RESUMO
BACKGROUND: Benzodiazepine use is associated with delirium, and guidelines recommend avoiding them in older and critically ill patients. Their perioperative use remains common because of perceived benefits. METHODS: We searched CENTRAL, MEDLINE, CINAHL, PsycInfo, and Web of Science from inception to June 2021. Pairs of reviewers identified randomised controlled trials and prospective observational studies comparing perioperative use of benzodiazepines with other agents or placebo in patients undergoing surgery. Two reviewers independently abstracted data, which we combined using a random-effects model. Our primary outcomes were delirium, intraoperative awareness, and mortality. RESULTS: We included 34 randomised controlled trials (n=4354) and nine observational studies (n=3309). Observational studies were considered separately. Perioperative benzodiazepines did not increase the risk of delirium (n=1352; risk ratio [RR] 1.43; 95% confidence interval [CI]: 0.9-2.27; I2=72%; P=0.13; very low-quality evidence). Use of benzodiazepines instead of dexmedetomidine did, however, increase the risk of delirium (five studies; n=429; RR 1.83; 95% CI: 1.24-2.72; I2=13%; P=0.002). Perioperative benzodiazepine use decreased the risk of intraoperative awareness (n=2245; RR 0.26; 95% CI: 0.12-0.58; I2=35%; P=0.001; very low-quality evidence). When considering non-events, perioperative benzodiazepine use increased the probability of not having intraoperative awareness (RR 1.07; 95% CI: 1.01-1.13; I2=98%; P=0.03; very low-quality evidence). Mortality was reported by one randomised controlled trial (n=800; RR 0.90; 95% CI: 0.20-3.1; P=0.80; very low quality). CONCLUSIONS: In this systematic review and meta-analysis, perioperative benzodiazepine use did not increase postoperative delirium and decreased intraoperative awareness. Previously observed relationships of benzodiazepine use with delirium could be explained by comparisons with dexmedetomidine. SYSTEMATIC REVIEW PROTOCOL: PROSPERO CRD42019128144.