Why the study?
Does automated propofol administration guided by the bispectral index improve the maintenance of adequate anesthesia depth compared to manual control in patients undergoing elective noncardiac surgery?
Does automated propofol administration guided by the bispectral index improve the maintenance of adequate anesthesia depth compared to manual control in patients undergoing elective noncardiac surgery?
Answers are grounded in this paper and cite their sources.
Automated closed-loop propofol administration outperforms manual control in maintaining targeted depth of anesthesia, demonstrating feasibility and reduced variability in clinical settings.
Closed-loop or automated anesthesia guided by electrocortical activity began in the United States >65 years ago.1 Automated anesthetic drug delivery has been used successfully during major surgery in patients with various comorbidities,2–8 in cardiac surgery in adult9 and pediatric patients,10 lung transplantation,11 pheochromocytoma resection,12 rigid bronchoscopy,13 obese patients,14 in deep sedation in intensive care units,15 and for sedation16 or anesthesia in pediatric patients.17 Several thousand patients have been anesthetized by the use of different closed-loop controllers worldwide. Before the clinical evaluation of a new automated controller, bench testing and validation are necessary, and ideally established engineering principles should be used to the best effect for this purpose18,19; however, the clinical interests of these tests are never demonstrated. Ultimately, clinical studies are the true test of controller performance because the relationship between in silico studies and an operation on a patient in a working clinical environment may not be straightforward. As a proof of concept or to test the basic feasibility of an automated controller, an observational clinical study with a small number of subjects in a limited clinical context without a control group is acceptable. A preliminary study such as this gives a first assessment of the stability of the controller in realistic working conditions. For confirmatory studies, the major trend is to perform large clinical trials that include a suitable control group, which provides the greatest level of evidence in research; large trials decrease the effect of random chance on the data, and the inclusion of multiple centers improves the applicability and reproducibility of the results.20 In this issue of the Journal, Puri et al.21 present a randomized controlled multicenter trial that describes the comparison between automated and manual propofol administration guided by the bispectral index (BIS). This trial was performed with several investigators in 242 patients with ASA physical status I and II who were undergoing elective, noncardiac surgery in 6 centers in India. The automated controller outperformed manual control in maintaining the BIS around a set point of 50 with satisfactory hemodynamic conditions, demonstrating that the controller is capable of operating effectively in typical working conditions. The current study especially highlights the high variability in the manual skill group, despite >10 propofol target modifications per hour by the anesthesia providers. The percentage of time of adequate anesthesia or when the BIS was in the range of 40 to 60 varied between 17% and 67% in the manual group, whereas in the closed-loop group, the BIS was in the correct range for 80% to 85% of the maintenance duration. Moreover, the automated controller had the same performance across the different centers and managed interindividual variation well, especially in patients at high altitudes.22 All automated controllers should be tested clinically in challenging conditions under close supervision and should include patients with comorbidities and during major surgery to ensure robustness across the full patient and case spectrum expected to be encountered during routine use. Because of the relatively low rate of adverse events related to the administration of anesthesia itself, outcomes studies likely will need to use large patient cohorts, although it seems all but impossible that the reduction in undesired variability in management attributable to the use of the closed loop would not benefit patients. With this new study, investigators of several randomized controlled trials have now reported that strict control or standardization of anesthesia depth by automated titration decreases the variability of anesthetic depth. The use of these automated controllers improves hemodynamic stability during vascular surgery,13 cardiac surgery,9 or during deep sedation in critically ill patients15 by the decrease in vasopressor use; outperforms manual control to maintain the BIS in the range of 40 to 60 during maintenance of general anesthesia2,4,6,8,23,24; decreases the workload during the induction period25; and enables faster recovery times.2,6,8 In addition, integrated multiple closed-loop controllers have been evaluated that combine the automated titration of propofol with remifentanil guided by electrocortical activity6 or by hemodynamic score8 and rocuronium guided by a neuromuscular blockade monitor.8 These studies have all reported that closed-loop systems work well in real patients. Future studies now need to demonstrate that the use of automated controllers has the potential to improve patient outcomes or reduce provider workload or errors.26 In clinical care, automated titration of anesthesia remains a research tool and is not encountered in routine use. The experience of the aviation industry with the introduction of autopilot may provide some lessons in this regard. Despite being invented >70 years earlier, it was not until 1988 that Airbus Industry introduced the A320, the first aircraft with a full automatic pilot. Cost-effectiveness was improved by reducing the cockpit crew from 3 to 2, through the simplification of training, and through improved fuel economy thanks to the consistency of the autopilot. The number of accidents, however, increased by 5% >5 years because of the conflict between the pilot and the autopilot, in part, because the pilots took more risks to demonstrate their capabilities, and in the same period, 40% of pilots aged >50 years failed to learn the new system. After the initial adjustment period, however, the accident rate decreased steadily despite an almost-yearly increase in passengers transported. This phenomenon (increased risk in the short run followed by long-term significant risk reduction) is a sort of risk-transfer that takes place with the introduction of a new form of automation and should be anticipated with clinical closed-loop systems. There will be errors related to the systems themselves, to misapplication of the closed loops to inappropriate patients or cases, and potentially even user−system interactions that result in harm. Good development and testing pathways can reduce these risks substantially, and in the best-case scenario, the total risks to patients with the system’s use will be lower even in the short term. During the long term, the benefits of closed-loop control are clear, if the airline industry is any guide. The use of autopilot is now ubiquitous and a major reason for the incredible safety record of air travel. It allows the same level of work to be performed more accurately, in more difficult conditions, with a smaller workforce, and with lower rates of adverse events. Given all of this, we believe that automation in anesthesia is not only feasible but also ultimately irrepressible, although probably categorical defenses may delay its widespread introduction. The benefits are numerous and clear, and its introduction into routine care is likely only a question of time. To that end, the work by Dr. Puri’s team21 in demonstrating the effectiveness of their closed-loop system across multiple hospitals, providers, and environments is an encouraging step toward the adoption of these closed-loop systems in the surgical environment. DISCLOSURES Name: Ngai Liu, MD, PhD. Contribution: This author helped in manuscript preparation Attestation: Ngai Liu approved the final manuscript. Conflicts of Interest: Ngai Liu is the inventor and co-owner of a patent for gain constant and control algorithm for a closed-loop anesthesia management system. He is the cofounder of MedSteer, a biomedical company, which promotes research and development in closed-loop anesthesia tools and control. Name: Joseph Rinehart, MD. Contribution: This author helped in manuscript preparation. Attestation: Joseph Rinehart approved the final manuscript. Conflicts of Interest: Joseph Rinehart is co-owner and coinventor of U.S. patent serial no. 61/432,081 for an intelligent, patient-adaptive, and case-based learning closed-loop fluid administration system based on the dynamic predictors of fluid responsiveness. This manuscript was handled by: Maxime Cannesson, MD, PhD.
No takes yet. Share an insight, caveat, or question.
Liu et al. (2015) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: