Abstract / Summary
Postoperative delirium (POD) is a common and serious complication after major orthopedic surgery in older adults. Several perioperative pharmacological strategies targeting neuroinflammatory and neuroprotective pathways have been tested, but their relative efficacy remains uncertain. We conducted a network meta-analysis (NMA) to compare these strategies and to assess the certainty of the resulting evidence. We systematically searched four electronic databases for randomized controlled trials (RCTs) evaluating perioperative pharmacological strategies relevant to postoperative delirium (POD) or postoperative cognitive outcomes in patients aged ≥ 60 years undergoing major orthopedic surgery. For the primary quantitative POD analysis, delirium had to be ascertained using a validated instrument. A frequentist random-effects NMA was performed. Active pharmacological comparators were modelled separately from inert placebo or standard-care controls. We report odds ratios (OR) with 95% confidence intervals (CI) and 95% prediction intervals (PI), P-scores, a league table, a contribution matrix, subgroup interaction tests by surgical context, and certainty of evidence using a GRADE/CINeMA-style approach. Eighteen RCTs (4,132 participants) were included in the systematic review. Thirteen trials (3,466 participants) formed the primary POD network, which contained 9 treatment nodes and no independent closed loops. Anti-inflammatory bundle (OR 0.21, 95% CI 0.09–0.51), dexmedetomidine (OR 0.37, 95% CI 0.27–0.51), dexamethasone (OR 0.36, 95% CI 0.15–0.84), and parecoxib (OR 0.46, 95% CI 0.30–0.72) had confidence and prediction intervals excluding 1. Dexmedetomidine was supported by the largest evidence base among these interventions. Esketamine was not significantly protective versus placebo (OR 0.77, 95% CI 0.33–1.76). Surgical context did not significantly modify treatment effects (dexmedetomidine p = 0.26; parecoxib p = 0.50; propofol p = 0.53). Dexmedetomidine is supported by the largest and most precise body of randomized evidence for POD prophylaxis after major orthopedic surgery, although certainty remains moderate and the network is sparse. Several other interventions show potentially beneficial estimates, but some are based on single small trials. No surgical-stress-dependent treatment hierarchy was demonstrated. Adequately powered, standardized head-to-head RCTs are needed.