Background: There is a concern that midazolam, when used as a component of sedation for colonoscopy, may impair cognition and prolong recovery. We aimed to identify whether midazolam produced short- and longer-term effects on multiple dimensions of recovery including cognition.
Methods: A 2-center double-blinded, placebo-controlled, parallel-group, randomized, phase IV study with a 1:1 allocation ratio was conducted in adults ≥18 years of age undergoing elective outpatient colonoscopy, with sufficient English language proficiency to complete the Postoperative Quality of Recovery Scale (PostopQRS). Participants were administered either midazolam (0.04 mg·kg) or an equivalent volume of 0.9% saline before sedation with propofol with or without an opiate. The primary outcome was incidence of recovery in the cognitive domain of the PostopQRS on day 3 after colonoscopy, which was analyzed using a χ test. Secondary outcomes included recovery in other domains of the PostopQRS over time, time to eye-opening, and hospital stay, and patient and endoscopist satisfaction. All hypotheses were defined before recruitment.
Results: During September 2015 to June 2018, 406 patients were allocated to either midazolam (n = 201) or placebo (n = 205), with one withdrawn before allocation. There was no significant difference in recovery in the cognitive domain of the PostopQRS on day 3 after colonoscopy (midazolam 86.8% vs placebo 88.7%, odds ratio, 0.838; 95% confidence interval [CI], 0.42-1.683; P= .625). Furthermore, there was no difference in recovery over time in the cognitive domain of the PostopQRS (P = .534). Overall recovery of the PostopQRS increased over time but was not different between groups. Furthermore, there were no differences between groups for nociceptive, emotive, activities-of-daily-living domains of the PostopQRS. Patient and endoscopist satisfaction were high and not different. There were no differences in time to eye-opening (midazolam 9.4 ± 12.8 minutes vs placebo 7.3 ± 0.7 minutes; P = .055), or time to hospital discharge (midazolam 103.4 ± 1.4 minutes vs placebo 98.4 ± 37.0 minutes; P = .516).
Conclusions: The addition of midazolam 0.04 mg·kg as adjunct to propofol and opiate sedation for elective colonoscopy did not show evidence of any significant differences in recovery in the cognitive domain of the PostopQRS, overall quality of recovery as measured by the PostopQRS, or emergence and hospital discharge times. The use of midazolam should be determined by the anesthesiologist.