Remimazolam lowers hypotension but does not prevent postoperative delirium in older adults

A focused narrative review of remimazolam in older surgical patients found that the drug is generally associated with fewer hypotensive episodes than propofol, but its effect on postoperative delirium remains inconsistent and cannot be presented as reliably protective. Randomized trials in orthopedic, hip, abdominal and thoracic surgery produced different neurocognitive signals: several found broadly similar rates of postoperative delirium, a small study in frail hip-surgery patients reported lower incidence with remimazolam, and a 2026 lung-resection trial found substantially more delirium during emergence from anesthesia. The review, which searched PubMed/MEDLINE and tracked key citations through 2 October 2026, assessed hemodynamic stability, emergence delirium and later postoperative delirium as related but distinct outcomes rather than combining them into a single claim.

Older adults form a particularly important surgical population because frailty, multimorbidity, reduced cardiovascular reserve and baseline cognitive vulnerability can increase susceptibility to both hypotension and delirium. Postoperative delirium is itself associated with prolonged hospitalization, functional decline and mortality, making anesthetic selection a potentially relevant part of perioperative care. However, the review emphasizes that drug effects cannot be separated easily from pain, hypoxemia, infection, sleep disruption and other factors that may contribute to neurocognitive complications. Guidance cited in the analysis therefore favors risk assessment, repeated screening with validated tools and multicomponent prevention, beginning in the recovery room or by postoperative day one, rather than relying on a single anesthetic to guarantee brain health.

Remimazolam is an ultra-short-acting intravenous benzodiazepine that positively modulates gamma-aminobutyric acid type A receptors, enhancing inhibitory signaling to produce hypnosis, anxiolysis and anterograde amnesia. Its defining chemical feature is an ester linkage that is rapidly hydrolyzed by carboxylesterases to CNS7054, a metabolite with much lower activity. This “soft-drug” design gives the drug high clearance, a short context-sensitive half-time and limited dependence on cytochrome P450 metabolism, supporting predictable offset and clinical titration. Age alone may not substantially alter its disposition, but frailty, hepatic dysfunction, body composition, pharmacodynamic sensitivity and opioid exposure can change the response. Conservative dosing and continued monitoring remain necessary because short duration does not eliminate oversedation, respiratory depression or the possibility of re-sedation.

The drug’s comparative hemodynamic profile is the most consistent finding in the literature reviewed. Across randomized surgical studies and meta-analyses, remimazolam was linked to fewer hypotensive events than propofol, although the size of the difference varied with induction and maintenance regimens, surgical populations and definitions of hypotension. This may be relevant in older adults with limited cardiovascular reserve, but lower blood-pressure events are an intermediate physiological outcome rather than proof of prevention of renal injury, myocardial injury or cognitive complications. The review also notes that much of the evidence connecting intraoperative hypotension with organ injury is observational, and trials specifically designed to test whether reduced hypotension mediates a lower risk of delirium have not established that pathway.

Delirium findings across the major randomized trials are notably heterogeneous. In a comparative orthopedic trial involving 300 analyzed patients older than 60, postoperative delirium on days one through three occurred in 15.6% of remimazolam recipients and 12.4% of propofol recipients, a nonsignificant difference, while recovery took longer in the remimazolam group and hypotension was less frequent. The multicenter REMPPOD non-inferiority trial randomized 728 adults aged 60 to 90 undergoing hip surgery and found similar postoperative delirium on days one through three, 11.5% with remimazolam versus 10.4% with propofol; it met its prespecified non-inferiority criterion, while post-anesthesia care unit agitation and hypotension were less frequent with remimazolam. By contrast, a small single-center trial of 136 frail patients aged 65 or older undergoing hip surgery reported delirium in 4.4% with remimazolam and 17.6% with propofol, alongside less post-induction hypotension and fewer EEG burst-suppression episodes. That potentially favorable result applies to a selected population and requires replication before it can support a general preventive claim.

Trials in abdominal and gastric surgery have generally produced neutral results. In 432 analyzed patients aged 65 or older undergoing gastrectomy, postoperative delirium was identical in the remimazolam and propofol groups, at 7.9%, although early post-anesthesia care unit delirium was not evaluated separately. A trial of 370 patients aged 65 to under 90 undergoing major abdominal surgery likewise found no significant difference in per-protocol postoperative delirium, 17.1% with remimazolam versus 19.7% with propofol, despite less hypotension with remimazolam. The repeated separation between hemodynamic findings and delirium outcomes argues against assuming that a more stable intraoperative blood pressure automatically translates into fewer neurocognitive events. Differences in frailty, surgical stress, one-lung ventilation, opioids, regional techniques, anesthetic depth and medication exposure may modify the effects in ways that existing trials were not designed to resolve.

The most consequential caution comes from a 2026 randomized non-inferiority trial in 455 patients aged 65 or older undergoing lung resection. In the modified intention-to-treat population, overall delirium was more frequent with remimazolam than with propofol, 48% versus 31%, an absolute difference of 16.5 percentage points with a 95% confidence interval of 7.7% to 25.3% and a risk ratio of 1.53. The trial therefore did not establish non-inferiority. Most of the difference arose during emergence in the post-anesthesia care unit, where delirium affected 46% of the remimazolam group and 28% of the propofol group. After discharge from the recovery unit through postoperative day three, delirium was uncommon and similar between groups, at 4% and 5%. This temporal pattern shows why assessment windows matter: a study that begins checking patients the following morning may miss an early emergence phenotype and reach a different conclusion from one that includes immediate recovery-room surveillance.

PACU delirium appears to be clinically meaningful even when episodes resolve before the next-day assessment, but its prognostic associations must not be mistaken for proof that remimazolam causes harm. In an observational cohort of 626 patients, PACU delirium was associated with hospital readmission, with an odds ratio of 2.17 and a 95% confidence interval of 1.13 to 4.17, and with approximately twice the likelihood of a hospital stay exceeding six days. A retrospective cohort of 3,781 patients aged at least 70 undergoing noncardiac, noncranial surgery detected PACU delirium in 9.1% and associated it with later ward delirium, falls and longer hospitalization. Other observational work has linked postoperative delirium generally to delayed cognitive recovery and faster cognitive decline, but these studies do not demonstrate that a short, isolated PACU episode causes persistent impairment. The evidence instead identifies early recovery delirium as a marker of perioperative vulnerability that warrants distinct assessment and follow-up.

Mechanistic explanations for remimazolam remain less secure than its basic pharmacology. A hip-replacement study found lower postoperative catecholamine and cortisol concentrations with remimazolam, but its design could not establish a drug-specific mechanism or show that the changes mediated clinical outcomes. Experimental cerebral and hepatic ischemia-reperfusion models have reported changes in inflammatory and cellular stress signaling, including NF-kappaB/NLRP3 and FOXO-related pathways. These findings are biologically plausible but derive from different organs and models, and they cannot be extrapolated to delirium prevention in older surgical patients. The same caution applies to the proposed links among reduced hypotension, preserved cerebral perfusion, lower neuroinflammation and better postoperative cognition: these are hypotheses until tested directly in appropriately designed human studies with neurocognitive endpoints.

On the evidence summarized in the review, remimazolam should not be selected routinely as a delirium-preventing drug. It may be reasonable when propofol-associated hypotension is a major concern or when benzodiazepine reversibility is advantageous, with conservative titration, appropriate anesthetic-depth management, respiratory and hemodynamic monitoring and continued observation after flumazenil. Delirium prevention remains a multimodal process involving adequate analgesia, oxygenation, medication review, sleep preservation, mobilization and structured screening. Future multicenter trials should prespecify PACU delirium and later postoperative delirium as separate endpoints, use repeated validated assessments, record baseline cognition and frailty, and document anesthetic dose, EEG metrics, burst suppression, hypotension burden, opioid exposure, regional techniques and flumazenil use. Analyses of hospital stay, readmission, functional recovery, cognition and the timing of assessments will be needed to determine whether differences during emergence have durable clinical meaning.

Subject of Research: Hemodynamic and neurocognitive outcomes associated with remimazolam versus propofol in older surgical patients

Article Title: Remimazolam in older surgical patients: postoperative delirium, hemodynamic stability, and mechanistic considerations

Article References: Peng, D., Zhong, Z., Zou, P., Yuan, L., Song, C., Zhou, B., Liao, B., Wang, S., & Zeng, J. (2026). Remimazolam in older surgical patients: postoperative delirium, hemodynamic stability, and mechanistic considerations. Frontiers in Medicine, 13, Article 1942031. https://doi.org/10.3389/fmed.2026.1942031

Image Credits: AI Generated

DOI: 10.3389/fmed.2026.1942031

Keywords: remimazolam, postoperative delirium, emergence delirium, hemodynamic stability, older adults, propofol, geriatric anesthesia, PACU, surgery, older, surgical, patients