Despite the widespread use of carbon dioxide insufflation (CDI) in cardiac surgery, there is still paucity of evidence to prove its benefit in terms of neurologic protection. Therefore, we conducted a meta-analysis of available randomized controlled trials comparing CDI vs standard de-airing maneuvers. Electronic searches were performed to identify relevant randomized controlled trials. Primary outcomes investigated were postoperative stroke, neurocognitive deterioration, and in-hospital mortality. Risk difference (RD) was used as summary statistic. Pooled estimates were obtained by means of random-effects model to account for possible clinical diversity and methodological variation between studies. Eight studies were identified with 668 patients randomized to CDI (n = 332) vs standard de-airing maneuvers (n = 336). In-hospital mortality was 2.1% vs 3.0% in the CDI and control group, respectively (RD 0%; 95% confidence interval [CI] -2% to 2%; P = 0.87; I2 = 0%). Incidence of stroke was similar between the 2 groups (1.0% vs 1.2% in the CDI and control group, respectively; RD 0%; 95% CI -1% to 2%; P = 0.62; I2 = 0%). Neurocognitive deterioration rate was 12% vs 21% in the CDI and control group, respectively, but this difference was not statistically significant (RD: -7%; 95% CI -0.22% to 8%; P = 0.35; I2 = 0%). The present meta-analysis did not find any significant protective effect from the use of CDI when compared with manual de-airing maneuvers in terms of clinical outcomes, including postoperative neurocognitive decline.
Keywords: carbon dioxide insufflation; cardiac surgery; meta-analysis; neurocognitive decline.
Copyright © 2017 Elsevier Inc. All rights reserved.