Abstract
Introduction:
Ureteroureterostomy is a viable option for the management of duplex ureters with associated obstruction, reflux, or incontinence in upper or lower systems. Recent advancements in pediatric urology have fostered a growing interest in comparing the safety, efficacy, and outcomes of the open, laparoscopic, and robotic techniques. Herein, this systematic review and network meta-analysis (NMA) aim to evaluate the safety, efficacy, and outcomes of these three surgical approaches in pediatric patients.
Methods:
A systematic literature review of Medline, Embase, and Web of Science identified studies comparing these surgical approaches in children (<18 years old). NMA assessed treatment effects using log odds ratios and mean differences with random effects. Heterogeneity and inconsistency were evaluated, and the risk of bias was assessed using the ROBINS-I tool.
Results:
Seven retrospective studies with 277 children (76 robotic, 124 laparoscopic, and 77 open) were included. There was no difference in age at operation across surgical techniques. No statistically significant difference in operative time was observed, but significant heterogeneity was noted (I2 = 75.2%, p = 0.0012). Overall complication rates ranged from 0% to 31.6% and were similar across all techniques. Clavien–Dindo ≥ 3 complications were low (0–11.1%) and did not significantly differ between approaches. Surgical success rates were comparable across techniques. Funnel plot demonstrated minimal likelihood of publication bias. ROBINS-I risk of bias evaluation demonstrated a serious risk of bias because of confounding from retrospective study designs.
Conclusion:
Based on our NMA, no surgical approach demonstrated clear superiority in operative time, complications, or success rates. Therefore, surgeons should choose the technique of ureteroureterostomy based on their expertise, comfort, and patient factors.
Introduction
Ureteroureterostomy is a viable approach for managing duplex ureters associated with obstruction, reflux, or incontinence. Recent advancements in pediatric urology have spurred interest in comparing the safety, efficacy, and outcomes of newer techniques, including conventional laparoscopic and robot-assisted laparoscopic ureteroureterostomy, against the gold-standard open approach. Although studies have demonstrated the feasibility of each technique, a consensus on their comparative effectiveness—particularly regarding operative time, complication rates, and surgical success—remains lacking.1–3
As with other procedures offering both open and minimally invasive options, each technique presents distinct advantages and limitations.1,4 Compared with open surgery, conventional laparoscopy provides improved cosmesis and reduced recovery time but demands advanced expertise and has a steep learning curve. 5 Robot-assisted laparoscopy, increasingly utilized in pediatric urology, offers enhanced dexterity, three-dimensional visualization, and greater precision. However, concerns persist regarding its higher cost and prolonged operative times. 6
Currently, there is no literature comparing the outcomes of all three techniques to each other. Therefore, we aim to compare the outcomes of robotic, laparoscopic, and open ureteroureterostomy in pediatric patients by performing a systematic review of the literature and network meta-analysis. We hypothesize that the robotic and laparoscopic surgeries will have longer operative times compared with open surgery, but all surgical approaches will have similar outcomes and complication rates.
Methods
The study protocol was registered on PROSPERO (CRD420250654377). Medline, Embase, and Web of Science were searched in February 2025. Search strategy included a combination and variation of search terms: ureteroureterostomy, robotic, and laparoscopy (Supplementary Appendix A).
The title/abstracts were independently screened by two authors. Full texts were reviewed by the first author and verified by a second author. Inclusion criterion included any studies comparing outcomes of any ureteroureterostomy techniques for duplex ureteral systems (open, laparoscopic, or robot-assisted laparoscopic) in children (<18 years of age). Studies of all languages and designs were included. Studies were excluded if involving an adult population, reporting outcomes of a single technique without comparison, or non-primary studies (e.g., reviews).
Baseline characteristics and results from included studies were extracted by the first author and verified by a second author. Any discrepancies were resolved through consensus between the two authors. The dataset included dichotomous outcome data, specifically the number of events (any reported complications, Clavien–Dindo ≥ 3 complications, operative success) in the experimental and control groups, along with their corresponding sample sizes, and continuous outcome data for time-related outcomes (operative time).
Among outcomes reported by 2 or more studies in each technique comparison, network meta-analysis was conducted using a random-effects model, accounting for between-study heterogeneity. The treatment effect estimates were measured using the log odds ratio or mean difference.
A network plot was generated to visualize the structure of treatment comparisons, where edges were weighted based on the number of studies contributing to each comparison. A series of forest plots was constructed to display the summary effect estimates for each treatment comparison. Heterogeneity was evaluated using I2 and tau. 2 The reference treatment was set to open surgery.
To assess the consistency and ranking of treatments, a net heat plot was created, highlighting the contribution of direct evidence (data from head-to-head comparison studies) and indirect evidence (derived from network analysis) to the overall estimates. To explore potential small-study effects and publication bias, a comparison-adjusted funnel plot was generated.
Risk of bias assessment was performed using Cochrane’s ROBINS-I (Intervention) tool by two independent authors. When there was a discrepancy, consensus was achieved as the final risk of bias assessment between the two authors. The robvis tool was used to create risk of bias plots. 7
All analyses were performed using R Studio (version 2024.04.0 + 735).
Results
A total of 1278 articles were identified (Fig. 1). Following removal of duplicate entries, 856 were screened. Following title/abstract screening, 41 full texts were reviewed. Thirty-four articles were excluded as they did not meet inclusion criterion. Seven retrospective studies were included in the systematic review and meta-analysis with 277 patients (76 robotic, 124 laparoscopic, 77 open).2–6,8,9
PRISMA flow diagram of the literature search.
Baseline characteristics are summarized in Table 1. Two studies compared robotic and laparoscopic approaches, two studies compared robotic and open approaches, and three studies compared laparoscopic and open approaches. The age at intervention was statistically similar across interventions (p = 0.28), although those undergoing open interventions appeared to be lower on qualitative assessment (Fig. 2).
Boxplot comparison of median age of surgery across each intervention group.
Summary of Study Characteristics
F = female; M = male.
Outcomes reported by each study are summarized in Table 2. Estimated blood loss, post-operative hospitalization duration, and proportion of stent, drain, or urethral catheterization were not meta-analyzed, as there were less than two studies in one or more comparison arms.
Summary of Reported Outcomes
EBL = estimated blood loss; NR = not reported.
Operative time
All seven studies included in this meta-analysis reported operative times (Fig. 3A). There was no statistically significant difference in operative times between the three surgical approaches (Fig. 4A), although there were slightly higher operative times for robotic and laparoscopic approaches on qualitative evaluation. There was significant heterogeneity with I2 of 75.2% (95% CI 44.0%−89.0%) and tau 2 of 385.7. The total Q-test for heterogeneity was significant (Q =20.18, df = 5, p = 0.0012) with significant heterogeneity within designs (Q = 18.38, df = 4, p = 0.0010), whereas inconsistency between designs was not significant (Q = 1.80, df = 1, p = 0.180).
Network plot of studies being evaluated in the network meta-analysis;
Forest plot comparing open approach to other approaches for
Overall complications
All seven studies contributed to the network meta-analysis for overall complications (Fig. 3A). The complication rates ranged from 0 to 31.6%. The overall complication rates were similar across all approaches on network meta-analysis (Fig. 4B). There was minimal heterogeneity with I2 of 0% (95% CI 0–74.6%) and tau 2 of 0. The total Q-test for heterogeneity was not significant (Q =0.57, df = 5, p = 0.985). The heterogeneity within designs (Q = 0.40, df = 4, p = 0.983) and inconsistency between designs were both not significant (Q = 0.27, df = 1, p = 0.604).
Clavien–Dindo ≥ 3 complications
Although all seven studies reported Clavien–Dindo ≥ 3 outcomes, two studies both reported no complications in both comparison groups. Therefore, five studies contributed to the network meta-analysis for Clavien–Dindo ≥ 3 complications (Fig. 3B). The rates of Clavien–Dindo ≥ 3 complications were lower than that of overall complications, ranging from 0 to 11.1%. Clavien–Dindo ≥ 3 complications were also similar across different interventions (Fig. 4C). There was minimal heterogeneity with I2 of 0% (95% CI 0–84.7%) and tau 2 of 0. The total Q-test for heterogeneity was not significant (Q =0.05, df = 3, p = 0.997). The heterogeneity within designs (Q = 0.03, df = 2, p = 0.986) and inconsistency between designs were both not significant (Q = 0.02, df = 1, p = 0.884). We present the type of complications by Clavien–Dindo classification, as well as interventions received for Clavien–Dindo ≥ 3 complications in Table 2.
Success
Although all seven studies reported surgery success rates, one study reported 100% success rate in both comparison groups; therefore, six studies contributed to the network meta-analysis (Fig. 3C). The likelihood of success was similar across the three interventions (Fig. 4D). There was minimal heterogeneity with I2 of 0% (95% CI 0–79.2%) and tau 2 of 0. The total Q-test for heterogeneity was not significant (Q =0.97, df = 4, p = 0.914). The heterogeneity within designs (Q = 0.78, df = 3, p = 0.855), and inconsistency between designs were both not significant (Q = 0.20, df = 1, p = 0.658). The definition of operative success used by each study is summarized in Supplementary Table S1.
Contribution of direct and indirect evidence
The contribution matrix was used to assess the contribution of direct and indirect evidence in the network meta-analysis and to evaluate potential inconsistencies. The matrix revealed that for operative time, open vs robotic comparison relied on the most indirect evidence, whereas other comparisons had a high proportion of direct evidence (Supplementary Fig S1A). For complications, the laparoscopic vs open comparison had the most indirect evidence (Supplementary Fig S1B/C). For success, the laparoscopic vs robotic comparison had the most indirect evidence (Supplementary Fig S1D).
Publication bias
There was minimal publication bias that could be evaluated based on a funnel plot with balanced distribution of studies close to the observed effect size for all evaluated outcomes (Fig. 5).
Funnel plot to evaluate publication bias based on pooled outcomes;
Risk of bias assessment
Risk of bias assessments using ROBINS-I tool showed a serious risk of bias in all studies (Supplementary Fig. S2), mostly because of bias due to confounding, as all included studies were retrospective cohort studies.
Discussion
This systematic review and network meta-analysis provide a comparative evaluation of robot-assisted laparoscopic ureteroureterostomy, conventional laparoscopic ureteroureterostomy, and open ureteroureterostomy in pediatric patients. Our findings from network meta-analysis of current evidence demonstrate that all three surgical approaches are effective in achieving high success rates.
Contrary to our hypothesis, there was no statistically significant difference in operative times between the three techniques, though robotic and laparoscopic approaches demonstrated slightly longer times on qualitative evaluation. It is likely that as robotic platforms become more integrated into the field of pediatric urology, improvements in efficiency may reduce operative duration. Nonetheless, there was significant heterogeneity identified for operative times, likely because of differences in how they are recorded across different institutions. Moreover, there are potential contributing factors to this variation including but not limited to case complexity, surgeon experience, and concomitant procedures that may not have been well captured in the data synthesis.
Overall complication rates were similar across all surgical techniques, with no statistically significant differences observed. However, the proportion of Clavien–Dindo ≥ 3 complications was low in all groups, suggesting that severe complications are uncommon, regardless of technique. Previous studies have suggested that robot-assisted surgery offers advantages in precision, reducing iatrogenic injury risk, but our findings do not indicate a clear superiority in safety outcomes. 1
Laparoscopic and robotic surgeries are well known for their cosmetic benefits, including smaller incisions, across various surgical procedures.10,11 However, this study did not assess cosmetic outcomes, as the included studies focused primarily on surgical success and clinical outcomes rather than patient-reported measures. Additionally, this study was not designed to compare upper vs lower ureteroureterostomy techniques and, therefore, cannot draw conclusions regarding the impact of anastomotic location on surgical outcomes. Although the location of ureteroureterostomy may be a potential confounder, most studies likely performed anastomosis at the mid-distal ureter. An exception is the study by Lee et al., which reported a higher prevalence of proximal anastomosis in robotic procedures compared with open surgery. Notably, prior research suggests that the success rates of upper and lower ureteroureterostomy techniques are comparable, indicating that anastomotic location may not significantly influence outcomes.12–14
This study is subject to several limitations. The included studies were retrospective in nature, with a high risk of bias from uncontrolled confounding. There may also be heterogeneity in study design, patient populations, and definitions of surgical success that may have influenced the results of network meta-analysis. Being a meta-analysis of published data, no individual-level data could be analyzed, and all studies identified were retrospective and non-randomized. Although robotic surgery is associated with higher costs, this meta-analysis was not designed to assess cost-effectiveness. Moreover, the wide confidence intervals in our pooled outcomes limit the ability to draw definitive conclusions, highlighting the need for higher-quality studies to control for potential confounding and bias. Nevertheless, funnel plots suggest minimal influence from confounding bias, with negligible heterogeneity in complication and success rates, indicating that reported outcomes across studies are likely consistent and reliable. Although reoperation rates themselves were not meta-analyzed, these would be part of the Clavien–Dindo ≥ 3 complications, and the confidence intervals would be smaller because of a smaller number of events, indicating that there is also a high likelihood of no difference in reoperation rates as well. Well-designed prospective studies are needed to better delineate the comparative advantages of each technique, particularly regarding long-term outcomes.
Conclusion
Our network meta-analysis of studies comparing robot-assisted, laparoscopy-assisted, and open surgical approaches to ureteroureterostomy suggests there is no significant benefit of one modality over another based on currently available studies. Based on current evidence, surgeons should determine the surgical approach based on their training and comfort to ensure the best outcomes for their patients.
Ethics
No patients were involved in this systematic review, and therefore ethical review was not required.
Availability of Data
Data will be made available to any individuals who establishes a data-sharing agreement.
Declaration of Generative AI and AI-Assisted Technologies in the Writing Process
Generative AI and AI-assisted technologies were not used in the preparation of this work.
Authors’ Contributions
J.K.K.: Conceptualization, methodology, formal analysis, investigation, data curation, writing—original draft, writing—review and editing, visualization, and project administration. N.V.B.: Conceptualization, methodology, and writing—review and editing. R.M.: Conceptualization, methodology, writing—review and editing, and project administration. K.M.S.: Conceptualization, methodology, writing—review and editing, and project administration. K.M.: Conceptualization, methodology, and writing—review and editing. M.K.: Conceptualization, methodology, writing—review and editing, and project administration. J.R.: Conceptualization, methodology, and writing—review and editing. R.C.R.: Conceptualization, methodology, and writing—review and editing. M.P.C.: Conceptualization, methodology, and writing—review and editing. P.D.: Conceptualization, methodology, and writing—review and editing. B.M.W.: Conceptualization, methodology, writing—original draft, writing—review and editing, project administration, and resources.
Footnotes
Author Disclosure Statement
The authors do not have any conflicts of interest to declare.
Funding Information
No funding was received for this study.
Supplemental Material
Supplemental Material
Supplemental Material
Supplemental Material
Abbreviations
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
