Thirty-day hospital admission following high-risk outpatient ERCP: incidence and analysis of risk factors based on a secondary analysis of the Stent Versus Indomethacin trial data set.

Objective: ERCP carries potential for serious adverse events, and there is risk for unplanned health encounters after the procedure. The primary objective of this study was to identify the rate of 30-day hospital admission after outpatient ERCP in patients at elevated risk for post-ERCP pancreatitis based on patient- and procedure-related risk factors. We also explored whether pancreatitis rates and lengths of hospital stay differ with delayed hospitalization (versus admission immediately after ERCP) and identified factors associated with 30-day hospital admission.

Methods: High-risk patients in the Stent Versus Indomethacin trial who underwent outpatient ERCP were identified. The primary outcome was hospital admission within 30 days after outpatient ERCP. Patient and procedural factors associated with 30-day hospital admission were explored by means of univariable analysis followed by multivariable analysis, controlling for potential known confounders.

Results: Among 768 patients at high risk for post-ERCP pancreatitis who underwent outpatient ERCP and were discharged home, 21.9% required hospitalization within 30 days. Compared with patients who underwent outpatient ERCP but were admitted immediately to the hospital after the procedure, patients who were discharged after ERCP and admitted later had a higher incidence of post-ERCP pancreatitis (40.5% vs 30.9%; P = .04), moderate to severe post-ERCP pancreatitis (24.4% vs 13.8%; P < .01), and longer length of hospital stay (6.2 d vs 3.7 d; P < .001). Factors associated with 30-day hospitalization included patient age (odds ratio [OR], 0.98; 95% CI, 0.97-0.99), biliary stent placement (OR, 1.62; 95% CI, 1.05-2.48), and attempt at pancreatic duct cannulation (OR, 1.64; 95% CI, 1.02-2.64).

Conclusions: High-risk patients who undergo outpatient ERCP and are discharged home are at considerable risk for hospital admission within 30 days and have an increased rate of pancreatitis and longer length of hospital stay compared with outpatients admitted to the hospital immediately after ERCP, and some factors may predict this outcome. These findings should influence decision making and informed consent when electing to proceed with high-risk ERCP, recognizing that patients with sphincter of Oddi dysfunction were over-represented in this cohort, which may limit generalizability to a typical ERCP practice.