International Journal of Surgery 68 (2019) 126–133

Contents lists available at ScienceDirect

International Journal of Surgery

journal homepage: www.elsevier.com/locate/ijsu

Review

Comparison of DJ stented, external stented and stent-less procedures for T

pediatric pyeloplasty: A network meta-analysis
Xu Liua,*,1, Chuiguo Huangb,**,1, Yin Guoc,1, Yiwei Yued, Jiawen Honge
a
Department of Urology, Henan Provincial Hospital, Zhengzhou, 450000, Henan Province, China
b
Department of Urology, The Second Affiliated Hospital of Zhengzhou University, Zhengzhou, 450000, Henan Province, China
c
Department of Obstetrics and Gynecology, Henan Provincial Hospital, Zhengzhou, 450000, Henan Province, China
d
College of Clinical Medicine, Zhengzhou University, Zhengzhou, 450000, Henan Province, China
e
Department of Urology, Yangxi County Hospital Group, Yangjiang, 529800, Guangdong Province, China

A R T I C LE I N FO A B S T R A C T

Keywords: Objective: To assess the efficacy and safety of double J (DJ) stented, external stented and stent-less procedures in
Ureteropelvic junction obstruction pediatric pyeloplasty by adopting a network meta-analysis (NMA).
Pediatric pyeloplasty Material and methods: Electronic databases including PubMed, Cochrane Library, Web of science and Embase
Double J database were retrieved. The trials that compared double J (DJ) stented, external stented or stent-less procedures
External stent
in pediatric pyeloplasty were identified. A network meta-analysis was conducted with the software of STATA
Stent-less procedure
Network meta-analysis
14.0. Probability-based ranking results were performed to identify the best treatment, and publication bias was
analyzed by funnel plots.
Results: 15 studies with 1731 participants were enrolled in the analysis, including 4 randomized controlled trials
(RCT) and 11 retrospective studies. The NMA results revealed that no significant differences were detected in the
outcomes of operative time, operative success, hospital stay, improvement of renal functions, overall compli-
cations and redo pyeloplasty. DJ stented and external stented procedures were associated with more post-
operative pain than that of stent-less procedures [DJ stented: OR = 4.47, 95%CI(1.05,19.08); external stented:
OR = 5.83, 95%CI(0.09,1.43)]. DJ stented procedure had a lower rate of urine leakage than those of external
stented procedure [OR = 0.18, 95%CI (0.04, 0.76)] and stent-less procedure [OR = 0.07, 95%CI=(0.01, 0.34)].
No significant difference was observed in other types of complications such as urinary tract infection (UTI), stent
migration, recurrent ureteropelvic junction obstruction (UPJO) and fever. The probabilities of ranking results
indicated that the DJ stented procedure was the best treatment in the outcomes of hospital stay, operative
success, improvement of renal functions, and the complication of urine leakage. Stent-less procedure showed its
advantages in the outcomes of operative time, flank pain and UTI. External stented procedure had the lowest rate
of overall complications and redo pyeloplasty.
Conclusions: There were no obvious differences in operative time, operative success, hospital stay, improvement
of renal functions, overall complications between external stented, DJ stented and stent-less procedures for
pediatric pyeloplasty. When considering the ranking results, the DJ stented procedure seemed to be more
beneficial for pediatric pyeloplasty than the other methods. However, with the limitation of our study, additional
high-quality studies are needed for further evaluation.

1. Introduction incidence of about 1 in 1000–2000 live births [2]. It is estimated that

about 20–50% of UPJO children end up with surgical treatments [3].
Ureteropelvic junction obstruction (UPJO) is defined as the failure Since its introduction in the 1940s, Anderson-Hynes dismembered
of urine outflow from the renal pelvis to the ureter, which may lead to pyeloplasty has remained the standard surgical treatment for UPJO [4].
progressive kidney damage [1]. The UPJO is one of the most common Laparoscopic pyeloplasty is the first minimally invasive surgery for
congenital causes of antenatally diagnosed hydronephrosis, with an pyeloplasty and has become a well-established treatment for UPJO in

*
Corresponding author. Department of Urology, Henan provincial hospital, Dongtinghu road, Zhengzhou city, 450000, China.
**
Corresponding author. Department of Urology, The Second Affiliated Hospital of Zhengzhou University, Jingba road, Zhengzhou city, 450000, China.
E-mail addresses: urologistliu@163.com (X. Liu), Huangcg0727@163.com (C. Huang).
1
These authors contributed equally to this study.

https://doi.org/10.1016/j.ijsu.2019.07.001
Received 20 February 2019; Received in revised form 12 June 2019; Accepted 3 July 2019
Available online 04 July 2019
1743-9191/ © 2019 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
X. Liu, et al. International Journal of Surgery 68 (2019) 126–133

the past decades [5,6]. In recent years, robot-assisted laparoscopic odds ratios (OR) or standardized mean difference (SMD).
pyeloplasty has also rapidly gained popularity [7]. Compared with the
open approach, laparoscopic pyeloplasty is associated with similar re- 2.3. Data extraction and quality assessment
sults, shorter hospitalization, and fewer complications [8–10]. Nowa-
days, Many hospital centers have promoted LP as the first-line treat- Two reviewers extracted the data from the included studies in-
ment for UJPO [11]. dependently. The extracted data were as follows: the first author's
Despite the rapid development of surgical techniques, there are still name, year of publication, country, study design, the total study
some debates about whether it is necessary to place a urinary stent to sample, interventions, samples of each group, mean age of patients,
drain, and which type of stent is better [12]. Many types of stents can be follow-up durations and main clinical outcomes. The main outcomes
used when performing a stented pyeloplasty. The commonly seen stents included operative time, operative success, hospital stay, improvement
can be divided into two categories: external stent and internal stent (DJ of renal functions in the follow-ups, overall complications, specific
stent). Both external and internal stents have been widely used for types of complications, and redo pyeloplasty due to reobstructions,
many years and have been proved effective in clinical practices operative failure, or complications. Operative success was defined as
worldwide. However, both external stents and DJ stents are associated unobstructed urine outflow from the renal pelvis to the ureter, or re-
with various complications, such as postoperative pain, urinary tract solved or improved hydronephrosis under ultrasound. Overall compli-
infection (UTI), etc. In recent years, stent-less pyeloplasty has been cations were the sum of all specific types of complications.
reported to have similar results and fewer complications than stented Two reviewers independently assessed methodological quality with
pyeloplasty [13–15]. It has gained much popularity among many sur- the assessment tool presented by the Cochrane Handbook for
geons worldwide. However, which is superior among DJ stented, ex- Systematic Reviews Interventions version 5.1.3. For included trials, the
ternal stented, and stent-less procedures remains unclear. Even though following criterions were evaluated and graded as having a low,
many studies have compared the differences between these three pro- medium, or high risk bias: random sequence generation, allocation
cedures, no clear conclusions have been drawn. Nowadays, Whether or concealment, blinding of participants and personnel blinding of out-
not to place stents and which type of stents to place in clinical practices come assessment, incomplete outcome data, selective reporting, and
mainly depends on the surgeons’ preference and experience. other bias. Disagreements were resolved through discussion by the two
With the network meta-analysis (NMA), we can make comparisons reviewers or consultation with a third team member.
with all direct and indirect evidence, and make the pooled results more
powerful and persuasive than the traditional meta-analysis. In addition, 2.4. Data synthesis and analysis
we can find out which method is the best treatment by the ranking
analysis. As far as we know, this is the first network meta-analysis We conducted a pair-wise meta-analysis to synthesize all direct
comparing three procedures for pediatric pyeloplasty. The main goal of evidence. The corresponding odds ratios(ORs) or weighted mean dif-
this study was to compare the efficacy and safety of DJ stented, external ference (WMD) and their 95% confidence intervals (95% CIs) were
stented and stent-less procedures, and to find out which one is the calculated. The Mantel–Haenszel Chi-square based test and I2 para-
optimal treatment for pediatric pyeloplasty. meter test were applied for evaluating the heterogeneity. The statistical
significance was defined as P < 0.05. The network meta-analysis was
2. Materials and methods performed based on the Bayesian framework model using STATA ver-
sion 14.0. The random-effects model was utilized in this study to cal-
2.1. Search strategy culate the evidence inconsistency. The relative ranking of each inter-
vention was presented as probability. In addition, publication bias was
A systematic review was performed according to the guidelines of evaluated via observing the symmetry characteristics in funnel-plots. A
Preferred Reporting Items for Systematic Reviews and Meta Analyses symmetrical and concentrated distribution of dots indicated no obvious
(PRISMA) [16] and Assessing the methodological quality of systematic deviation.
reviews (AMSTAR) [17]. Computer-based retrieval of PubMed, Co-
chrane Library, Embase and Web of Science databases was conducted. 3. Results
The MeSH terms and related synonyms including “pyeloplasty”,
“double J”, “DJ”, “stent*“, “catheter”, “tube”, “catheter”, “double-J”, 3.1. Included studies characteristics
“children”, “pediatric”, etc. were combined in the search strategy. We
also manually searched reference lists of related publications including The entire process of literature retrieval and screening for NMA was
reviews, meta-analyses and other types of articles. After importing the illustrated in Fig. 1. A total of 1896 publications were obtained. After
retrieved results into the endnote library, the eligibility of these pub- removing the duplicates, 1430 studies were screened by reading titles
lications were carefully checked. All duplicates and irrelevant pub- and abstracts. 42 articles were performed with full-text assessment.
lications were removed after scanning the title and abstract. The re- Finally, 15 [13,18–31] studies were enrolled in our analysis. The
maining publications were further accessed by full-text scanning. baseline characteristics of the included studies were presented in
Table 1.
2.2. Selection criteria According to the Cochrane Collaboration tool for assessing the risk
of bias, the qualities of the included trials were assessed in seven as-
The studies were considered eligible if they met the following in- pects. 4 of 15 included studies had adequate randomization. The re-
clusion criteria: (1) clinical trials. (2) Patients were diagnosed with maining trials were the retrospective study. 2 article reported the detail
UPJO and underwent pyeloplasty. (3) Patients were infants or young information about allocation sequence concealment. Blind design was
children. (4) Patients were treated with DJ stented, external stented, or hard to perform in surgical treatments. Attrition bias and reporting bias
stent-less procedures. were generally well performed in included trials. The risk-of-bias as-
The following criteria were conducted for study exclusion: (1) In sessment of the included trials was presented in Fig. 2. The network
addition to UPJO, patients were accompanied by other diseases. (2) plots of three procedures was shown in Fig. 3.
Besides pyeloplasty, some other surgeries or treatments unrelated to
UPJO were performed. (3) Patients over 18 years old were involved in 3.2. Network meta-analysis of operative time
studies. (4) Sufficient data of necessary information such as treatments
or outcomes were not provided. (5) Data were unavailable to calculate 7 trials with 637 patients described the outcomes of operative time,

127
X. Liu, et al. International Journal of Surgery 68 (2019) 126–133

Fig. 1. PRISMA flow diagram of the study selection process for network meta-analysis.

Table 1
Characteristics of the included studies in meta-analysis.
Authors, country Study design, year Sample Inventions Groups sample, Mean age Follow-up Outcomes

Nagdeve et al. India RCT, 2018 n = 42 Group 1: DJ stenteded Group 1: n = 21, age 4.227 y 3 mos a, c, d, e
Group 2: Stent-less Group 2: n = 21, age 4.257 y
Chu et al. USA Retrospective study, 2018 n = 64 Group 1: DJ stented Group 1: n = 47 age:6.5 y 12.3 mos a, b, c, e
Group 2: External stented Group 2: n = 17 age:7.7 y
Nasser et al. Egypt RCT, 2017 n = 30 Group 1: Stent-less Group 1: n = 15, age 4mos 22/21.46mos a, b, c, d, e,f
Group 2: External stented Group 2: n = 15, age 3.1 mos
Garg et al. India RCT, 2015 n = 40 Group 1: External stented Group 1: n = 20, age 2.7y ≥3 mos b, c,e,f
Group 2: DJ stented Group 2: n = 20, age 3.67y
Lee et al.Canada Retrospective study, 2015 n = 62 Group 1: External stented Group 1: n = 24; age:40 mos 23.8/21.1 mos a,b, c, e
Group 2: DJ stented Group 2: n = 38; age:80 mos
Zoeller et al. Germany Retrospective study, 2014 n = 86 Group 1: DJ stented Group 1: n = 48, age:5.6 y 12mos b,e,f
Group 2: External stented Group 2: n = 38, age:5.6 y
Kocvara et al. Czech Reprospective study, 2014 n = 70 Group 1: Stent-less Group 1: n = 34, age: 35 mos 36.2mos a,b,d,e
Group 2: DJ stented Group 2: n = 21, age: 46 mos
Group 3: External stented Group 3: n = 15, age: 34 mos
Kim et al. Korea Reprospective study, 2012 n = 70 Group 1: DJ stented Group 1: n = 22 29.6mos c,e
Group 2: Stent-less Group 2: n = 54
Helmy et al. France Retrospective study, 2011 n = 22 Group 1: External stented Group 1: n = 11, age:31 mos; 34/35mos a, c, d, e, f
Group 2: DJ stented Group 2: n = 11, age 37mos
Son et al. Vietnam Retrospective study, 2011 n = 155 Group 1: Stent-less Group 1: n = 33 6mos-4y e,f
Group 2: External stented Group 2: n = 122
Bayne et al. US Retrospective study, 2011 n = 367 Group 1: Stent-less Group 1: n = 231, age 4.486 y 480.98 d e,f
Group 2: External stented Group 2: n = 136, age 5.012 y
Braga et al. Canada Retrospective study, 2008 n = 470 Group 1: DJ stented Group 1: n = 242, age 19 y 41/39 mos b,c, d,e
Group 2: External stented Group 2: n = 228, age 18 y
Elmalik et al. UK Retrospective study, 2008 n = 105 Group 1: Stent-less Group 1: n = 47, age 65.0mos 40.3mos a,b, c,d, e
Group 2: DJ stented Group 2: n = 58, age 53.8mos
Arda et al. Turkey RCT, 2002 n = 31 Group 1:External stented Group 1: n = 15, age 31.6mos 35 mos c, e
Group 2: Stent-less Group 2: n = 16, age 38.9mos
Smith et al. US Retrospective study, 2002 n = 117 Group 1: Stent-less Group 1: n = 65 unclear c,e,f
Group 2: External stented Group 2: n = 52

RCT: randomized controlled trial; a: operative time (mins); b: operative success; c:hospital stay (days); d:renal functions; e:postoperative complications; f: redo
pyeloplasty.

128
X. Liu, et al. International Journal of Surgery 68 (2019) 126–133

Fig. 2. Risk of bias graph and summary of the included studies.

Table 2
The network meta-analysis results for surgical outcomes.
Outcomes DJ External Stent-less

Operative time [SMD(95%Cl)]

DJ 1 −0.17(-1.05, 0.71) −0.08(-1.38, 1.21)
External 0.17(-0.71, 1.05) 1 0.09(-1.22, 1.39)
Stent-less 0.08(-1.21, 1.38) −0.09(-1.39, 1.22) 1
Success rate [OR(95%Cl)]
DJ 1 0.83(0.7, 1.22) 0.74(0.64, 1.14)
External 1.20 (0.82, 1.43) 1 0.86(0.75, 1.32)
Stent-less 1.36 (0.88, 1.57) 1.16 (0.76, 1.33) 1
Hospital stay [SMD(95%Cl)]
DJ 1 0.67(-0.83, 2.61) 1.64(-0.12, 3.41)
External −0.67(-2.16, 0.83) 1 0.98(-0.66, 2.61)
Stent-less −1.64(-3.41, 0.12) −0.98(-2.61, 0.66) 1
Renal functions improvement [SMD(95%Cl)]
DJ 1 0.73(0.38, 1.22) 0.64(0.15, 1.09)
External 1.37 (0.82, 2.60) 1 0.62(0.31, 1.12)
Stent-less 1.56 (0.92, 6.80) 1.62 (0.89, 3.20) 1
Overall complications [OR(95%Cl)]
DJ 1 1.67(0.71, 4) 1.49(0.6, 3.7)
External 0.60(0.25,1.40) 1 0.89(0.32, 2.5)
Stent-less 0.67(0.27,1.66) 1.12(0.40,3.13) 1
Redo pyeloplasty [OR(95%Cl)]
Fig. 3. Network plots of the comparisons between DJ stent, external stent and DJ 1 0.19(0.02, 1.69) 0.21(0.02, 2)
External 5.26(0.59,50) 1 1.11(0.69, 1.79)
stent-less.
Stent-less 4.78(0.50,45.52) 0.90(0.56,1.45) 1

and the time between studies varied greatly. The average operative DJ: double-J; OR: odds ratio; SMD: standardized mean difference; CI: credible
time of DJ stented, external stented and stent-less groups in the 7 stu- intervals.
dies were 147.23min (range: 60.4–190min), 154.96min (ran-
ge:78.9–205min), 123.47min (range:85.3–162min). In the pooled 3.4. Network meta-analysis of hospital stay
meta-analysis results shown in Table 2, we didn't find significant dif-
ferences between three groups. The probability-based ranking results As for the outcome of hospital stay, 9 studies with 940 patients were
were shown in Table 4. Stent-less group ranked 1st (39.5%), DJ stented taken into analysis. The average hospital stay of DJ stented, external
group ranked 2nd (38.2%), external stented group ranked 3rd (40.6%). stented and stent-less procedures were 2.93d (range: 1.0–5.15d), 3.80d
In this result, ranking 1st meant the shortest operative time, and the last (range: 1.0–11.95d), and 4.2d (range: 2.6–5.9d), respectively. The
meant the longest time. pooled results revealed that there were no significant differences in the
comparisons of DJ stented, external stented, and stent-less procedures
(Table 2). Probabilities of the ranking results were as follows: DJ
3.3. Network meta-analysis of operative success stented group ranked 1st (81.4%), external stented group ranked 2nd
(75.2%), stent-less group ranked 3rd (91.6%) (Table 4). In this out-
8 of 15 studies with 839 patients reported the outcomes of surgery. come, ranking 1st had the minimal length of hospital stay, and the 3rd
All groups achieved a high success rate, and the average success rate of had the maximal.
DJ stented, external stented and stent-less groups were 93.2% (range:
88%–95%), 92.6% (range: 86%–94.7%), 89.2% (range: 85%–93.4%),
respectively. The pooled results showed that the success rates in three 3.5. Network meta-analysis of the improvement of renal functions
groups were similar, and no significant differences were detected
(Table 2). The ranking results with probability were as follows: DJ Among all included studies, 6 trials with 457 patients described the
stented group was the best (76.3%), external stented group was the improvement of renal functions. Compared with preoperative renal
second (63.7%), and stent-less group was the last (56.6%) (Table 4). function, all groups were significantly improved during follow-up
(P < 0.05). However, the comparisons of the improvement of renal

129
X. Liu, et al. International Journal of Surgery 68 (2019) 126–133

Table 3 Table 4
The meta-analysis results of all types of postoperative complications. Ranking results based on simulations.
Complications DJ vs external DJ vs stent-less External vs stent-less Endpoints (%) Ranks DJ External Stent-less

Urine leakage 0.18(0.04,0.76) 0.07(0.01, 0.34) 0.36(0.09,1.43) Operative time (rank 1 is the best, rank 3 Rank 1 38.4 21.2 39.5
Pain 0.78(0.57, 4.49) 4.47(1.05,19.08) 5.83(1.21,27.98) is the worst) Rank 2 23.9 38.2 37.9
UTI 2.63(0.75,9.09) 3.37(0.92,12.32) 1.28(0.50,3.23) Rank 3 37.7 40.6 22.7
Stent migration 2.17(0.68,6.88)
Recurrent UPJO 0.85(0.43,1.66) Success rate Rank 1 76.3 23.1 18.5
Fever 0.39(0.06, 2.70) 2.02(0.61, 6.73) 5.14(0.66, 40.20) Rank 2 16.0 63.7 24.9
Pyelonephritis P = 0.24 Rank 3 7.7 13.2 56.6
Bladder spasms P = 0.03
Urinoma P = 0.65 Hospital stay Rank 1 81.4 17.4 1.2
Hematuria P = 0.03 P = 0.743 Rank 2 17.6 75.2 7.2
Hydronephrosis P = 0.78 P = 0.73 Rank 3 1.0 7.3 91.6

DJ: double-J; UTI: urinary tract infection; UPJO: Ureteropelvic junction ob- Renal functions improvement Rank 1 71.4 23.9 19.1
struction. Rank 2 18.9 50.3 65.7
Rank 3 9.7 45.8 16.2

functions between three groups did not show significant differences. Overall complications Rank 1 9.0 73.6 17.4
The results were presented in Table 2. In the ranking results with Rank 2 35.0 22.2 42.8
probabilities, DJ stented group ranked 1st(71.4%) in the improvement Rank 3 56.0 4.2 39.8
of renal function, followed by stent-less group (65.7%) and external
Redo pyeloplasty Rank 1 6.3 62.5 31.2
stented group(45.8%). Rank 2 2.9 35.2 61.9
Rank 3 90.8 2.3 6.9
3.6. Network meta-analysis of overall complications
Urine leakage Rank 1 99.0 1.0 0.1
Rank 2 1.0 91.7 7.3
11 of 15 studies reported overall complications, recruiting of 765 Rank 3 0.0 7.4 92.6
patients. The average overall complication rate of DJ stented, external
stented, and stent-less procedures were 34.8%, 20.2%, and 27.1%, re- Pain Rank 1 2.1 1.1 96.8
spectively. The NMA results revealed that there were no significant Rank 2 61.9 35.3 2.8
Rank 3 35.9 63.7 0.4
differences in three groups (Table 2). The ranking results based on
probabilities were as follows: external stented group ranked 1st UTI Rank 1 2.1 30.6 67.3
(77.3%), stent-less group ranked 2nd (42.8%), DJ stented group ranked Rank 2 5.7 63.9 30.4
3rd (56.0%) (Table 4). For this outcome, rank 1st had the lowest overall Rank 3 92.2 5.5 2.3
complications, the 3rd one was associated with the highest overall
UTI: urinary tract infection.
complications rate.
The specific types of complications were also analyzed in our ana-
redo pyeloplasty. The average rate of DJ stented, external stented, and
lysis, including urine leakage, pain, UTI, stent migration, fever, pyelo-
stent-less procedures were 8.5% (range: 5.3%–9.1%), 7.6% (range:
nephritis, recurrent UPJO, bladder spasms, urinoma, hematuria, and
0%–13.2%), and 7.2% (range: 4.6%–13.4%), respectively. NMA results
hydronephrosis. Data integration was performed when complications
suggested that there were no significant differences between the com-
were described with a sufficient number of studies. The analysis results
parisons of three groups (Table 2). The probabilities of ranking results
of different complications were presented in Table 3. From pooled re-
were as follows: external stented group ranked 1st (62.5%), stent-less
sults, we found that the urine leakage rate of DJ stented group was
group ranked 2nd (61.9%), DJ stented group ranked 3rd (90.8%)
lower than that of the external stented group [OR = 0.18, 95%CI(0.04,
(Table 4). Among all interventions, rank 1st was associated with the
0.76)] and stent-less group [OR = 0.07, 95%CI=(0.01, 0.34)]. The DJ
lowest redo surgery rate, and the last one was the highest.
stented group and external stented group were associated with more
postoperative flank pain than stent-less did [DJ stented: OR = 4.47,
95%CI(1.05, 19.08); external stented: OR = 5.83, 95%CI=(1.21, 3.8. Consistency and convergence analysis
27.98)]. No significant differences were observed between three groups
in the outcome of UTI, stent migration, recurrent UPJO, and fever. The node-splitting analysis was applied to evaluate inconsistency by
Based on the ranking results shown in Table 4, we could find that DJ comparing the differences between direct and indirect evidence. No
stented group ranked 1st (99%), external stented group ranked second significant inconsistency was detected among the various treatments
(91.7%), then stent-less (92.6%) was the least in urine leakage. Stent- with the P-value were lower than 0.05. That meant the consistency
less was the best in postoperative pain (96.8%), DJ stented group was model was reliable. The publication bias was analyzed by the funnel
second (61.9%) and external stented group was the last(63.7%). The plot, and no obvious publication bias was detected in most outcomes.
sequence of UTI from best to worst was stent-less group (67.3%), ex- The funnel plots of all outcomes were shown in Fig. 4.
ternal stented group (63.9%), and DJ stented group (92.2%). In the
complications of pyelonephritis, bladder spasms, hematuria, and hy- 4. Discussion
dronephrosis, the relevant data were limited and could not be in-
tegrated. We listed the P value of comparisons reported in the studies in For external stented, DJ stented, and stent-less pyeloplasty, each
Table 3. According to the results, we could find that most of the P va- procedure is associated with its advantages and disadvantages. The
lues were ≥0.05, which suggested that there were no significant dif- external stented procedure allows for the assessment of the repair in
ferences in the comparison. pyeloplasty and can be simply removed without sedation. But it has
several potential unfavorable conditions such as increased risk of renal
3.7. Network meta-analysis of redo pyeloplasty parenchyma damage, bleeding, flank pain, UTI and reduced quality of
life [32]. DJ stented procedure can provide support and prevent edema
A total of 7 studies with 1114 participants reported the outcome of in the anastomotic site for the patients who undergoing pyeloplasty

130
X. Liu, et al. International Journal of Surgery 68 (2019) 126–133

Fig. 4. Funnel plots of all outcomes: (A) operative time; (B) operative success (C) hospital stay; (C) improvement of renal functions; (D) overall complications; (E)
redo pyeloplasty.

[22,24]. However, it is related to a higher risk of UTI and requires vs external stented [24,27], DJ stented vs stent-less [29,37], and ex-
additional general anesthesia to remove the stent [14,22]. Moreover, ternal stented vs stent-less procedures [18]. Elmalik et al. [22] found
this internal approach forms artificial vesicoureteric reflux that may that DJ stented patients had a shorter hospital stay than that of un-
result in persistent kidney damage [33]. The stent-less procedure can stented patients. As pyeloplasty can eliminate the obstruction of the
avoid stent-related complications after pyeloplasty, without the need of ureteropelvic junction, renal functions will be improved in most pa-
instrument tube and second anesthesia. However, it is associated with tients. It is especially obvious for patients with poor renal functions
some common complication such as urine leakage. before surgery. Nagdeve et al. [31] compared DJ stented and stent-less
In our study, 15 studies recruiting 1731 were included and ana- procedures in a randomized controlled study, and conducted a 3-month
lyzed. According to the network meta-analysis results, we found that follow up. He found that the renal functions in DJ stented group im-
there were no significant differences in the outcomes of operative time, proved from 30.41% preoperatively to 32.78% postoperatively, and
operative success, length of hospital stay, improvement of renal func- increased from 28.71% preoperatively to 30.8% postoperatively in
tions, overall complications and redo surgery rate. DJ stented proce- stent-less group. In another randomized comparative study, Nasser
dure had a lower rate of urine leakage than that of external stented and et al. [28] found that the mean renal functions improved from 26.7% to
stent-less procedures. The stent-less procedure was associated with less 32.7% in external stented patients, and increased from 32.6% to 42.4%
postoperative pain than that of DJ stented and external stented proce- in stent-less patients. The improvements in renal functions were not
dures. The ranking results showed that DJ stented pyeloplasty was the different significantly between the two groups.
best in the outcomes of hospital stay, operative success, improvement of As for the outcome of overall complications, we didn't find statis-
renal functions, and the complication of urine leakage. Stent-less pye- tical differences between the external stented, DJ stented, and stent-less
loplasty was the best in the outcomes of operative time, flank pain and procedures. Several previous studies [20,22,27,29] also proved this
UTI. External stented pyeloplasty had the lowest rate of overall com- result. As for the outcome of redo pyeloplasty, our results showed that
plications and redo pyeloplasty. there were no obvious differences between the three procedures. Si-
For the outcome of operative time, many previous publications have milar results were reported in other studies [13,19,20,38]. When it
drawn similar conclusions with our results. Braga et al. [20] compared came to specific complications, many studies had similar results to our
the operative time of DJ stented and external stented procedures, and findings. The stent-less group was associated with more urinary leakage
did not find significant differences. Similar results could be seen in [19,20,29], and less flank pain than stented groups [20,37,39]. Nag-
Nasser's study [28]. Of course, there were different consequences. deve et al. [31] concluded that the DJ stented patients were more
Helmy et al. [22] reported that external stented pyeloplasty had a symptomatic than unstented patients in the postoperative period, but
slightly shorter operative time than that of the DJ stented pyeloplasty. the difference was not statistically significant.
As for the operative success rate, Braga et al. [20] analyzed 470 pa- Although no significant differences found in main surgical out-
tients, and found that the operative success rate of DJ stented and ex- comes, the ranking results identified the best treatment for all out-
ternal stented groups were 95% and 94.7% respectively. Similar results comes. According to the ranking results, the DJ stented procedure had
were found in other studies [21,28]. The operative success was greatly the highest operative success, renal function improvement, and the
affected by the surgery. The Anderson-Hynes dismembered pyeloplasty shortest hospital stay. External stented method had the lowest rate of
was reported with a high success rate of more than 90% in many studies overall complications and redo pyeloplasty. Stent-less procedure had
[12,34–36]. For the hospital stay, some studies concluded that the the shortest operative time. As for specific types of complications, DJ
length of hospital stay was similar in the comparisons with DJ stented stented procedure had a lower rate of urine leakage than the external

131
X. Liu, et al. International Journal of Surgery 68 (2019) 126–133

stented procedure. When compared with DJ stented and external Research registration number
stented procedures, the stent-less procedures had less postoperative
flank pain. Based on all analysis results, it seems to be that DJ stented Reviewregistry591
pyeloplasty has more benefits than external stented and stent-less https://www.researchregistry.com/browse-the-registry#
pyeloplasty. While in clinical practice, the choice of DJ stented, ex- registryofsystematicreviewsmeta-analyses/
ternal stented and stent-less procedures still require the consideration registryofsystematicreviewsmeta-analysesdetails/
of the patients' willing, surgeons' experience and the individuals’ con- 5b6f11f3e19cd95541adda1e/.
dition.
As far as we know, our study is the first comprehensive meta-ana- Guarantor
lysis that evaluated the difference of external stented, DJ stented and
stent-less procedures in pediatric pyeloplasty. We performed data Xu Liu(urologistliu@163.com).
consolidations with all included studies in the main outcomes, making Chuiguo Huang(huangcg0727@163.com).
the summary results accurate and convincing. However, our analysis
does have certain limitations. Firstly, only 15 studies were included, so Data statement
it might not be sufficient to conduct a powerful meta-analysis.
Furthermore, only 4 RCTs were included and the rest were mostly The raw data of this study are derived from the included studies,
retrospective study. These retrospective studies might be associated which are available in public. All detailed data included in the study are
with a high risk of bias, and affect the reliability of pooled results. available upon request by contact with the corresponding author.
Secondly, the external stented procedures might have several different
types, the combination of different external stents into one group might Provenance and peer review
be inappropriate and could increase the heterogeneity of the results.
The last, limited by lacking sufficient data in included studies, some Not commissioned, externally peer-reviewed.
important outcomes were not analyzed. This might make our evalua-
tion incomplete and unsystematic. Appendix A. Supplementary data

Supplementary data to this article can be found online at https://

5. Conclusion
doi.org/10.1016/j.ijsu.2019.07.001.
In conclusion, there were no significant differences in operative
References
time, operative success, hospital stay, improvement of renal functions,
overall complications for the external stented, DJ stented and stent-less
[1] S.A. Koff, Requirements for accurately diagnosing chronic partial upper urinary
procedures in pediatric pyeloplasty. DJ stented procedure seemed to tract obstruction in children with hydronephrosis, Pediatr. Radiol. 38 (Suppl 1)
have more advantages than external stented and stent-less procedures (2008) S41–S48.
when considering ranking results. However, due to the limitations of [2] H.A. Damen-Elias, T.P. De Jong, R.H. Stigter, G.H. Visser, P.H. Stoutenbeek,
Congenital renal tract anomalies: outcome and follow-up of 402 cases detected
our analysis, additional high-quality studies are needed to further
antenatally between 1986 and 2001, Ultrasound Obstet. Gynecol. 25 (2) (2005)
evaluate the outcomes. 134–143.
[3] B. Chertin, A. Pollack, D. Koulikov, et al., Conservative treatment of ureteropelvic
junction obstruction in children with antenatal diagnosis of hydronephrosis: lessons
Ethical approval learned after 16 years of follow-up, Eur. Urol. 49 (4) (2006) 734–738.
[4] P. O'reilly, P. Brooman, S. Mak, et al., The long‐term results of Anderson–Hynes
pyeloplasty, BJU Int. 87 (4) (2001) 287–289.
All analyses were based on previous published studies, thus no [5] R. Munver, R.E. Sosa, J.J. del Pizzo, Laparoscopic pyeloplasty: history, evolution,
ethical approval and patient consent are required. and future, J. Endourol. 18 (8) (2004) 748–755.
[6] A. Bachmann, R. Ruszat, T. Forster, et al., Retroperitoneoscopic pyeloplasty for
ureteropelvic junction obstruction (UPJO): solving the technical difficulties, Eur.
Source of funding Urol. 49 (2) (2006) 264–272.
[7] L.H. Olsen, T.M. Jorgensen, Computer assisted pyeloplasty in children: the retro-
peritoneal approach, J. Urol. 171 (6 Pt 2) (2004) 2629–2631.
The authors declare no sources of funding. [8] H. Mei, J. Pu, C. Yang, et al., Laparoscopic versus open pyeloplasty for ureteropelvic
junction obstruction in children: a systematic review and meta-analysis, J.
Endourol. 25 (5) (2011) 727–736.
Author contribution [9] R.V. Clayman, Comparison of open versus laparoscopic pyeloplasty techniques in
treatment of uretero-pelvic junction obstruction, J. Urol. 173 (6) (2005) 2023.
[10] F. van der Toorn, J. van den Hoek, K.P. Wolffenbuttel, J.R. Scheepe, Laparoscopic
Xu Liu: Design the protocol, acquisition and interpretation of data,
transperitoneal pyeloplasty in children from age of 3 years: our clinical outcomes
technical procedures, drafted the initial manuscript, and approved the compared with open surgery, J. Pediatr. Urol. 9 (2) (2013) 161–168.
final manuscript, supervised all phases of the study. [11] F. Atug, S.V. Burgess, E.P. Castle, R. Thomas, Role of robotics in the management of
Chuiguo Huang: Design the protocol, acquisition and interpretation secondary ureteropelvic junction obstruction, Int. J. Clin. Pract. 60 (1) (2006) 9–11.
[12] C.K. Yeung, Y.H. Tam, J.D. Sihoe, K.H. Lee, K.W. Liu, Retroperitoneoscopic dis-
of data, technical procedures, drafted the initial manuscript, and ap- membered pyeloplasty for pelvi-ureteric junction obstruction in infants and chil-
proved the final manuscript, supervised all phases of the study. dren, BJU Int. 87 (6) (2001) 509–513.
Yin Guo: Design the protocol, acquisition and interpretation of data, [13] J. Kim, S. Park, H. Hwang, et al., Comparison of surgical outcomes between dis-
membered pyeloplasty with or without ureteral stenting in children with ureter-
technical procedures, drafted the initial manuscript, and approved the opelvic junction obstruction, Korean J Urol 53 (8) (2012) 564–568.
final manuscript. [14] A.L. Shalhav, A.A. Mikhail, M.A. Orvieto, et al., Adult stentless laparoscopic pye-
Yiwei Yue: Revised the manuscript, modified the language issues, loplasty, J. Soc. Laparoendosc. Surg. 11 (1) (2007) 8–13.
[15] C.Y. Bilen, Y. Bayazit, A. Gudeloglu, et al., Laparoscopic pyeloplasty in adults:
and approved the final manuscript. stented versus stentless, J. Endourol. 25 (4) (2011) 645–650.
Jiawen Hong: Acquisition and interpretation of data, drafted the [16] K. Knobloch, U. Yoon, P.M. Vogt, Preferred reporting items for systematic reviews
initial manuscript, and approved the final manuscript. and meta-analyses (PRISMA) statement and publication bias, J. Cranio-Maxillo-Fac.
Surg. 39 (2) (2011) 91–92.
[17] B.J. Shea, B.C. Reeves, G. Wells, et al., AMSTAR 2: a critical appraisal tool for
Conflicts of interest systematic reviews that include randomised or non-randomised studies of health-
care interventions, or both, BMJ 358 (2017) j4008.
[18] I.S. Arda, P. Oguzkurt, S. Sevmis, Transanastomotic stents for dismembered
The authors declare no conflict of interest.

132
X. Liu, et al. International Journal of Surgery 68 (2019) 126–133

pyeloplasty in children, Pediatr. Surg. Int. 18 (2–3) (2002) 115–118. 1127–1130.

[19] A.P. Bayne, K.A. Lee, E.D. Nelson, et al., The impact of surgical approach and ur- [30] C. Zoeller, M. Lacher, B. Ure, C. Petersen, J.F. Kuebler, Double J or transrenal
inary diversion on patient outcomes in pediatric pyeloplasty, J. Urol. 186 (4 Suppl) transanastomotic stent in laparoscopic pyeloplasty in infants and children: a com-
(2011) 1693–1698. parative study and our technique, J. Laparoendosc. Adv. Surg. Tech. 24 (3) (2014)
[20] L.H. Braga, A.J. Lorenzo, W.A. Farhat, et al., Outcome analysis and cost comparison 205–209.
between externalized pyeloureteral and standard stents in 470 consecutive open [31] N.G. Nagdeve, P.D. Bhingare, P. Sarawade, A randomized control trial comparing
pyeloplasties, J. Urol. 180 (4 Suppl) (2008) 1693–1698 ; discussion1698-1699. outcome after stented and nonstented anderson-hynes dismembered pyeloplasty, J.
[21] D.I. Chu, D. Shrivastava, J.P. Van Batavia, et al., Outcomes of externalized pye- Indian Assoc. Pediatr. Surg. 23 (4) (2018) 186–191.
loureteral versus internal ureteral stent in pediatric robotic-assisted laparoscopic [32] H.B. Joshi, A. Stainthorpe, R.P. MacDonagh, et al., Indwelling ureteral stents:
pyeloplasty, J. Pediatr. Urol. 14 (5) (2018) 450.e1–450.e6. evaluation of symptoms, quality of life and utility, J. Urol. 169 (3) (2003)
[22] K. Elmalik, M.M. Chowdhury, S.N. Capps, Ureteric stents in pyeloplasty: a help or a 1065–1069 ; discussion 1069.
hindrance? J. Pediatr. Urol. 4 (4) (2008) 275–279. [33] M.L. Godley, P.G. Ransley, Vesicoureteral reflux: pathophysiology and experimental
[23] R.K. Garg, P. Menon, K.L. Narasimha Rao, S. Arora, Y.K. Batra, Pyeloplasty for studies, Pediatric urology: Elsevier (2010) 283–300.
hydronephrosis: issues of double J stent versus nephrostomy tube as drainage [34] M. Iwamura, S. Soh, A. Irie, et al., Laparoscopic pyeloplasty for ureteropelvic
technique, J. Indian Assoc. Pediatr. Surg. 20 (1) (2015) 32–36. junction obstruction: outcome of initial 12 procedures, Int. J. Urol. 11 (7) (2004)
[24] T. Helmy, T. Blanc, A. Paye-Jaouen, A. El-Ghoneimi, Preliminary experience with 449–455.
external ureteropelvic stent: alternative to double-j stent in laparoscopic pyelo- [35] B.J. Minnillo, J.A. Cruz, R.H. Sayao, et al., Long-term experience and outcomes of
plasty in children, J. Urol. 185 (3) (2011) 1065–1069. robotic assisted laparoscopic pyeloplasty in children and young adults, J. Urol. 185
[25] R. Kocvara, J. Sedlacek, M. Drlik, et al., Unstented laparoscopic pyeloplasty in (4) (2011) 1455–1460.
young children (1-5 years old): a comparison with a repair using double-J stent or [36] B.K. Varda, E.K. Johnson, C. Clark, et al., National trends of perioperative outcomes
transanastomotic externalized stent, J. Pediatr. Urol. 10 (6) (2014) 1153–1159. and costs for open, laparoscopic and robotic pediatric pyeloplasty, J. Urol. 191 (4)
[26] T.S. Le, C.T. Le, T.H. Le, et al., Transpelvic anastomotic stenting: a good option for (2014) 1090–1095.
diversion after pyeloplasty in children, J. Pediatr. Urol. 7 (3) (2011) 363–366. [37] A.R. Khawaja, T.I. Dar, F. Bashir, et al., Stentless laparoscopic pyeloplasty: a single
[27] L.C. Lee, N. Kanaroglou, J.M. Gleason, et al., Impact of drainage technique on pe- center experience, Urol. Ann. 6 (3) (2014) 202–207.
diatric pyeloplasty: comparative analysis of externalized uretero-pyelostomy versus [38] A. Yarmohamadi, P. Saeedi, A. Hoghabrosadat, Surgical results of anderson-hynes
double-J internal stents, Can Urol Assoc J 9 (7–8) (2015) E453–E457. dismembered pyeloplasty without internal drainage in adults with ureteroplevic
[28] F.M. Nasser, A.M. Shouman, M.S. ElSheemy, et al., Dismembered pyeloplasty in junction obstruction, Nephro-Urol. Mon. 7 (3) (2015) e21800.
infants 6 Months old or younger with and without external trans-anastomotic ne- [39] M. Castagnetti, A. Berrettini, M. Cimador, et al., Complications of trans-anastomotic
phrostent: a prospective randomized study, Urology 101 (2017) 38–44. externalised stents in open pyeloplasty: influence of the method of placement, the
[29] K.E. Smith, N. Holmes, J.I. Lieb, et al., Stented versus nonstented pediatric pyelo- duration of stenting, and the associated bladder drainage, Pediatr. Surg. Int. 26 (3)
plasty: a modern series and review of the literature, J. Urol. 168 (3) (2002) (2010) 309–313.

133