Shin_2026_Medicine.(Baltimore)_105_e46373

BACKGROUND: Postoperative delirium (POD) frequently occurs in elderly surgical patients and is associated with adverse outcomes. This review aimed to evaluate whether sugammadex, a selective neuromuscular blockade reversal agent, reduces postoperative delirium-incidence (POD-I) compared with neostigmine. METHODS: PubMed, EMBASE, and Cochrane CENTRAL were searched up to May 2025 for randomized controlled trials (RCTs) and non-randomized controlled trials (NRCTs) comparing sugammadex and neostigmine regarding to POD-I. Primary outcome was POD-I, with subgroup analyses according to postoperative period (within 24 hours vs after postoperative 1 day) and the type of anticholinergic coadministered with neostigmine (atropine vs glycopyrrolate). Secondary outcome was perioperative changes in Mini-Mental State Examination scores (preoperative vs postoperative). Analyses were performed separately for RCTs and NRCTs using RevMan 5.4. RESULTS: Twelve studies (7 RCTs, n=759; 5 NRCTs, n=50,115) were included. In RCTs, sugammadex significantly reduced POD-I compared with neostigmine (risk ratio (RR), 0.67; 95% confidence interval (CI), 0.49-0.91; I2=0%; P=.008). Subgroup analysis of RCTs on POD-I within the 1st 24 hours postoperatively showed a significantly lower incidence in the sugammadex group (RR, 0.64; 95% CI, -0.45 to 0.90; P=.010). Subgroup analysis of RCTs by coadministered anticholinergics did not significantly influence POD-I (atropine: RR, 0.67; P=.06; glycopyrrolate: RR, 0.66; P=.06). In contrast, subgroup analysis of NRCTs on POD-I showed no significant difference between the groups (RR, 1.15; 95% CI, 0.88-1.50; I2=82%; P=.33). Preoperative and postoperative Mini-Mental State Examination scores in RCTs were similar between groups (preoperative: mean difference 0.03, P=.78; postoperative: mean difference 0.75, P=.12). CONCLUSION: Sugammadex is associated with a reduced incidence of POD compared to neostigmine, particularly within the 1st 24 hours after surgery, based on consistent findings from RCTs. This benefit was not observed in NRCTs. Further well-designed RCTs using standardized delirium assessments are required.