To evaluate the association between midazolam premedication and postoperative delirium in a large retrospective cohort of patients ≥70 years.

Retrospective cohort study.

A single tertiary academic medical center.

Patients ≥70 years having elective non-cardiac surgery under general anesthesia from 2020 to 2021.

Midazolam premedication, defined as intravenous midazolam administration prior to induction of general anesthesia.

The primary outcome, postoperative delirium, was a collapsed composite outcome including at least one of the following: a positive 4A's test during post-anesthesia care unit stay and/or the initial 2 postoperative days; physician or nursing records reporting new-onset confusion as captured by the CHART-DEL instrument; or a positive 3D-CAM test. The association between midazolam premedication and postoperative delirium was assessed using multivariable logistic regression, adjusting for potential confounding variables. As secondary analysis, we investigated the association between midazolam premedication and a composite of other postoperative complications. Several sensitivity analyses were performed using similar regression models.

In total, 1973 patients were analyzed (median age 75 years, 47% women, 50% ASA score ≥ 3, 32% high risk surgery). The overall incidence of postoperative delirium was 15.3% (302/1973). Midazolam premedication was administered to 782 (40%) patients (median [IQR] dose 2 [1Fowler A.J., Abbott T.E.F., Prowle J., Pearse R.M. Age of patients undergoing surgery Br J Surg 2019 ;  106 : 1012-101810.1002/bjs.11148 [cross-ref]

Cliquez ici pour aller à la section Références,2Levant S., Chari K., De Frances C.J. Hospitalizations for patients aged 85 and over in the United States, 2000–2010 Signif. Heal. Stat. Sel. Reports from Fed. Agencies :  (2016).  81-89

Cliquez ici pour aller à la section Références] mg). After adjustment for potential confounding variables, midazolam premedication was not associated with increased odds of postoperative delirium, with adjusted odds ratio of 1.09 (95% confidence interval 0.82–1.45; P = 0.538). Midazolam premedication was also not associated with the composite of other postoperative complications. Furthermore, no association was found between midazolam premedication and postoperative delirium in any of the sensitivity analyses preformed.

Our results suggest that low doses of midazolam can be safely used to pre-medicate elective surgical patients 70 years or older before non-cardiac surgery, without significant effect on the risk of developing postoperative delirium.