Abstract
BACKGROUND: With the increasing complexity and prolonged duration of endoscopic retrograde cholangiopancreatography (ERCP) procedures, sedation shifted from conscious sedation with benzodiazepines to deep sedation with propofol. We assessed the technical success and adverse event rates of ERCP with deep versus conscious sedation. METHODS: Consecutive patients treated with ERCP in the University Medical Center Utrecht over a 7-year period (2010-2016) were screened for eligibility. Gastroenterologist-administered conscious sedation with midazolam was used from 2010-2013, whilst anesthesiology-administered deep sedation with propofol was used from 2013-2016. Data were retrospectively collected from electronic medical records. Outcomes were technical success and procedure-related adverse events within 30 days after ERCP. Associations of sedation type with outcomes were analyzed in univariable and multivariable analyses. RESULTS: A total of 725 patients were included: 336 (46%) with conscious sedation and 389 (54%) with deep sedation. Technical success was significantly higher when propofol-based sedation was used (317 [82%] vs. 252 [75%], P=0.034). Adverse events also occurred significantly more often in the propofol group (77 [20%] vs. 38 [11%], P=0.002), due to higher rates of post-ERCP cholangitis (21 [5%] vs. 8 [2%], P=0.039), and post-ERCP pancreatitis (29 [7%] vs. 11 [3%], P=0.014). After adjustment, propofol-based sedation remained significantly associated with technical success and adverse events, with odds ratios of 1.53 (95% confidence interval [CI] 1.05-2.21) and 1.95 (95% CI 1.25-3.04), respectively. CONCLUSION: Propofol-based sedation resulted significantly more often in technical success of ERCP compared with midazolam-based sedation, but adverse events were almost twice as common, with higher rates of post-ERCP pancreatitis and cholangitis.