When addressing an urgent plea as presented here, first one has to explain what is an independent holistic anaesthesia delivery system (IHADS), and second why there is such a need for it. To get started, here is a description of the IHADS, which does not yet exist as a term but is the most concise summarising title that I can imagine. An IHADS is a computerised drug delivery system that controls actuators, which in turn may be any kind of anaesthetic drug delivery unit such as syringe pumps, vapourisers, injectors, and so on. All delivered anaesthetic drugs (and here I emphasise that this includes hypnotics, sedatives, tranquilisers, analgesics and supplemental medication, but not muscle relaxants) have to be incorporated into the system. At the core of the ‘system’ is a central computer that contains the pharmacokinetic profiles of all these drugs and is able to calculate in real-time their individual plasma and target concentrations along with their mutual interactions. The term ‘holistic’ stands for ‘all-encompassing’, and, therefore, a holistic drug delivery system would simultaneously control and monitor the application of all involved anaesthetic drugs, which act at the same time in a patient. Instead of what we do now, namely, having separately running state-of-the-art computerised infusion pumps which have to be individually controlled, an IHADS would consist of a central processor (computer) unit that communicates with several drug delivery units (such as infusion pumps, vapourisers; Fig. 1).
By working in such a holistic anaesthesia system, the infusion pumps are actuators only, in cybernetics called ‘slaves’. Such a system would be indeed holistic if it is able to deal with all kinds of anaesthetic drugs that have either a sedative/hypnotic or an analgesic effect, independently of their route of administration. Therefore, the system should be able to deal with all intravenous anaesthetics, as well as all inhaled agents. The basis of the calculations is the pharmacokinetic profile of the involved drugs, as well as knowledge of their mutual interaction. The most advanced existing system of this kind is the Draeger SmartPilot View (Drägerwerk AG, Lübeck, Germany).1 However, this system is implemented in Draeger's own anaesthesia architecture and cannot be viewed as independent. Some additional limitations of this existing system are discussed below.
In the first instance, the system must know which drug is being delivered at which dosage or rate. This data input can be provided in three different ways:
1. Direct control of injection pumps that are part of, or are connected to, the system. This would be the main feature of the device: a group of injection pumps that are controlled by a central computer. The pumps must not have an inbuilt processor; they should only report to the steering unit information on the inserted syringe type and size, the resistance in the infusion system and the position of the plunger.
2. Manual input of drugs directly delivered by the user by choosing from a dropdown menu or by entering data via a keyboard. This would be suitable for drugs that have been given as a bolus directly by the user. However, it would be preferable to avoid manual interference and to administer even small volume bolus doses by electronically controlled syringe pumps.
3. Data from external delivery units such as vapourisers or resulting drug levels from analysers such as end-expiratory volatile agent concentrations.
The resulting plasma-site and effect-site levels of all involved drugs should be calculated and monitored over time and displayed in a manner such that their interaction is comprehensively visible to the user. Thus, the core part of the system, which is a computer bestowed with suitable software, represents both a display of the ongoing pharmacological process and the interface with which at least certain parts of the drug delivery (in particular, the injection pumps) can be controlled and manipulated. It is evident that the system initially needs to be informed about all the characteristics of the patient that are relevant to perform the pharmacokinetic calculations. In an advanced system, these data might be automatically retrieved via a network connection from the patient data management system (PDMS) of the hospital. In an ideal configuration, the system is always informed about the drug delivery systems running and the history of each delivered drug. Additionally, it is also able to display the future plasma-site and effect-site drug level courses, as they would develop if the dosage remains unchanged.2,3
Being a computer-based drug delivery system, it would be easy to update software with improved or augmented data, to add new drugs and to modify existing algorithms according to changes in knowledge about their properties. Such a system is also expandable to acquire inputs from various monitoring devices and to generate output orders to various drug delivery systems. In any case, an IHADS would be very flexible and could be operated over a long period of time, even though the technical environment might change. For example, existing or future input from depth of anaesthesia or analgesia intensity measurement devices could be incorporated into the display of anaesthetic activity. The control of delivery and measurement of neuromuscular blockade is, at this early stage, not as urgent as the hypnotic/analgesic medication, but at a later stage, a separate ‘channel’ could be assigned to this.
As soon as such a system becomes available in clinical practice, one could include a partial or totally closed loop control of drug delivery, provided that precise and reliable monitoring of the underlying measurements is available in real-time. However, the existing monitoring of the depth of hypnosis and even more so of the intensity of analgesia is still not reliable enough to leave these inputs alone to supervise and to modify drug dosages. Therefore, closed loop configurations cannot yet be recommended, and instead we should leave the anaesthetist in charge to decide which pattern of hypnosis/analgesia interaction he chooses to be administered.
A user-friendly display of the actual running drug dosages would encompass at least two essential windows (Fig. 2). One would be a dynamic chart indicating the three relevant drug levels (for example the target, plasma and effect-site levels) as well as the flow generated by the delivery unit (the pump infusion rate or vapour concentration setting) over time. This kind of display is very similar to that implemented in the minuscule monitors of infusion pumps or in the popular pharmacokinetic software ‘TIVAtrainer’.4 These multiple representations of simultaneously running drugs are arranged above each other, so that they all have the same time frame on the horizontal axis (that can be stretched or compressed according to the user's needs). The actual time is indicated by a vertical line that divides them into the ‘past’ (left of the line) and ‘future’ (right of the line). The ‘future’ – which should appear in somewhat paler colours – predicts the drug ‘levels’ as they would appear if the dosage remained unchanged. The representation of the main hypnotic agent should also contain a colour-coded segment of dosages that show the resulting interaction zone caused by the concomitant additionally delivered analgesic drugs, as is implemented in the ‘TIVAtrainer’.
The second part of the screen should be dedicated to a display that I call ‘hypnosis–analgesia interaction matrix’ (in short ‘matrix’), which is a two-dimensional co-ordinate system with the x-axis indicating the total hypnotic effect (of all active hypnotic drugs) and the y-axis indicating the total analgesic effect (of all active analgesic drugs). It is a well known fact that there is a wide range of possible combinations of these two components that lead to acceptable general anaesthesia; from an anaesthetic with deep hypnotic and weak analgesic actions to the opposite, an anaesthetic with superficial hypnotic and strong analgesic action. With the current usual dosage regimens, only a few anaesthetists are aware of the actual configuration of their anaesthetic dosage in this respect, not only because this circumstance might be ignored, but also because of the unavailability of a quantitative anaesthesia steering system that accurately indicates the actual state of the interaction. By increasing the analgesic component, one can reduce the hypnotic one and vice versa, without changing the overall ‘anaesthetic depth’. Very few anaesthetists have a clear understanding where in the wide range of this hypnotic/analgesic spectrum of interactions their anaesthetic is, and if they make dosage changes in which direction it is heading. Nevertheless, a growing concern for this problem is seen in studies in which a close view of this two-dimensional drug interaction is presented.5 The scaling and the units of the two axes have yet to be determined, for example by using appropriate surrogate parameters. For the overall analgesic effect, this can be represented, for example, by morphine or, even better, by remifentanil equivalents (expressed in ng ml−1 effect-site concentration). For the overall sedative-hypnotic effect, one might employ propofol equivalents (expressed in μg ml−1 effect-site concentration). It is important that all relevant drug effects are summarised in these two dimensions. When drugs with both properties, such as ketamine, are used their hypnotic and analgesic components should be proportionally adjusted. Within the surface between the two axes, a flashing point indicates the actual position of the ‘holistic’ anaesthetic activity. The left low corner at the intersection of the two axes (the level zero) indicates a completely awake state that also has no analgesic action. Increasing sedative/hypnotic drug activity moves the point upwards in the matrix, whereas increasing analgesic drug activity drags the point to the right. As a consequence, a point position far up and to the right shows a deep combined hypnotic/analgesic state, which is characteristic of profound anaesthesia, while all possible other hypnosis/analgesia combinations would be instantly visible by recognising the location of the point. A dotted curve in the matrix delineates the probable threshold to respiratory depression, at which one would expect apnoea if the flashing point is further right and above it. To implement a ‘historical’ third dimension, the path that has been passed by the point in a certain period of time might be displayed as a track on the surface of the matrix. This track might fade by going back in time or might show a colour change according to the time that has passed. If a change in dosage has been made, an arrow emerging from the flashing point indicates its impending move towards a new location. The speed of the change can be coded in the size or colour intensity of the arrow. Within the surface of the matrix, one could include spots that identify zones for certain surgical interventions that usually need a specific combination of hypnotic and analgesic activity.
Another very useful feature of this matrix would be if it could be used as a two-way interface between user and drug delivery. One way would be to show the matrix position as the result of the actual drug actions and the other one would be to enable the user to point to a new position for the flashing point on the matrix where he would like it to be (for example by ‘clicking’ there with the computer mouse). This move would induce respective output changes in the delivery units, resulting in modifications of the drug dosages till the flashing point arrives at the desired new position. The speed and dynamics of the dose change should be variable and set by the user.
The question is whether such a system is really necessary, and as this article is intended to be a ‘plea’, it is clear that in my opinion it is. Not only are all necessary components for an IHADS already available, it is curious that they have not yet been assembled a long time ago. As a natural development in general technology, as soon as the theoretical and technological prerequisites for a novel technique become available and are recognised as such, it is only a matter of time until a new invention or development appears on the scene. Our recent understanding for the combination of hypnotic and analgesic drug activity has been augmented considerably since potent and controllable agents of both kinds have been available (in particular, propofol and remifentanil). In addition, the mutual interaction of these drugs for general anaesthesia or analgesia-sedation has become clearer to the general anaesthesiologist, and there is an increasing awareness that these qualities can be quantified and used accordingly via a controlling unit that depicts their prevailing interactions.
The theoretical and technical prerequisites for such an IHADS are all available. Infusion pumps that can be controlled by a central computer and thus would act as ‘slaves’ are on the market. Meanwhile, such pumps automatically recognise the inserted syringes by their type, size and filling volume. The infusion pumps not only follow delivery orders from a central processor, but also report their actual activity, plunger pressure and residual volumes. With this, warnings for obstructed infusion lines as well as impending refills of the syringe can be issued on time. The pharmacokinetic algorithms for most intravenous anaesthetics and analgesics, which are at the core of the IHADS, were determined more than a decade ago and are freely available. An advanced drug level calculation and display is available (for example, the TIVAtrainer software) which also calculates hypnotic/analgesic drug interactions and projects the interaction zone onto the course of the main hypnotic agent. With this we already have the first half of the monitor screen display described above. A similar display, as suggested above, for the hypnosis–analgesia interaction matrix is, in principle, already implemented in the SmartPilot View by Draeger, as an isobologram for propofol activity and a surrogate analgesic drug action calculated in remifentanil equivalents. However, to be a truly holistic system, both the hypnotics and the analgesic components of all drugs have to be included into the calculation. Most importantly, these effects have to be summarised with suitable surrogates for a overall hypnotic activity as well as the overall analgesic level, for example as the total hypnotic activity of propofol that is administered together with a benzodiazepine or an inhaled anaesthetic agent. Otherwise drugs that have both properties, such as ketamine, will need to be broken down into their hypnotic and analgesic activities, in order to be calculated separately.
The pharmacokinetic profile of inhaled anaesthetics is well known and has been included in simulation software such as ‘GasMan’,6 although interactions with other drugs are not implemented there. To my knowledge, appropriate algorithms for all possible interactions with these drugs have not yet been calculated or published. A somewhat limited system that controls up to two pumps is available with the ‘Orchestra’ anaesthesia drug delivery system (Fresenius SE, Bad Homburg, Germany). However, this device is based on independent pumps that are connected to a central controlling system, thus resembling primitive creatures with a developed peripheral nervous system but an undersized brain. In particular, this system is unable to accept information about a manually administered drug dose and cannot process data from other drug delivery systems such as vapourisers or drug level monitors such as gas analysers. In contrast, the proposed IHADS concept is based on a powerful central controlling system with dependent delivery systems arranged around it. The strict centralisation of the control unit makes repeated programming of several single delivery units unnecessary. Naturally, all actuators that deliver anaesthetic drugs have to be connected to the central control system by standardised interfaces that enable communication in both directions. Only such an overall controlling unit would be able to acquire various data from multiple sources and process all available information to calculate total hypnotic and analgesic drug activities and interactions.
To summarise, this plea contains both the urgent expression of the need for a comprehensive anaesthesia delivery system that encompasses all available pharmacological knowledge about currently used anaesthetic drugs, that should be delivered by a central control and command unit, with a display of their mutual interactions, as well as my naive astonishment that such equipment is not yet available, although al the prerequisites have been present for a long time. Maybe it is inappropriate to expect this editorial to move the industry in the desired direction, but at least it might ignite a discussion among anaesthesiologists of the necessity and potential for such a device.
Assistance with the editorial: Thomas Schnider, Kantonsspital St Gallen, Switzerland, assisted with the editorial.
Financial support and sponsorship: none declared.
Conflicts of interest: the author has no commercial relationship to companies working in the field of drug delivery systems. However, over the last decade, he has had many discussions with representatives of such companies to gauge their ability and interest in developing an independent holistic anaesthesia delivery system, not yet leading to substantial results.
Comment from the Editor: this editorial was checked and accepted by the editors, but was not sent for external peer-review.
1. Ostrovsky G. SmartPilot View: an anesthesiologist's flight simulator. Med Gadget 2010 (http://medgadget.com/2010/01/smartpilot_view_an_anesthesiologists_flight_simulator.html)
2. Kennedy R. Seeing the future of anesthesia drug dosing: moving the art of anesthesia from impressionism to realism. Anesth Analg 2010; 111:252–255.
3. Struys MM, De Smet T, Mortier EP. Simulated drug administration: an emerging tool for teaching clinical pharmacology during anesthesiology training. Clin Pharmacol Ther 2008; 84:170–174.
4. Enlund M. TCI: target controlled infusion, or totally confused infusion? Call for an optimised population based pharmacokinetic model for propofol. Ups J Med Sci 2008; 113:161–170.
5. Christensen RE, Gholami AS, Reynolds PI, Malviya S. Anaesthetic management and outcomes after noncardiac surgery in patients with hypoplastic left heart syndrome: a retrospective review. Eur J Anaesthesiol 2012; 29:425–430.
6. Van Zundert T, Hendrickx J, Brebels A, et al. Effect of the mode of administration of inhaled anaesthetics on the interpretation of the F(A)/F(I) curve: a GasMan simulation. Anaesth Intensive Care 2010; 38:76–81.