Impact of atrial fibrillation on in-hospital outcomes following endoscopic retrograde cholangiopancreatography: a propensity score-matched analysis of the National Inpatient Sample (2016-2020)

Abdulrahim Y. Mehadia, Bekure B. Sirawb, Parth Patelb, Eli A. Zaherb, Ebrahim A. Mohamedb, Shahin Ishab, Abel Tenaw Tasammac, Yordanos T. Tafesseb, Yonas Gebrecherkosb, Juveriya Yasmeenb, Mouaz Oudihb, Mohammed Harounb

John H. Stroger Hospital of Cook Country, Chicago, IL, USA; Ascension Saint Joseph Hospital, Chicago, IL, USA; Addis Ababa University, Ethiopia

aDepartment of Internal Medicine, John H. Stroger Hospital of Cook County, Chicago, IL, USA (Abdulrahim Y. Mehadi); bDepartment of Internal Medicine, Ascension Saint Joseph Hospital, Chicago, IL, USA (Bekure B. Siraw, Parth Patel, Eli A. Zaher, Ebrahim A. Mohamed, Shahin Isha, Yordanos T. Tafesse, Yonas Gebrecherkos, Juveriya Yasmeen, Mouaz Oudih, Mohammed Haroun); cDepartment of Internal Medicine, Addis Ababa University, Addis Ababa, Ethiopia (Abel Tenaw Tasamma)

Correspondence to: Bekure B. Siraw, MD, MPH, Department of Internal Medicine, Ascension Saint Joseph Hospital, 2900 N Lakeshore Dr., Chicago, IL, 60657, USA, e-mail: bekuresiraw.eras@gmail.com
Received 12 December 2024; accepted 1 April 2025; published online 25 April 2025
DOI: https://doi.org/10.20524/aog.2025.0964
© 2025 Hellenic Society of Gastroenterology

Abstract

Background Endoscopic retrograde cholangiopancreatography (ERCP) is a critical tool in managing hepatobiliary and pancreatic diseases. Atrial fibrillation (AF) has been associated with greater morbidity in patients undergoing ERCP. This study compared in-hospital ERCP outcomes in patients with and without AF.

Methods This retrospective cohort study utilized data from the National Inpatient Sample (2016-2020). Patients who underwent ERCP during hospitalization were included. Patients with AF were matched 1:1 to those without AF, based on demographic and clinical variables. The primary outcome was all-cause in-hospital mortality. Secondary outcomes included procedure-related and non-procedure-related complications, hospitalization cost and length of stay.

Results The final matched sample consisted of 29,942 patients, with 14,971 in each group (AF and non-AF). Patients with AF demonstrated significantly higher in-hospital mortality compared to those without AF (3.6% vs. 1.9%; odds ratio [OR] 1.87, 95% confidence interval [CI] 1.62-2.17). The AF group had a significantly longer median length of stay (8.1 vs. 6.4 days; β 1.7; 95%CI 1.5-1.8) and incurred higher hospitalization costs ($111,000 vs. $87,255; β $23,745; 95%CI $20,783-26,708). In terms of complications, patients with AF had significantly higher rates of acute kidney injury (OR 1.33, 95%CI 1.27-1.40) and sepsis (OR 1.38, 95%CI 1.30-1.48). However, the rates of procedure-specific complications, including biliary perforation, post-ERCP pancreatitis and post-ERCP cholangitis, were similar between the 2 groups.

Conclusion Patients with AF undergoing ERCP have higher in-hospital mortality, longer stays, greater costs, and higher rates of acute kidney injury and sepsis, although procedure-specific complication rates remain unaffected.

Keywords Endoscopic retrograde cholangiopancreatography, atrial fibrillation, mortality, in-hospital outcome, national inpatient sample

Ann Gastroenterol 2025; 38 (3): 345-352

Introduction

Endoscopic retrograde cholangiopancreatography (ERCP) is an essential tool in managing hepatobiliary and pancreatic pathology. Since its development in the 1960s, ERCP has evolved from a purely diagnostic procedure into a primarily therapeutic intervention [1]. Today, it is considered the gold standard for managing these conditions, offering a high success rate with lower morbidity and mortality compared to surgical alternatives [2,3]. However, ERCP remains associated with significant procedural risks, including post-ERCP pancreatitis, bleeding and infections, making it one of the higher-risk endoscopic procedures [4,5]. Recent studies continue to highlight its complication rates, underscoring the need for careful patient selection and risk mitigation strategies [6,7].

The importance of understanding and mitigating the risks associated with ERCP cannot be overstated, especially given the complex nature of the conditions it treats. Preceding studies have highlighted various factors that can influence ERCP outcomes [8]. Comorbidities such as atrial fibrillation (AF) have been shown to increase adverse outcomes in patients undergoing ERCP, as presented in an analysis of the National Inpatient Sample (NIS) in 2016 [9]. Given the significant implications of AF for in-hospital patient outcomes, it is crucial to examine its impact on those undergoing ERCP.

To date, comprehensive studies employing robust methodologies have not been conducted to compare the in-hospital outcomes of ERCP in patients with and without AF. To address this gap, our study leveraged propensity score matching to minimize confounding variables and gain a more precise understanding of the risks associated with AF in this context. This study sought to elucidate the varying impact of AF on key outcomes, including all-cause in-hospital mortality, procedure-related and non-procedure-related complications, length of hospital stay, and total hospitalization costs.

Materials and methods

Data source

Data for this retrospective cohort study were collected from the NIS database provided by the Healthcare Cost and Utilization Project (HCUP) from January 1, 2016, to December 31, 2020. NIS is the largest, all-payer, inpatient care database publicly available in the United States and represents approximately 20% of all discharges from United States (US) hospitals [10]. It stores de-identified information about individual hospitalizations, including demographic characteristics, diagnoses and procedures, in the form of International Classification of Diseases (ICD)-10 codes [11]. The internal validity of the database is maintained by annual quality assessments, and its external validity is assessed and ensured by comparison with the National Hospital Discharge Survey from the National Center for Health Statistics, the American Hospital Association Annual Survey Database, and the Medicare Provider Analysis and Review inpatient data from the Centers for Medicare and Medicaid Services [12].

Study population

Using the NIS data, patients 18 years and above who had undergone ERCP during their hospital stay between January 1st, 2016, and December 31st, 2020, were included in this study. The identification of ERCP was made based on ICD-10 PCS codes (Supplementary Table 1). ICD-10 codes were used to identify patients with atrial fibrillation (Supplementary Table 1). Discharge weights in the NIS were used to compute national estimates. Employing 1:1 propensity score matching, 14,971 patients with atrial fibrillation were matched with an equivalent number of patients without atrial fibrillation. The matching variables were age, sex, race, region, location and teaching status of the hospital, payer, Charlson comorbidity index, body mass index (BMI), immunosuppressive medications (chemotherapy or steroids), anticoagulant and antiplatelet use before admission, pacemaker and implantable defibrillator cardioverter implant history before admission, the presence or absence of hyperlipidemia, diabetes or hypertension, alcohol or nicotine dependence history, as well as heart failure, malignancy, chronic kidney disease (CKD), chronic liver disease (CLD), chronic obstructive pulmonary disease (COPD), asthma, pulmonary arterial hypertension (PAH), indication for ERCP (acute cholangitis, acute gallstone pancreatitis, choledocholithiasis, pancreaticobiliary mass, obstructive jaundice with no clear cause, and miscellaneous), and procedure performed during the ERCP (biliary and pancreatic stenting).

Study objective

The primary objective of this study was to elucidate the impact of atrial fibrillation on in-hospital outcomes among patients undergoing ERCP during their hospitalization. The primary outcome assessed was all-cause in-hospital mortality during the index hospitalization. Secondary outcomes encompassed the length of stay, the total cost of hospitalization, and procedure-related and non-procedure-related complications during the same hospitalization.

Study variables

Demographic variables such as age, sex and race/ethnicity were directly extracted from the NIS data. Age was categorized into 4 ranges: 18-44, 45-64, 65-84 and 85+ years. BMI categories were defined using ICD-10-CM diagnostic codes for overweight (BMI 25-29.9 kg/m2), class I and II obesity (BMI 30-39.9 kg/m2), and Class III obesity (BMI >40 kg/m2). Patients who did not fall into these 3 categories were classified as having a BMI <25 kg/m2. AF was identified using ICD-10-CM diagnosis codes (Supplementary Table 1). The presence or absence of hyperlipidemia, diabetes, hypertension, alcohol dependence history, nicotine dependence history, heart failure, malignancy, CKD, COPD, CLD, asthma and PAH was determined using corresponding ICD-10-CM diagnosis codes (Supplementary Table 1). The indication for the ERCP was evaluated based on the presence of 1 or more of the following conditions: acute cholangitis, acute gallstone pancreatitis, choledocholithiasis, pancreaticobiliary mass, obstructive jaundice with no clear cause; all these were coded using the corresponding ICD-10 CM codes. When none of these conditions were present during the index hospitalization, the indication for the ERCP was labeled as miscellaneous. Pancreatic and biliary stenting during the ERCP were coded using corresponding ICD-10 PCS codes. Antiplatelet or anticoagulant use, and pacemaker and implantable cardioverter-defibrillator implants before admission were also coded using corresponding ICD 10 codes. A literature review guided the identification of potential procedure-related and non-procedure-related complications following ERCP, and new variables were created based on the corresponding ICD-10-CM diagnostic codes (Supplementary Table 1). For acute post-ERCP pancreatitis, since no specific ICD-10 code was available, a previously well-validated combination of existing ICD-10 codes was used [13-15]. Inpatient mortality and length of stay data were extracted directly from the NIS dataset.

Statistical analysis

In light of potential dissimilarities between patients with atrial fibrillation and those without, we implemented a propensity score matching technique to address potential confounding factors [16]. The propensity score was computed using a non-parsimonious logistic regression model incorporating the relevant matching variables reported above. Employing a 1:1 matching scheme without replacement, nearest-neighbor matching, and a caliper width of 0.001 on the probability scale, we sought to ensure the robustness of the matching process. The standardized mean differences (SMD) in the sample population pre- and post-matching are shown in Table 1. Instances of hospitalizations that fell beyond the specified caliper width were excluded to minimize bias, as their comorbidities were disproportionate to the remainder of the cohort. Continuous variables were succinctly summarized using means and standard deviations, with inter-group comparisons conducted through t-tests. Categorical variables, on the other hand, were summarized using proportions and subjected to inter-group comparisons utilizing the chi-square test. The odds ratios (OR) of atrial fibrillation for each procedure-related and non-procedure-related complication were assessed using the Cochrane-Mantel-Haenszel test. Simple linear regression models were used to evaluate differences in length of stay and cost of hospitalization. The output of all tests was reported with a 95% confidence interval at a 0.05 level of significance. All data cleaning and analysis were performed using R statistical software version 4.1.3 (2023) [17]. Propensity score matching was performed using the “MatchIt” package in R [18].

Table 1 Baseline patient characteristics in the study population

thumblarge

Results

A total of 29,942 patients underwent ERCP from 2016-2020; of these, 14,971 patients had AF, and they were matched with the same number of patients without AF. The mean age was 76.8±11 years and 54% of patients were male. The majority of patients were White (80%). There was no significant difference between the groups as regards the baseline characteristics, including comorbidities, Charlson comorbidity index, indications for ERCP, hospital setting where ERCP was performed, and long-term anticoagulation or antiplatelet use.

The most common indication for ERCP was choledocholithiasis (50%), followed by pancreaticobiliary mass (15.7%), acute gallstone pancreatitis (14.8%), obstructive jaundice (11%), acute cholangitis (5%), and other (3%). Biliary stents were placed in 47% of the patients, while pancreatic duct stents were placed in 6.6% (Table 1). The most common procedure-related complication was gastrointestinal/hepatobiliary bleeding (1.2%), followed by post-ERCP pancreatitis (0.6%), biliary perforation (0.5%) and post-ERCP cholangitis (0.2%) (Table 2).

Table 2 Frequency of in-hospital complications and outcomes

thumblarge

Outcomes

A total of 827 patients (2.8%) died during their hospital stay (1.9% in the non-AF group vs. 3.6% in the AF group). Mortality was significantly higher among the AF cohort in comparison to the non-AF cohort (OR 1.87, 95% confidence interval [CI] 1.62-2.17). The overall cohort’s median length of stay was 7.3 days. Mean hospital stay was slightly longer in the AF group (8.1 vs. 6.4 days; β 1.7; 95%CI 1.5-1.8). The median cost of hospitalization in the entire cohort was $99,134. The AF group had a higher cost of hospitalization ($111,000 vs. $87,255; β $23,745; 95%CI $20,783-26,708).

There was a greater risk of periprocedural gastrointestinal or hepatobiliary bleeding in the AF group (1.3% vs. 1%; OR 1.33, 95%CI 1.08-1.65; P=0.008). The incidence of acute kidney injury (OR 1.33, 95%CI 1.27-1.40; P<0.001) and sepsis (OR 1.38, 95%CI 1.3-1.48; P<0.001) was also significantly higher in the AF group. In addition, patients in the AF group received more blood transfusions than the non-AF group (OR 1.35, 95%CI 1.21-1.5; P<0.001). Notably, the 2 groups had no significant discrepancies in the occurrence rates of biliary perforation, post-ERCP pancreatitis, or post-ERCP cholangitis. Moreover, the incidence of deep vein thrombosis (DVT), pulmonary embolism (PE), and periprocedural cardiac arrest was comparable across both groups (Table 3).

Table 3 Odds ratio for in-hospital complications in the AF group compared to the non-AF group

thumblarge

Discussion

To the best of our knowledge, this is the first propensity score-matched study to compare the outcome of patients with and without AF undergoing ERCP. According to our study, those with AF had longer mean hospital stays, a higher incidence of in-hospital mortality, AKI, sepsis, periprocedural gastrointestinal and hepatobiliary bleeding, and need for blood transfusion, whereas rates of various periprocedural complications, such as cardiac arrest, pancreatitis, biliary perforation, cholangitis and DVT, were similar between the 2 groups.

AF has been established as a risk factor for greater in-hospital mortality in various patient groups, including patients with known cardiovascular disease, critically ill Intensive Care Unit patients, elderly patients, patients with pulmonary hypertension, and hospitalized COVID-19 patients [19-22]. However, whether AF affects inpatient outcomes in patients undergoing ERCP is largely unexplored. The overall in-hospital mortality following ERCP ranges from 0.1-3.3% in various studies, depending on the study population, procedural indications and patient comorbidities [23-25]. The 2.8% overall mortality rate observed in our study falls within this range, but is on the higher end compared to some earlier reports. One major factor contributing to this discrepancy is the inclusion of an older patient population with significant comorbidities, as our study specifically focused on patients with and without AF, a population inherently at higher risk for adverse outcomes. Studies reporting lower mortality rates often include a broader ERCP population, including elective outpatient cases with lower acuity, which may underestimate the true mortality risk in hospitalized patients undergoing the procedure. Notably, the higher mortality rate in patients with AF observed in our study aligns with the findings of Sharma et al, who analyzed NIS data from 2016 and found a greater risk of adverse outcomes in AF patients undergoing ERCP [9]. Our study expands on these findings by utilizing a multi-year dataset and propensity score matching to provide a more robust comparison. The similar incidence of periprocedural cardiac arrest between groups suggests that immediate procedural complications are unlikely to be the primary driver of higher mortality in the AF cohort. Instead, the significantly higher rates of sepsis and acute kidney injury in the AF group suggest that late postprocedural complications play a crucial role in driving mortality differences.

With regard to the utilization of hospital resources, patients with AF had a longer hospital stay and higher costs of hospitalization, a finding that is in line with the study by Sharma et al [9]. Both length of stay and hospital charges are directly related to the presence and severity of in-hospital complications, as more time and resources are needed to manage them. The higher incidence of several inpatient complications observed in the AF cohort could partially explain the longer length of stay and increased hospital charges observed in the same group. In addition, the level of anticoagulation has to be monitored and necessary adjustments made in patients with AF, further lengthening their hospital stay.

While the overall incidence of periprocedural pancreatitis was remarkably lower than reports from prior studies, the rate of periprocedural bleeding was similar to previous reports [26-31]. Most previously identified risk factors for post-ERCP pancreatitis are procedure-related factors, such as difficult and repeated cannulation of the main pancreatic duct, multiple contrast material injections into the main pancreatic duct, pancreatic sphincterotomy, endoscopic snare papillectomy, and balloon dilation of an intact biliary sphincter [32-36]. The presence of these factors is largely influenced by the indication for the procedure, and it is not expected to be affected by the concomitant presence of AF. Other patient-related factors, such as age and sex, were also similar between the 2 groups following propensity score matching, mitigating their increased risk of post-ERCP pancreatitis observed in prior studies [37-39]. A statistically significant higher incidence of periprocedural gastrointestinal and hepatobiliary bleeding in the AF group is an expected finding, as these patients tend to be on chronic anticoagulation before undergoing ERCP.

Concerning the thrombotic events, the risk of developing PE was significantly lower in patients with AF. Similarly, although it did not reach statistical significance, there was a trend toward a lesser risk of DVT in this group. This is despite patients with AF having a longer mean hospital stay (8.2 vs. 6.1 days), a well-established risk factor for the development of DVT and PE. A possible explanation for this finding is the fact that after undergoing ERCP, AF patients would be put on therapeutic anticoagulation as soon as feasible, while those without AF would be kept on prophylactic anticoagulation for the duration of their hospitalization.

Our study had several strengths, including a large sample size, which provides a robust and comprehensive overview of in-hospital outcomes across a wide variety of US hospital institutions. Moreover, the use of propensity matching mitigates confounding variables and allows for a more balanced comparison between groups. However, several limitations should be acknowledged. First, the retrospective design limits inference, and causation cannot be established. Additionally, while propensity score matching helps reduce confounding, residual confounding may persist despite our best efforts to account for relevant factors. Second, the reliance on administrative data coded via ICD-10 introduces the potential for misclassification bias and limits the granularity of certain variables, particularly when assessing complications and comorbidities. Furthermore, the ICD-10 codes used to identify atrial fibrillation and procedural complications, such as biliary perforation, post-ERCP pancreatitis, and gastrointestinal or hepatobiliary bleeding, may not fully capture the clinical complexity or severity of these conditions. Additionally, administrative databases like the National Inpatient Sample do not capture data regarding the cause of death or prolonged hospital stay, limiting the interpretation of some of our findings. These limitations should be considered when interpreting our findings, highlighting the need for prospective studies with more precise data collection to validate and expand upon these results.

In summary, our study highlights that AF significantly impacts hospital outcomes in patients undergoing ERCP. Greater mortality, longer hospital stays, and higher complication rates were seen in patients with AF. Vigilant perioperative monitoring and careful management of anticoagulation are crucial in mitigating these risks. Future prospective studies should attempt to confirm the associations observed, and to explore the underlying mechanisms contributing to the increased risk of in-hospital mortality and complications in patients with AF undergoing ERCP. Moreover, studies should investigate the optimal management strategies for anticoagulation in these patients, attempting to balance the risk of bleeding and thrombotic events.

Summary Box

What is already known:

  • Endoscopic retrograde cholangiopancreatography (ERCP) is the gold standard for treating hepatobiliary and pancreatic conditions but has high morbidity

  • Atrial fibrillation (AF) is linked to worse outcomes, including in-hospital mortality and complications

  • The impact of AF on ERCP outcomes remains unclear

  • Periprocedural complications in ERCP are influenced by multiple factors, but the role of AF has not been well studied

What the new findings are:

  • AF patients undergoing ERCP had higher in-hospital mortality (3.6% vs. 1.9%) and longer stays (8.1 vs. 6.4 days)

  • AF was associated with higher hospitalization costs ($111,000 vs. $87,255) and more complications, including bleeding and acute kidney injury

  • No significant differences were observed in post-ERCP pancreatitis, biliary perforation, or cardiac events

References

1. Kröner PT, Bilal M, Samuel R, et al. Use of ERCP in the United States over the past decade. Endosc Int Open 2020;8:E761-E769.

2. Targarona EM, Bendahan GE. Management of common bile duct stones:controversies and future perspectives. HPB (Oxford) 2004;6:140-143.

3. Ramchandani M, Pal P, Reddy DN. Endoscopic management of acute cholangitis as a result of common bile duct stones. Dig Endosc 2017;29(Suppl 2):78-87.

4. Tarikci Kilic E, Kahraman R, Ozdil K. Evaluation of safety and outcomes of endoscopic retrograde cholangiopancreatography in 1337 patients at a single center. Medeni Med J 2019;34:290-296.

5. Disario JA. Hospital volume and ERCP outcomes:the writing is on the wall. Gastrointest Endosc 2006;64:348-350.

6. Guda NM, Reddy DN, Kumar A. Complications of ERCP. Indian J Gastroenterol 2014;33:1-9.

7. Talukdar R. Complications of ERCP. Best Pract Res Clin Gastroenterol 2016;30:793-805.

8. Katsinelos P, Lazaraki G, Chatzimavroudis G, et al. Risk factors for therapeutic ERCP-related complications:an analysis of 2,715 cases performed by a single endoscopist. Ann Gastroenterol 2014;27:65-72.

9. Sharma S, Acharya A, Aziz M, et al. S3237:Presence of atrial fibrillation adversely affects outcomes in patients undergoing ERCP:analysis of nationwide inpatient sample. Am J Gastroenterol 2020;115:S1694.

10. HCUP-US NIS Overview. Available from:https://hcup-us.ahrq.gov/nisoverview.jsp [Accessed 12 April 2025].

11. Steiner C, Elixhauser A, Schnaier J. The healthcare cost and utilization project:an overview. Eff Clin Pract 2002;5:143-151.

12. NIS related reports. Available from:https://hcup-us.ahrq.gov/db/nation/nis/nisrelatedreports.jsp [Accessed 12 April 2025].

13. Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res 2011;46:399-424.

14. R:The R Project for Statistical Computing. Available from:https://www.r-project.org/[Accessed 12 April 2025].

15. Ho D, Imai K, King G, Stuart EA. MatchIt:nonparametric preprocessing for parametric causal inference. Journal of Statistical Software 2011;42:1-28.

16. Abbas A, Sethi S, Vidyarthi G, Taunk P. Predictors of postendoscopic retrograde cholangiopancreatography pancreatitis, analysis of more than half a million procedures performed nationwide over the last 15 years. JGH Open 2020;4:736-742.

17. Schulman AR, Abougergi MS, Thompson CC. Assessment of the July effect in post-endoscopic retrograde cholangiopancreatography pancreatitis:Nationwide Inpatient Sample. World J Gastrointest Endosc 2017;9:296-303.

18. Inamdar S, Berzin TM, Sejpal DV, et al. Pregnancy is a risk factor for pancreatitis after endoscopic retrograde cholangiopancreatography in a national cohort study. Clin Gastroenterol Hepatol 2016;14:107-114.

19. Wolf PA, Mitchell JB, Baker CS, Kannel WB, D'Agostino RB. Impact of atrial fibrillation on mortality, stroke, and medical costs. Arch Intern Med 1998;158:229-234.

20.  Terwangne C, Lelubre C, Hanotier P, et al. Prevalence and impact of atrial fibrillation on intra-hospital mortality in patients aged ≥75 years. Am J Cardiol 2022;177:40-47.

21. Rubens M, Saxena A, Barclay GJ, et al. Atrial fibrillation- a risk factor for in-hospital mortality of patient with pulmonary hypertension. J Heart Lung Transplant 2022;41:S35.

22. Spruit JR, Jansen RWMM, de Groot JR, et al. Does atrial fibrillation affect prognosis in hospitalised COVID-19 patients?A multicentre historical cohort study in the Netherlands. BMJ Open 2023;13:071137.

23. Siiki A, Tamminen A, Tomminen T, Kuusanmäki P. ERCP procedures in a Finnish community hospital:a retrospective analysis of 1207 cases. Scand J Surg 2012;101:45-50.

24. Kapral C, Mühlberger A, Wewalka F, Duller C, Knoflach P, Schreiber F;Working Groups Quality Assurance and Endoscopy of Austrian Society of Gastroenterology and Hepatology (OeGGH). Quality assessment of endoscopic retrograde cholangiopancreatography:results of a running nationwide Austrian benchmarking project after 5 years of implementation. Eur J Gastroenterol Hepatol 2012;24:1447-1454.

25. Glomsaker T, Hoff G, Kvaløy JT, Søreide K, Aabakken L, Søreide JA;Norwegian Gastronet ERCP Group. Patterns and predictive factors of complications after endoscopic retrograde cholangiopancreatography. Br J Surg 2013;100:373-380.

26. Andriulli A, Loperfido S, Napolitano G, et al. Incidence rates of post-ERCP complications:a systematic survey of prospective studies. Am J Gastroenterol 2007;102:1781-1788.

27. Williams EJ, Taylor S, Fairclough P, et al. Risk factors for complication following ERCP;results of a large-scale, prospective multicenter study. Endoscopy 2007;39:793-801.

28. Suissa A, Yassin K, Lavy A, et al. Outcome and early complications of ERCP:a prospective single-center study. Hepatogastroenterology 2005;52:352-355.

29. Watanabe H, Yoneda M, Tominaga K, et al. Comparison between endoscopic papillary balloon dilatation and endoscopic sphincterotomy for the treatment of common bile duct stones. J Gastroenterol 2007;42:56-62.

30. Nakaji S, Hirata N, Matsui H, et al. Hemodialysis is a strong risk factor for post-endoscopic sphincterotomy bleeding in patients with choledocholithiasis. Endosc Int Open 2018;6:E568-E574.

31. Kochar B, Akshintala VS, Afghani E, et al. Incidence, severity, and mortality of post-ERCP pancreatitis:a systematic review by using randomized, controlled trials. Gastrointest Endosc 2015;81:143-149.

32. Parekh PJ, Majithia R, Sikka SK, Baron TH. The “scope“of post-ERCP pancreatitis. Mayo Clin Proc 2017;92:434-448.

33. Dumonceau JM, Kapral C, Aabakken L, et al. ERCP-related adverse events:European Society of Gastrointestinal Endoscopy (ESGE) Guideline. Endoscopy 2020;52:127-149.

34. Chen JJ, Wang XM, Liu XQ, et al. Risk factors for post-ERCP pancreatitis:a systematic review of clinical trials with a large sample size in the past 10 years. Eur J Med Res 2014;19:26.

35. Mariani A, Giussani A, Di Leo M, Testoni S, Testoni PA. Guidewire biliary cannulation does not reduce post-ERCP pancreatitis compared with the contrast injection technique in low-risk and high-risk patients. Gastrointest Endosc 2012;75:339-346.

36. Nakai Y, Isayama H, Sasahira N, et al. Risk factors for post-ERCP pancreatitis in wire-guided cannulation for therapeutic biliary ERCP. Gastrointest Endosc 2015;81:119-126.

37. Maitin-Casalis N, Neeman T, Thomson A. Protective effect of advanced age on post-ERCP pancreatitis and unplanned hospitalisation. Intern Med J 2015;45:1020-1025.

38. Finkelmeier F, Tal A, Ajouaou M, et al. ERCP in elderly patients:increased risk of sedation adverse events but low frequency of post-ERCP pancreatitis. Gastrointest Endosc 2015;82:1051-1059.

39. Matsubayashi H, Fukutomi A, Kanemoto H, et al. Risk of pancreatitis after endoscopic retrograde cholangiopancreatography and endoscopic biliary drainage. HPB (Oxford) 2009;11:222-228.

Notes

Conflict of Interest: None