Endoscopic retrograde cholangiopancreatography (ERCP) routinely used as an inpatient procedure, due to post-procedural complications (1,2). The inpatient setting allows close observation but also increases healthcare costs and may not be necessary for all patients (1). Improved procedural safety, sedation practices, and early recovery protocols have raised the possibility of performing ERCP on an outpatient or ambulatory basis, allowing same-day discharge (3). The present study aimed to evaluate the feasibility and safety outcomes of ambulatory ERCP performed in a tertiary referral center. 

A retrospective study of consecutive ERCP between June 2019 and October 2025 was performed. Safety was defined as the presence of early adverse events (≤72 hours). Adverse events were categorized as minor and serious. The latter included those that required new hospitalization after discharge (outpatients) or prolonged hospitalization (inpatients). Feasibility was defined as successful same-day discharge without unplanned admission or observation ≥24 hours. Ambulatory patients followed the center's local protocol and were advised to immediately report if pain, bleeding, dizziness, nausea, vomiting, or any alarm symtom arised.

A total of 6000 ERCPs were performed from 2019 to 2025. 1000 patients were randomly selected and reviewed, and 206 were excluded from the analysis. A total of 794 patients were enrolled and distributed into outpatient [382/794 (48,1%)] and inpatients [423/794 (53.2%)]. Overall median age was 54.0 (35 – 68) (p=.005) and 54.9% were female (p=.267). Previous ERCP procedures were more common in the outpatient group (20.7% vs 10.7%, p<.001), previous cholecystectomy was most common in the outpatient group (31.2% vs 23.3%, p>.05) (Table 1).

The most common ERCP indication was choledocholithiasis in the outpatient (74.9%) and inpatient (65.3%) groups, followed by tumor suspicion in the outpatient group (16.0%) and biliary sludge in the inpatient group (18.9%). Difficult cannulation in the outpatient and inpatient groups was 7.6% and 7.0%, respectively (p=.764). Sphincterotomy was performed in 90.8% of outpatient and 92.0% of inpatient, respectively (p=.562), and biliary stenting was reported in 28.3% and 24.5%, respectively (p=.229).

Complication rate was similar among groups with 14.4% in the outpatient and 9.7% in the inpatient group (p>.05). Bleeding was the most common adverse, being mild in forty-two outpatient cases (11.0%) and 28 inpatient cases (7.0%), managed with sclerotherapy; however, 2 cases in outpatient and 1 case in the inpatient group required hospitalization. A sub-analysis was conducted considering the serious adverse events that required hospitalization after discharge or extended hospital stay. There were 13 cases among the outpatients who required a new hospitalization after discharge, and 12 inpatients who required prolonged hospital stay (p>.05). Post-ERCP pancreatitis was higher in the outpatient groups (2.4%), compared to 2.2% of inpatients (p>.05). Most inpatient cases achieved technical success (99.5%, p<.001) and less clinical success (88.8%, p<.001), compared to outpatient's (86.4% and 93.7%), respectively. No mortality was perceived in either group.

Ambulatory ERCP proved to be a feasible and safe strategy, with complication and serious adverse event rates comparable to those of inpatient ERCP. Most adverse events were mild, and only a small proportion of patients required hospitalization or extended observation.