S. Loperfido et al., MAJOR EARLY COMPLICATIONS FROM DIAGNOSTIC AND THERAPEUTIC ERCP - A PROSPECTIVE MULTICENTER STUDY, Gastrointestinal endoscopy, 48(1), 1998, pp. 1-10
Background: There is a lack of multicenter prospective studies on comp
lications of diagnostic and therapeutic endoscopic retrograde cholangi
opancreatography (ERCP). Methods: We studied 2769 consecutive patients
undergoing ERCP at nine centers in the Triveneto region of Italy over
a 2-year period. Six centers performed ERCP on less than 200 patients
per year (small centers). General and ERCP-specific major complicatio
ns were predefined. Data were collected at the time of ERCP, before di
scharge, and in cases of readmission within 30 days. ERCP was defined
as therapeutic when endoscopic sphincterotomy (n = 1583), precut (n =
419), or drainage (n = 701) had been carried out, singularly or in com
bination. Results: One hundred eleven major complications (4.0%) were
recorded: moderate-severe pancreatitis 36 (1.3%), cholangitis 24 (0.87
%), hemorrhage 21 (0.76%), duodenal perforation 16 (0.58%), others 14
(0.51%). Among 942 diagnostic ERCPs there were 13 major complications
(1.38%) and 2 deaths (0.21%), whereas among 1827 therapeutic ERCPs the
re were 98 major complications (5.4%) and 9 deaths (0.49%).The differe
nce in the incidence of complications between diagnostic and therapeut
ic ERCPs was statistically significant (p < 0.0001). Small center and
precut were recognized as independent risk factors for overall major c
omplications of therapeutic ERCP, whereas the following risk factors w
ere identified in relation to specific complications: (1) pancreatitis
: age less than 70 years, pancreatic duct opacification, and nondilate
d common bile duct; (2) cholangitis: small center, jaundice; (3) hemor
rhage: small center; and (4) retroperitoneal duodenal perforation: pre
cut, intramural injection of contrast medium, and Billroth II gastrect
omy. Conclusions: Major complications are mostly associated with thera
peutic procedures and low case volume. Present data support a policy o
f centralization of ERCP in referral centers. A more selected and safe
r use of precut may be expected to further limit the adverse events of
ERCP.