Dexmedetomidine versus propofol for intraoperative sedation and postoperative delirium in elderly patients undergoing non-cardiac surgery under regional anesthesia: a systematic review and meta-analysis

Background Whether intraoperative sedation with dexmedetomidine (DEX) reduces postoperative delirium (POD) compared with propofol (PRO) in elderly patients undergoing non-cardiac surgery under regional anesthesia remains unclear. We addressed this question in a systematic review and meta-analysis. Methods We searched PubMed, Embase, the Cochrane Library, CINAHL, and four Chinese databases from inception to August 12, 2026 (Embase: from 2000) for randomized controlled trials (RCTs) comparing DEX with PRO for intraoperative sedation, with POD as the primary outcome. Two reviewers independently screened studies and extracted data; risk of bias was assessed with the Cochrane RoB 2.0 tool and certainty of evidence with GRADE. The DerSimonian-Laird random-effects model was primary; Mantel-Haenszel fixed-effect, leave-one-out, and trial sequential analysis (TSA) served as sensitivity analyses. Non-poolable secondary outcomes were described qualitatively. Results Six RCTs (2,527 patients) were included. POD incidence was 6.3% (80/1,269) with DEX versus 13.0% (164/1,258) with PRO (random-effects odds ratio [OR] 0.45, 95% confidence interval [CI] 0.33–0.60, P < 0.001; fixed-effect OR 0.44, 95% CI 0.33–0.58), with low heterogeneity (I² ≈ 3%). Absolute risk reduction was 6.7% (number needed to treat 15; relative risk reduction 52%). Leave-one-out analysis yielded pooled ORs of 0.41–0.48, all remaining statistically significant. Prespecified TSA (80% power, two-sided alpha 0.05, 20% relative risk reduction, control-event proportion 13.04%) estimated a required information size (RIS) of 5,010 participants. The cumulative sample (2,527) did not reach this RIS, although the Z-curve crossed the monitoring boundary, providing sequential evidence against the null without attaining the prespecified information size. The GRADE certainty was moderate. Bradycardia could not be pooled (inconsistent definitions). Conclusions In elderly patients undergoing non-cardiac surgery under regional anesthesia, intraoperative DEX sedation was associated with a lower incidence of POD than PRO. However, safety outcomes were incompletely and inconsistently reported, all included studies had some concerns regarding risk of bias, and the prespecified TSA did not reach the required information size. These findings do not establish DEX as the preferred sedative for all patients; treatment selection should remain individualized, with attention to baseline cardiovascular status and the potential for bradycardia and other adverse events. Trial registration PROSPERO CRD420261474862 (registered August 10, 2026).