Background: Electroencephalogram burst suppression can be associated with postoperative delirium; however, the results of relevant studies are discrepant. This systematic review and meta-analysis aimed to assess the association between intraoperative burst suppression and postoperative delirium in adult surgical patients.
Methods: PubMed, MEDLINE, Embase, Google Scholar, and the Cochrane Central Register of Controlled Trials were systematically searched and updated in May 2023. We included cohort studies, case-control studies, and randomized-controlled studies reporting on postoperative delirium incidence with documented intraoperative burst suppression in adults receiving general anesthesia for any surgery. The primary outcome was the pooled odds ratio (OR) for postoperative delirium in cases with intraoperative burst suppression compared to those without burst suppression, calculated using a random-effects model. Two independent investigators extracted the data. The protocol was prospectively registered in PROSPERO (registration number: CRD42022326479); the results were reported according to PRISMA guidelines.
Results: Fourteen studies (6435 patients) were included in the analysis. The overall incidence of postoperative delirium was 21.1% (1358/6435). Patients with intraoperative burst suppression had a higher incidence of postoperative delirium than those without burst suppression (pooled OR, 1.492; 95% confidence interval (CI) [1.022-2.178]; I2 =44%; 95% CI [0%-75%]; τ2 = 0.110). The intraoperative duration of burst suppression was significantly longer in patients who developed postoperative delirium (standardized mean difference [SMD] 0.462 [95% CI, 0.293-0.632]; I2 = 63%; 95% CI [16%-84%]; τ2 = 0.027). The burst suppression ratio was significantly higher in the delirium group (SMD 0.150; 95% CI [0.055-0.245]; I2 = 0%; 95% CI [0%-85%]; τ2 = 0.00).
Conclusion: Our meta-analysis suggests an association between intraoperative burst suppression and postoperative delirium; however, the quality of evidence was very low. The limited number of studies and substantial heterogeneity across them emphasize the need for further high-quality studies to establish a more robust conclusion.