The efficacy and safety of robotic, laparoscopic, and open surgery for pediatric choledochal cysts: a systematic review and network meta-analysis

Abstract

Background

Pediatric choledochal cysts (CCs) are typically managed via surgical treatment, and a growing number of pediatric patients with CC are undergoing robotic surgery. However, whether it offers superior safety and efficacy compared to open surgery and laparoscopic surgery remains unclear. The primary aim of this study was to compare the efficacy and safety of three surgical approaches [open cyst excision with a Roux-en-Y hepaticojejunostomy (OP), laparoscopic cyst excision with a Roux-en-Y hepaticojejunostomy (LA), and robotic cyst excision with a Roux-en-Y hepaticojejunostomy (RO)] for CC in children.

Methods

PubMed, Embase, Cochrane Library, and Web of Science databases were systematically searched. Two independent reviewers assessed the studies for eligibility and quality and then extracted the relevant data. Confidence in Network Meta-Analysis (CINeMA) and the Newcastle-Ottawa Scale (NOS) were applied for assessing quality. Direct statistical analyses were carried out via Stata 14.0 (StataCorp), while a network meta-analysis was performed through use of using ADDIS 1.16.18 (GetReal Initiative, Innovative Medicines Initiative).

Results

A total of 18 studies were included for analysis. OP had the significantly shortest operative time; RO had the significantly shortest hospital stay; and LA had the significantly lowest incidence of postoperative bowel obstruction.

Conclusions

RO is effective and safe for the management of CC. The opportunity to convert to open surgery should not be missed due to fears of excessive bleeding or increased postoperative bile leakage.

Highlight box.

Key findings

• Robotic surgery for pediatric choledochal cysts (CCs) is both effective and safe.

What is known and what is new?

• Avoiding conversion to open surgery for fear of excessive bleeding or increased bile leakage after surgery may be unnecessary and represent a missed opportunity.

• Enhanced technical training could help reduce the incidence of postoperative bowel obstruction in robotic surgery.

What is the implication, and what should change now?

• This study provides evidence-based support for the selection of robotic surgery for the treatment of pediatric CCs.

Introduction

Pediatric choledochal cyst (CC) is a common clinical biliary malformation and congenital abnormality of children. In North America, the incidence of CC is estimated to be 1 in 150,000 (1), but it is more common in Asian countries, including China, Japan, and South Korea, where the incidence is 1,000 cases per one million people (2-4). CC has a high risk of progressing to severe hepatobiliary complications such as cholangitis, pancreatitis, and even cancer, and thus prompt treatment is essential.

The primary treatment for CC is complete resection of the cyst via a Roux-en-Y hepaticojejunostomy (HJ), with three surgical approached being available: open cyst excision with a Roux-en-Y HJ (OP), laparoscopic cyst excision with a Roux-en-Y HJ (LA), and robotic cyst excision with a Roux-en-Y HJ (RO). A host of studies have compared the efficacy of these three surgical approaches, but only one systematic review and meta-analysis (5) has compared the efficacy between open cyst excision and laparoscopic cyst excision with a hepaticoduodenostomy (HD) or HJ. In one systematic review and meta-analysis (6), it was found that HD led to a greater higher of postoperative reflux/gastritis than did HJ but was associated with a shorter hospital stay, and thus the results were inconclusive.

Significant controversies persist regarding the comparative efficacy and safety of the three surgical methods for managing CC. The available evidence indicates that open surgery achieves the shortest operative time (7-9); however, it is associated with significantly greater intraoperative blood loss (9,10) and higher bile leakage rates (11) compared to minimally invasive techniques. The findings regarding whether robotic or laparoscopic surgery offers superior outcomes in terms of reduced the length of hospital (12,13) and minimized blood loss (9,12) are conflicting. Furthermore, direct comparative studies between robotic and laparoscopic approaches remain limited. Therefore, we conducted network meta-analysis to comprehensively evaluate the safety and efficacy profiles of OP, LA, and RO for CC. We present this article in accordance with the PRISMA NMA reporting checklist (available at https://tp.amegroups.com/article/view/10.21037/tp-2025-102/rc) (14).

Methods

This meta-analysis study was registered in the PROSPERO database (registration No. CRD42021285421). All stages of study identification, selection, quality assessment, and data extraction were carried out independently by two reviewers (H.C. and X.W.). Any discrepancies were resolved through consultation with a third reviewer (S.S.).

Study identification

The PubMed, Embase, Cochrane Library, and Web of Science databases were systematically searched for relevant studies. Reference lists of these studies were searched manually via the “related articles” function in PubMed. Moreover, in order to improve literature searches in the network meta-analyses, we consulted an information specialist (15). In PubMed and Cochrane Library, the search strategy included combining text words and medical subject heading (MeSH) terms. Meanwhile, for the Cochrane Library, the strategy included text words and Emtree. However, for Web of Science, only text words were applied. Search terms included “choledochal cyst”, “laparoscopy”, “robotic surgical approaches”, and “operative surgical approaches”, which were all used in combination with the Boolean operators AND, OR, and NOT. The literature search was completed in June 2025. The detailed search strategy for PubMed is summarized in Table S1.

Study selection

We employed the population, intervention, comparison, outcome, and study design (PICOS) framework for study inclusion as follows: the population included children with CC; the intervention included surgical approaches, specifically OP, LA, and RO; the comparison involved these surgical approaches acting as comparators to one another; outcomes included studies reporting at least one of operative time, length of hospital stay, intraoperative blood loss, postoperative bile leakage rates, or postoperative intestinal obstruction rates; and the study design included randomized controlled trials (RCTs) or cohort studies.

Articles in which the outcomes of interest could not be calculated, nonhuman studies, editorials, and comments were excluded. Full-text articles were obtained if the inclusion criteria were fulfilled.

Risk of bias assessment

The risk of bias for cohort studies enrolled in the systematic review and meta-analysis was assessed according to the Newcastle-Ottawa Scale (NOS).

Quality of evidence assessment

Confidence in Network Meta-Analysis (CINeMA) version 1.9.1 (https://cinema.ispm.unibe.ch/) based on grading of recommendations assessment, development, and evaluation (GRADE) methodology was applied for assessing quality of evidence and reported in the results (16).

Outcomes measures

We defined the primary outcomes as operation time, length of hospital stay, and intraoperative blood loss. The secondary outcomes measured were complications that included the number and rates of bile leakage and intestinal obstruction after operation.

Data extraction

Data collected for each included article comprised the following predefined items: (I) study identifier (first author and year of publication); (II) type of surgical treatment and number of participants; (II) baseline characteristics of study participants (sex, mean age, and weight); and (IV) treatment outcomes, as described above.

Statistical analysis

Direct meta-analysis

The statistical analyses were carried with Stata software version 14.0 (StataCorp, College Station, TX, USA). For each study, the odds ratio (OR) with 95% confidence interval (CI) was calculated for dichotomous variables, and the mean difference (MD) with 95% CI was calculated for continuous variables. The point estimate of the OR value was considered to be statistically significant at a P level of less than 0.05 if the 95% CI did not exceed 1. The point estimate of the MD value was considered statistically significant at a P level of less than 0.05 if the 95% CI did not exceed 0.

Network meta-analysis

The statistical analyses were performed via ADDIS software 1.16.8 (GetReal Initiative, Innovative Medicines Initiative).

Network meta-analysis using Bayesian methods was based on Markov chain Monte Carlo (MCMC) simulation. The network meta-analysis employed Bayesian computation with four Markov chains. According to a 20,000-iteration burn-in period, posterior sampling comprised 50,000 iterations with a thinning interval of 10. Inference was derived from 10,000 posterior samples generated with a variance scaling factor of 2.5 (17). The Brooks-Gelman-Rubin statistical method was used to evaluate the convergence of data. Data achieved good convergence, if the potential scale reduction factors (PSRFs) were close to 1 (18).

Because the structure of all comparisons forms a closed loop, rather than a simple star, testing for inconsistency is necessary (19,20). For comparisons between two interventions with both direct and indirect evidence, the consistency between these types of evidence was identified via node-split analysis. In the absence of significant inconsistency, a Bayesian consistency model was employed to synthesize relative outcome data. Bayesian network meta-analysis is fundamentally a framework integrating direct and indirect evidence. If significant inconsistency occurred—potentially attributable to variations in study design or effect modifiers causing discordance between direct and indirect evidence—direct evidence estimates for relative outcomes were prioritized.

Results

There were 7,384 references identified from PubMed (n=2,529), Embase (n=517), Cochrane Library (n=17), and Web of Science (n=4,321) in the initial retrieval, but no additional records were identified through other sources. Initially, 1,388 duplicate records were removed, and another 5,943 studies that deviated from the inclusion criteria were excluded after a review of the titles and abstracts. From the remaining 53 studies, the full-texts were reviewed, and finally, 18 studies (7-13,21-31) were included, comprising 2,199 participants. The process of identification and selection of studies is depicted in Figure 1.

Figure 1.

PRISMA flowchart depicting the search strategy and selection of articles for the review.

Characteristics of eligible studies

The general characteristics of patients as reported in the studies are summarized in Table 1. One study (9) compared the three examined intervention methods, three (9,12,13) studies compared RO with LA, 15 studies (8-11,21-31) compared LA with OP, and two studies (7,9) compared RO with OP. All of the included studies were cohort studies.

Table 1. Characteristics of the included studies.

Study	Intervention	Cases, n	Males, n	Females, n	Age†	Weight (kg)†	Type of study	NOS score
Kim NY (7)	RO	36	6	30	4.79±4.63 y	19.4±14.2	Cohort study	9
	RO	36	6	30	4.79±4.63 y	19.4±14.2
Lee C (8)	LA	76	22	56	3.42±3.28 y	14.2±9.4	Cohort study	8
	OP	109	20	89	3.9±3.48 y	13.9±10.5
Xie X (9)	RO	41	10	31	4 [2.54–6.46] y	18.74±11.44	Cohort study	9
	LA	104	25	79	2.33 [0.73–4.42] y	13.06±6.06
	OP	226	52	174	2.79 [1.48–5] y	14.48±8.05
Yu BH (10)	LA	70	39	31	5.6±3.3 y	NR	Cohort study	8
	OP	86	42	44	5.6±3.3 y	NR
van den Eijnden MHA (11)	LA	12	NR	NR	1.1 [0.03–8.9] y	9.0 [2.8–28]	Cohort study	8
	OP	79	NR	NR	2.3 [0.06–17.7] y	11.0 [3.7–52.5]
Chi SQ (12)	RO	70	22	48	NR	NR	Cohort study	8
	LA	70	22	48	NR	NR
Koga H (13)	RO	10	NR	NR	5.6±3.4 y	18.7±8.2	Cohort study	8
	LA	27	NR	NR	5.2±3.8 y	18.5±11.6
Aspelund G (21)	LA	4	NR	NR	4.42±3.5 y	NR	Cohort study	9
	OP	12	NR	NR	5.5±4.5 y	NR
Urushihara N (22)	LA	10	4	6	117 [20–268] d	6.4 [2.8–8.7]	Cohort study	8
	OP	11	2	9	39 [8–270] d	4.1 [2.9–8.8]
Miyano G (23)	LA	27	4	23	3.21 [0.17–2.05] y	12.7 [3.0–30.0]	Cohort study	8
	OP	31	6	25	3.52 [0.08–15.83] y	13.3 [2.9–39.6]
Guo WL (24)	LA	23	NR	NR	NR	NR	Cohort study	7
	OP	42	NR	NR	NR	NR
Liem NT (25)	LA	115	NR	NR	NR	NR	Cohort study	6
	OP	261	NR	NR	NR	NR
Matsumoto M (26)	LA	6	2	4	39 [8–270] d	3.35 [3.09–3.70]	Cohort study	8
	OP	7	2	7	34 [8–550] d	3.5 [3.27–3.85]
Diao M (27)	LA	218	56	162	4.16 y [7 d–18 y]	NR	Cohort study	8
	OP	200	51	149	4.59 y [13 d–17 y]	NR
Cherqaoui A (28)	LA	9	NR	NR	53.71 [12–156] d	NR	Cohort study	9
	OP	10	NR	NR	62.5 [12–192] d	NR
Ng JL (29)	LA	13	5	8	3.04 y	NR	Cohort study	9
	OP	22	3	19	3.04 y	NR
Ryu HS (30)	LA	22	3	19	14 [7–22] d	3.35 [3.09–3.70]	Cohort study	8
	OP	21	4	17	13 [9.5–21] d	3.5 [3.27–3.85]
Liuming H (31)	LA	39	NR	NR	5 [0.25–13] y	13.5 [5.1–37]	Cohort study	8
	OP	38	NR	NR	4 [0.16–15] y	12 [4.6–43]

^†, data are presented as mean ± SD or mean [range]. d, days; LA, laparoscopic cyst excision with a Roux-en-Y hepaticojejunostomy; OP, open cyst excision with a Roux-en-Y hepaticojejunostomy; RO, robotic cyst excision with a Roux-en-Y hepaticojejunostomy; n, number; NOS, Newcastle-Ottawa Scale; NR, not reported; SD, standard deviation; y, years.

Risk-of-bias, heterogeneity, and quality of evidence assessment

Risk-of-bias assessments were performed, as displayed in Table 1. The NOS scores of all the included studies were between six to nine (details in Figure S1). The results of direct comparison and heterogeneity analysis are shown in Table 2.

Table 2. Results of the direct meta-analysis.

Items	Experimental group vs. control group	K	Effect	Effect value (95% CI)	I2 (%)	Z	P value
Operation time	OP vs. LA	7	Direct estimate MD	−1.22 (−1.56 to −0.88)	95.7	7.07	<0.001
	RO vs. LA	2	Direct estimate MD	−0.54 (−0.67 to −0.41)	0.0	8.07	<0.001
	OP vs. RO	2	Direct estimate MD	−2.13 (−4.22 to −0.03)	97.6	1.99	0.047
Length of hospital stay	LA vs. OP	6	Direct estimate MD	−1.40 (−2.29 to −0.51)	87.4	3.09	0.002
	RO vs. LA	3	Direct estimate MD	−1.41 (−2.82 to −0.00)	95.1	1.96	0.050
	RO vs. OP	2	Direct estimate MD	−1.77 (−3.93 to 0.39)	84.3	1.60	0.11
Intraoperative blood loss	LA vs. OP	3	Direct estimate MD	−117.14 (−241.44 to 7.15)	99.5	1.85	0.07
	RO vs. LA	2	Direct estimate MD	−8.50 (−24.22 to 7.21)	98.5	1.06	0.27
	RO vs. OP	2	Direct estimate MD	47.84 (−114.02 to 209.70)	99.2	0.58	0.56
Bile leakage after operation	LA vs. OP	9	Direct estimate OR	1.35 (0.48 to 3.82)	32.3	0.56	0.58
	RO vs. LA	2	Direct estimate OR	0.31 (0.04 to 2.69)	0.0	1.06	0.29
	RO vs. OP	2	Direct estimate OR	2.41 (0.21 to 27.76)	–	0.71	0.48
Intestinal obstruction after operation	LA vs. OP	5	Direct estimate OR	0.38 (0.10 to 1.40)	0.0	1.45	0.15
	RO vs. LA	1	Direct estimate OR	2.58 (0.16 to 42.17)	–	0.66	0.51
	RO vs. OP	2	Direct estimate OR	2.32 (0.36 to 14.99)	0.0	0.88	0.38

CI, confidence interval; K, number of included studies; LA, laparoscopic cyst excision with a Roux-en-Y hepaticojejunostomy; MD, mean difference; OP, open cyst excision with a Roux-en-Y hepaticojejunostomy; OR, odds ratio; RO, robotic cyst excision with a Roux-en-Y hepaticojejunostomy.

As shown in Figures 2-4, the quality of evidence of network meta-analysis was low for the comparison of operation time and bile leakage after operation in the LA group and the RO group, low for the comparison of bile leakage after operation in the LA group and the OA group, and high for the comparison of operation time in the OP group and the RO group. The outcomes for other comparisons had a moderate quality of evidence (details are provided in Figures S2-S7).

Figure 2.

Strength of evidence and network estimate MD of operation time (OP, LA, and RO). The vertical data for a given intervention represent results obtained when it served as the experimental arm and was compared against other interventions, while the horizontal data represent results obtained when the intervention served as the control arm. Red indicates low quality of evidence; yellow indicates moderate quality of evidence; and green indicates high quality of evidence. LA, laparoscopic cyst excision with a Roux-en-Y hepaticojejunostomy; MD, mean difference; OP, open cyst excision with a Roux-en-Y hepaticojejunostomy; RO, robotic cyst excision with a Roux-en-Y hepaticojejunostomy.

Figure 3.

Strength of evidence and network estimate MD of length of hospital stay and intraoperative blood loss (OP, LA, and RO). The vertical data for a given intervention represent results obtained when the intervention served as the experimental arm and was compared against other interventions, while the horizontal data represent results obtained when the intervention served as the control arm. Yellow indicates moderate quality of evidence; the upper triangle represents outcomes of length of hospital stay; and the lower triangle represents outcomes of intraoperative blood loss. LA, laparoscopic cyst excision with a Roux-en-Y hepaticojejunostomy; MD, mean difference; OP, open cyst excision with a Roux-en-Y hepaticojejunostomy; RO, robotic cyst excision with a Roux-en-Y hepaticojejunostomy.

Figure 4.

Strength of evidence and network estimate OR of length of bile leakage and intestinal obstruction after operation (OP, LA, and RO). The vertical data for a given intervention represent results obtained when it served as the experimental arm and was compared against other interventions, while the horizontal data represent results obtained when the intervention served as the control arm. Red indicates low quality of evidence; yellow indicates moderate quality of evidence; the upper triangle represents outcomes of bile leakage after operation; and the lower triangle represents outcomes of intestinal obstruction after operation. LA, laparoscopic cyst excision with a Roux-en-Y hepaticojejunostomy; OP, open cyst excision with a Roux-en-Y hepaticojejunostomy; OR, odds ratio; RO, robotic cyst excision with a Roux-en-Y hepaticojejunostomy.

Statistical analysis

Nine studies (7-9,13,21,24,25,30,31) reported the operation time. The PSRFs were 1.00, confirming complete model convergence, optimal iteration effect, and stable results. However, node-splitting analysis identified significant differences between the direct and indirect treatment effect estimates for two treatment comparisons (the comparison of LA and OP and the comparison of RO and LA) (Table S2). Compared with the other two surgical approaches (RO and LA), OP had the shortest operation time (network estimate MD: −1.61 hours, 95% CI: −2.86 to −0.30, P<0.05; direct estimate MD: −1.22 hours, 95% CI: −1.56 to −0.88, P<0.001) (Table 2, Figure 2). Moreover, the RO group and LA group were significantly different in terms of operation time (direct estimate MD: −0.54 hours, 95% CI: −0.67 to −0.41, P<0.001) (Table 2).

The length of hospital stay was reported by nine studies (7-9,12,13,21,27,30,31). The PSRFs were 1.00, confirming complete model convergence, optimal iteration effect, and stable results. Meanwhile, node-splitting analysis did not identify any significant differences between the direct and indirect treatment effect estimates for any of the treatment comparisons (Table S3). Among the surgical approaches, the RO technique had the highest probability of obtaining the highest rank (network estimate MD: −1.26 days, 95% CI: −3.19 to 0.57, P<0.05; network estimate MD: −2.19 days, 95% CI: −4.12 to −0.15, P<0.05) (Figure 3). The other two methods had lower probabilities of being the optimal strategy, with LA having a higher probability than OP (Figure S8).

Five studies (7,9,10,12,31) reported intraoperative blood loss. All the PSRFs were 1.00, indicating that the model was convergent and that the result was stable. Furthermore, node-splitting analysis did not identify any significant differences between the direct and indirect treatment effect estimates for any of the treatment comparisons (Table S4). Among the surgical approaches, LA had the highest probability of obtaining the highest rank (network estimate MD: −64.56 mm, 95% CI: −245.04 to 106.05, P>0.05; network estimate MD: −79.56 mm, 95% CI: −232.37 to 74.19, P>0.05) (Figure 3). The other two approaches had lower probabilities of being the best ranking strategy, with RO having a higher probability than OP (Figure S8). However, overall the difference between the three approaches was not statistically significant (Figure 3).

Postoperative bile leakage was reported by eleven studies (7,9-12,22,23,27-29,31). The PSRFs were 1.00, confirming complete model convergence, optimal iteration effect, and stable results. However, node-splitting analysis identified significant differences between the direct and indirect treatment effect estimates for two treatment comparisons (the comparison of LA and OP and the comparison of RO and LA) (Table S5). Among the surgical approaches, the RO technique had the highest probability of obtaining the highest rank (network estimate OR: 0.43, 95% CI: 0.01 to 9.23, P>0.05; direct estimate OR: 0.31, 95% CI: 0.04 to 2.69, P=0.29) (Table S2, Figure 4). The other two approaches had lower probabilities of being the best ranking strategy, with OP having a higher probability than LA (Figure S8).

Six studies (7,9,11,22,26,30) reported postoperative intestinal obstruction after. All the PSRFs were 1.00, which proved that the model was convergent and that the result was stable. Moreover, node-splitting analysis did not identify any significant differences between the direct and indirect treatment effect estimates for any of the treatment comparisons (Table S6). Among the surgical approaches, LA had the highest probability of obtaining the highest rank (network estimate OR: 0.03, 95% CI: 0.00 to 0.93, P<0.05; network estimate OR: 0.07, 95% CI: 0.00 to 0.62, P<0.05) (Figure 4). The other two approaches had lower probabilities of being the best ranking strategy, with RO having a higher probability than OP (Figure S8).

Discussion

To our knowledge, this study was the first to compare OP, LA, and RO for the treatment of CC via network meta-analysis. A total of 18 cohort studies were included, comprising 2199 participants. Previous research has only compared the efficacy of LA to that of RO in treating CC, and most of the related studies have been single-arm in nature, reporting only one surgical approach. In a previous meta-analysis that included 40 studies, 36 reported data on LA, while eight reported data on RO (32). Compared to single-arm studies, two-arm or multiarm studies have lower risks of bias, better result interpretation, easier data comparison, fewer limitations in statistical analysis, and more accurate estimates of treatment effects. Thus, the studies included in our analysis were all two-arm studies.

Although OP is generally less favored due to its association with longer hospital stays, greater trauma, larger incisions, slower recovery, and poorer cosmesis, its inclusion in our analysis was justified for two key reasons. First, both LA and RO for CC may require conversion to OP. Therefore, we examined the advantages and disadvantages of OP to provide evidence for optimizing surgical plans in the case of conversion, as the challenges encountered during conversion to OP are also encountered in OP. Second, due to the limited number of studies directly comparing LA and RO, including OP allowed us to conduct a network meta-analysis with more included articles, reducing bias and enhancing the credibility of the comparison between LA and RO.

Additionally, compared to other approaches, LA for CC involves technical challenges such as difficulties with HJ and suturing, resulting in longer operative times. These technical hurdles likely explain LA’s higher conversion rate to OP (5.9%; 95% CI: 3.4–9.0%) as compared to RO (3.4%; 95% CI: 0.2–10.5%) (32). Research has shown that using an endoscope to clear protein plugs and stones from the distal common channel during cyst removal can reduce the conversion rate (13). When the surgery time exceeds 2 hours, the risk of complications are significantly elevated; specifically, with prolonged surgery time, the complication rate roughly doubles, and the likelihood of complications increased with each additional 30 minutes of operating time (33). In LA and RO, the bleeding in the operative field is difficult to control laparoscopically, anatomical structures are not clear (e.g., inflammatory adhesions), and biliary reconstruction is too difficult, all of which increase the operative time for both of these approaches, and such challenges may lead to intermediate laparotomy. Crucially, the operation time could be shortened after a midstage open abdomen, as the OP group had the shortest operation time, and there were no significant differences in intraoperative blood loss or postoperative bile leakage rates across the three approaches (LA, RO, and OP). Therefore, avoiding conversion to open surgery due to concerns of increased bleeding or bile leakage appear unwarranted.

Our results demonstrated that OP was significantly associated with reduced postoperative hospital stay and the incidence of postoperative bowel obstruction. There are two primary reasons for this. First, cyst anatomy and biliary reconstruction are the most critical parts of the surgery. The more severe the biliary dilation and obstruction are, the more difficult the surgery, leading surgeons to prefer OP for a smoother procedure. Second, parents may implement incorrect postoperative care. Due to the presence of wound dressings, they may be unwilling to allow the child get out of bed and hesitant to initiate early feeding, which can hinder recovery of intestinal function. Therefore, conversion to open surgery itself is unlikely to increase the incidence of postoperative bowel obstruction but rather reflect more severe conditions requiring OP. Educating parents on postoperative care is essential to reducing complications, expediting discharge, and lowering medical costs. For CC, the primary causes of postoperative intestinal obstruction are the formation of intra-abdominal adhesions and complications related to the Roux-en-Y HJ, specifically internal herniation and volvulus involving the Roux limb. Secondary contributing factors include postoperative inflammation, infection (particularly peritonitis and abscesses), and postoperative pancreatitis. Meticulous application of surgical techniques including careful tissue handling, precise hemostasis, thorough closure of mesenteric defects (especially Petersen’s defect), and minimization of contamination, is critical for prevention. Compared to laparoscopic techniques, robotic surgery offers enhanced instrument articulation, superior tremor filtration, and a wider range of motion for suturing, which likely leads to better prevention of postoperative intestinal obstruction. However, the finding in our study of a significantly lower incidence of intestinal obstruction following laparoscopic surgery may be attributed to the inherent advantages of robotic technology enabling surgeons to manage more complex cases in children. Furthermore, studies indicate that increasing surgeon proficiency with robotic techniques can significantly reduce intra-abdominal bleeding (34), thereby decreasing the incidence of intestinal obstruction caused by intra-abdominal adhesions.

Due to the limited reported data, our analysis did not compare overall postoperative complication rates or perform detailed cost analyses, and only postoperative intestinal obstruction was evaluated as a long-term complication. The studies revealed no significant difference in the incidence of overall postoperative complications (9,12) or long-term postoperative complications—such as anastomotic stricture or bile duct stones (8,9,12)—across the three surgical groups of CC. However, we found that the operative time for RO was on average 32.4 minutes less than that of LA. The shorter RO operative times observed here potentially translate to a reduced complication risk, given evidence that complication likelihood increases with operative duration (33). Studies indicate that robotic-assisted CC resection for pediatric patients incurs significantly higher costs compared to laparoscopic approaches (9,35). A model-based cost-effectiveness analysis by Martin et al. indicated that although robotic-assisted surgery involves higher direct costs, they may be offset by the higher hospital stay costs from avoiding open surgery, suggesting robotic-assisted surgery is economically advantageous (36). In the current clinical practice for CC excision, two laparoscopic approaches are used: multiport laparoscopic CC excision with Roux-en-Y hepaticoenterostomy (MPCH) and single-port laparoscopic CC excision with Roux-en-Y hepaticoenterostomy (SPCH). The current evidence indicates no statistically significant differences in operative time, length of hospital stay, intraoperative blood loss, or complication rates between these two techniques (37). However, further research is required to evaluate the efficacy and safety profiles of SPCH as compared to those of OP and RO.

Limitations

None of the eligible studies were RCTs, and thus a relevant RCT is needed to validate the findings of our study. Moreover, given the small number of included studies, the meta-analysis was limited to assessing a single long-term outcome.

Conclusions

Open, laparoscopic, and robotic approaches demonstrated comparable rates of postoperative bile leakage and intraoperative blood loss; open surgery was associated with a significantly shorter operative time; and conversion to open surgery during minimally invasive approaches may expedite completion. Therefore, it seems unnecessary for surgeons to miss the opportunity to switch to open surgery due to concerns of excessive bleeding or increased postoperative bile leakage. Although robotic surgery possesses technical advantages over laparoscopy in reducing postoperative ileus, surprisingly, our statistical analysis indicated a higher incidence of ileus with the robotic approach. This discrepancy may be attributed to selection bias, whereby surgeons preferentially utilized robotic assistance for managing infants with more complex disease. In conclusion, robotic surgery appears to be an effective and safe modality for treating CC in the pediatric population.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.