Endourological versus open cystolithotomy for bladder stone management among children: A systematic review and meta-analysis

Eligibility criteria

The inclusion criteria in this study were as follows: 1) All randomised controlled trials (RCTs) and comparative non-randomised studies (NRSs) comparing open cystolithotomy and endourological procedures for bladder stones; 2) full-text articles published in English; 3) children (male and female participants) <18 years of age of any ethnicity with bladder stones (single/multiple) were included. Patients with a history of bladder augmentation or diversion were excluded. All studies were grouped into three groups: group A (TUCL versus PCCL), group B (TUCL versus CL) and group C (PCCL versus CL).

Information sources

We conducted the systematic review in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) checklist²⁶ and Cochrane Handbook for Systematic Reviews of Interventions. This study followed the SAGER guidelines for reporting sex and gender.²⁷ The MEDLINE (RRID:SCR_002185), EMBASE (RRID:SCR_001650), and Cochrane Central Register of Controlled Trials (RRID:SCR_006576) databases and ClinicalTrials.gov (RRID:SCR_002309) were searched for relevant English-language publications from 1 to 30 August 2022.

Search strategy

We used keywords adjusted to search engine specification in the form of (bladder stone OR bladder calculi) AND (transurethral cystolithotomy OR percutaneous cystolithotripsy OR open cystolithotomy) AND (children OR paediatric).

Selection process

The PICOS framework was used to trace studies and identify the suitability of any we found.¹⁰ Study selection was carried out independently and duplicated by each author, referring to inclusion and exclusion criteria. The decision to study eligibility was determined by each author independently. Any disagreement was resolved by discussion.

Data collection process

Data collection was carried out by each author independently and in duplication. We extracted the study’s primary characteristics, including the first author, location, sample size, and publication year. We also extracted patient baseline data and postoperative data, including a SFR, complication rate, procedure duration and length of stay. We used a 2x2 contingency table to obtain each study’s odds ratio (OR) and pooled the overall ORs using RevMan (RRID:SCR_003581) version 5.3.

Data items

The following data were extracted from each study included in the review: study methodology (study design, year); participant characteristics (age, sex, stone size, lithotripsy energy, SFR, complication, procedure duration, catheterisation, hospital stay and follow up); intervention and comparator description (TUCL, PCCL and CL); and outcome measures.

Study risk of bias assessment

Risk of bias assessment was independently assessed by each author using the risk of bias assessment tool by Higgins et al., (2011)¹¹ including: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting and other bias, which were produced via RevMan (RRID:SCR_003581) version 5.3.

Effect measures

The odds ratio (OR) with 95% CI was used to assess the success of treatment (SFR and complication). The mean duration and mean hospital stay were assessed using mean difference (MD) with 95% confidence intervals (CI).

Synthesis methods

The meta-analysis was performed using RevMan (RRID:SCR_003581) version 5.3. The results are described as odds ratios (OR) with 95% confidence intervals (CI) for dichotomous variables and as mean difference with 95% CI for continuous variables. A chi-squared statistic (Cochrane Q) was used to evaluate the level of heterogeneity. The I² statistic was used to determine the percentage of the total variation in the estimated effect across studies. The data was analysed using the random effect model when I² >25%, and fixed effect model when I² is less than 25%. Statistical significance was set at values of p<0.05. For studies that provided the minimum and maximum value instead of standard deviation (SD) for the mean difference analysis, estimated SD were calculated with the formula derived from a study by Walter and Yao (2007).¹² Sensitivity analyses were not conducted to explore the effect of study quality as there were too few studies and some studies used different criteria for measuring outcomes.

Reporting bias assessment

We could not use funnel plots of between-treatment effect and its precision on individual studies for publication bias due to the small numbers of studies.

Certainty assessment

To assess certainty (or confidence) in the body of evidence for an outcome, this meta-analysis used GRADE score, which has four level of evidence very low, low, moderate, and high.¹³

Study selection

A total of 2,975 articles were initially retrieved from the database. After removing duplicates and evaluating the title and abstracts, 28 articles were subjected to full-text review. From the 28 article studies screened, 23 studies were excluded due to the following reasons: they were studies on adults (n=16), had an insufficient number of patients (n=2), or did not focus on endourological and open cystolithotomy (n=5). Finally, five studies were included in the analysis: four NRS and one RCT. The article selection process was performed according to the PRISMA statement (Figure 1).

Figure 1. PRISMA flowchart detailing the article identification, screening, and inclusion process and results.

Study characteristics

Three of the five studies included in this systematic review were non-randomised studies, retrospective, and single centre.^14–16 The other two were prospective, but one was randomised and the other one was not randomised.^17–18 The numbers of patients enrolled in the studies did not differ between studies. The total number of patients in this meta-analysis was 436 patients. Two studies compared endourological (TUCL; PCCL) versus open cystolithotomy, two studies compared PCCL versus TUCL, and the last one compared PCCL versus open cystolithotomy. Measure characteristics included intervention type, sex, age, stone size, and lithotripsy energy. Outcomes (SFR, complications, procedure duration, and length of hospital stay) and follow-up were also measured (Table 1).

Risk of bias in studies

Of the five studies included in this meta-analysis, three had high risk of bias because assessor outcome blinding was not performed.^14,15,18 Selection bias in some studies was also considered high because no allocation concealment was performed. In general, the quality of the studies in this meta-analysis varied from low to high. The bias assessment was carried out independently by each author and in duplication, in which all authors were resolved by discussion or consensus. Figure 2 visualizes the summary of bias risk.

Figure 2. Risk of bias assessment summary using traffic light and summary plots.

(A) Traffic light plot and (B) summary plot.

Results of individual studies

Stone-free rate

Group A (TUCL versus PCCL)

Mahran et al., 2000, Al-Marhoon et al., 2009 and Javanmard et al., 2018 reported SFRs of 100%. All stones were removed successfully.^14–16 Mishra et al., 2020 reported both groups (TUCL and PCCL) were comparable with good outcomes. The clearance rate was 100% in the PCCL group, while in the TUCL group it was 86.6%. However statistical significance was not reached.¹⁸ Shahat et al., 2022 reported TUCL SFR was 96%, while PCCL 98%. There were no significant differences between both groups (p=1).¹⁷

Group B (TUCL versus CL) and Group C (PCCL versus CL)

Mahran et al., 2000, Al-Marhoon et al., 2009 and Javanmard et al., 2018 reported SFRs of 100%.^14–16

Complication

Group A (TUCL versus PCCL)

Mahran et al., 2000 reported one (4%) patient who was observed for urethral rupture and extravasation during TUCL and four (15%) patients who developed early and late complications (persistent leakage of urine, acute abdomen, and stricture urethra) for PCCL.¹⁶ Shahat et al., 2022 reported that complication for TUCL were persistent haematuria, urethral abrasion, difficulty after catheter removal, and bladder perforation, while PCCL had fever, persistent haematuria, stone entrapped, subcutaneous fluid collection, and intraperitoneal collection.¹⁷ Javanmard et al., 2018 reported complications in PCCL included surgical site infection,² urinary leakage,¹ and irritative symptoms,² while six patients from TUCL group developed haematuria.¹⁵ Mishra et al., 2000 reported zero complications in the PCCL group and one patient conversion into open surgery in the TUCL group.¹⁸ Al-Marhoon et al., 2009 showed urethral rupture (1 (1.4%)) in the TUCL group and bladder perforation (1 (1.4%)) in the PCCL group.¹⁴

Group B (TUCL versus CL)

Mahran et al., 2000 reported one (4%) patient who was observed for urethral rupture and extravasation during TUCL and one (4%) patient who had small intestinal injury in the CL group.¹⁶ Javanmard et al., 2018 reported six patients from the TUCL group who developed haematuria and one patient who underwent CL had urinary leakage and perivesical fluid collection.¹⁵ Al-Marhoon et al., 2009 showed one (1.4%) patient in the TUCL group who had a urethral rupture and one (1.9%) patient who had peritoneal perforation and small intestinal injury during tube drain fixation in the CL group.¹⁴

Group C (PCCL versus CL)

Mahran et al., 2000 reported four (15%) patients who developed early and late complications (persistent leakage of urine, acute abdomen, and stricture urethra) for PCCL and one (4%) patient who had small intestinal injury in the CL group.¹⁶ Javanmard et al., 2018 reported complications in PCCL included surgical site infection,² urinary leakage,¹ and irritative symptoms² and one patient who underwent CL had urinary leakage and perivesical fluid collection.¹⁵ Al-Marhoon et al., 2009 reported one (1.4%) patient had bladder perforation in the PCCL group and one (1.9%) patient had peritoneal perforation and small intestinal injury during tube drain fixation in the CL group.¹⁴

Procedure duration

Group A (TUCL versus PCCL)

Mahran et al., 2000 reported no significant differences in procedure duration between two groups. It was 46 ± 14 minutes for the TUCL or PCCL group.¹⁶ Shahat et al., 2022 reported PCCL had significantly faster procedure duration than TUCL (p=0.012). It was 34.2 ± 24.8 minutes for TUCL group and 25.3 ±18.7 minutes for PCCL group.¹⁷ Javanmard et al., 2018 reported significantly longer procedure duration in the TUCL group (36.3 ± 5.9 minutes) than the PCCL group (30.5 ± 5.2 minutes) (p=0.000).¹⁵ Mishra et al., 2000 reported faster procedure duration for the PCCL group (33.5 ± 8.4 minutes) than the TUCL group (38.2 ± 6.7 minutes).¹⁸ Al-Marhoon et al., 2009 reported that the procedure duration was comparable in both groups (46 ± 14 minutes).¹⁴

Group B (TUCL versus CL)

Mahran et al., 2000 reported no significant differences in procedure duration between the two groups. It was 46 ± 14 minutes and 38 ± 12 minutes in the TUCL and CL groups, respectively.¹⁶ Javanmard et al., 2018 reported that the CL group had significantly faster procedure duration (26.0 ± 6.3 minutes) than the TUCL group (36.3 ± 5.9 minutes) (p=0.000).¹⁵ Al-Marhoon et al., 2009 reported that procedure duration was comparable in both groups, 46 ± 14 and 38 ± 12 minutes for TUCL and CL, respectively.¹⁴

Group C (PCCL versus CL)

Mahran et al., 2000 reported comparable procedure duration in both groups. It was 46 ± 14 minutes in the PCCL group and 38 ± 12 minutes in the CL group.¹⁶ Javanmard et al., 2018 reported that the CL group (26.0 ± 6.3 minutes) had significantly faster procedure duration than the PCCL group (30.5 ± 5.2 minutes) (p=0.000).¹⁵ Al-Marhoon et al., 2009 reported that procedure duration was comparable in both groups, 46 ± 14 and 38 ± 12 minutes in the PCCL and CL groups, respectively.¹⁴

Length of hospital stay

Group A (TUCL versus PCCL)

Mahran et al., 2000 reported that length of stay was comparable in both groups (1.6 ± 0.8 days).¹⁶ Shahat et al., 2022 reported no significant difference in terms of length of stay for both groups, 3.2 ± 1.4 and 3.2 ± 1.4 days (p=0.13) in the TUCL and PCCL groups, respectively.¹⁷ Javanmard et al., 2018 reported that the TUCL group (1.3 ± 0.6 days) had shorter length of stay than the PCCL group (2.4 ± 0.7 days) (p=0.001).¹⁵

Group B (TUCL versus CL)

Javanmard et al., 2018 reported that the TUCL group (1.3 ± 0.6 days) had shorter length of stay than the CL group (3.5 ± 1 days) (p=0.001).¹⁵ Mahran et al., 2000 reported significant differences regarding length of stay. It was shorter in the TUCL group (1.6 ± 0.8 days) than the CL group (3.9 ± 1.2 days) (p<0.05).¹⁶

Group C (PCCL versus CL)

Javanmard et al., 2018 reported that the PCCL group (2.4 ± 0.7 days) had a shorter length of stay than the CL group (3.5 ± 1 days) (p=0.001).¹⁵ Mahran et al., 2000 reported significant differences in terms of length of stay. It was shorter in the PCCL group (1.6 ± 0.8 days) than the CL group (3.9 ± 1.2 days) (p<0.05).¹⁶

Results of syntheses

Stone-free rate

All five studies reported the SFRs for all techniques. A meta-analysis for group A (TUCL versus PCCL) revealed no significant intergroup differences in SFR (OR 0.31; 95% CI: 0.05-2.00; p = 0.22; Figure 3). Meta-analysis was not deemed appropriate for other groups (TUCL versus PCCL and TUCL versus CL) because the heterogeneity test was not applicable (Figure 3).

Figure 3. Forest plot of TUCL vs. PCCL stone-free rates.

TUCL, transurethral cystolithotripsy; PCCL, percutaneous cystolithotomy.

Complications

Complication rates were not significantly different among all studies in group A (TUCL versus PCCL) (OR: 1.59; 95% CI: 0.81-3.12; p = 0.18) (Figure 4). Group B (TUCL versus CL) showed that the complication rate was lower among the CL group, but the intergroup differences were not significant (OR: 2.32; 95% CI: 0.90-5.99; p = 0.08) (Figure 5). For group C (PCCL versus CL), the complication rate seemed lower among the CL patients, but the difference was not significant (OR: 2.11; 95% CI: 0.90-4.95; p = 0.09) (Figure 6).

Figure 4. Forest plot of group A (TUCL vs. PCCL) complications.

TUCL, transurethral cystolithotripsy; PCCL, percutaneous cystolithotomy.

Figure 5. Forest plot of group B (TUCL vs. CL) complications.

TUCL, transurethral cystolithotripsy; CL, cystolithotomy.

Figure 6. Forest plot of group C (PCCL vs. CL) complications.

PCCL, percutaneous cystolithotomy; CL, cystolithotomy.

Procedure duration

For group A (TUCL vs. PCCL), the procedure duration was reported in all studies, and a meta-analysis showed a significant difference in favour of PCCL (Mean difference: 5.08; 95% CI: 2.87-7.28; p < 0.00001) (Figure 7). Group B showed that the procedure duration of the CL group was shorter than that of the TUCL group (Mean Difference 9.83; 95% CI: 7.50-12.17; p < 0.00001) (Figure 8). Group C reported that the CL group had a significantly shorter procedure duration than the PCCL group (Mean difference 4.91; 95% CI: 2.90-6.92; p < 0.0001) (Figure 9).

Figure 7. Forest plot of group A (TUCL vs. PCCL) procedure durations.

TUCL, transurethral cystolithotripsy; PCCL, percutaneous cystolithotomy.

Figure 8. Forest plot of group B (TUCL vs. CL) procedure durations.

TUCL, transurethral cystolithotripsy; CL, cystolithotomy.

Figure 9. Forest plot of group C (PCCL vs. CL) procedure durations.

PCCL, percutaneous cystolithotomy; CL, cystolithotomy.

Length of hospital stays

Figure 10 shows that TUCL resulted in a significantly shorter length of hospital stay than PCCL (Mean difference -0.71; 95% CI: -0.96-(-0.46); p < 0.00001). For group B, the length of hospital stay was shorter in the TUCL group than the CL group (Mean difference -2.22; 95% CI: -2.50-(-1.93); p < 0.00001) (Figure 11). Group C showed that the PCCL group had a significantly shorter hospital stay than the CL group (Mean difference: -1.35; 95% CI: -1.62-(-1.07); p < 0.00001) (Figure 12).

Figure 10. Forest plot of TUCL vs. PCCL lengths of hospital stay.

TUCL, transurethral cystolithotripsy; PCCL, percutaneous cystolithotomy.

Figure 11. Forest plot of TUCL vs. CL lengths of hospital stay.

TUCL, transurethral cystolithotripsy; CL, cystolithotomy.

Figure 12. Forest plot of PCCL vs. CL lengths of hospital stay.

PCCL, percutaneous cystolithotomy; CL, cystolithotomy.

Reporting biases

As a rule of thumb, tests for funnel plot asymmetry should be used only when there are at least 10 studies included in the meta-analysis, because when there are fewer studies the power of the tests is too low to distinguish chance from real asymmetry. Thus, in this meta-analysis we did not proceed with reporting bias assessment.

Certainty of evidence

This systematic review and meta-analysis have several limitations. In general, the quality of the studies according to the GRADE score in this meta-analysis varied from low to moderate. There were some studies with serious limitations and no reporting bias assessment was reported (Table 2).