Morbidity and mortality of endoscopist-directed nurse-administered propofol sedation (EDNAPS) in a tertiary referral center

BACKGROUND AND STUDY AIMS Endoscopist-Directed Nurse-Administered Propofol Sedation (EDNAPS) has been evaluated in community settings rather than tertiary referral centers. PATIENTS AND METHODS A hospital-wide prospectively collected database of Medical Emergency Team Calls (METCALL), emergency responses triggered by medically unstable patients, was reviewed. Responses that followed EDNAPS were extracted and compared with a prospectively entered database of all endoscopies performed using EDNAPS over the same period. RESULTS A total of 33,539 endoscopic procedures (16,393 gastroscopies, 17,146 colonoscopies) were performed on 27,989 patients using EDNAPS. Intravenous drugs included midazolam (0 - 5 mg), fentanyl (0 - 100 mcg), and propofol (10 - 420 mg). Of 23 METCALLs (18 gastroscopies and 5 colonoscopies), there were 16 with ASA scores of III or higher. Indications for gastroscopy were gastrointestinal (GI) hemorrhage (n = 11; 8 variceal, 3 nonvariceal), dysphagia (n = 5), PEG removal (n = 1), and dyspepsia (n = 1). Fifteen of 22 patients, including all of those who had a colonoscopy, made a full recovery and returned to the ward or were discharged home. In the gastroscopy group, seven were intubated and admitted to Intensive Care, of whom six were emergency cases for gastrointestinal bleeding (n = 4 variceal, n = 2 non variceal) and one in which the indication was PEG removal. Two deaths occurred in the intubated group. CONCLUSIONS In a tertiary referral center, EDNAPS for low-to-moderate risk (ASA ≤ 2) patients undergoing gastroscopy and colonoscopy is very safe. Gastroscopy is associated with greater anesthetic risk than colonoscopy and those with high ASA scores needing urgent endoscopy for upper gastrointestinal hemorrhage are at particular risk of cardiorespiratory decompensation.