Purpose: Hypoxemia is the most common cardiopulmonary event during endoscopic sedation. The risk factors for hypoxemia in a healthy ASA I/II patients have not been extensively studied in a large population. Methods: A post hoc analysis from a multicenter randomized controlled trial of subjects undergoing routine colonoscopy or EGD was performed. 1000 ASA I/II/III subjects with a BMI < 35 were randomized to receive either midazolam or propofol-mediated sedation by an endoscopist/nurse team. 494 ASA I/II subjects (357 colon, 137 EGD) received midazolam-mediated sedation. 478 subjects ASA I/II (347 colon, 131 EGD) received propofol-mediated sedation using the SEDASYS® System, an investigational device. Both groups also received a narcotic (fentanyl for propofol-mediated; fentanyl or meperidine for midazolam-mediated). The sedation target for both groups was minimal to moderate sedation. Pulse oximetry, capnography, BP and ECG data points were captured for both groups every second using a computerized digital storage device. Hypoxemia was defined as an O2 saturation <90% for >15 seconds. Potential risk factors analyzed were age ≥ 55 years, gender, BMI ≥ 30, BMI ≤ 20, and a history of cardiovascular or neurological/psychiatric disorders. The factors were analyzed using a multivariate ANOVA. Results: The number of subjects with hypoxemic events was significantly higher (p < 0.001) in the midazolam-mediated arm (76 subjects, 170 events) than in the propofol-mediated arm (32 subjects, 42 events). The risk factors for hypoxemia for the midazolam-mediated group that had statistical significance were subjects with a BMI ≥ 30 or a previous history of cardiovascular disease (Table 1). The propofol-mediated group showed no statistically significant risk factors (p > 0.15 for all factors), though the most influential risk factors mirrored those of the midazolam group, BMI ≥ 30 or a previous history of cardiovascular disease. Also of note, for the midazolam group gender as a risk factor (p=0.067) almost had statistical significance. There were 49% (243/494) females in the midazolam group. In the midazolam group 61% (46/76) of the subjects who experienced hypoxemia were female (49%). For the subjects who experienced hypoxemia the gender breakdown for the midazolam group showed a bias for hypoxemia for females.Table 1: Hypoxemia risk factors for Midazolam and propofol groupsConclusion: Even healthy subjects as defined by ASA I/II are prone to hypoxemia during endoscopic sedation. Propofol-mediated sedation using the SEDASYS System resulted in a significantly lower incidence of hypoxemia and appeared to mitigate the risk factors associated with midazolam-mediated sedation. Disclosure: John J. Vargo, MD, MPH, FACG, Consultant,Ethicon Endo-Surgery, Cininnati, OH Daniel J. Pambianco, MD, FACG Consultant,Ethicon Endo-Surgery, Cininnati, OH Paul Niklewski, BS Employee, Ethicon Endo-Surgery, Cininnati, OH.