Appraisal of Postoperative Outcomes of Volatile and Intravenous Anesthetics: A Network Meta-Analysis of Patients Undergoing Cardiac Surgery.

ABSTRACT

OBJECTIVES: To determine the relative efficacy of specific regimens used as primary anesthetics, as well as the potential combination of volatile and intravenous anesthetics among patients undergoing cardiac, thoracic, and vascular surgery. DESIGN: This frequentist, random-effects network meta-analysis was registered prospectively (CRD42022316328) and conducted according to the PRISMA-NMA framework. Literature searches were conducted up to April 1, 2022 across relevant databases. Risk of bias (RoB) and confidence of evidence were assessed by RoB-2 and CINeMA, respectively. Pooled treatment effects were compared with propofol monotherapy. SETTING: Fifty-three randomized controlled trials (N = 8,085) were included, of which 46 trials (N = 6,604) enrolled patients undergoing cardiac surgery. PARTICIPANTS: Trials enrolling adults (≥18) undergoing cardiac, thoracic, and vascular surgery, using the same induction regimens, and comparing volatile and/or total intravenous anesthesia for the maintenance of anesthesia. Given that the majority of trials focused on those undergoing cardiac surgery and the heterogeneity, analyses were restricted to this population. MEASUREMENT AND MAIN RESULTS: Outcomes of interest included intensive care unit (ICU) length of stay (LOS), myocardial infarction, in-hospital and 30-day mortality, stroke, and delirium. Across 19 trials (N = 1,821; 9 arms; I2 = 64.5%), sevoflurane combined with propofol decreased ICU LOS (mean difference [MD] -18.26 hours; 95% CI -34.78 to -1.73 hours), whereas midazolam with propofol (MD 17.51 hours; 95% CI 2.78-32.25 hours) was associated with a significant increase in ICU LOS, when compared with propofol monotherapy. Among 27 trials (N = 4,080; 10 arms; I2 = 0%), midazolam was associated with significantly greater risk of myocardial infarction versus propofol (risk ratio 1.94; 95% CI 1.01-3.71). There were no significant differences across other outcomes. CONCLUSION: In patients undergoing cardiac surgery, sevoflurane with propofol was associated with decreased ICU LOS compared with propofol monotherapy. Midazolam with propofol increased ICU LOS compared with propofol alone. The combined use of intravenous and volatile anesthetics should be explored further. Future trials in thoracic and vascular surgery are warranted.