ERCP-INDUCED BILI-ENTERIC PERFORATIONS: GOOD OUTCOMES EVEN FOR THOSE REQUIRING SURGICAL INTERVENTION
Rogeh Habashi*1,2,3, Kevin Schuster1,2
1Surgery, Yale University - Surgery, New Haven, CT; 2Yale School of Public Health, New Haven, CT; 3McMaster University , Hamilton, ON, Canada
Background: Bili-enteric perforation (BEP) is a rare (0.09-1.67%) but potentially devastating complication of endoscopic retrograde cholangiopancreatography (ERCP). BEPs may be managed surgically (S), percutaneously (IR), endoscopically (EN) or without invasive intervention (NO). This study examines the BEP risk factors and outcome differences between the management strategies.
Methods: Using the National Inpatient Sample database (NIS), the records of 261, 072 patients undergoing ERCP in a 7-year period (2006-2012) were retrospectively analyzed. Univariate analyses were performed on demographics and interventions of ERCP using Mann-Whitney and Chi-Square Tests. Log-Transformed Generalized Linear Modelling, Multinomial Logistic Regression and Chi-Square Tests were used to investigate differences between the management groups in length of stay and in-hospital mortality.
Results: The cumulative incidence of BEP is 1.36% (n=3,546), predominantly enteric (77%, n=2,778). On univariate analyses, age above 60, Asian-decent, chronic anemia, liver disease and malnourishment were associated with increased risk of BEP (p<0.0001). Duodenal diverticulum, biliary and pancreatic stenting were also associated with increased risk of perforation (p<0.0001).
70.39% (2,496) of the BEP were managed without invasive intervention and 29.61% (n=1050) required surgery. Endoscopy comprised 6.03% (214) and percutaneous drainage represented 8.18% (290) with 1.07% (38) undergoing both endoscopy and percutaneous drainage. Endoscopy and percutaneous drainage failed in 92 and 136 patients, respectively, necessitating surgery. Using multinomial regression, enteric perforations were associated with higher odds of a surgical management relative to biliary perforations (OR = 4.69 [3.67-5.98], p<0.0001).
There were no differences in in-hospital mortality between the non-invasively, surgically and endoscopically managed groups. However relative to surgery, the percutaneous drainage group was associated with higher odds of in-hospital mortality (2.08 [1.32-3.28, p=0.0014). Although the surgery group was associated with longer length of stay than the non-invasively managed group (median=11 days, [IQR=11] vs. 7 days, [7], p<0.0001), there was no difference in LOS between the surgery and endoscopy groups (11 days, [11], vs. 10 days [12], p=0.23). Finally relative to surgery, the percutaneous drainage group was associated with longer LOS (15 days [14], p<0.0001).
Conclusion: Surgery for BEP is associated with (1) no increased in-hospital mortality relative to non-invasive management and (2) shorter length of stay relative to percutaneous drainage.
Back to 2020 Abstracts