Issue 26, December 2008 Targeting Anesthesia Care

Computer-Assisted Intravenous Anesthesia: From Theory to Clinical Practice

Correspondence: Prof. L. Barvais MD, PhD, Free University of Brussels, Brussels, Belgium . (E-mail and other contact info can be obtained from CWWJ’s Editor-in-Chief).

Key Words: Computer, Intravenous Anesthesia, Target Controlled Infusions.
Running title: Computer-Assisted Intravenous Anesthesia.

Clinical Window Web Journal #26: Computer-Assisted Intravenous Anesthesia: From Theory to Clinical Practice (December 2008). ISSN 1795–6269.

Introduction

Total intravenous anesthesia (TIVA) uses intravenous (IV) agents exclusively to ensure general anesthesia. It allows rapid, smooth induction and arousal by independent titration of unconsciousness and analgesia. Furthermore, modern TIVA targets hemodynamic stability, and if needed, muscle relaxation.

Since contemporary anesthesia has achieved a very high level of safety, additional improvements may be achieved by optimizing anesthetic drug administration, adapting and titrating it against individual anesthetic and analgesic needs. This may be achieved by the interaction and integration of latest-generation, fast-acting IV agents, new drug delivery systems targeting plasma and/or effect-site concentrations, and new monitors for depth of anesthesia.

Target-controlled infusion in anesthesia: Basics

Since first described by Schüttler and colleagues in 1983, target-controlled infusion (TCI) delivering systems have become widespread, and different systems have been commercialized [1]. Indeed, TCI drug administration offers several advantages compared to manual infusion: better control of anesthetic depth, quicker recovery, and less hemodynamic instability [2]. Current TCI of anesthetic agents relies on real-time simulation of drug plasma concentrations, based on conventional or population pharmacokinetic (PK) models. The conventional models propose mean pharmacokinetic parameters of a given drug, whereas population PK sets propose adjusted pharmacokinetic parameters for some of the most common components of variability, such as age, gender, and/or lean body weight. After the anesthesiologist has selected a given target plasma concentration, a computer calculates the bolus and the infusion rate necessary to obtain and to maintain the theoretical calculated target, continually adjusting the pump-infusion rate.

Despite an approximate 30 percent inaccuracy of calculated-target concentration compared to blood-drug measurements (due to interindividual variability and/or model performance), TCI allows more precise titration to a given clinical effect, as it makes it easier to achieve steady-state drug-blood concentrations. By contrast, manual adjustment of the drug continuous-infusion rates result in more unstable drug concentrations. Knowledge of plasma concentration is essential, as it represents the driving force toward the drug effect-site, i.e., the central nervous system for anesthetics and analgesics.

However, the clinical effect of a drug not only depends on the concentration gradient between the plasma and the effect-site, but is also influenced by drug characteristics such as protein binding, non-ionized fraction, and liposolubility on membrane permeability and drug-receptor interactions. To further optimize TCI’s performance, it is thus necessary to integrate the drug’s effect-site pharmacodynamics (PD) to compute the target. As the central nervous system is the major target (or effect-site) of anesthetic drugs, integration of PK and PD data utilizing information derived from EEG and clinical surrogates of brain function allows for better definition of the population PK/PD for TCI systems.

Effect-site TCI systems may more accurately predict the time course to clinical effect and may shorten the time to reach the drug’s peak effect by allowing plasma concentrations first to rise above the target [3]. Moreover, TCI not only offers stable plasma and effect-site concentrations of anesthetics and analgesics, it can also predict duration to reach the new level, when decreasing the target drug concentration, thus allowing better “landing” conditions at the end of surgery.

Finally, a synergistic interaction between opioids and hypnotics has been described and some studies have quantified these interactions to allow better understanding of drug combinations in different contexts (e.g., induction, laryngoscopy, skin incision, recovery) [4,5]. These studies offer information for improved use of TCI.

TCI systems: From theory to clinical practice

Several pharmacokinetic and pharmacodynamic model-driven infusion software tools have been developed during the last two decades. Initially reserved for strict research purposes, TCI has entered everyday clinical practice as software-driven infusion systems became commercially available. We use software developed at the Free University of Brussels called Infusion ToolBox (ITB). This experimental, in-house product controls and monitors several intravenous drug-infusion pumps, using a built-in set of predefined pharmacokinetic and pharmacodynamic models [6].

Several departments and agencies collaborated in the development of ITB, including the Department of Computer Science, Faculty of Medicine, Free University of Brussels and the Department of Anesthesiology of the Erasme University Hospital. Since its first clinical use in the early nineties, its developers have adapted ITB in attempts to offer optimal drug delivery and patient safety. The software has evolved over time from a simple infusion tool to a more comprehensive system: by including a computer-monitor interface, ITB now collects patient vital signs, enabling permanent feedback control of drug-infusion rates based on predefined algorithms.

TCI has become everyday anesthesia practice at the Erasme University Hospital, and since 2005 we have 15 workstations (Figure 1) available for residents and senior anesthesiologists. In addition to a basic default TCI mode, advanced users may program special sessions, with personalized sets of drugs, PK/PD models, and monitoring tools.

Users at our institution find ITB a user-friendly graphical interface. Prior to drug delivery, the anesthesiologist enters patient data and selects the drugs to be used. By default, ITB proposes propofol and remifentanil, using PK/PD sets established by Schnider and Minto, respectively [7,8]. Numerous other hypnotic/opioid combinations may be chosen, as well as corresponding drug dilutions and population models. For each drug selected, ITB proposes a dynamic control panel: after the setting and confirmation of the desired target concentration, the software calculates and controls the infusion rate required to achieve and maintain the target. ITB then displays the calculated plasma and/or effect-site target concentration numerically and graphically (Figure 2).

The most recent versions of ITB now propose default induction values based on patient characteristics and ASA physical status score, as well as type of surgery. Target concentration is adjusted to the individual patient; the use of present-generation, rapid- and short-acting drugs allows close titration to clinical response, much like halogenated gases.

The use of simplified neuromonitors using processed EEG, (e.g., Bispectral index or Entropy), has become widespread in recent years, improving individualized titration of anesthetic depth. Electroencephalographic monitoring has been shown to reduce drug consumption, thus causing less hemodynamic instability and faster arousal, while reducing perioperative awareness. Typically, we increase the hypnotic target concentration (mostly propofol) by small increments during induction, allowing plasma and/or effect-site concentrations to stabilize before any further increase. This slow, stepwise titration technique is particularly appropriate for frail and/or elderly patients and assures hemodynamic tolerance during induction and maintenance. After loss of consciousness, the target is adjusted according to the physician’s clinical judgment and clinical neuromonitoring.

Analgesic requirements (usually remifentanil or sufentanil) are adjusted depending on the anticipated or actual nociceptive stimulus and effect.
Finally, all TCI-relevant data are recorded during the session, and results can be displayed at the end of surgery in a tabular or graphical view (Figure 3). We have used the various generations of ITB in clinical practice for over ten years. For example, in 2005 more than 5000 patients were routinely anesthetized with this system.

TCI systems: Future developments

The recent introduction of a new monitor/ITB interface, called DataLogger, opens new possibilities in both experimental and clinical anesthesia at the Erasme University Hospital. The DataLogger continuously collects online data from patient monitors, allowing a record of all vital parameters. We anticipate continued development of automated closed-loop control of the target infusions by the ITB software.

The DataLogger module now allows us to introduce the techniques of data mining and predictive modeling in anesthetic research. Briefly, data mining aims to identify patterns and establish relations to create a statistical model of future behavior. We recently launched a multicenter study designed to sift through the ITB database for patient characteristics, infusion schemes, and vital parameters to analyze trends and differences in our practices. By identifying certain predictors common to a given type of patient, surgery, and anesthesia, predictive analytics could indeed help us define an even more individualized infusion scheme for each patient.

The connection between the TCI system and the patient monitor may also enable our team to develop experimental closed-loop systems, allowing the anesthesiologist to switch the infusion system to autopilot mode. Liu et al. [10] recently implemented a BIS based closed-loop system in our ITB software. They showed that compared to manual control, automated TCI control of consciousness using BIS was not only feasible, but it also outperformed manual control in terms of adequate anesthesia maintenance [10]. More complex algorithms taking into account more than one clinical parameter and respecting the synergistic action between drugs will now be developed at our institution.

A leap into the future

Computer-assisted TCI infusion has become available for everyday anesthesia delivery and allows effect-site titration similar to volatile anesthesia, not only for hypnotics but also for opioids. Clinical research now focuses on on-line interaction between monitoring and drug-delivery systems, allowing the development of semi-automated drug delivery computer systems requiring the anesthetist confirmation or even closed-loop systems of drug delivery. Though benefits in terms of morbidity and mortality will be difficult to establish, we believe that future computer-assisted anesthesia practices will allow the anesthesiologist to refocus attention on patient care, such as airway fiberoscopy, transesophageal echocardiography, and hemodynamic management. We believe that computer-assisted anesthesia offers one way to implement more safety and efficacy in everyday anesthetic practice. We find ITB easy to use and fully flexible as a TCI tool, and it helps us understand and improve our way of thinking and managing IV anesthesia delivery.

References

[1] Schuttler J, Schwilden H, Stoekel H. Pharmacokinetics as applied to total intravenous anesthesia. Practical implications. Anesthesia 1983; 38 Suppl: 53–56.

[2] Struys MM, De Smet T, Versichelen LF, Van De Velde S, Van den Broecke R, Mortier EP. Comparison of closed loop controlled administration of propofol using a bispectral index as the controlled variable versus "standard practice" controlled administration, Anesthesiology 2001; 95: 6–17.

[3] Struys MM, De Smet T, Depoorter B, Versichelen LF, Mortier EP, Dumortier FJ, Shafer SL, Rolly G. Comparison of plasma compartment versus two methods for effect compartment target-controlled infusion for propofol. Anesthesiology 2000; 92: 399–406.

[4] Vuyk J, Mertens M, Olofsen E, Burm A, Bovill J. Propofol anaesthesia and rational opioid selection. Determination of optimal EC50-EC95 propofol-opioid concentrations that assure adequate anesthesia and a rapid return to consciousness. Anesthesiology 1997 Dec; 87(6): 1549–62.

[5] Bouillon T, Bruhn J, Radulescu L, Andresen C, Shafer T, Cohane C, Shafer S Pharmacodynamic Interaction between Propofol and Remifentanil Regarding Hypnosis, Tolerance of Laryngoscopy, Bispectral Index, And Electroencephalographic Approximate Entropy. Anesthesiology. 2004 Jun; 100(6):1353–72.

[6] Cantraine FR, Coussaert EJ. The first object oriented monitor for intravenous anesthesia. J Clin Monit Comput. 2000; 16(1): 3–10.

[7] Schnider TW, Minto CF, Gambus PL, Andersen C, Goodale DB, Shafer SL, Youngs EJ. The influence of method of administration and covariates on the pharmacokinetics of propofol in adult volunteers. Anesthesiology 1998; 88: 1170-1182.

Schnider TW, Minto CF, Shafer SL, Gambus PL, Andersen C, Goodale DB, Youngs EJ. The influence of age on propofol pharmacodynamics. Anesthesiology 1999; 90:1502–1516.

[8] Minto CF, Schnider TW, Shafer SL.. Pharmacokinetics and pharmacodynamics of remifentanil. II. Model application. Anesthesiology. 1997 Jan;86(1):24–33.

[9] Gepts E, Shafer SL, Camu F, Stanski DR, Woestenborghs R, Van Peer A, Heykants JJ. Linearity of pharmacokinetics and model estimation of sufentanil. Anesthesiology 1995;83:1194-1204.

[10] Liu N, Chazot T, Genty A, Landais A, Restoux A, McGee K, Laloe PA, Trillat B, Barvais L, Fishler M. Titration of propofol for anesthetic induction and maintenance guided by the bispectral index: closed-loop versus manual control. Anesthesiology 2006; 104: 686-95.

Clinical Window Web Journal #26: Computer-Assisted Intravenous Anesthesia: From Theory to Clinical Practice (December 2008). ISSN 1795–6269.

© 2006-2008 GE Healthcare Finland Oy doing business as GE Healthcare. All rights reserved. The copyright, any and all trademarks and trade names and other intellectual property rights subsisting in or used in connection with and related to this publication are, unless another owner is specified, the property of GE Healthcare. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of GE Healthcare.

Last updated: 30 December 2008
Legal notice	© GE Healthcare 2008 ISSN 1795-6269 (Web) ISSN 1795-6277 (CD)	Webmaster