The effect of remimazolam versus conventional anesthesia on postoperative delirium in frail patients: a prospective, controlled cohort study.

Background: Frail elderly patients are at high risk for postoperative delirium (POD), a serious complication associated with poor outcomes. The choice of anesthetic agent may represent a modifiable risk factor. This study aimed to compare the effect of anesthesia involving remimazolam versus conventional general anesthesia without remimazolam on the incidence of POD in this vulnerable population. Methods: We enrolled frail elderly patients (defined as Clinical Frailty Scale ≥ 5) scheduled for elective non-cardiac surgery. Patients received either anesthesia involving remimazolam (R group, n = 301) or conventional general anesthesia without remimazolam (C group, n = 305), based on the attending anesthesiologist's clinical decision. The primary outcome was the incidence of POD within the first three postoperative days, assessed by trained, blinded researchers using the 3-min Confusion Assessment Method (3D-CAM). Secondary outcomes included postoperative recovery quality (agitation, sleep, pain) and intraoperative safety parameters. Results: Between June 2024 and June 2025, a total of 606 patients were enrolled and analyzed (R group: n = 301; C group: n = 305). The overall incidence of POD was significantly lower in the R group (119/301, 39.5%) compared to the C group (143/305, 46.9%) (Relative Risk 0.84; 95% CI 0.71-0.99; P = 0.038). The remimazolam group also experienced less emergence agitation and, by postoperative day 3, reported significantly better sleep quality and lower pain scores (P < 0.01). A multivariate logistic regression identified the use of remimazolam as an independent protective factor against POD (Adjusted Odds Ratio 0.68, 95% CI 0.47-0.98, P = 0.041), while advanced age and longer anesthesia duration were independent risk factors. Conclusion: In this cohort of frail elderly patients, the use of remimazolam for general anesthesia was associated with a lower incidence of postoperative delirium compared to conventional anesthesia. This choice represents a promising, modifiable strategy for improving neurological outcomes in this high-risk group.