Safety of Intermittent Parenteral Midazolam Injection in Hospitalized Patients Admitted to Non-Intensive Care Nursing Units: A Retrospective Matched-Cohort Study.

Background: The prescribing information for parenteral midazolam contains a Boxed Warning stating its use may precipitate respiratory arrest, and its safety in non-intensive care unit (ICU) nursing units has been understudied.

Objective: To characterize the safety of intermittent midazolam injections in patients admitted to non-ICU nursing units relative to lorazepam injections.

Methods: This single-center, retrospective, matched-cohort study included patients ≥18 years of age who received intermittent midazolam in non-ICU nursing units. Midazolam administrations were matched 1:1 to lorazepam administrations. Safety outcomes included hypotension, bradycardia, bradypnea, and escalation in level of nursing unit care or oxygen requirement.

Results: A total of 94 midazolam administrations were matched to lorazepam administrations. Demographic data were similar between groups. Midazolam was more commonly given intravenously (95.7% vs 85.1%, P = 0.02) and in an intermediate care nursing unit (31.9% vs 14.9%, P = 0.009). The lorazepam cohort had significantly fewer concomitant respiratory depressants administered, and there were no differences in the use of beta-blockers or antihypertensives. There were no differences in the incidence of hypotension (1.1% vs 2.1%, P = 1), bradycardia (5.3% vs 2.1%, P = 0.44), bradypnea (1.1% vs 0%, P = 1), escalation in level of care (2.1% vs 2.1%, P = 1), or escalation in oxygen requirement (5.3% vs 3.2%, P = 0.72).

Conclusion and relevance: The administration of parenteral midazolam in non-ICU nursing units resulted in a similar incidence of hypotension, bradycardia, bradypnea, escalation in oxygen requirement, and escalation in level of care compared with lorazepam. These results suggest that midazolam may have a similar safety profile to lorazepam, supporting its use in non-ICU nursing units.