Teletherapeutic drug administration by long distance closed-loop control of propofol
Br J Anaesth 2007;98:189-95
Ihmsen H, Naguib K, Schneider G, Schwilden H, Schüttler J, Kochs E
Purpose The purpose of this pilot study was to explore the performance characteristics of a propofol infusion system controlled via computer from a distance of approximately 120 miles.
Background Advancements in telecommunication technology have made medical care possible even when the providers and patients are separated by long distance, such as during oceanic or space travel. Although robotically-assisted ‘telesurgery’ has been demonstrated, there have been no published reports of ‘teleanesthesia.’
Methodology A high speed fiberoptic virtual private network (VPN) with the capability for data transmission rates up to 1000 Mbits/s connected investigators at two sites in Germany with physician anaesthetists present at both sites. At the patient’s location, a 1-channel monitor recorded EEG data and transmitted the information via computer to the distant location where a second computer analyzed incoming signals, computed the median edge frequency (MEF) from 8 second epochs of EEG data, derived propofol infusion rates using an adaptive feedback control algorithm, and then transmitted control information to a propofol infusion pump at the patient location. Updated instructions were sent to the infusion pump from the control site computer every 8 seconds. In the event of interrupted communication between sites, propofol infusion could be continued by the local computer based on the last valid information transmitted, or the patient-site anaesthetist could assume manual control of the propofol administration. Information about the patient’s status and surgical progress was communicated between anaesthetists using a text message box.
Eleven adult patients, ASA I or II, scheduled for elective general surgical procedures were recruited to participate in the investigation. General anesthesia was initiated in a traditional fashion in the institution’s induction room, using boluses of propofol, sufentanil, and atracurium. After tracheal intubation, anesthesia was continued with propofol, sevoflurane, or desflurane while the patient was transferred to the operating room and EEG electrodes were applied. Once computer communication between the patient and distant control sites was established, propofol infusion was initiated as a target controlled infusion and administration of the volatile anesthetics was discontinued. Following surgical incision, the propofol infusion was converted from target control to closed-loop control with propofol infusion rates adjusted automatically by the computer at the distant site to maintain a MEF between 1.5 and 2 Hz. Intraoperative sufentanil and atracurium were administered as necessary at the discretion of the anesthetist. Approximately 20 minutes prior to the end of surgery, propofol infusion control reverted from closed-loop back to target control. The propofol infusion was discontinued when the surgical procedure was completed. Variables measured included standard clinical data (vital signs, hemoglobin saturation, and end-tidal CO2), performance error estimates (differences between the actual and desired MEF values), and amount of data lost (difference between data stored on the patient location computer versus control site data).
Result Control of propofol administration via Internet was achieved in all 11 cases. Infusion times ranged from 49 to 338 minutes, with an average time of 133 min. Infusion control was incomplete during 3 cases (27%). Infusions were discontinued after 1 hour in 2 cases due to a software malfunction, and in 1 case because of an interruption in the Internet connection. Distant site closed-loop control of propofol administration was achieved during 65% of all anesthesia time. The remaining 35% of the time included induction and emergence phases, time periods when the EEG signal was interrupted by electrocautery interference, or the infusion was target controlled with the infusion parameters determined at the distant site and transmitted to the infusion pump via Internet.
Vital signs, hemoglobin saturation, and end-tidal CO2 remained stable. The mean (standard deviation) MEF was 1.8 (.2) Hz during closed-loop control and average BIS value was 42 (2). Total propofol delivered to patients during closed-loop control averaged 5.9 (1.4) mg/kg/hr. Emergence time to extubation ranged from 5 to 10 min with a mean of 7.5 min. Intra-patient EEG fluctuations (wobble) averaged 18.8% for MEF and 6.2% for BIS. Of the 10,905 epochs of EEG data transmitted, only 5 (0.05%) epochs were lost.
Conclusion Long distance administration of anesthetic drugs is possible. Although it may not become routine, the teletherapeutic system explored in this investigation illustrates the potential for remote control of infusions in unique situations.
At an AANA Annual Meeting in the mid-1990s, Executive Director John Garde encouraged all CRNAs to read Daniel Burrus’s book, Technotrends.1 Burrus uses the metaphor of a card game to develop a series of axioms about the future. Among these axioms are “What can be dreamed can be done” and “Once a card is in the deck, it will be played.” It is through the Technotrends lens, that I have watched the progression of technological advances in health care with great curiosity.
A key surgical milestone was Operation Lindbergh, the first successful demonstration of transoceanic telesurgery, as surgeons operating from New York performed a laparoscopic cholecystectomy on a patient hospitalized in France.2 This surgical feat has not yet been matched by anesthesia providers. Cone and colleagues reported the use of various communication methods to “telemonitor” patients and “telementor” Ecuadorian anesthesiologists during surgical procedures, but the actual administration of anesthesia was not controlled from the United States.3,4
Ihmsen et al have moved the dream of teleanesthesia one step closer to reality. The report of this pilot investigation was candid: distant control of anesthesia was limited only to the maintenance phase of the anesthetics and needed to be discontinued in 3 of 11 procedures. However, when the process worked, it worked well. Patients were remarkably stable, although physiological stability is not too surprising because bedside computer-controlled administration of propofol has been demonstrated previously. Limitations of the investigators’ approach were noted, although with more subtlety. For example, the use of a virtual private network to transmit data at speeds up to 1000 Mbits/s far exceeds current commonly-available capabilities, so matching the communications speed reported by the investigators would require a dedicated private network or access to the Abilene Network (through Internet2). The authors speculate that their system might be useful in isolated locations; however, it would seem unlikely that most remote locations have the necessary state-of-the-art fiberoptic Internet capability. Wireless communication, as a possible alternative, introduces a new set of problems, such as data security. An interesting consequence of the investigators’ approach to teleanesthesia is that manpower requirements doubled because of the need for anesthesia providers at both the control and patient locations. If teleanesthesia becomes commonplace, then perhaps it will become the frontier for discussions about implementation of the anesthesia care team, direction of anesthesia care, and reimbursement for anesthesia services.
Daniel Burrus advises us that today’s dreams are tomorrow’s realities. Unquestionably computer and communications technologies will continue to influence the science of anesthesia, but how these advancements will affect day-to-day clinical practice is less clear. The challenge becomes how we can best use technological achievements to improve the delivery of, and access to, high quality anesthesia care. Burrus’s other axioms encourage readers to overcome the complacency of past successes and “re-become” experts. I wonder what tools of the trade we CRNAs will need to become future experts.
Alfred E. Lupien, Ph.D., CRNA
1. Burrus D. Technotrends: How to Use Technology to Go Beyond Your Competition. New York, NY:Collins; 1994.
2. Marescaux J, Leroy J, Gagner M, et al. Transatlantic robot-assisted telesurgery. Nature 2001;413:379-380.
3. Cone SW, Gehr L, Hummel R, Rafiq A, Doarn CR, Merrell RC. Case report of remote anesthetic monitoring using telemedicine. Anesth Analg 2004;98:386-8.
4. Cone SW, Gehr L, Hummel R, Merrell RC. Remote anesthetic monitoring using satellite telecommunications and the Internet. Anesth Analg 2006;102:1463-7.
AORN Journal’s Editor-in-Chief Dr. Nancy Girard’s recently published editorial on how perioperative nursing will be the influenced of technology Girard NJ. Science fiction comes to the OR. AORN J 2007;86:351-3. The editorial is also available on-line at: http://download.journals.elsevierhealth.com/pdfs/journals/0001-2092/PIIS000120920700508X.pdf.
An interesting project to follow is the HAPsMRT (High Altitude Platforms for Mobile Robotic Telesurgery) in which an unmanned airborne vehicle (UAV) is used as a communications link between an operating surgeon and remotely-located patient. For more information, readers can search on-line using “HAPsMRT” as a keyword or refer to a press release at the following URL: http://www.auvsi.org/media/pressreleases/2006/aerovironment.pdf.