Background: Because of possible complications, it has been common prac
tice to admit most if not all patients undergoing therapeutic ERCP. Th
erefore, little descriptive data exist on the safety of outpatient the
rapeutic ERCP. Methods: We assessed 262 consecutive ERCPs in 209 patie
nts undergoing outpatient therapeutic ERCP lover a 5-year period, with
particular attention to the development of complications. All outpati
ent endoscopic sphincterotomies and stent placements performed over a
5-year period were prospectively entered into an ongoing data base tha
t was used for the analysis. In addition, hospital and office records
for all patients were retrospectively reviewed, including a 30 to 45 d
ay follow-up in a private office setting. Results: Suspected or docume
nted choledocholithiasis was the most common indication for ERCP and w
as present in 132 (50%), followed by malignant obstruction in 77 (29%)
, type I sphincter of Oddi dysfunction (on the basis of symptoms, live
r test abnormalities, and bile duct dilatation) in 36 (14%), chronic p
ancreatitis in 10 (3.8%), HIV cholangiopathy in 4 (1.5%), and other co
nditions in 3 (1.1%). Overall, 181 patients (69%) underwent a sphincte
rotomy. The 30-day post-ERCP complication rate was 5.7% (95% CI: 3.2%
to 9.3%), occurring in 15 of 262 cases. Complications necessitating ho
spitalization developed in 9 of the 262 ERCPs for a rate of 3.4% (95%
CI: 1.6% to 6.4%). The mean duration of hospital stay among patients a
dmitted for a complication was 2.7 +/- 1.8 days (range, 1 to 7 days).
All patients were discharged without permanent sequelae. No 30-day pro
cedure-related fatalities were reported. Conclusion: In this selected
series of 262 consecutive cases, endoscopic sphincterotomy and stent p
lacement were safely performed in an ambulatory setting. Prior to reco
mmending a generalized change in existing practice, however, this find
ing requires validation with larger series of cases, including the per
formance of other outpatient therapeutic ERCP techniques.