Efficacy and Safety of Remimazolam Compared to Midazolam for Sedation During Endoscopic Ultrasonography: A Single‐Center Retrospective Cohort Study

TL;DR

Remimazolam provides faster recovery and better sedation success than midazolam for endoscopic ultrasonography in outpatients.

Contribution

First comparison of remimazolam and midazolam for sedation in pancreatobiliary endoscopic ultrasonography.

Findings

Remimazolam showed significantly higher rapid recovery rates compared to midazolam.

Sedation success was higher with remimazolam, with shorter recovery times and less flumazenil use.

Hypoxemia occurred slightly more with remimazolam but was mild and reversible.

Abstract

Remimazolam is a novel ultra–short‐acting benzodiazepine that may offer advantages over conventional sedatives in endoscopic procedures. Evidence for its use in pancreatobiliary endoscopic ultrasonography (EUS) is limited. We compared the efficacy and safety of remimazolam and midazolam in outpatient pancreatobiliary EUS.

This retrospective study included outpatients undergoing diagnostic pancreatobiliary EUS between July 2024 and July 2025. Patients received either remimazolam (initial 0.2 mg/kg, top‐up 0.1 mg/kg [high‐risk patients: 0.16 mg/kg, 0.08 mg/kg]) or midazolam. The target sedation depth was Modified Observer's Assessment of Alertness/Sedation ≤3, procedures were performed without supplemental oxygen, and pentazocine was co‐administered. The primary outcome was rapid recovery (meeting all criteria: modified Aldrete score ≥6 at 5 min, ≥9 at 30 min, ambulation ≥3 m at 30 min). Secondary outcomes included sedation success (all: EUS completion, without >2 top‐ups within 10 min, without rescue sedation, without agitation requiring interruption), recovery time, flumazenil use, and adverse events.

We analyzed 139 patients (remimazolam, n = 75; midazolam, n = 64). Rapid recovery was more frequent with remimazolam (70.7% vs. 25.0%, p < 0.001). Sedation success was higher (92.0% vs. 73.4%, p = 0.005), recovery was shorter (median 34 vs. 55 min, p < 0.001), and flumazenil use was lower (4.0% vs. 45.3%, p < 0.001). Hypoxemia (SpO2 <90% lasting ≥10 s) occurred slightly more often with remimazolam (30.6% vs. 26.6%, p = 0.707), but all episodes were mild and reversible with airway support/oxygen. Hypotension was rare and comparable.

Remimazolam provided faster recovery and higher sedation success than midazolam in outpatient pancreatobiliary EUS. Supplemental oxygen before sedation is a reasonable option to enhance safety.

Figures