What Would You Like to Start With? Choice of Initial Dose of Moderate Sedation and Impact on Time to Endoscope Insertion

Abstract

Introduction: Midazolam and fentanyl are the most frequently used agents for endoscopy sedation. Guidelines recommend slow incremental dosing to prevent oversedation, however the choice of initial sedative and dose is at the discretion of the endoscopist. We aimed to describe practice patterns in the choice of initial dose and determine whether initial dose is associated with time to endoscope insertion (induction time). Methods: We performed a single-center retrospective cohort study on adults who underwent upper endoscopy and lower endoscopy under moderate sedation from 2008-2016. Multiple procedures on the same day were excluded. The choice of sedative agents, doses, and times of administration were noted. The initial dose was defined as the combination of sedatives given within the first minute of sedation. Induction time was defined as minutes from initiation of sedation to the insertion of the endoscope. Reversal agent use was noted. Patient age, sex and race were recorded and adjusted for. Results: A total of 14,631 upper and 24,281 lower endoscopies were included (Table 1). Midazolam and fentanyl were used in over 95% of cases. A variety of initial sedation combinations were prevalent (Figure 1). The combination of midazolam 2 mg and fentanyl 50 mcg was the most common initial sedation for all age groups. Induction time was lowest for midazolam 2 mg and fentanyl 75 mg across all ages. Induction time for “2 and 50” was consistently longer than for “1 and 75” and for “2 and 75.” Single-agent midazolam 2 mg was associated with the longest induction times of 6 minutes or more. Reversal agents were required in 14 patients who had received “2 and 50” versus 7 patients who had received “2 and 75” (0.08% vs 0.08%, p=0.99). After adjusting for demographics, initial choice of “1 and 75” or “2 and 75” resulted in 1.1 and 1.5 minute decreases respectively in induction time compared to initial use of midazolam 2 mg and fentanyl 50 mcg (p < 0.01, Table 2).Figure: Bars: Initial sedation dose combinations, in order of frequency, stratified by age group. Lines: mean induction time for each initial sedation dose combination.Table: Table. Characteristics of 38,912 Patients Who Underwent Upper and Lower Endoscopy under Moderate SedationTable: Table. Linear Regression Model for Sedative Induction Time in Minutes, with Most Frequent Initial Sedative Doses Combinations, Adjusted for Patient Demographics.Conclusion: There is significant variability in the choice of initial sedative dose, even within age categories. Midazolam appears to have slower onset compared with fentanyl, as evidenced by longer induction times for midazolam-predominant combinations. The initial use of midazolam 2 mg and fentanyl 75 mg was associated with shorter time to endoscope insertion than other combinations, and was not associated with an increased use of reversal agents.