV group (83.33%) than in the EV group (94.45%), although the difference was not statistically significant. Differences in reported success rates may reflect heterogeneity in follow-up duration, inclusion of recurrent or complex fistulas, and variability in definitions of success. Overall, although OT may be longer in robotics-assisted series, perioperative morbidity and fistula closure outcomes appear comparable with established laparoscopic techniques. However, both laparoscopic and robotics-assisted studies predominantly consist of small, retrospective cohorts with variable case complexity and surgeon experience, which limits direct comparison and precludes definitive conclusions regarding the superiority of one minimally invasive approach over another. Taken together, these findings suggest that robotics-assisted VVF repair offers outcomes comparable with established laparoscopic techniques, with potential advantages related to ergonomics and suturing dexterity, although current evidence remains insufficient to support routine adoption over laparoscopy in the absence of prospective, comparative, and cost-effectiveness data.

The transvaginal approach has been consistently associated with primary closure rates of 82–98% across multiple single-center series [35,36,37,38,39], with higher success rates reported as surgical experience increases. Regarding its comparison with RAS in the treatment of VVF, a recent systematic review and meta-analysis by Srivastava et al. [40] specifically analyzed transvaginal repair in a series of surgically challenging cases, such as supratrigonal, juxtacervical, vault, and apical VVF, and reported a pooled success rate of 86.3%, notably lower than the 97.1% success rate observed in the present review of robotics-assisted repairs (defined as lack of recurrence in each follow-up period). These findings suggest that although the transvaginal approach may be highly effective in selected patients, cases involving certain fistula characteristics, such as high-lying or complex ones, may potentially benefit from a TA, and specifically robotics-assisted, approach. Nonetheless, further prospective and comparative evidence is required to establish the optimal surgical route in such patient subgroups.

The findings of this review are consistent with those reported by Lecoanet et al. [25], who conducted the first retrospective multicenter study comparing robotics-assisted TV and EV approaches to VVF repair. Both studies demonstrated no statistically significant differences between the two techniques across multiple outcomes, including OT, POC rate, and recurrence rate. Although not reaching statistical significance in either study, there was a tendency toward shorter OTs in the EV group, which may reflect the less invasive nature of this approach, as it avoids bladder bivalving and allows for more limited dissection. Conversely, both the present review and the study by Lecoanet et al. [25] observed a numerically higher rate of POCs in the EV group, including one case of ureteral injury in their series. This could be explained by the greater technical challenge of dissecting the vesicovaginal space without direct visualization of the bladder interior or ureteral orifices, particularly in patients with prior pelvic surgery. These procedural differences may account for the subtle trends observed in perioperative outcomes, despite the overall similarity in clinical efficacy between the two techniques. These findings are in accordance with previous evidence of surgical outcome similarity between both approaches in the context of laparoscopic repair of VVF: a systematic review in 2015 by Miklos et al. [5] compared the EV and TV approaches in the context of minimally invasive TA surgery (only 12.84% were RAS) and also found no statistically significant difference in outcomes between techniques. Since then, a recent (2024) comparative analysis study by Tilala et al. [34] comparing the two techniques in a conventional laparoscopy setting, found the EV technique to be associated with significantly shorter OT, lower EBL, reduced LOS, and fewer POCs. Although this study was not randomized and did not utilize RAS, its findings suggest that some of the outcome trends observed in the present review, specifically regarding OT, might reach statistical significance as further comparative evidence becomes available. Nevertheless, given the current evidence, it is our view that the choice between an EV or TV approach to VVF repair should primarily be guided by the surgeon’s experience and familiarity with the technique. That said, patient- and fistula-specific factors, such as a history of prior pelvic or fistula surgery, as well as fistula location near the ureteric orifices, should also be taken into account, given that in these situations, the TV approach could offer an easier and potentially safer pathway, avoiding ureteric injury. This is illustrated by the findings of Zhou et al. [10], who compared laparoscopic EV and TV approaches in a cohort of patients with dense adhesions between the bladder and vagina; in their study, OTs were significantly longer in the EV group (p < 0.001), suggesting that the TV technique might have been more suitable for these patients.

Despite the promising results observed in this review, the widespread implementation of RAS for VVFs remains limited, particularly in low-resource settings. In these regions, where the burden of obstetric fistulas is highest, access to robotics-assisted platforms is virtually non-existent owing to infrastructure limitations, cost constraints, and a lack of trained personnel [41, 42]. This is especially relevant given that the majority of VVFs in these underserved regions are still of obstetric origin, which are typically larger, more complex, and associated with significant tissue loss owing to prolonged obstructed labor [43, 44]. Additionally, there is to our knowledge currently no large-scale cost-effectiveness analysis published for robotics-assisted repair of VVF, despite the well-recognized high capital and maintenance costs associated with RAS [21, 25]. Although robotics-assisted approaches may reduce morbidity, hospital stay, and potentially long-term costs by minimizing complications and recurrence, these theoretical benefits must be weighed against the substantial upfront investment. This economic factor remains a significant barrier to broader adoption, particularly in public health systems and resource-constrained environments. Future research should explore not only clinical outcomes (particularly in comparison with conventional laparoscopy) but also cost-effectiveness metrics to better inform policy decisions and global applicability.

This review has some limitations, such as the fact that most included studies were retrospective case series with small sample sizes, which may introduce selection and reporting biases. There was considerable heterogeneity among studies regarding surgical techniques, patient characteristics, and outcome definitions, as well as small sample sizes, particularly in the EV group. Furthermore, the absence of long-term follow-up in many cases limits the ability to assess the durability of fistula repair over time. Several potentially relevant outcomes, including EBL, LOS, and urethral CT, could not be statistically compared owing to incomplete or inconsistent reporting of summary statistics. Although methods exist to estimate means and SDs from medians and associated measures of dispersion, these were not applied given the small study sizes and substantial heterogeneity across studies, and this represents an important limitation of the present review. Last, publication bias cannot be excluded, as studies with negative outcomes may be underrepresented in the literature.