Pediatric endoscopic retrograde cholangiopancreatography: A systematic review of indications, technical success, adverse events, and complication prevention

Abstract

BACKGROUND

Pediatric endoscopic retrograde cholangiopancreatography (ERCP) has evolved from a mainly diagnostic test into a predominantly therapeutic intervention for pancreatobiliary disease. Yet children differ from adults in body size, anesthetic requirements, radiation sensitivity, pancreaticobiliary anatomy, and access to appropriately sized equipment. Current evidence remains fragmented across indications, techniques, adverse-event definitions, and prevention strategies, and reporting of post-ERCP pancreatitis (PEP) is inconsistent. We reasoned that a descriptive, endpoint-specific synthesis of recent studies would characterize practice, outcomes, and safety more faithfully than pooled estimates derived from incompatible denominators.

AIM

To synthesize recent evidence on pediatric ERCP indications, technical success, adverse events, PEP, and its prevention.

METHODS

PubMed/MEDLINE, EMBASE, Web of Science, and Cochrane Library were searched. After deduplication and two-stage screening, 356 full-text reports were assessed, 144 underwent data extraction, and 128 studies were retained following post-extraction quality verification. Given the clinical and methodological heterogeneity, no meta-analysis was performed; instead, study-level outcomes were summarized descriptively with endpoint-specific denominators, reporting medians, interquartile range (IQR), and counts.

RESULTS

The 128 studies were published from 2017 to 2026, with 101 (78.9%) appearing since 2020; single-center retrospective series predominated (70 studies, 54.7%). The United States and China were the leading data sources (32 studies each, 25.0%). Pancreatitis or pancreatic-duct disease (82 studies, 64.1%) and choledocholithiasis (81 studies, 63.3%) were the dominant indications, and ERCP was primarily therapeutic (88 studies, 68.8%). Median study-level technical success was 94.7% (IQR: 89.8%-97.2%; 49 studies), median PEP rate was 5.2% (IQR: 3.0%-9.5%; 39 studies), and median overall adverse-event rate was 8.9% (IQR: 3.2%-14.2%; 89 studies).

CONCLUSION

Pediatric ERCP is generally technically successful, and PEP remains the principal procedure-related adverse event. Standardized definitions, prospective registries, and pediatric-specific PEP prevention trials are needed.

Key Words: Children; Endoscopic retrograde cholangiopancreatography; Post-endoscopic retrograde cholangiopancreatography pancreatitis; Adverse events; Biliary tract diseases; Pancreatic diseases; Choledocholithiasis; Systematic review

Core Tip: This systematic review synthesizes recent evidence on pediatric endoscopic retrograde cholangiopancreatography from a finalized 128-study dataset. Rather than pooling event rates across incompatible definitions, it emphasizes endpoint-specific denominators. Contemporary studies show high study-level technical success (median 94.7%), whereas post-endoscopic retrograde cholangiopancreatography pancreatitis (PEP; median 5.2%) remains the principal procedure-related adverse event. The review highlights persistent reporting gaps and the need for standardized pediatric definitions, prospective multicenter registries, and pediatric-specific PEP prevention trials, including those of rectal nonsteroidal anti-inflammatory drugs and pancreatic-duct stents.

INTRODUCTION

Endoscopic retrograde cholangiopancreatography (ERCP) has evolved from a primarily diagnostic procedure into a predominantly therapeutic intervention for pancreatobiliary disease. In children, the decision to proceed with ERCP is governed by a different balance of benefits and risks than in adults. Pediatric patients vary widely in body size, airway and anesthetic requirements, radiation sensitivity, pancreaticobiliary anatomy, and access to pediatric-sized accessories, and many programs rely on close collaboration among pediatric gastroenterologists, advanced endoscopists, surgeons, interventional radiologists, and anesthesiologists[1,2]. With growing pediatric expertise and the advent of dedicated registries, ERCP is now offered across a broader range of ages and indications, including in infants and very young children[1,3,4].

The clinical indications for pediatric ERCP are diverse. Common scenarios include choledocholithiasis and biliary obstruction, acute recurrent and chronic pancreatitis, pancreatic-duct anomalies and pancreas divisum, pancreaticobiliary maljunction, congenital biliary dilatation or choledochal cyst, biliary strictures after liver transplantation, bile leaks, and pancreatic trauma[5-9]. Several contemporary series report high technical success, yet the evidence base remains fragmented across single-center cohorts, administrative datasets, multicenter collaborations, case series, and disease-specific studies[5,10-13]. Multicenter collaboratives such as the Pediatric ERCP Database Initiative and the INternational Study group of Pediatric Pancreatitis: In search for a cuRE have begun to describe practice patterns, technical outcomes, and safety prospectively across many centers[5,14,15].

Safety assessment is equally heterogeneous. Post-ERCP pancreatitis (PEP) is usually the most conspicuous adverse event, but studies differ in how they define it, whether they distinguish transient hyperamylasemia from pancreatitis, how long they follow patients, and whether they count bleeding, perforation, infection, pain, anesthesia-related events, readmission, or reintervention as procedure-related complications. Earlier syntheses, including a systematic review with quantitative meta-analysis of pediatric ERCP complications and a 21-year tertiary-center series with literature review, helped frame these questions but were constrained by the available primary data[16,17]. Adult-derived definitions and prevention strategies are frequently invoked, yet direct pediatric risk-factor evidence remains limited and at times conflicting[18-20]. Pediatric prevention evidence is likewise still emerging, resting on early trials and observational studies of rectal or systemic nonsteroidal anti-inflammatory drugs[21-24].

For these reasons, a synthesis of recent pediatric ERCP evidence should not simply pool event rates across incompatible denominators. Using a descriptive, endpoint-specific approach, this review summarizes a finalized 128-study analysis set to characterize recent study designs, indications, interventions, technical success, adverse events, and PEP reporting, and to define priorities for future pediatric-specific research.

MATERIALS AND METHODS

Search strategy and study identification

A literature search identified 1774 records across four databases: (1) PubMed/MEDLINE (n = 481); (2) EMBASE (n = 912); (3) Web of Science Core Collection (n = 330); and (4) Cochrane Library (n = 51). After removal of 599 duplicates by normalized DOI, PMID, and title matching, 1175 unique records underwent title and abstract screening. At this stage, 765 records were excluded as not pediatric (n = 275), not ERCP-focused (n = 471), or not addressing adverse events, PEP prevention, safety, technical outcomes, or other clinically relevant ERCP outcomes (n = 19). The remaining 410 records proceeded to full-text retrieval.

Among these 410 records, 356 had a matched full-text PDF, whereas 54 lacked usable full text or were triaged before full-text assessment. The 54 records without usable full text comprised unavailable full texts (n = 23), conference abstracts without complete articles (n = 14), protocol or registry records (n = 10), guideline or consensus records (n = 5), and records retained only as background citations (n = 2). The 356 full-text reports were then assessed for eligibility: (1) 144 were included for data extraction; and (2) 212 were excluded or retained only as background citations. The grouped primary reasons were not ERCP-focused or insufficient ERCP data (n = 79), case report or very small case series (n = 49), non-pediatric population or pediatric data not separable (n = 46), secondary literature, commentary, guideline, or survey (n = 36), non-English full text (n = 1), and another full-text exclusion reason (n = 1).

After extraction, the 144 records underwent post-extraction quality verification before the final descriptive synthesis. Nine were excluded in the first quality-control round and seven more in the second. The 16 post-extraction exclusions were attributable to adult or non-pediatric populations or non-separable pediatric data (n = 8), review or meta-analysis articles without original cohort data (n = 4), conference-abstract compilations (n = 2), ERCP outcomes that were not separately extractable (n = 1), or severe PDF encoding damage (n = 1). The final descriptive synthesis therefore comprised 128 studies (Figure 1).

Figure 1 Preferred Reporting Items for Systematic reviews and Meta-Analyses flow diagram of study identification, screening, eligibility, and inclusion. The search identified 1774 records; 1175 remained after deduplication; 410 were selected for full-text retrieval; 356 full-text reports were assessed; 144 records underwent data extraction; and 128 studies were included in the final descriptive synthesis.

Eligibility criteria

Eligible records were pediatric or predominantly pediatric ERCP studies reporting indications, technical outcomes, adverse events, PEP, risk factors, PEP prevention strategies, peri-procedural care, or other clinically relevant therapeutic outcomes. Studies were excluded when the population was adult or could not be stratified by age, when ERCP was absent or only incidental, when the record lacked original data extractable for the present synthesis, or when pediatric ERCP data could not be separately verified.

Data extraction and variables

For each included study, the extracted variables comprised publication year, country or region, study design, sample size, age, ERCP volume, indications, treatment or intervention type, technical success, complications or adverse events, PEP, follow-up, and main conclusions. Sample size and ERCP volume were retained as reported, because studies variously used patients, procedures, admissions, database records, or strategy groups rather than a uniform patient-level denominator (Figures 2, 3, 4, and 5).

Figure 2 Publication trend of included studies. Annual number of studies included in the final 128-study descriptive synthesis, by publication year.

Figure 3 Country or region distribution of included studies. Leading data-source countries or regions among the 128 included studies.

Figure 4 Indications and therapeutic interventions among included studies. Study-level keyword classification; categories are not mutually exclusive.

Figure 5 Study-level outcome distributions. Endpoint-specific descriptive distributions for technical success, post-endoscopic retrograde cholangiopancreatography pancreatitis, and overall complications or adverse events. ERCP: Endoscopic retrograde cholangiopancreatography; IQR: Interquartile range.

Definitions and denominator handling

Technical success was defined pragmatically as completion of the intended ERCP-related technical objective as reported by each source study, such as successful cannulation, therapeutic completion, stone clearance, drainage, stent placement, stricture treatment, or another intended procedural endpoint. Percentages referring to diagnostic accuracy, prediction-model performance, symptom improvement, or other non-technical outcomes were excluded from the technical-success summary.

Overall complications or adverse events comprised ERCP or endoscopy-related events reported by the source studies, namely, PEP, bleeding, perforation, cholangitis or infection, stent-related events, reintervention, anesthesia-related events, and death when described as procedure-related or temporally associated. PEP was counted only when explicitly reported as PEP or ERCP-related pancreatitis. Because of heterogeneity in populations, indications, definitions, ascertainment, and follow-up, no pooled meta-analysis was performed. Continuous study-level variables were summarized as medians, interquartile range (IQR), and ranges, and categorical variables as counts and percentages of the 128 included studies. Indication and intervention categories were not mutually exclusive.

Statistical analysis

This systematic review used descriptive study-level statistics only. No inferential hypothesis testing, comparative analysis, or pooled meta-analysis was performed; accordingly, no formal biostatistical review was applicable. Continuous variables were summarized as medians with IQR and ranges, and categorical variables as counts and percentages, all derived from the extracted study-level data.

RESULTS

Study selection

The database search yielded 1774 records, from which 599 duplicates were removed. Title and abstract screening of the 1175 unique records selected 410 for full-text retrieval. A usable full-text PDF was available for 356 records, whereas 54 lacked usable full text or were triaged before full-text assessment. Full-text assessment of the 356 reports yielded 144 records for data extraction and 212 that were excluded or retained only as background citations. After post-extraction quality verification, 16 further records were excluded, leaving 128 studies in the final descriptive synthesis (Figure 1).

Characteristics of included studies

The 128 included studies were published between 2017 and 2026, with 101 (78.9%) appearing in 2020 or later, reflecting a marked recent increase in pediatric ERCP reporting (Table 1). The evidence base was geographically diverse: (1) The United States and China were the most frequent data sources (32 studies each, 25.0%); (2) Multinational or international collaborative studies (12 studies, 9.4%); (3) Türkiye (9 studies, 7.0%); (4) Argentina (5 studies, 3.9%); (5) India (4 studies, 3.1%); and (6) Italy (4 studies, 3.1%).

Table 1 Study characteristics in the final 128-study analysis set, n (%).

Characteristic	Finding
Included studies	128
Publication years	2017-2026
Published in 2020 or later	101 studies (78.9)
Leading data-source countries or regions	United States 32 (25.0); China 32 (25.0); multinational or international 12 (9.4); Türkiye 9 (7.0)
Most common design	Single-center retrospective study: 70 (54.7)
Other designs	Multicenter retrospective 14 (10.9); national or administrative database 12 (9.4); case series 10 (7.8); prospective, randomized, or prospective database 6 (4.7)
Extractable sample size	127 studies; median 66 (IQR: 31-147; range, 1-1124)
Extractable ERCP volume	123 studies; median 68 (IQR: 30-150; range, 1-1124)

ERCP: Endoscopic retrograde cholangiopancreatography; IQR: Interquartile range.

Most studies were observational. Single-center retrospective studies accounted for 70 (54.7%), followed by multicenter retrospective studies (14, 10.9%), national or administrative database studies (12, 9.4%), case series (10, 7.8%), prospective studies, randomized trials, or prospective database studies (6, 4.7%), case reports (4, 3.1%), and conference-abstract-based studies (2, 1.6%); a further 10 studies (7.8%) required manual classification. Large national administrative datasets and multicenter registries contributed several of the highest-volume analyses[5,13,20]. Sample size could be extracted from 127 studies, with a median study-level sample size of 66 patients or records (IQR: 31-147; range, 1-1124). ERCP volume could be extracted from 123 studies, with a median of 68 procedures or ERCP-related records per study (IQR: 30-150; range, 1-1124).

Indications and interventions

Pediatric ERCP indications were heterogeneous and frequently overlapping. Pancreatitis or pancreatic-duct disease was represented in 82 studies (64.1%) and choledocholithiasis or common bile duct stones in 81 (63.3%) (Table 2)[25]. Congenital biliary dilatation, choledochal cyst, or pancreaticobiliary maljunction appeared in 26 studies (20.3%), as did biliary stricture or post-transplant biliary disease (26 studies, 20.3%)[7,8]. Less common categories included pancreatic trauma or injury, biliary leak or bile-duct injury, biliary atresia, primary sclerosing cholangitis or other cholestatic disorders, and other or mixed indications[9,26].

Table 2 Study-level classification of indications and interventions, n (%).

Category	Studies
Pancreatitis or pancreatic-duct disease	82 (64.1)
Choledocholithiasis or common bile duct stones	81 (63.3)
Congenital biliary dilatation, choledochal cyst, or pancreaticobiliary maljunction	26 (20.3)
Biliary stricture or post-transplant biliary disease	26 (20.3)
Sphincterotomy or papillotomy	72 (56.3)
Stent placement or removal	69 (53.9)
Stone extraction, lithotripsy, or duct clearance	52 (40.6)
Balloon dilation or sweeping	33 (25.8)
Biliary or nasobiliary drainage	23 (18.0)

Categories are not mutually exclusive; percentages use the 128 included studies as the denominator.

Therapeutic ERCP predominated. By study-level keyword classification, 88 studies (68.8%) primarily involved therapeutic ERCP, 11 (8.6%) combined diagnostic and therapeutic ERCP, and 29 (22.7%) did not specify the ERCP role clearly enough for automated classification. The most frequently reported interventions were sphincterotomy or papillotomy (72 studies, 56.3%), stent placement or removal (69 studies, 53.9%), stone extraction, lithotripsy, or duct clearance (52 studies, 40.6%), balloon dilation or sweeping (33 studies, 25.8%), and biliary or nasobiliary drainage (23 studies, 18.0%).

Technical success

Technical success was recorded in the extraction table for all 128 studies, but many entries were narrative or referred to outcomes other than ERCP technical success. A numeric percentage could be parsed automatically from 106 studies, and after semantic filtering, 49 were deemed suitable for descriptive analysis. Among these, the median study-level technical success rate was 94.7% (IQR: 89.8%-97.2%; range, 24%-100%) (Table 3). This aligns with large contemporary multicenter experiences that report high overall technical success while noting lower rates in younger or smaller children and in pancreatic indications[5,6,10-12].

Table 3 Endpoint-specific outcome summaries.

Endpoint	Usable studies (n)	Study-level summary
Technical success	49	Median 94.7% (IQR: 89.8%-97.2%; range, 24%-100%)
PEP	39	Median 5.2% (IQR: 3.0%-9.5%; range, 0%-20%)
Overall complications or adverse events	89	Median 8.9% (IQR: 3.2%-14.2%; range, 0%-50%)

Missing, ambiguous, or non-endpoint-specific text was not treated as zero-event data. PEP: Post-endoscopic retrograde cholangiopancreatography pancreatitis; IQR: Interquartile range.

Adverse events and PEP

PEP was the most consistently described procedure-related adverse event. A numeric PEP percentage could be parsed from 83 studies, but only 39 were semantically suitable for PEP-specific analysis after entries referring to overall adverse events, non-PEP complications, or non-specific text were excluded. Among these 39 studies, the median study-level PEP rate was 5.2% (IQR: 3.0%-9.5%; range, 0%-20%), in line with dedicated pediatric PEP analyses[18-20].

Complication data were available in 123 of 128 studies. A numeric complication or adverse-event rate could be extracted from 89, among which the median study-level complication rate was 8.9% (IQR: 3.2%-14.2%; range, 0%-50%). Besides PEP, reported adverse events included bleeding, cholangitis or infection, perforation, stent-related events, need for reintervention, and rare deaths[27]. Most studies graded adverse events as mild to moderate, but definitions, follow-up intervals, and reporting granularity varied substantially.

Pediatric-adult comparison and peri-procedural reporting

Direct pediatric-adult evidence was limited. In an American Society for Gastrointestinal Endoscopy complexity-matched comparison, general anesthesia was used in 84% of pediatric vs 54% of adult ERCPs (P < 0.0001); pediatric and adult PEP rates were 5.2% and 6.9%, respectively (P = 0.4), and overall adverse-event rates were 6.1% and 10.0% (P = 0.1)[12]. Most pediatric studies reported only an anesthesia category rather than reproducible drug-level doses. One small post-liver-transplant cohort of 7 children reported conscious sedation with midazolam 5 mg plus pethidine 50 mg, with occasional fentanyl; this isolated fixed-dose regimen was not regarded as a general pediatric dosing recommendation[28].

Radiation and procedure-time data were likewise study-specific. The Pediatric ERCP Database Initiative reported a median fluoroscopy time of 120 seconds (IQR: 60-240 seconds) across 1073 procedures, whereas a separate 385-procedure cohort reported a median of 4.85 minutes and identified endoscopist volume as the strongest predictor of prolonged exposure[29,30]. In a 186-procedure series, 119 procedures (64%) lasted less than 60 minutes and 67 (36%) lasted 60 minutes or longer, and another cohort reported a mean ERCP duration of 54.2 minutes (SD = 18.7 minutes)[31,32]. These estimates were not pooled because studies variously reported procedure time, anesthesia time, cannulation time, combined operative time, or total room time.

PEP-prevention doses were available from a small number of pediatric studies. A 906-procedure cohort used rectal indomethacin 50 mg in patients weighing less than 30 kg and 100 mg in those weighing more than 30 kg; PEP occurred after 3.0% of procedures with indomethacin vs 9.5% without it (adjusted odds ratio = 0.28), without any increase in bleeding or acute kidney injury[23]. A feasibility trial used intravenous ibuprofen 10 mg/kg (maximum 800 mg), and an observational cohort used intravenous ketorolac 0.5 mg/kg (maximum 30 mg), with a lower PEP rate after pancreatic-duct manipulation (11% vs 25%)[21,22].

PEP severity was reported inconsistently. Severe PEP occurred in 3 of 232 pediatric procedures (1.2%) in the complexity-matched study[12], in 1 of 92 (1.1%) in a separate cohort with 19 PEP events[19], and in 3 of 298 (1.0%) in the ketorolac cohort[22]. These study-specific proportions were not combined, because the severity criteria, denominators, and follow-up windows were not uniform. The pediatric-adult comparison and reporting limitations are summarized in Table 4.

Table 4 Pediatric-adult comparison and reporting boundaries.

Domain	Pediatric evidence	Adult comparison and interpretation
Indications	Pancreatitis or pancreatic-duct disease appeared in 82/128 studies (64.1%), choledocholithiasis in 81/128 (63.3%), and congenital biliary disease or post-transplant/stricture disease in 26/128 each (20.3%); malignant indications were uncommon	Adult ERCP includes a greater malignant-obstruction burden. Differences in case mix affect procedural intent, repeat intervention, and interpretation of adverse-event rates
Anesthesia, sedation, and analgesia	A complexity-matched study used general anesthesia in 84% of pediatric vs 54% of adult ERCPs. Drug-level doses were rarely reported; one 7-child transplant cohort used midazolam 5 mg plus pethidine 50 mg, with occasional fentanyl	The isolated fixed-dose regimen is not a general pediatric dosing recommendation. The available literature does not support a pooled sedative or analgesic dose
Radiation exposure	Median fluoroscopy time was 120 seconds (IQR: 60-240) in 1073 pediatric ERCPs and 4.85 minutes in a separate 385-procedure cohort; lower operator volume predicted longer exposure	Fluoroscopy time, kerma-area product, cumulative air kerma, and image count are not interchangeable. ALARA, low-dose pulsed fluoroscopy, collimation, and dose-metric documentation are appropriate safeguards
Procedure duration	One 186-procedure series reported 119 procedures (64%) lasting < 60 minutes and 67 (36%) lasting ≥ 60 minutes. Another cohort reported a mean duration of 54.2 minutes (SD = 18.7)	Definitions varied among procedure time, anesthesia time, cannulation time, and combined operative time; a pooled pediatric-adult duration comparison would therefore be invalid
PEP prophylaxis	Reported pediatric regimens included rectal indomethacin 50 mg below 30 kg and 100 mg above 30 kg (PEP 3.0% vs 9.5% without treatment), intravenous ibuprofen 10 mg/kg (maximum 800 mg), and intravenous ketorolac 0.5 mg/kg (maximum 30 mg)	Adult guidance uses a fixed 100-mg rectal NSAID dose. Pediatric evidence is weight-sensitive and study-specific; a universal pediatric regimen cannot yet be inferred
Adverse events and PEP severity	Review medians were 5.2% for PEP (39 studies) and 8.9% for overall adverse events (89 studies). In the matched study, pediatric vs adult PEP was 5.2% vs 6.9% and adverse events were 6.1% vs 10.0%. Severe PEP was approximately 1.0%-1.2% of procedures in cohorts that graded severity	Severe PEP was uncommon in reporting cohorts, but no dataset-wide incidence was calculated because severity definitions, denominators, and follow-up windows were inconsistent; unreported events were not treated as zero

Values are study-specific unless identified as review-level medians. Adult comparisons are not pooled into the pediatric distributions. ALARA: As Low As Reasonably Achievable; ERCP: Endoscopic retrograde cholangiopancreatography; IQR: Interquartile range; PEP: Post-endoscopic retrograde cholangiopancreatography pancreatitis; NSAID: Nonsteroidal anti-inflammatory drug.

DISCUSSION

Across these 128 studies, pediatric ERCP is largely a therapeutic procedure, is performed in a growing number of countries, and achieves high technical success in experienced hands. Most reports remain retrospective single-center series, but several large multicenter datasets and prospective registry analyses now describe practice patterns, technical outcomes, and safety in broader populations[5,6,11,12]. The aggregate rates discussed below should be read in light of differences in case mix and reporting quality rather than treated as fixed benchmarks.

The pediatric ERCP disease profile differs materially from adult practice. In the present 128-study dataset, pancreatitis or pancreatic-duct disease appeared in 82 studies (64.1%), choledocholithiasis in 81 (63.3%), congenital biliary dilatation, choledochal cyst, or pancreaticobiliary maljunction in 26 (20.3%), and biliary stricture or post-transplant disease in 26 (20.3%); these categories were not mutually exclusive (Table 4). Malignancy was not a recurring pediatric study-level indication, whereas malignant biliary obstruction constitutes a larger share of adult ERCP practice. Such differences in case mix influence procedural intent, recurrence, the need for repeat intervention, and the interpretation of adverse-event rates. For biliary stones, the recurring questions are whether to follow an ERCP-first or surgery-first pathway, when to perform cholecystectomy, and how to avoid unnecessary ERCP; several surgical groups now favor transcystic laparoscopic common bile duct exploration or concurrent single-anesthesia approaches as alternatives or complements to ERCP[33-36]. Pancreatic indications typically involve acute recurrent or chronic pancreatitis, pancreatic-duct anomalies, pancreas divisum, or trauma, for which repeated interventions and PEP risk weigh more heavily[14,15,37-40].

High technical success must likewise be interpreted in light of case mix and peri-procedural care. The complexity-matched pediatric-adult comparison found similar technical and safety outcomes but more frequent general anesthesia in children (84% vs 54%)[12]. Most pediatric primary studies did not report weight-adjusted or age-adjusted sedative and analgesic doses. The only explicit fixed regimen in the extraction matrix came from a 7-child post-transplant cohort (midazolam 5 mg plus pethidine 50 mg, with occasional fentanyl), which is too small and context-specific to support a general dosing recommendation[28]. Radiation and time measures were similarly heterogeneous: Median fluoroscopy time was 120 seconds (IQR: 60-240) in 1073 procedures and 4.85 minutes in a separate 385-procedure cohort[29,30], whereas reported ERCP duration comprised 64% of procedures lasting less than 60 minutes in one series and a mean of 54.2 minutes (SD = 18.7) in another[31,32]. Collectively, these study-specific values support the As Low As Reasonably Achievable principle, pre-procedure magnetic resonance cholangiopancreatography when appropriate, avoidance of diagnostic ERCP, low-dose pulsed fluoroscopy, tight collimation, avoidance of unnecessary spot images, documentation of fluoroscopy time and dose metrics, and referral of complex cases to experienced centers[41].

PEP remained the safety outcome of greatest concern. The median study-level PEP rate was 5.2% (IQR: 3.0%-9.5%; 39 studies), and the median overall adverse-event rate was 8.9% (IQR: 3.2%-14.2%; 89 studies); these are medians of study-level rates rather than pooled patient-level incidences. In the complexity-matched comparison, pediatric and adult PEP rates were 5.2% vs 6.9% (P = 0.4) and overall adverse events 6.1% vs 10.0% (P = 0.1)[12]. Severe PEP accounted for approximately 1.0%-1.2% of all procedures in the several cohorts that explicitly graded severity[12,19,22]. This does not establish a review-wide severe-PEP incidence, because most studies did not apply uniform severity criteria, denominators, or follow-up windows; unreported severe PEP was not treated as zero.

PEP prevention is an evolving area. Adult guidance recommends a fixed 100-mg rectal dose of indomethacin or diclofenac in eligible adults[42], but pediatric dosing is weight-sensitive. A 906-procedure pediatric cohort used rectal indomethacin 50 mg below 30 kg and 100 mg above 30 kg and reported PEP after 3.0% of procedures with prophylaxis vs 9.5% without, without an increase in bleeding or acute kidney injury[23]. A randomized feasibility study used intravenous ibuprofen 10 mg/kg (maximum 800 mg) but was underpowered for PEP, whereas an observational study used intravenous ketorolac 0.5 mg/kg (maximum 30 mg) and found a lower PEP rate among children undergoing pancreatic-duct manipulation (11% vs 25%)[21,22]. These data should be read as study-specific regimens rather than a universal pediatric algorithm; prospective validation must account for weight, renal function, bleeding risk, and local anesthesia and endoscopy protocols.

ERCP also has well-defined roles beyond stone and pancreatic disease. In infants and neonates, it is technically feasible in experienced hands and can accurately exclude biliary atresia in selected cholestatic infants, albeit with particular anesthetic and equipment demands[1,3,4,26]. After pediatric liver transplantation, endoscopic treatment of anastomotic biliary strictures is an established, graft-preserving strategy[8]. ERCP further contributes to the management of pancreaticobiliary maljunction and congenital biliary dilatation and to the diagnosis and drainage of pancreatic-duct injury after trauma[7,9]. How these procedures are delivered is also changing: Single-anesthesia ERCP combined with laparoscopic cholecystectomy, single-use duodenoscopes, and outpatient pediatric ERCP all reflect evolving care models[43-45].

For programs, the practical message is to invest in structured risk stratification before the procedure rather than rescue afterward. A pre-procedural assessment should weigh age and body size, native-papilla status, the biliary vs pancreatic indication, planned pancreatic-duct manipulation, the likelihood of difficult cannulation, altered or post-surgical anatomy, transplant history, the anticipated need for repeat ERCP, and local access to pediatric anesthesia, surgery, interventional radiology, and post-procedure observation. Context matters too: Secondary analyses suggest that ERCP for acute pancreatitis and procedures in lower-volume settings may carry different outcome profiles, and that complication rates track with center experience[46-50]. For suspected choledocholithiasis, recent surgical-pathway studies suggest that selected children fare better with improved case selection[51-53] and, in suitable centers, a surgery-first approach[34,35].

Reporting gaps were a recurring problem. Many records gave numeric percentages that were not valid technical-success or PEP endpoints, merged PEP with all adverse events, used inconsistent denominators, or omitted follow-up windows. Sedative and analgesic doses, radiation dose metrics, procedure-time definitions, preventive-agent dosing, and PEP severity were especially incomplete. Future studies should distinguish procedure time from anesthesia or room time; report fluoroscopy time together with kerma-area product or cumulative air kerma when available; state drug dose, timing, route, weight adjustment, and contraindications; specify whether denominators refer to patients or procedures; separate PEP from asymptomatic hyperamylasemia; grade severity by consensus criteria; and define a minimum follow-up window. Without such standardization, pooled estimates and direct pediatric-adult comparisons will remain unreliable.

CONCLUSION

Recent evidence supports the diagnostic and therapeutic role of ERCP in selected pediatric pancreatobiliary diseases, in which study-level technical success is generally high. PEP remains the principal procedure-related adverse event and should sit at the center of risk stratification, prevention, and post-procedure observation. Because the literature is still heterogeneous and largely retrospective, the priorities for future work are standardized pediatric ERCP definitions, prospective multicenter registries, pediatric-specific prevention trials, device and training pathways for smaller children, and long-term outcome assessment.