Background and aims: Remimazolam is an ultrashort-acting benzodiazepine.
Methods: We performed a randomized double-blind comparison of remimazolam to placebo for outpatient colonoscopy. This study design was a requirement of the U.S. Food and Drug Administration. An additional group was randomized to open-label midazolam administered according to its package insert instructions (the randomization ratio for remimazolam:placebo:midazolam was 30:6:10). Study medications were administered under the supervision of the endoscopist, without any involvement of an anesthesia specialist. Patients were given 50 to 75 μg of fentanyl before receiving study medications. Patients who failed to achieve adequate sedation in any arm were rescued with midazolam dosed at the investigator's discretion. The primary endpoint was a composite that required 3 criteria be met: completion of the colonoscopy, no need for rescue medication, and ≤5 doses of remimazolam or placebo in any 15-minute interval (≤3 doses of midazolam in any 12-minute interval in the open-label midazolam arm).
Results: There were 461 randomized patients in 12 U.S. sites. The primary endpoint was met for remimazolam, placebo, and midazolam in 91.3%, 1.7%, and 25.2% of patients, respectively (P < .0001 for remimazolam vs placebo). Patients administered remimazolam received less fentanyl, had faster recovery of neuropsychiatric function, were ready for discharge earlier, and felt back to normal sooner than patients with both placebo and midazolam. Hypotension was less frequent with remimazolam, and hypoxia occurred in 1% of patients with remimazolam or midazolam. There were no treatment-emergent serious adverse events.
Conclusion: Remimazolam can be administered safely under the supervision of endoscopists for outpatient colonoscopy, and it allows faster recovery of neuropsychiatric function compared with placebo (midazolam rescue) and midazolam. (Clinical trial registration number: NCT02290873.).
Copyright © 2018 American Society for Gastrointestinal Endoscopy. Published by Elsevier Inc. All rights reserved.