ERCP for the Surgeon — Prove the Stone, Earn the Cannulation, and Own the Duct Afterwards

The surgeon meets ERCP at three decision points — whether the stone is proven enough to justify the scope, how the duct gets cleared in one sitting or two, and who owns the gallbladder afterwards — because the scope is minimally invasive yet carries 6 to 15% adverse events, diagnosis alone never earns those odds, and the operation that follows decides whether the patient returns. Prove the stone with EUS or MRCP, cannulate by protocol with early precut when access is difficult, protect every pancreas with rectal NSAID and high-risk ones with a stent, open large stones with sphincterotomy plus balloon, choose single-stage exploration where expertise allows, drain cholangitis by severity, stent malignant strictures by resectability, and rescue bleeding and perforation without delay — with every number taken from the papers named beside it.^{[1][6][18][24]}

A 55-year-old woman with gallbladder stones, a dilated duct and a bilirubin of 5 mg/dL; a 70-year-old man with ascending cholangitis and a stone seen on ultrasound; and a 32-year-old woman at 18 weeks' gestation with choledocholithiasis and gallstone pancreatitis. One needs confirmation before therapy, one needs drainage on the day, and one needs the stone gone with the fetus shielded. The examiner will watch you refuse diagnostic ERCP, quote EUS-versus-MRCP accuracy, defend early precut with pancreatitis numbers, price NSAID-plus-stent prophylaxis, argue single-stage against two-stage with clearance and pancreatitis trade-offs, time the cholecystectomy within 72 hours, and stent the malignant stricture by resectability — with every number taken from the papers named beside it.^{[6][8][13][17][27]}

ERCP is the primary treatment of choledocholithiasis — minimally invasive, but with adverse events in 6 to 15% — and the American guideline built on GRADE methods sets the whole game: confirm before you scope, dilate after cutting for big stones, operate on the same admission for gallstone pancreatitis, and never spend a diagnostic ERCP where EUS or MRCP would do.^[1] The strategic arc fits one sentence: prove the stone noninvasively, cannulate with few attempts and early precut, medicate every pancreas, open the papilla to the stone size, clear duct and gallbladder in the strategy your unit does best, drain sepsis by severity, stent cancer by resectability, and watch for bleeding and perforation after every cut.^{[6][2][15][5][24]}

Prove the stone before the scope

High-risk criteria still find stones in 87.1% of patients, yet even there the tandem strategy pays: of 138 patients without ascending cholangitis who met the other two high-risk criteria and went to EUS first, 21 cases (15.2%) avoided a diagnostic ERCP — because diagnostic ERCP should be avoided, full stop.^[6][3] Cholangitis itself is the strongest single predictor, with a specificity of 67.7% and a positive predictive value of 90.2%, while bilirubin above 4 mg/dL with a dilated duct is far softer at 74.2% specificity and only 55.6% predictive value — so the jaundiced, dilated patient without cholangitis is exactly who needs EUS first.^[6] Cochrane pooling puts both confirmatory tests high: across 13 EUS studies sensitivity runs 0.95 with specificity 0.97, against 0.93 and 0.96 across seven MRCP studies, with no evidence of a difference between them — so a positive test of either kind earns extraction, and a negative one needs no further invasive test.^[7] Head-to-head pooling sharpens the edge slightly toward ultrasound: sensitivity 0.97 against 0.87 with specificity near-equal, and a diagnostic odds ratio of 162.5 against 79.0 — yet the randomised intermediate-likelihood trial found both sensitivities similarly high at 92 to 98% with few wrong decisions either way, and concluded the choice rests on local expertise, availability and patient preference.^[8][9] The European rule for that intermediate patient is explicit: EUS or MRCP first to make the indication for therapeutic ERCP.^[9]

Cannulate cleanly, precut early when stuck

Cannulation and sphincterotomy are the cornerstones on which almost every therapeutic ERCP stands, and the world guideline names the quality contract plainly: a high cannulation rate and a low pancreatitis rate are the indicators every endoscopist is judged by.^[4] The first discipline is restraint — keep the number of cannulation attempts as low as possible — because each blind prod at the papilla buys pancreatitis.^[2] When access is difficult, early precut beats persistence: across six randomised trials with 966 subjects, cannulation sat at 90% in both arms, yet pancreatitis fell to 2.5 % with early precut against 5.3 % with persistence — and the seven-trial update in 999 difficult-access patients confirmed it with a risk ratio of 0.57, identical cannulation, and no excess bleeding or perforation.^[13][14] The current synthesis is therefore conditional, not absolute: similar cannulation rates, less pancreatitis, no change in overall complications — so in experienced hands, cut early rather than prod longer.^[13][14]

Prevent pancreatitis by the book

Rectal NSAID is the floor for everyone: 100 mg of diclofenac or indomethacin immediately before or after ERCP in all patients without contraindication, with a 5-Fr prophylactic pancreatic stent strongly considered on top whenever risk is high.^[2] The adverse-events guideline repeats the pair with the high-risk triggers spelled out — inadvertent guidewire insertion or opacification of the pancreatic duct, and double-guidewire cannulation — and adds what not to do routinely: no salvage pancreatic stenting once pancreatitis has declared itself, and no routine antibiotic prophylaxis before ERCP.^[3] The stent question was settled by 1,950 high-risk patients across 20 North American centres: the trial demanded the upper bound of the pancreatitis difference to sit below a 5% non-inferiority margin, and indomethacin alone failed it — 145 cases (14.9%) against 110 (11.3%), a 3.6% risk difference — so the stent stays on top of the suppository in high-risk ducts.^[10] Network pooling of 55 trials in 17,062 patients ranks the options: 5-to-7 Fr stents, rectal diclofenac 100 mg, 3 Fr stents and rectal indometacin 100 mg all beat placebo, diclofenac 100 mg is the best performing rectal NSAID, and the background pancreatitis rate without prophylaxis sits at 12.2%.^[11] Hydration has earned a place beside the suppository: lactated Ringer's at 3 mL/kg/h during the procedure with a 20 ml/kg bolus after and 3 mL/kg/h for eight more hours proved noninferior to rectal indomethacin — six pancreatitis cases overall (1.7%), one with hydration against five with indomethacin — so volume is a legitimate alternative where NSAIDs are contraindicated.^[12]

Open the papilla for the stone in front of you

Balloon dilation is not a routine substitute for cutting — but beyond routine stones the combination wins: for large duct stones the network of 16 studies in 2,545 patients gives the sphincterotomy-plus-balloon arm a higher index clearance than cutting alone with less need for mechanical lithotripsy, and no excess bleeding, perforation, pancreatitis, recurrence or death.^[2][19] The large-stone threshold in that pooling is 10 mm and above, and the lithotripsy saving is concrete: less need with the combination (relative risk 0.48) and with balloon alone (0.58) against cutting.^[19] Small-balloon technique tells the same story at smaller scale: adding a cut to balloon dilation shortens stone removal time and lifts the initial clearance rate across 497 against 548 patients, with total clearance, lithotripsy use and complications equal.^[20] When the stricture — not the stone — is the question, the single-operator cholangioscope earns its keep: in 165 patients across 179 procedures, accuracy for malignancy ran 87% with adequate biopsies in 72%, and difficult stones cleared in 73% — with cholangitis, one case fatal, as the price of looking inside.^[25] And before any stent goes through the papilla, know the cutting trade: across seven randomised trials in 1,022 stented patients, adding sphincterotomy did not move pancreatitis either way but multiplied bleeding sevenfold — so the guideline stance holds, and routine pre-stent cutting is declined for single plastic or uncovered and partially covered metal stents.^[21][3]

One stage or two: the surgeon's choice

Eleven randomised trials in 1,513 patients give the single-stage operation the technical edge: fewer failures and a shorter stay by over a day and a half, with mortality, morbidity, cost and retained stones equal — superior in good-risk hands where expertise, time and instruments allow.^[15] The 2025 update across 14 studies and 1,849 patients softens that verdict to equipoise: success 872 of 919 against 866 of 930, with residual stones, mortality, morbidity and stay all statistically tied — both techniques viable, with multicentre trials still wanted.^[16] The preoperative-ERCP comparison explains the tension: across 12 trials and 1,545 patients, ERCP first clears more ducts with fewer bile leaks, but buys that clearance with significantly more pancreatitis — while the single-stage operation shortens the overall stay with morbidity and mortality tied.^[17] Whichever route clears the duct, the gallbladder must follow fast: eight randomised trials in 576 patients show operating within 72 hours probably cuts conversion to open surgery and recurrent biliary events by an order of magnitude, and may shorten operating time, complications and stay — with the rare-duct-injury evidence still too thin to price.^[18]

Drain cholangitis now, stage the rest

Mild cholangitis usually needs no drain at all — antibiotics and supportive care suffice in most cases, with drainage held for non-responders — while moderate disease earns early endoscopic or percutaneous drainage, with sphincterotomy and stone clearance staged once the patient has improved.^[5] Severe disease reverses the order: resuscitate the circulation and breathing first, then drain as soon as the patient is stabilised — never scope a crashing patient before resuscitation.^[5] Antibiotics ride with drainage, not before the scope as ritual: routine prophylaxis before ERCP is discouraged, held for anticipated incomplete drainage, severe immunocompromise, and cholangioscopy cases.^[3]

Stent the malignant stricture, resect when resectable

Covered and uncovered metal stents palliate distal malignant obstruction equally: across seven studies and 1,070 patients, patency, failure, survival and adverse events match — covered stents migrate more (relative risk 2.34) with more overgrowth, uncovered ones grow tumour inward instead — so cover choice is a trade of failure modes, not of survival.^[22] Before resection the calculus changes toward thrift: in 86 randomised jaundiced patients, covered metal and plastic matched on procedure events, reintervention, time to surgery and surgical complications — so with similar outcomes, plastic is preferred for its cost-effectiveness.^[23]

Name the bleed and the hole

Overall ERCP complications run 5 to 10%, led by pancreatitis and clinically significant post-sphincterotomy bleeding — and the bleeding that will not stop has a named rescue: a fully covered metal stent, temporarily placed, when standard haemostasis fails.^[24][3] Perforation is rarer at 0.14 to 1.6% and deadlier at 4.2 to 29.6% mortality, classified by the instrument that caused it — and the evidence is blunt that delayed diagnosis and treatment worsen prognosis, so immediate recognition and early management by perforation type is what minimises death.^[24]

Special ducts: bypassed stomachs and pregnancies

Roux-en-Y anatomy defeats the enteroscope more often than the ultrasound-guided shortcut: in 60 patients the gastrogastrostomy-assisted route succeeded technically in 100% against 60.0% for enteroscopy, in roughly half the procedure time and with days less hospital stay — at a similar adverse-event rate.^[26] Pregnancy compresses fluoroscopy to seconds: across 17 ERCPs the mean exposure ran 14 seconds for an estimated fetal dose of 40 mrad, with one bleed clipped and one pancreatitis admitted three days — modified techniques safe, dosimetry routinely recorded.^[27]

The cohorts behind the numbers run Buxbaum ASGE with GRADE methods, Dumonceau ESGE twice over, Crinò world consensus across three continents, Miura Tokyo flowchart, Sirinawasatien 240-patient tandem cohort, Giljaca Cochrane with 18 studies and 2,366 participants, Meeralam five head-to-head studies, Jagtap 224-patient randomisation, Elmunzer 1,950 high-risk patients at 20 centres, Akshintala 55 trials with 17,062 patients, Guha 352-patient hydration trial, Cennamo six trials with 966 subjects, Tang seven trials with 999 difficult-access patients, Singh 11 trials with 1,513 patients, Chaouch 14 studies with 1,849 patients, Lyu 12 trials with 1,545 patients, Khan eight trials with 576 patients, Aziz 16 studies with 2,545 large-stone patients, Hu nine studies with 497 against 548 patients, Martins seven trials with 1,022 stented patients, Guzmán-Calderón seven studies with 1,070 stented patients, Song 86-patient preoperative randomisation, Prachayakul perforation review, Kalaitzakis 165-patient cholangioscopy series, Bukhari 60-patient bypass comparison, and Kahaleh 17 pregnancy ERCPs — randomised where randomisable, pooled where small, guideline where judgment is needed.^{[1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27]}

Pregnancy, bypass and the re-stenosed stent close the traps: the gravid stone former gets minimal-fluoroscopy ERCP with recorded dosimetry and her cholecystectomy timed obstetrically; the bypass patient gets the ultrasound-created shortcut, not the long enteroscope; the resectable cancer gets plastic, not metal; the difficult cannulation gets an early precut, not a tenth blind attempt; and the post-ERCP abdomen with pain out of proportion gets imaged for perforation now, since delay is what kills.^{[27][26][23][13][24]}

The intermediate-risk randomisation ran 224 patients at 49.6% prevalence with 92-98% sensitivity in both arms and few wrong calls.^[9] Stent the pancreas 12-24 hours when cutting conventionally.^[2] Lactated Ringer 3 mL/kg/h plus a 20 ml/kg bolus plus 8 hours was noninferior to indomethacin — volume where NSAIDs are out.^[1]

References27ShowHide

1. [1]Buxbaum JL, et al. ASGE guideline on the role of endoscopy in the evaluation and management of choledocholithiasis. Gastrointest Endosc, 2019.PMID 30979521
2. [2]Dumonceau JM, et al. Prophylaxis of post-ERCP pancreatitis: European Society of Gastrointestinal Endoscopy (ESGE) Guideline - updated June 2014. Endoscopy, 2014.PMID 25148137
3. [3]Dumonceau JM, et al. ERCP-related adverse events: European Society of Gastrointestinal Endoscopy (ESGE) Guideline. Endoscopy, 2020.PMID 31863440
4. [4]Crinò SF, et al. World Endoscopy Organization guidelines on endoscopic retrograde cholangiopancreatography biliary cannulation and sphincterotomy techniques. Dig Endosc, 2025.PMID 40518920
5. [5]Miura F, et al. Tokyo Guidelines 2018: initial management of acute biliary infection and flowchart for acute cholangitis. J Hepatobiliary Pancreat Sci, 2018.PMID 28941329
6. [6]Sirinawasatien A, et al. The use of endoscopic ultrasound in tandem with endoscopic retrograde cholangiopancreatography in the 2019 American Society for Gastrointestinal Endoscopy guideline for patients at high risk of choledocholithiasis can help to avoid diagnostic endoscopic retrograde cholangiopancreatography in individuals without ascending cholangitis. DEN Open, 2025.PMID 39845698
7. [7]Giljaca V, et al. Endoscopic ultrasound versus magnetic resonance cholangiopancreatography for common bile duct stones. Cochrane Database Syst Rev, 2015.PMID 25719224
8. [8]Meeralam Y, et al. Diagnostic accuracy of EUS compared with MRCP in detecting choledocholithiasis: a meta-analysis of diagnostic test accuracy in head-to-head studies. Gastrointest Endosc, 2017.PMID 28645544
9. [9]Jagtap N, et al. EUS versus MRCP to perform ERCP in patients with intermediate likelihood of choledocholithiasis: a randomised controlled trial. Gut, 2022.PMID 35144973
10. [10]Elmunzer BJ, et al. Indomethacin with or without prophylactic pancreatic stent placement to prevent pancreatitis after ERCP: a randomised non-inferiority trial. Lancet, 2024.PMID 38219767
11. [11]Akshintala VS, et al. Non-steroidal anti-inflammatory drugs, intravenous fluids, pancreatic stents, or their combinations for the prevention of post-endoscopic retrograde cholangiopancreatography pancreatitis: a systematic review and network meta-analysis. Lancet Gastroenterol Hepatol, 2021.PMID 34214449
12. [12]Guha P, et al. An Open-Label Randomized Controlled Trial Comparing Effectiveness of Aggressive Hydration Versus High-dose Rectal Indomethacin in the Prevention of Postendoscopic Retrograde Cholangiopancreatographic Pancreatitis (AHRI-PEP). J Clin Gastroenterol, 2023.PMID 35470300
13. [13]Cennamo V, et al. Can early precut implementation reduce endoscopic retrograde cholangiopancreatography-related complication risk? Meta-analysis of randomized controlled trials. Endoscopy, 2010.PMID 20306386
14. [14]Tang Z, et al. Early precut sphincterotomy does not increase the risk of adverse events for patients with difficult biliary access: A systematic review of randomized clinical trials with meta-analysis and trial sequential analysis. Medicine (Baltimore), 2018.PMID 30200135
15. [15]Singh AN, et al. Single-stage laparoscopic common bile duct exploration and cholecystectomy versus two-stage endoscopic stone extraction followed by laparoscopic cholecystectomy for patients with gallbladder stones with common bile duct stones: systematic review and meta-analysis of randomized trials with trial sequential analysis. Surg Endosc, 2018.PMID 29603004
16. [16]Chaouch MA, et al. Meta-analysis of randomized controlled trials comparing single-stage laparoscopic versus two-stage endoscopic management followed by laparoscopic cholecystectomy of preoperatively diagnosed common bile duct stones. Medicine (Baltimore), 2025.PMID 40101031
17. [17]Lyu Y, et al. Laparoscopic common bile duct exploration plus cholecystectomy versus endoscopic retrograde cholangiopancreatography plus laparoscopic cholecystectomy for cholecystocholedocholithiasis: a meta-analysis. Surg Endosc, 2019.PMID 30511313
18. [18]Khan QI, et al. Early versus delayed laparoscopic cholecystectomy after endoscopic retrograde cholangiopancreatography (ERCP) for choledocholithiasis: a systematic review and meta-analysis of randomised controlled trials. Surg Endosc, 2026.PMID 42414771
19. [19]Aziz M, et al. Endoscopic sphincterotomy vs papillary large balloon dilation vs combination modalities for large common bile duct stones: a network meta-analysis. Endosc Int Open, 2022.PMID 36531684
20. [20]Hu J, et al. Comparing the efficacy of endoscopic balloon dilation alone and combined with endoscopic sphincterotomy for common bile duct stone: a systematic review and meta-analysis. Ann Palliat Med, 2022.PMID 35144408
21. [21]Martins OC, et al. The addition of endoscopic sphincterotomy to biliary stent placement and its impact on the risk of adverse events: an updated meta-analysis of randomized controlled trials. Gastrointest Endosc, 2024.PMID 38782136
22. [22]Guzmán-Calderón GE, et al. Covered Versus Uncovered Metal Stents for the Drainage of the Malignant Distal Biliary Obstruction With ERCP: A Systematic Review and Meta-Analysis. J Clin Gastroenterol, 2025.PMID 38995024
23. [23]Song TJ, et al. Metal versus plastic stents for drainage of malignant biliary obstruction before primary surgical resection. Gastrointest Endosc, 2016.PMID 27109456
24. [24]Prachayakul V, et al. Endoscopic retrograde cholangiopancreatography-related perforation: Management and prevention. World J Clin Cases, 2014.PMID 25325062
25. [25]Kalaitzakis E, et al. Diagnostic and therapeutic utility of single-operator peroral cholangioscopy for indeterminate biliary lesions and bile duct stones. Eur J Gastroenterol Hepatol, 2012.PMID 22433791
26. [26]Bukhari M, et al. An international, multicenter, comparative trial of EUS-guided gastrogastrostomy-assisted ERCP versus enteroscopy-assisted ERCP in patients with Roux-en-Y gastric bypass anatomy. Gastrointest Endosc, 2018.PMID 29730228
27. [27]Kahaleh M, et al. Safety and efficacy of ERCP in pregnancy. Gastrointest Endosc, 2004.PMID 15278066