Editorial
June 10, 2024
Anesthesia Dose and Delirium—A Picture Coming Into Focus
Miles Berger, Mark D. Neuman
JAMA. Published online June 10, 2024. doi:10.1001/jama.2023.26819
In this issue of JAMA,1 Deschamps and colleagues report results of the Electroencephalographic Guidance of Anesthesia to Alleviate Geriatric Syndromes-Canada (ENGAGES-Canada) trial. ENGAGES-Canada compared rates of postoperative delirium among 1140 patients 60 years or older undergoing cardiac surgery who were randomly assigned to 1 of 2 strategies to prevent excessive dosing of inhaled general anesthetics: electroencephalogram (EEG)-guided dosing vs usual care, in which anesthesia dose was guided by expired anesthesia gas levels and hemodynamic measurements. Although some previous studies have found an association between deep general anesthesia and postoperative delirium,2,3 it remains unknown whether using EEG guidance to minimize time with intraoperative EEG suppression—a marker of deep anesthesia—might reduce delirium compared with usual care.
Conducted at 4 Canadian hospitals, ENGAGES-Canada achieved an 18% reduction in median inhaled anesthesia dose with an EEG-guided strategy, accompanied by a reduction of total time with EEG suppression from 12 to 4 minutes. However, the study found no difference in delirium incidence by group, with 18% of patients in each group experiencing this outcome; other outcomes—including awareness under anesthesia, length of hospital stay, ICU utilization, and mortality—also did not differ by group.
These findings add detail to a growing literature investigating the relationship between general anesthesia dose and postoperative delirium. The results of ENGAGES-CANADA mirror those of the earlier ENGAGES trial.4 This prior study enrolled patients undergoing both cardiac and noncardiac surgical procedures and found that an EEG-guided strategy led to an approximately 14% reduction in anesthesia dose and a reduction in EEG suppression from 13 to 7 minutes, but no change in delirium incidence compared with usual care.
In contrast, a substudy of the international Anesthetic Depth and Complications After Surgery (BALANCED)2 trial compared 2 EEG-guided anesthesia dosing strategies among 655 older adults undergoing primarily noncardiac surgical procedures. Patients were assigned to receive either deep vs light general anesthesia, based on 2 different target values of the bispectral index, a numerical EEG parameter that can be used to guide anesthesia dosing. BALANCED successfully reduced the median anesthesia dose by 25% between the 2 study groups, along with a reduction in EEG suppression time from 5 minutes to 2 minutes. Yet, unlike in ENGAGES-Canada, these changes in the BALANCED trial were associated with a decrease in postoperative delirium, from 28% to 19% among patients assigned to deeper vs lighter anesthesia. Similarly, 2 randomized trials published in 2013—the Cognitive Dysfunction after Anesthesia (CODA) study3 and the study by Radtke and colleagues5—also found that EEG-guided anesthetic titration vs routine care (without EEG guidance) led to a significant reduction in delirium incidence following noncardiac surgery in older adults.
What can explain these differences? One possibility could relate to the nature of the characteristics of patients enrolled in each study. ENGAGES-CANADA only enrolled patients undergoing cardiac surgery, and the sample carried a high level of comorbidity; 92% of all patients enrolled in ENGAGES-Canada had an American Society of Anesthesiologists Physical Status Classification of IV or greater, indicating severe systemic disease that is a constant threat to life. In contrast, more than 99% of the participants enrolled in BALANCED, CODA, and the Radtke study underwent noncardiac surgery. The older adults enrolled in these 3 studies were also healthier overall than in the ENGAGES-Canada or the original ENGAGES samples; for example, only 2% of patients enrolled in BALANCED received an American Society of Anesthesiologists classification of IV. Whereas in the healthier samples of BALANCED, CODA, and the Radtke study, EEG-guided anesthetic titration reduced postoperative delirium, it is possible that any benefits of such titration in the ENGAGES studies were diminished by the cumulative impact of other delirium risk factors.6,7
A second possibility is that the differences in outcome findings between these trials may simply reflect differences in the relative amount of anesthesia administered in the treatment and comparison groups of the ENGAGES trials vs other studies. Measured in terms of minimum alveolar concentration (MAC), a dosing unit specific to inhaled anesthesia medications, the reduction in anesthesia dose in ENGAGES and ENGAGES-Canada was 14% and 18%, respectively, vs 25% in BALANCED and 39% in CODA. Patients assigned to receive usual care in ENGAGES-Canada received a similar dose of anesthesia as patients assigned to the deep anesthesia group in BALANCED (0.8 vs 0.79 MAC). Yet the amount of anesthesia delivered in the EEG-guided group in ENGAGES-Canada was higher than the analogous group of the BALANCED study (0.66 vs 0.59 MAC).
Notably, the ENGAGES studies did reduce the extent of intraoperative EEG suppression to a greater extent than BALANCED. ENGAGES and ENGAGES-Canada reduced absolute suppression time between usual care vs EEG-guided groups by 6 minutes and 8 minutes, respectively. In comparison, the reduction between groups in BALANCED was 3 minutes (CODA and the study by Radtke et al did not report EEG suppression time). One interpretation that reconciles the findings of ENGAGES and ENGAGES-Canada with these other 3 studies is that employing EEG monitoring with the principal goal of reducing EEG suppression time with anesthesia is unlikely to prevent delirium. On the other hand, simply reducing the dose of anesthesia to the lowest safe level possible based on EEG guidance—as was done in BALANCED, CODA, and the study by Radtke et al—very well might. Indeed, debate persists as to whether intraoperative EEG suppression plays a causal role in delirium pathogenesis vs simply serving as a surrogate marker for underlying brain health.8 As such, a finding that reducing EEG suppression time does not impact rates of delirium after a surgical procedure does not on its own exclude the possibility that certain degrees of anesthesia dose reduction may be beneficial.
Considered through this lens, ENGAGES-Canada highlights a key question about postoperative delirium of relevance to patients undergoing both cardiac and noncardiac surgical procedures. Why, at a given anesthetic dose, do certain patients demonstrate EEG suppression while others do not? And why do patients who are more sensitive to developing EEG suppression in response to anesthetics more often develop postoperative delirium?9,10Answering these questions could help to identify individuals at greatest postoperative delirium risk and target these patients for delirium prevention interventions. Such efforts may also help us understand what patient groups may be most likely to show a reduction in delirium risk due to EEG-guided anesthetic titration. More broadly, deepening our understanding of why some patients and not others have a propensity toward EEG suppression under anesthesia may shed light on the mechanisms of delirium itself, which fundamentally constitutes a brain state that is accompanied by generalized EEG slowing and suppression.11
So what does this all mean for clinical practice? Ultimately, ENGAGES-Canada, interpreted in the context of the prior ENGAGES study and the BALANCED, CODA, and Radtke et al trials, tells us less about whether intraoperative EEG monitoring should be used to guide anesthetic administration and more about how, and perhaps also for whom, it should be used. Validation studies may help by confirming the promising findings of BALANCED, CODA, and the Radtke et al trial by testing anesthesia dose-minimization strategies among those categories of patients not well represented in their samples, namely patients undergoing cardiac surgery and those with very high comorbidity burdens. In the meantime, the findings of the ENGAGES studies allow us to confidently rule out major benefits of an intraoperative EEG monitoring strategy based primarily on preventing EEG suppression, while still leaving the door open to potential benefits from other intraoperative EEG monitoring regimens.
For older adults and their families, the collective findings from these trials should offer hope that strategies focused on better tailoring anesthesia delivery to the needs of each individual patient may still hold potential to help prevent delirium. Indeed, as the most common postoperative complication among older adults6,7 and a harm of medical care that can impair the well-being of patients far beyond the resolution of the initial episode,12the problem of postoperative delirium demands serious, continued efforts to discover effective methods for prevention. The findings of ENGAGES-Canada, seen in context of other important work in this area, both reflects the progress we have made over the past decade in understanding the relationship between anesthesia dose and delirium risk, while also highlighting the need for further work to bring this complex picture more fully into focus.
