Background and aims: Although moderate (conscious) sedation is intended during elective gastrointestinal endoscopy, unintended levels of deep sedation occur. The aims of this study were to prospectively evaluate the incidence and risk factors of deep sedation during elective endoscopy with meperidine and midazolam intended to maintain a level of moderate sedation.
Methods: Eighty American Society of Anesthesiology class 1-2, outpatients presenting for elective esophagogastroduodenoscopy (EGD), colonoscopy, endoscopic retrograde cholangiopancreatography (ERCP), and endoscopic ultrasonography (EUS) were offered enrollment. Intravenous meperidine and midazolam were administered according to a standardized protocol. Hemodynamic parameters and levels of sedation were assessed and recorded by a single observer at 3-min intervals. The Modified Observer's Assessment of Alertness/Sedation (MOAA/S) scale (ranging 1-5) is a subjective sedation assessment scale used to assess sedation levels. Occurrence of deep sedation, defined by MOAA/S 1-2, was recorded. Univariable and multivariable analyses were used to assess predictors of deep sedation.
Results: Deep sedation occurred in 54/80 (68%) patients for a total of 204/785 (26%) of total sedation assessments. The percentage of deep sedation episodes of all sedation-level observations by procedure was 26% for EGD, 11% for colonoscopy, 35% for ERCP, and 29% for EUS. Deep sedation occurred at least once in 60% of EGD, 45% of colonoscopy, 85% of ERCP, and 80% of EUS. Multivariable analysis showed that only ERCP and EUS were independent risk factors of deep sedation.
Conclusions: Deep sedation occurs frequently during elective endoscopy with meperidine and midazolam used with the intent of moderate sedation. ERCP and EUS are risk factors for the occurrence of deep sedation, independent of sedation dose or length of procedure.