Tu1258 Incidence of Sedation Related Adverse Events (Srae) During Endoscopy in Patients Supported by Continuous Flow Left Ventricular Assist Devices (LVAD)

Incidence of Sedation Related Adverse Events (Srae) During Endoscopy in Patients Supported by Continuous Flow Left Ventricular Assist Devices (LVAD) Vladimir M. Kushnir, Laura M. De Mondesert*, Susan Joseph, Gregory A. Ewald, Scott Silvestry, Steven A. Edmundowicz Gastroenterology, Washington University School of Medicine, St. Louis, MO; Cardiology, Washington University School of Medicine, St Louis, MO; Cardiothoracic Surgery, Washington University School of Medicine, St Louis, MO Background: Continuous flow LVADs have revolutionized management of advanced heart failure in the past decade. Gastrointestinal complications requiring diagnostic and therapeutic endoscopy are common during long term LVAD support. Patients with advanced heart failure are considered high risk for SRAE during endoscopy. In patients supported by LVADs the risk is compounded by lack of pulsatile blood flow; which makes vital sign monitoring challenging. Aim: Define the incidence of SRAE during endoscopy in the LVAD population. Methods: Consecutive patients supported by the HeartMate II LVAD undergoing endoscopy with conscious sedation (CS) or propofol based sedation under monitored anesthesia care (MAC) at a tertiary care center between 2005-12 were retrospectively identified. Clinical records were interrogated and the incidence of SRAE was recorded. SRAE were defined as: hypoxemia (SaO2 90%), respiratory failure requiring intubation, new onset arrhythmia, hypotension (mean arterial pressure 60mmHg), low LVAD flow ( 4 L/min) and premature procedure termination. Logistic regression analysis was used to identify independent risk factors for SRAE. Results: 74 patients underwent 190 endoscopic procedures [median 2 per patient (range 1-10)] while supported by LVAD. Procedures included: 61 upper endoscopies, 71 colonoscopies, 24 small bowel enteroscopies (SBE), 34 upper endoscopy colonoscopy (during same sedation). 110 (57.8%) procedure were performed under MAC. There were no significant differences in baseline patient or procedure characteristics between MAC & CS groups (table 1). SRAE occurred more frequently with MAC [53/110] then CS [24/80], p 0.01 (Table 2). One procedure performed under CS was terminated prematurely due to a 10 second episode of ventricular tachycardia, which resolved spontaneously. The reminders of SRAE were successfully treated without procedure interruption. There were no episodes of respiratory failure requiring tracheal intubation. All episodes of hypotension were successfully treated with fluid (n 40, 21.1%) and/or neosynephrine bolus (n 39, 20.1%); none of the patients required continuous vasopressor infusion. In a linear regression model containing demographics and procedural factors sedation related AE were significantly associated with the use of MAC [OR 2.0(95%CI 1.13.9), p 0.04)] and performing SBE [OR 5.6(95%CI 1.9-16.9), p 0.002). Conclusions: SRAE occur in 40% of LVAD supported patients during endoscopy; however the vast majority of SRAE are transient and can be treated successfully without procedure interruption. Hypotension is the most frequent SRAE and the need for inter-procedural vasopressor administration should be anticipated, particularly when patients are sedated with MAC. Independent predictors of SRAE were the use of MAC and performing SBE.