Endoscopy 2019; 51(04): S203-S204
DOI: 10.1055/s-0039-1681775
ESGE Days 2019 ePosters
Friday, April 5, 2019 09:00 – 17:00: Clinical Endoscopic Practice ePosters
Georg Thieme Verlag KG Stuttgart · New York

PREDICTIVE MODEL TO DETERMINE THE NEED OF REPEATING ERCP AFTER ENDOSCOPIC TREATMENT OF BILIARY LEAKS

JC Silva
1   Gastroenterology, Centro Hospitalar de Vila Nova de Gaia e Espinho, Vila Nova de Gaia, Portugal
,
J Rodrigues
1   Gastroenterology, Centro Hospitalar de Vila Nova de Gaia e Espinho, Vila Nova de Gaia, Portugal
,
R Pinho
1   Gastroenterology, Centro Hospitalar de Vila Nova de Gaia e Espinho, Vila Nova de Gaia, Portugal
,
L Proença
1   Gastroenterology, Centro Hospitalar de Vila Nova de Gaia e Espinho, Vila Nova de Gaia, Portugal
,
C Fernandes
1   Gastroenterology, Centro Hospitalar de Vila Nova de Gaia e Espinho, Vila Nova de Gaia, Portugal
,
M Sousa
1   Gastroenterology, Centro Hospitalar de Vila Nova de Gaia e Espinho, Vila Nova de Gaia, Portugal
,
AC Gomes
1   Gastroenterology, Centro Hospitalar de Vila Nova de Gaia e Espinho, Vila Nova de Gaia, Portugal
,
J Carvalho
1   Gastroenterology, Centro Hospitalar de Vila Nova de Gaia e Espinho, Vila Nova de Gaia, Portugal
› Author Affiliations
Further Information

Publication History

Publication Date:
18 March 2019 (online)

 

Aims:

Endoscopic retrograde cholangiopancreatography (ERCP) is the first-line procedure to iatrogenic biliary leaks approach. In patients who have undergone biliary stenting, the timing and optimal method of stent removal (ERCP/esophagogastroduodenoscopy) is controversial. The present study aimed to evaluate ERCP efficacy in biliary leaks treatment and identify patients in whom repetition of ERCP may be unnecessary.

Methods:

Retrospective, unicentric analysis of patients who underwent ERCP with sphincterotomy and biliary stent placement between 2008 – 2017 due to iatrogenic biliary leaks. All patients were repeated ERCP with removal of the stent(s). Factors associated with the outcome, resolution of the biliary leak and absence of another pathology in reevaluation ERCP, were identified.

Results:

43 patients were included, 62.8% (n = 27) female, mean age 58.2 ± 17.2years. Most common etiology of biliary leaks was laparoscopic cholecystectomy (56.8%) and the most common location the cystic duct stump (53.5%). Technical success was 93.3%, with resolution of the biliary leak in 92.9%. On multivariate analysis, elective iatrogenic procedure (OR = 209.1, 95% CI 2.18 – 2050.8), normal total bilirubin (OR = 138.9, 95% CI 1.19 – 1627.2), ERCP performed in ≤7 days (OR = 32.9, 95% CI 1.08 – 1004.8) and removal of the stent in ≤12 weeks (OR = 40.7, 95% CI 1.11 – 1634.9) were independently associated with resolution of the biliary leak and absence of another pathology (p < 0.05;r2= 0.71). The area under the ROC curve of these criteria for outcome prediction was 0.93 (p < 0.001). When ≥3 criteria were present (42.9%), the model presented specificity of 100%, sensitivity of 58.1%, positive predictive value of 100% and negative predictive value of 45.8% in outcome prediction.

Conclusions:

We identified criteria that allow selection of 43% of patients in whom repetition of ERCP may be unnecessary. Biliary stents can be removed by esophagogastroduodenoscopy, increasing safety and efficiency of healthcare resources utilization.