Background: Deep general anesthesia has been associated with increased mortality in 5 observational studies. The association may be causal or an epiphenomenon due to increased anesthetic sensitivity in high-risk patients. We conducted a pilot study to assess the feasibility of performing a definitive randomized controlled trial. The aims of the study were to determine whether anesthetic depth targeting in a high-risk group was feasible and to document anesthetic doses and arterial blood pressures associated with "deep" and "light" general anesthesia.
Methods: ASA physical status III and IV patients, aged ≥60 years, having surgery lasting ≥2 hours, with expected hospital stay ≥2 days, and receiving general anesthesia were randomly allocated to a Bispectral Index (BIS) or spectral entropy (SE) target of 35 ("low" group) or 50 ("high" group). The primary end point was mean BIS or SE. Secondary end points were postanesthesia care unit length of stay and pain scores, quality of recovery score, hospital length of stay, postoperative complications, and death. A composite end point of postoperative complications (pneumonia, myocardial infarction, stroke, pulmonary embolism, heart failure, and death) was determined at 1 year.
Results: One hundred twenty-five patients were recruited. The mean of the median BIS/SE values for each patient during the maintenance phase of anesthesia in the low and high groups was significantly different: 39 vs 48 (mean difference 8 [95% confidence interval {CI95}, 6 to 10], P < 0.001). There was also a significant difference in mean volatile anesthetic administration (minimum alveolar concentration): 0.98 vs 0.64 (mean difference -0.35 [CI95, -0.44 to -0.26], P < 0.001) and target propofol concentrations: 4.0 vs 3.1 μg/mL (mean difference -0.8 [CI95, -1.2 to -0.3], P = 0.004). Intraoperative mean arterial blood pressures were similar (85 vs 87 mm Hg; mean difference 2 [CI95, -2 to 6], P = 0.86), and there were no differences in short-term recovery characteristics or hospital length of stay. There was a significant difference in the incidence of wound infection at 30 days (13% vs 3%; risk difference -10% [CI95, -21 to -0.1], P = 0.04). At 1 year, the composite rates of complications in the low and high groups were 28% and 17% (risk difference -11 [CI95, -25 to 4], P = 0.15) and mortality rates were 12% and 9%, respectively (risk difference -2 [CI95, -14 to 9], P = 0.70).
Conclusions: This pilot study demonstrated that depth of anesthesia targeting with BIS or SE was achievable in a high-risk population with adequate separation of processed electroencephalogram monitor targets. The expected incidence of postoperative complications and mortality occurred. We conclude that a large, multicenter, randomized controlled trial is feasible.