From Invention to Innovation: Bringing Perioperative Physiological Closed-Loop Systems to the Bedside.

From the Food and Drug Administration (FDA) report a public workshop the FDA held on October 13 and 14, 2015, regarding Physiological Closed-Loop Controlled Devices. The material they are reporting on is derived from the white paper they prepared for this workshop.2 This event was widely attended by various organizations, including the American Society of Anesthesiologist, the Department of Defense Office of Naval Research, academia, industry representatives, and the FDA. The significance of this event was that closed-loop systems used in anesthesia and the critical care setting are an emerging technology that impacts the future of anesthesia practice. The article summarizes regulatory considerations for closed-loop critical care devices for readers in the areas of critical care medicine, anesthesiology, systems engineering, physiological modeling, and human factors. As an executive editor of Anesthesia & Analgesia since 2012, I am honored that the FDA decided to submit this very important article to our Journal. Anesthesia & Analgesia is the official journal of the International Anesthesia Research Society (IARS), whose mission statement is “to encourage, stimulate, and fund ongoing anesthesia-related research projects that will enhance and advance the specialty, and to disseminate current, state-of-the art, basic and clinical research data in all areas of clinical anesthesia, including perioperative medicine, critical care, and pain management. The IARS is focused solely on the advancement and support of education and scientific research related to anesthesiology.” As such, our Journal has played a pivotal role in the publication of several high-level research manuscripts dedicated to physiological closed-loop systems in perioperative settings. A simple search on PubMed using the search query: (anesth analg and “closed loop”) provides interesting insights regarding the development of this topic in our profession. The oldest reference was published in 1994, reporting on the interaction of rocuronium (Org 9426) with etomidate, fentanyl, midazolam, propofol, thiopental, and isoflurane using closed-loop feedback control of rocuronium infusion.3 Twenty-nine references were found between 1994 and today, encompassing all fields of physiological closed-loop controlled systems in anesthesiology: muscle relaxation,4 pharmacological modeling,5,6 reinforcement learning techniques,7,8 large clinical studies evaluating closed-loop coadministration of remifentanil and propofol based on the bispectral index monitor,9 fluid management and hemodynamic optimization systems,10,11 arterial blood pressure optimization using closed-loop phenylephrine administration,12,13 closed-loop neonatal oxygen therapy,14 use of closed-loop anesthesia systems during high-risk procedures such as cardiac surgery15 or moderate sedation for transcatheter aortic valve implantation,16 and a multicenter trial in India evaluating closed-loop anesthesia management versus standard of care.17 Of these 29 articles, 9 have been published between 1994 and 2012, and the remaining 20 have been published since 2012. Anesthesia & Analgesia has been a leader in this field and is especially proud to be able to publish this report from the FDA in this month’s issue of the Journal. What did we learn regarding the development of these systems over the past 10 years? First, closed-loop anesthesia is a polarizing topic: people either love it or hate it. There is a strong push back from some anesthesiologists, who argue that these systems could hurt patients and take over our jobs.18 However, this remains an oversimplification of what anesthesiologists perform well on a daily basis. These systems can only be accepted if they improve the workflow and quality of care. They also need to be developed by clinicians who are the only one to know how these systems can best benefit our patients. The goal of these devices is not to replace anesthesiologists, but to help us focus on higher levels of care to improve quality. Second, developing these systems requires a specific set of engineering skills and a very thorough approach in technological development. Contrary to popular belief, computers may crash and thus do not always do what they have been asked to do. Closed-loop softwares need to be tested extensively in silico before being used in vivo in animals and subsequently in patients. Robustness and stability have to be evaluated for all of these softwares, and any change made to the computer code before or after use in patients will be considered a new device.