Endoscopy 2007; 39(9): 793-801
DOI: 10.1055/s-2007-966723
Original article

© Georg Thieme Verlag KG Stuttgart · New York

Risk factors for complication following ERCP; results of a large-scale, prospective multicenter study

E.  J.  Williams1 , S.  Taylor2 , P.  Fairclough3 , A.  Hamlyn4 , R.  F.  Logan5 , D.  Martin6, 7 , S.  A.  Riley8 , P.  Veitch9, 10 , M.  L.  Wilkinson11 , P.  R.  Williamson2 , M.  Lombard1
  • 1Department of Gastroenterology, Royal Liverpool University Hospital, Liverpool, UK
  • 2Centre for Medical Statistics and Health Evaluation, School of Health Sciences, University of Liverpool, Liverpool, UK
  • 3Department of Gastroenterology, St Bartholomew’s Hospital, London, UK
  • 4Department of Gastroenterology, Russell’s Hall Hospital, Dudley, UK
  • 5Division of Public Health and Epidemiology, Queens Medical Centre, Nottingham, UK
  • 6Department of Radiology, Wythenshawe Hospital, Manchester, UK
  • 7Royal College of Radiologists
  • 8Department of Gastroenterology, Northern General Hospital, Sheffield, UK
  • 9Department of Surgery, Royal Free Hospital, London, UK
  • 10Association of Upper GI Surgeons
  • 11Department of Gastroenterology, St Thomas’ Hospital, London, UK
Further Information

Publication History

submitted 30 November 2006

accepted after revision 11 June 2007

Publication Date:
17 August 2007 (online)

Background and study aims: Analyses of endoscopic retrograde cholangiopancreatography (ERCP) complication are often constrained by the number of endpoints observed. This large-scale study aimed to identify the principal risk factors for ERCP complication.

Patients and methods: This was a prospective multicenter study of ERCP complications, based in five English regions. An exploratory univariable analysis of patients’ first recorded procedures identified potentially important patient- and procedure-related factors. For overall complications and pancreatitis, variables significant in univariable analysis were included in multiple regression.

Results: A total of 66 centers collected data on 5264 ERCPs, performed on 4561 patients. A therapeutic intervention was attempted in 3447/4561 (76 %) of patients as part of their first recorded ERCP. Following first recorded ERCP, 230 patients (5.0 %) suffered ≥ 1 complication: pancreatitis in 74 (1.6 %), cholangitis in 48 (1.0 %), hemorrhage in 40 (0.9 %), perforation in 20 (0.4 %), and miscellaneous in 54 (1.2 %). Significant factors from multiple regression were included in a multi-level analysis, which incorporated variables measured at the level of the endoscopist and hospital. For overall complication, risk factors (P value, odds ratio [OR], 95 % confidence interval [CI]) were: cannulation attempts > 1 (P = 0.094, OR 1.32, 95 % CI 0.95 - 1.83), precut (P = 0.033, OR 1.55, 95 % CI 1.04 - 2.32), and suspected sphincter of Oddi dysfunction (P = 0.121, OR 1.97, 95 % CI 0.84 - 4.64). For pancreatitis, risk factors (Pvalue, OR, and 95 % CI) were: cannulation attempts > 1 (P = 0.0001, OR 3.14, 95 % CI 1.74 - 5.67), female sex (P < 0.001, OR 2.22, 95 % CI 1.43 - 3.45), age (P < 0.002, OR 1.09 per 5 year decrease, 95 % CI 1.03 - 1.15), and performance in a district (as opposed to university) hospital (P = 0.034, OR 2.41, 95 % CI 1.08 - 5.41).

Conclusion: Careful patient selection combined with skilled cannulation minimizes complications. Higher-risk procedures should be performed in specialist centers.

References

M. Lombard, MD

Department of Gastroenterology, Royal Liverpool University Hospital

Prescot Street

Liverpool, L7 8XP

UK

Fax: +44-151-7065832

Email: martin.lombard@rlbuht.nhs.uk