Introduction Midazolam is a short acting benzodiazepine that is commonly used for sedation during colonoscopy. There is no standard dose of midazolam; however, British Society of Gastroenterology guidelines suggest a maximum of 5 mg with lower doses for elderly patients. Caecal intubation rate (CIR) is a commonly used performance indicator for colonoscopy. Data exploring the relationship between midazolam dose and CIR is limited.

Methods A retrospective cohort study of all patients who had undergone a colonoscopy at Milton Keynes General hospital between January 2010 and December 2012. Patients were identified from the Endoscopy Unit database and their records were reviewed. Patient details, midazolam dose and depth of insertion were extracted into a standardised form. Caecal intubation was defined as insertion of the colonoscope to a point proximal to the ileocaecal valve so that the entire caecum could be visualised.

Results 6200 patients were included for analysis. The median age was 62 years and 49.4% were male. The mean midazolam dose was 1.9 mg. 1004 patients had a low dose of midazolam (<2 mg), 4618 a standard dose (2 mg) and 578 a high dose (>2 mg). The CIR in the low dose cohort was 83.6%, in the standard dose cohort was 91.3% and in the high dose cohort was 78.7%. Procedural discomfort was significantly greater in the high dose cohort. When patients with poor bowel preparation were removed from the cohort (n = 5534), CIR was 85.2% in the low dose cohort vs. 92.1% in the standard dose cohort. Patients who received doses of midazolam <2 mg or > 2 mg had a significantly lower CIR than patients who received 2 mg, with P < 0.001 in both comparisons.

Conclusion In this study we demonstrate that a standard (2 mg) dose of midazolam is associated with a significantly higher caecal intubation rate than lower doses. Following stratification of patients by adequacy of bowel preparation, CIR remained lower in the low dose midazolam group. Appropriate patient selection for standard dose midazolam is important to avoid respiratory and cardiovascular compromise. Further confirmatory prospective studies are warranted.

Disclosure of Interest None Declared.