Secondary Logo

Journal Logo

Technology, Computing, and Simulation: Review Article

Closed-Loop Systems in Anesthesia

Is There a Potential for Closed-Loop Fluid Management and Hemodynamic Optimization?

Rinehart, Joseph, MD*; Liu, Ngai, MD, PhD; Alexander, Brenton, MS*; Cannesson, Maxime, MD, PhD*

Author Information
doi: 10.1213/ANE.0b013e318230e9e0
  • Free
  • SDC
  • Take the CME Test

Although few people have heard or understand the term “closed-loop control,” it is ubiquitous in modern life. A closed loop is simply a system wherein a controller monitors one or more system variables and adjusts one or more interventions to control that system in response1 (Fig. 1). A prime example is central air-conditioning: the main unit pumps cool air throughout the house, the temperature is monitored by sensors in the home, and these data are then compared with the settings on the thermostat and used to increase or decrease the cooling so that the temperature hovers near the set point. Although it would be possible to manually adjust the flow of cold air, the odds are that you would often overshoot or undershoot your desired temperature and the system would require many adjustments every hour to even modestly approximate the performance that an automated controller easily achieves. Other common examples are the cruise-control systems in automobiles, electric ovens, and the operation of elevators. Moreover, virtually every system in living organisms is under physiologic closed-loop control. Anesthesiologists themselves constitute closed-loop controllers (Fig. 1): patient output is monitored in the form of vital signs and this information is used to tailor interventions by the practitioner. In control engineering terminology, this arrangement constitutes a closed-loop control system with a human controller in the loop. One noteworthy feature of this particular arrangement is that the attention and actions of the human controller are intermittent.

Figure 1
Figure 1:
Generic closed-loop control scheme. The controller is givenreference settings, which are then compared with the feedback data (the system output). The difference between the two is the error and this is used to generate a new system input. In the case of a Bispectral Index (BIS)-propofol controller, the output would be the measured patient BIS level, the input would be the rate of propofol administration, and the reference setting would be the target BIS.

The benefits of a well-designed automated controller are obvious: automation of previously manual tasks and improvement in stability and accuracy of the controlled variables. Furthermore, a closed-loop system is not distractible from its task and the implemented algorithm is perfectly repeatable, meaning it can be tested and steadily improved upon in practice. A medical example of this behavior is the action of implantable pacemakers and defibrillators, which run for years making decisions without the need for manual validation.

With the benefits of closed-loop systems being what they are, one might question why there are not more used in anesthesia. Obstacles are numerous but include the complexity of biologic systems, uncertainty in measurements and clinical care, regulatory approval and safety, and acceptance by practitioners.24 The most important element of an effective closed-loop system is the accuracy and validity of the variable(s) monitored by the system; without good feedback, the controller cannot respond appropriately to changes. Nevertheless, this does not mean that closed-loop systems are unrealistic or even impractical goals, especially with the rapid advances in monitoring and computing power that we are now witnessing.

Recent improvements in hemodynamic monitoring, and in particular the dynamic variables predictive of fluid responsiveness, suggest that closed-loop management of fluid resuscitation in trauma, operating rooms, and intensive care settings is rapidly becoming an achievable goal. In this article, we review the literature on the predictive dynamic variables of fluid responsiveness, the value of and challenges facing closed-loop controllers in clinical care, and finally suggest approaches in which closed-loop fluid management controllers might be implemented and what features they will likely need to be considered safe, robust, and effective.


Figure 2 illustrates schematically 2 of the most frequently used control methods in clinical devices. Table 1 lists definitions of the most important vocabulary used in control engineering. The type of control method used will have a significant impact on the reliability, benefits, and limitations of the loop, and in large part is dependent on the characteristics of the target system. In the following sections, we describe the most frequently used types of control in health care systems: proportional-integral-derivative (PID), model-based, rule-based, artificial neural networks (ANNs), and fuzzy logic controllers.

Figure 2
Figure 2:
Examples of controller types. A, Classic proportional-integral-derivative (PID) controller. The calculated error is sent to the controller where the proportional (present), integral (past), and derivative (future) error values are combined to create a new input to the system. The PID controller is one of the most widely used controllers in engineering applications. B, Representation of an artificial neural network. One or more network inputs are received and processed through a series of “nodes” in hidden layers, ultimately resulting in one or more network output signals. In a recurrent neural network, nodes in hidden layers are also able to feedback to nodes in the same or earlier layers.
Table 1
Table 1:
Definitions of Engineering Control Terminology

The most generic and widely used controller type is the PID controller (Fig. 2A). Differences between the output value and the set point are calculated and used to adjust the input value. Each of the 3 terms in the name accounts for a different component of the error: the present (proportional), past (integral), and future (derivative) error values. To prevent system oscillations, tuning must take place to adjust the individual gains of each of these 3 components.57 PID controllers are often used as the sole control method in clinical applications, particularly in linear systems, but can also be used in combination with other controller types that have further advantages over PID control. In anesthesiology, PID control has frequently been used to control depth of hypnosis.810

The model-based approach monitors how accurately a designed model predicts the observed response and uses that information to improve its future processes. Training of the system with simulated data (reference model) or real-time responses (system identification) is required to establish a model's precision before it can be used in clinical situations.1113 In medical systems, the model itself is normally based on a combination of established physiologic variables (i.e., pharmacokinetics, circulatory responses, and target site concentrations). Such models help to account for interpatient variability and improve robustness of the controller. Therefore, when a clinical scenario frequently exhibits large differences among patients, a model-based approach is typically advantageous over an exclusively PID-based controller. To improve model parameters from a standard population-based response, a Bayesian approach can also be used to optimize individual parameters within the system.14 Model-based approaches have frequently been used for control of neuromuscular blockade11,15,16 and control of hypnosis.1720

The rule-based controller implements a set of user-defined rules analyzing variables of interest to organize and execute specific controller actions. These rules are generally in an “if-then” format and can be used to interpret quantitative data, qualitative data, or a combination of the two. The manual nature of this design allows for a high level of specification and control, which can be valuable in complex systems. Obviously, a rule-based controller is only as effective as the governing rule set and requires extensive testing for stability, but again these controller types are often combined with other control schemes. Combined approaches like these have been used successfully for hemodynamic management21 and depth of hypnosis.22

ANNs are another approach gaining popularity (Fig. 2B). The basis of this approach is to mimic a biologic neural network: there are multiple layers of “neurons” (nodes with individual functions and weights), which manipulate input data into an appropriate output signal. A learning strategy is required to set the various functions and weights to produce a network that correctly interprets specific data.23 These learning strategies can be automated or supervised, depending on the desired function of the network, but once optimized, these systems can be very robust.24,25 An ANN is frequently used when there are multiple inputs and outputs and when a function needs to be inferred from observed data. Simulation and animal studies have demonstrated the feasibility of control of arterial blood pressure (in 6 rabbits and simulation) and sevoflurane concentration using ANNs.24,26,27 The clinical application of ANNs in humans is limited to 7 anesthetized patients,26 thus the robustness of these controllers in medical care awaits further clinical validation.28

Uncertainty and gray areas are both common in medical diagnosis and treatment.29,30 Some form of probabilistic decision-making will be required in most systems to account for this fact. “Fuzzy logic,” as Zadeh31 termed it in 1965, is one mathematical model that allows computer programs to work with uncertainty. Instead of allowing for only the logical values true and false, degrees of truth are allowed. Linguistically, we are used to dealing with these partial truths; clinicians have little trouble understanding what is meant by “his blood pressure is a little low.” Fuzzy logic allows this same approach to be used in computer systems, and because of the utility of this approach, there has been an exponential growth in the medical and bioinformatics literature on fuzzy logic in the past two decades.32

One drawback to fuzzy logic is that because of the relation to linguistic thinking, there is a degree of arbitrariness that goes into the design of the controller. Even when experts are used in the design process, there may be some disagreement about set boundaries and significance (for example, what constitutes “significant” hypoxia, a saturation of 90%, 88%, or 86%?). For this reason, fuzzy sets, once designed, need to be rigorously tested to verify their appropriateness across a range of clinical scenarios.3


Automated systems offer clear benefits to clinicians. The first, touched on in the introductory text, is the freeing of limited human resources. In medical settings such as surgery, trauma, and intensive care, this typically means the time and attention of the highly trained specialists providing patient care, a scarce resource in the best of times, which becomes less available as the criticality and intensity of the scenario increases.33 By shifting some of the decision-making to automated systems, particularly systems that have been proven to be effective and reliable, the attention of providers can be focused on other aspects of care.3436

Additionally, autonomous devices can be used in areas where immediate access to experts may not be available. If a closed-loop controller is shown to be as effective as a clinician in its management of a particular variable, it becomes a sort of portable expert. It can then be deployed in combat zones, remote rural locations, or even outer space if need be, and can act in place of experts until they become available.33,37,38 This does not mean automated systems are not without limitations, however. A brief review of some of the described systems will elucidate the benefits and possibilities of automated systems as well as the challenges faced in their implementation.

Neuromuscular Blockade

One requirement of automated controllers is a reliable measure of the system's desired effect. Neuromuscular blockade is a natural target for closed-loop controllers because the effects of these drugs are relatively easily measured by noninvasive monitors. In 1976, Cass et al.39 reported on the use of computer-controlled infusions of d-tubocurarine, gallamine, alcuronium, and pancuronium in sheep. After that, the first closed-loop muscle-relaxant infusion systems were reported on beginning in the mid-1980s,15,4046 with later comparison studies showing a steadier level of neuromuscular blockade than was achieved with intermittent bolus or manually set infusion.45 Later systems became more sophisticated, including adaptive elements that could self-adjust in response to patient-specific factors.4749 On the whole, use of closed-loop control of muscle relaxants for surgery has been shown to provide “stable surgical operating conditions over a wide range of patient sensitivities while infusing the minimum amount of drug.”50

It should be noted that closed-loop controllers have less utility when overdosing of the controlled treatment carries little risk and the effect can be rapidly reversed. For this reason, the recent availability of drugs such as cyclodextrin that can immediately reverse deep residual blockade now limit the clinical interest of muscle-relaxant closed-loop control.51

Insulin Therapy

The chronic nature of diabetes mellitus and the need for continuous real-time adjustment of insulin levels in response to fluctuating glucose levels has made insulin therapy another natural target for closed-loop control systems. Computer-assisted dose calculations and modeling began in the 1980s along with early attempts at closed-loop control.5258 Continuous wearable closed-loop infusion pumps were also first tested in this era.59,60 Although truly portable, closed-loop, glucose-sensing insulin infusion pumps are not yet widely available, there is rapid progress in this area as glucose monitoring continues to become more sophisticated.6165

The safety record of computer-assisted and computer-controlled insulin infusions is very good, with modern systems consistently reporting tighter glycemic control with fewer complications and less hypoglycemia than standard strict protocols.6669 The direct relationship between insulin and glucose levels has been partly responsible for the success of these systems despite the interpersonal and even interdose response variability encountered. The chief difficulty in creating continuous portable closed-loop insulin management systems is that reliable measurement of blood glucose level still requires invasive testing.7074 This may soon be overcome; a subcutaneous enzymatic sensor is now clinically available, for example, and has been successfully used in a closed-loop insulin controller.73

Sedation and Anesthesia

Automated administration of anesthetics began in the United States. Mayo and Bickford developed a prototype automatic anesthetic delivery system that regulated the administration of ether or thiopental using electroencephalographic (EEG) activity. This research group first successfully closed the loop on 50 human patients back in 1950!75 Closed-loop research continued at the University of Utah during the 1980s and focused on control of respiration and inhaled volatile anesthetics.7678 Moreover, closed-loop systems for nitroprusside infusions, muscle relaxants, and even the first closed-loop fluid-infusion system based on urine output were reported.15,7986 These early systems, as well as other closed-loop anesthetic delivery systems developed around the same time, were found to be safe and to reliably deliver the desired concentration of anesthetic over long time spans.76,87 Different inputs have been used to control anesthetic drugs. Examples include median EEG activity for methohexital administration17 and auditory evoked potentials for propofol.88 Interest in closed-loop control was renewed with the introduction of the Bispectral Index (BIS®) monitor, which allowed for the continuous monitoring of electrocortical activity in the operating room. Table 2 summarizes studies that included the BIS monitor in the closed-loop controller. Since 1998, approximately 400 patients have been anesthetized using a BIS-propofol closed-loop PID model (83%), rule-based (15%), or neural network controller (2%), demonstrating the feasibility of automated propofol administration during induction and maintenance of general anesthesia.8,10,18,8993 The BIS monitor has also been used to automate delivery of isoflurane,94,95 and still other EEG monitors can be used to control depth of hypnosis.96

Table 2
Table 2:
Summary of Studies Using Bispectral Index–Propofol Closed-Loop Controllers for Anesthesia Management

These systems are effective but do have limitations because the target return variables are only moderately accurate surrogates for depth of anesthesia.97,98 Even given this limitation, modern closed-loop IV anesthesia infusion systems have been able to outperform anesthesiologists in randomized clinical trials.8,10,22,93,94

The selection of the controlled variable remains difficult when discussing automated administration of an opioid. Mean arterial blood pressure and heart rate were previously used to control alfentanil administration,99,100 but electrocortical activity has also been used to control alfentanil94,101 and remifentanil.93,102 Recently, the Sedasys® system was made commercially available in Australia and Canada.103 This controller allows the automated titration of propofol and oxygen using hemodynamic and oxygen saturation variables during endoscopic procedures. Overall, the automated titration of propofol has been shown to be safe in patients with severe comorbidities,89,91 to improve hemodynamic stability,91 and to decrease the time to tracheal extubation8,93 without increasing the occurrence of adverse events. Curiously, the only commercial automated propofol administration system currently in use is designed for nonanesthesia providers.103

Ventilation and Oxygenation

Ventilator management is an especially advanced area showing great promise for closed-loop systems. The first automatic control of ventilation was described in 1957 by Saxton and Myers,104 who used a servo to control negative pressure ventilation. Beginning with the description by Hewlett et al.105 of “mandatory minute ventilation” in 1977, multiple adaptive modes of ventilation and fraction of inspired oxygen titration have been proposed that rely on closed-loop controllers for their operation.106111 A particular focus area of closed-loop ventilators is weaning, with clinical studies showing shorter times to extubation when these modes are used.112,113 Some authors have suggested that closed-loop ventilator management has advanced to the point where it may be an emerging standard of care.114,115

Other Systems

Closed-loop systems have been reported for many other clinical applications: induction of labor with Pitocin,116,117 nitroprusside in hypertensive crises,82,83,118 vasopressive drugs,119 phenylephrine during spinal anesthesia for cesarean delivery,120 postoperative autotransfusion,121 and even intraaortic balloon pump function.122


Goals of Fluid Management and Hemodynamic Optimization

As stated by Arthur Guyton: “The primary function of circulation is to supply body tissues with nutrients, to eliminate waste products, to transfer hormones between different body parts, and, in general to maintain “homeostasis”—an appropriate tissue environment for optimal cell function and survival.”123,124 To achieve this goal, two physiologic objectives are required: adequate perfusion pressure to drive blood into organ capillaries, and adequate cardiac output (CO) to deliver oxygen.123,124 A variety of studies have demonstrated that meeting these goals by CO maximization during high-risk surgery results in an improved postoperative outcome with concomitant savings on costs.125130 Although the ideal goal of fluid administration would be to perfectly match the oxygen delivery to the oxygen demand (optimization), the lack of clinically reliable oxygen uptake monitoring131 makes CO maximization the most logical approach until such monitoring becomes available.

Historical Approaches to Hemodynamic Optimization

That hypovolemia induces hypotension, oliguria, and tachycardia is obvious. These signs are inadequate in themselves, however, because they only appear in severe hypovolemia132 and are not specific.133135 Invasive central venous pressure (CVP) and pulmonary capillary wedge pressure have been used for years to monitor patients' intravascular volume status. However, the assumption that CVP and pulmonary capillary wedge pressure reflect ventricular preload (or preload dependence) is erroneous, and studies have failed to demonstrate any accuracy of these variables for predicting the effects of intravascular volume expansion on CO,133,135137 or to improve patients' outcomes.138145

Studies focusing directly on CO optimization use a CO monitor (often esophageal Doppler) to titrate fluid administration to the point at which CO does not further increase after intravascular volume expansion. Reports have shown the clinical benefits of this approach.146,147 Very few centers have adopted goal-directed fluid administration protocols in their daily clinical practice, however, likely because of the traditional difficulty in monitoring CO. The clinical “gold standard” for CO measurement is still an invasive pulmonary artery catheter with intermittent thermodilution.148151

Recent Advances in Perioperative Hemodynamic Monitoring and Fluid Management

What anesthesiologists and/or intensivists want to know before they administer fluids is: “will my patient's CO increase in response to intravascular volume expansion?” In other words, is the patient preload dependent? Instead of simply monitoring a given hemodynamic variable, functional hemodynamic monitoring addresses this question by assessing how the system responds to an induced change. In mechanically ventilated patients under general anesthesia or moderate sedation, the effects of positive pressure ventilation on preload and stroke volume can be used to detect fluid responsiveness: if mechanical ventilation induces respiratory variations in stroke volume or in arterial pulse pressure variation (PPV), it is more likely that the patient is preload dependent and CO will increase if fluid is given. Figure 3 shows the Frank-Starling relationship with the associated arterial waveform tracings, and Figure 4 shows how PPV is calculated.

Figure 3
Figure 3:
Frank-Starling relationship with corresponding respiratory variations in the arterial pressure waveform. High respiratory variation in the arterial pressure or in the plethysmographic waveforms indicates that the patient is on the steep portion of the Frank-Starling relationship. Low variation indicates that the patient is on the plateau. Cardiac output maximization concepts aim at increasing cardiac output until it reaches the plateau of the Frank-Starling relationship. This goal, as demonstrated on this figure, could be achieved by minimizing respiratory variation in arterial pressure or plethysmographic waveform.
Figure 4
Figure 4:
Pulse pressure variation (PPV) and systolic pressure variation (SPV) calculation from the arterial pressure waveform. PPmax and PPmin = maximum and minimum arterial pulse pressure over a single respiratory cycle. PPV is then calculated as: (PPmax − Ppmin)/[(PPmax + PPmin)/2].

The functional hemodynamic variables have been developed and validated in clinical practice over the last 4 decades. Respiratory variations in the arterial pressure were found to be related to patients' fluid status, and systolic pressure variation (SPV) is frequently >10 mm Hg higher in hypovolemic patients compared with normovolemic patients.152 In 1983, Coyle et al.153 described the Delta up and Delta down components of SPV in an abstract, but this work was never released as a full paper. In 1987, Perel et al.132 showed that SPV was related to the intravascular volume status in an animal model, that it was an early detector of hypovolemia, and that it reacted earlier than CVP. Several studies were then able to demonstrate that SPV was an accurate predictor of fluid responsiveness in adult patients undergoing surgery154 and in patients treated in the intensive care unit (ICU).155 More recently, PPV has been shown to be superior to SPV in mechanically ventilated patients with septic shock.156

In the same time period, noninvasive assessment of fluid responsiveness based on the analysis of respiratory variations using the plethysmographic waveform was investigated. Respiratory variations in the plethysmographic waveform have been shown to be related to the patient's fluid status,157 and there is a strong relationship between SPV and respiratory variations in the peak of the plethysmographic waveform before and after hemorrhage.158 In 2007, a study showed that the respiratory variations of the plethysmographic waveform amplitude were able to predict fluid responsiveness in mechanically ventilated patients.159 A multitude of other studies have been published demonstrating that the invasively and noninvasively assessed functional hemodynamic variables are useful for the prediction of fluid responsiveness in an operating room and an ICU setting, even after considering their potential limitations.160165

These dynamic variables have several limitations. First, patients are required to be sedated and their lungs mechanically ventilated. Studies have thus far failed to demonstrate that these variables can predict fluid responsiveness in spontaneously breathing patients.166168 Moreover, tidal volume affects the predictive value of the variables. A minimal tidal volume of 8 mL/kg of body weight is necessary.169,170 Patients must be in sinus rhythm,135 the chest and pericardium must be closed,171 and intraabdominal pressure should be within normal ranges.172 Apart from these cardiopulmonary limitations that are common to any dynamic variable of fluid responsiveness, variables derived from the plethysmographic waveform have additional limitations. The plethysmographic waveform analysis is limited by vasomotor tone, which strongly affects the waveform.173,174 Thus, this technique may only be used during profound general anesthesia159,175,176 and it seems less stable in the ICU setting.177

One advantage of the functional variables of fluid responsiveness is that they can be derived from a noninvasive arterial pressure waveform or from the plethysmographic waveform135,178 (Video 1, see Supplemental Digital Content 1, Recently, several new methods have been developed to automatically and continuously calculate these indices.178184 These continuous measures make optimization strategies that incorporate them feasible,178 and studies suggest that this approach has the ability to improve postoperative outcomes.128,185188

From Goal-Directed Fluid Therapy to Closed-Loop Hemodynamic Management

Closed-loop fluid management is in its infancy. Previously, urine output and mean systemic filling pressures were both used as feedback variables for automated fluid administration systems,79,189 but beyond the initial study, these systems were not further reported on. More recently, in 2008, Kramer et al.190192 published a series of articles evaluating different aspects of closed-loop resuscitation using arterial blood pressure, CO, or skeletal muscle oxygenation as the feedback variable. Using a PID algorithm, they were able to show more stable urine output rates than were achieved with manual hourly adjustments.193

Despite the modest successes of these studies, closed-loop fluid resuscitation has been hampered by the limited quality of measures such as urine output and arterial blood pressure in predicting fluid responsiveness (Table 3). The proliferation of noninvasive CO monitors, along with the development and validation of the dynamic predictors of fluid responsiveness, however, has made closed-loop fluid management truly practical. Because invasive monitors are no longer necessary, CO monitoring can be performed continuously across a broad range of patients. The evidence supporting goal-directed fluid management is strong enough that its routine use should be adopted in moderate- to high-risk patients during both surgery and in the ICU.128,130,178,185,194196

Table 3
Table 3:
Advantages and Limitations of Potential Inputs for a Closed-Loop Fluid Management System

The dynamic predictors of fluid responsiveness have made it possible to predict in advance of a bolus whether the volume is likely to improve CO (i.e., whether a patient is preload dependent). Moreover, they also indicate the degree of preload dependence present in a given patient, providing further invaluable information about intravascular volume status. Thus, despite the limitations in the clinical use of these measures (for example, the requirement for mechanical ventilation), they are nevertheless essential measures for automated systems to use in guiding fluid therapy and could be used in up to 40% of all patients undergoing anesthesia.197 The possible advantage of automated fluid administration is even more pronounced in ICU settings where physician and nursing attention is often spread over the entire unit; a system capable of intelligently and continuously maintaining an optimal CO could have a profound impact on outcomes. Simulation data comparing a novel model- and rule-based closed-loop fluid management system with anesthesiologist management was recently presented showing a higher and steadier CO in the closed-loop group.198

As a first step toward clinical application of closed-loop systems, decision support systems may be developed and implemented. Although decision support algorithms also standardize care and make outcome-based analysis more practical, widespread adoption of these protocols is challenging. Clinical protocols require time and often significant effort with frequent interventions from caregivers; thus, the time spent in applying protocols will most likely be the main barrier to implementation. Closed-loop systems can provide a bridge to the acceptance of clinical protocols by performing the many required tasks of the protocol in a less demanding manner.

Despite the potential utility, there are myriad challenges facing the clinical application of closed-loop fluid management. The first and most obvious consideration with all automated systems is safety. Closed loops must be “fail safe”; if they malfunction or receive bad data, they must behave in a way that does not harm the patient. In the case of fluid administration, this means not overloading (or severely underresuscitating) the patient. Similarly, these systems must be tolerant of the artifact and error in clinical monitors and again must be capable of filtering this noise without jeopardizing patient care.173 Food and Drug Administration approval and clinician acceptance may also prove difficult.

An important point is that these systems are not meant to run in the absence of a supervising clinician, although the general public and indeed many physicians may believe that these devices will “take over.” It cannot be overstated that this is not the case. Closed-loop systems are meant to be supervised by experts while helping to implement clinical protocols and standardize patient care. In addition, they may help bring the level of safety provided by anesthesiologists to an outside location where no physicians are available, but they will still require oversight by a trained specialist in these settings to be operated safely.

There are challenges with monitoring devices themselves, principally with the fidelity of monitors in measuring the target variable and the ability of the closed-loop controller to account for the uncertainty involved in all monitoring devices. Additionally, there will be tremendous interpatient variability in a given set of monitored values because of disease states, cardiac and pulmonary status, and even baseline physiology. Similarly, a physiologic model may not be applicable to or consider all possible circumstances a controller might encounter. For this reason, a robust and specific measure of the effect is necessary to allow for interpatient or intrapatient variability along with extensive testing in simulation and supervised clinical trials.

Infusion pumps used to administer fluids will have to be precise and will need several layers of fail-safes for use with a closed-loop system should the system malfunction or become disconnected from the pumps. The closed loop itself needs to be intelligent enough to detect situations in which it cannot handle fluid and hemodynamic management on its own. Of course, the supervisor will always have the ability to “open the loop” and take over for the algorithm, providing an important additional layer of protection.

Another challenge may be the complexity of hemodynamic physiology. Unlike insulin systems that target glucose level, or anesthetic closed loops that target BIS, hemodynamic management does not lend itself to management using only a single feedback variable. An ideal closed-loop system for fluid management will use both CO and the dynamic predictors of fluid responsiveness to tailor fluid therapy, along with heart rate and arterial blood pressure at the very minimum. These challenges will, in all probability, require a sophisticated controller algorithm to manage safely.


Anesthesiologists are often compared to airline pilots: induction, maintenance, and emergence from anesthetics have many parallels with the takeoff, cruising, and landing of aircraft. In aviation, automatic pilots, although unquestionably very sophisticated, are nothing more than closed-loop systems that use data garnered from myriad sensors throughout the aircraft to control engine power and flight controls to provide a safe, stable flight to the destination. Likewise, closed-loop controllers can act as copilots in clinical settings, providing safe and stable control of one or more aspects of care under the supervision of a clinician.

With the development and refinement of the dynamic predictors of fluid responsiveness, fluid management can now be included with hypnosis, ventilation, and a host of other applications for which closed-loop control is feasible. In time, these individual systems will likely be combined into comprehensive controllers capable of fully integrated anesthesia management, with depth of anesthesia, ventilation, temperature, fluid management, and more, all controlled by a single overarching system. This arrangement will allow for cooperation and coordinated action among the various individual components.

We postulate that the future of anesthesia will rely on two separate but interdependent technologies: the sensors and monitors on one side, and the controllers on the other. A great deal of work remains to be done, for example: controller design, clinical application, and outcomes-based research. Nevertheless, if we can build automated systems to reliably keep 350,000-pound aircraft in the air for hours at a time, there should be little doubt we can design automated systems to safely and reliably manage complicated aspects of medical care.


Name: Joseph Rinehart, MD.

Conflict of Interest: Dr. Rinehart is coowner and coinventor of US patent serial no. 61/432,081 for intelligent, patient-adaptive, and case-based learning closed-loop fluid administration system based on the dynamic predictors of fluid responsiveness.

Attestation: Dr. Rinehart participated in manuscript drafting and final approval of the manuscript.

Name: Ngai Liu, MD, PhD.

Conflict of Interest: Dr. Liu is a patent holder in France for the gain constants and control algorithm for a closed-loop anesthesia management system (no. BFF08P669, Institut National de la Propriété Industrielle, France).

Attestation: Dr. Liu participated in manuscript drafting and final approval of the manuscript.

Name: Brenton Alexander, MS.

Conflicts of Interest: Dr. Alexander has no conflicts of interest to declare.

Attestation: Dr. Alexander participated in manuscript drafting and final approval of the manuscript.

Name: Maxime Cannesson, MD, PhD.

Conflict of Interest: Dr. Cannesson is a consultant for Fresenius Kabi, Edwards Lifesciences, Masimo Corp., Covidien, ConMed, BMeye, CNSystem, and Philips Medical Systems. Dr. Cannesson is coowner and coinventor of US patent serial no. 61/432,081 for intelligent, patient-adaptive, and case-based learning closed-loop fluid administration system based on the dynamic predictors of fluid responsiveness. Dr. Cannesson is inventor and coowner of US patent appl. no. 13/095,827 for assessing fluid responsiveness using the EKG waveform analysis.

Attestation: Dr. Cannesson participated in manuscript drafting and final approval of the manuscript.

This manuscript was handled by: Dwayne R. Westenskow, PhD.


Video 1.

Demonstration of the evolution in arterial pressure variation and hemodynamic variables before and after intravascular volume expansion in an illustrative patient.


1. Ogata K. Modern Control Engineering. 5th ed. Upper Saddle River: Pearson Education, Inc., 2010
2. Manberg PJ, Vozella CM, Kelley SD. Regulatory challenges facing closed-loop anesthetic drug infusion devices. Clin Pharmacol Ther 2008;84:166–9
3. Jaulent MC, Joyaux C, Colombet I, Gillois P, Degoulet P, Chatellier G. Modeling uncertainty in computerized guidelines using fuzzy logic. Proc AMIA Symp 2001:284–8
4. Ellwein LB, Eckman MH. Biologic complexity in mathematical modeling. Med Decis Making 1989;9:38–9
5. Malwatkar GM, Sonawane SH, Waghmare LM. Tuning PID controllers for higher-order oscillatory systems with improved performance. ISA Trans 2009;48:347–53
6. Kumar Padhy P, Majhi S. Improved automatic tuning of PID controller for stable processes. ISA Trans 2009;48:423–7
7. Harinath E, Mann GK. Design and tuning of standard additive model based fuzzy PID controllers for multivariable process systems. IEEE Trans Syst Man Cybern B Cybern 2008;38:667–74
8. Liu N, Chazot T, Genty A, Landais A, Restoux A, McGee K, Laloe PA, Trillat B, Barvais L, Fischler M. Titration of propofol for anesthetic induction and maintenance guided by the bispectral index: closed-loop versus manual control: a prospective, randomized, multicenter study. Anesthesiology 2006;104:686–95
9. Absalom AR, Sutcliffe N, Kenny GN. Closed-loop control of anesthesia using bispectral index: performance assessment in patients undergoing major orthopedic surgery under combined general and regional anesthesia. Anesthesiology 2002;96:67–73
10. Puri GD, Kumar B, Aveek J. Closed-loop anaesthesia delivery system (CLADS) using bispectral index: a performance assessment study. Anaesth Intensive Care 2007;35:357–62
11. Schumacher PM, Stadler KS, Wirz R, Leibundgut D, Pfister CA, Zbinden AM. Model-based control of neuromuscular block using mivacurium: design and clinical verification. Eur J Anaesthesiol 2006;23:691–9
12. Martinoni EP, Pfister ChA, Stadler KS, Schumacher PM, Leibundgut D, Bouillon T, Bohlen T, Zbinden AM. Model-based control of mechanical ventilation: design and clinical validation. Br J Anaesth 2004;92:800–7
13. Sieber TJ, Frei CW, Derighetti M, Feigenwinter P, Leibundgut D, Zbinden AM. Model-based automatic feedback control versus human control of end-tidal isoflurane concentration using low-flow anaesthesia. Br J Anaesth 2000;85:818–25
14. De Smet T, Struys MM, Greenwald S, Mortier EP, Shafer SL. Estimation of optimal modeling weights for a Bayesian-based closed-loop system for propofol administration using the bispectral index as a controlled variable: a simulation study. Anesth Analg 2007;105:1629–38
15. Jaklitsch RR, Westenskow DR. A model-based self-adjusting two-phase controller for vecuronium-induced muscle relaxation during anesthesia. IEEE Trans Biomed Eng 1987;34:583–94
16. Olkkola KT, Schwilden H, Apffelstaedt C. Model-based adaptive closed-loop feedback control of atracurium-induced neuromuscular blockade. Acta Anaesthesiol Scand 1991;35:420–3
17. Schwilden H, Schuttler J, Stoeckel H. Closed-loop feedback-control of methohexital anesthesia by quantitative EEG analysis in humans. Anesthesiology 1987;67:341–7
18. De Smet T, Struys MM, Neckebroek MM, Van den Hauwe K, Bonte S, Mortier EP. The accuracy and clinical feasibility of a new Bayesian-based closed-loop control system for propofol administration using the bispectral index as a controlled variable. Anesth Analg 2008;107:1200–10
19. 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 Bispectral Index as the controlled variable versus “standard practice” controlled administration. Anesthesiology 2001;95:6–17
20. Mortier E, Struys M, De Smet T, Versichelen L, Rolly G. Closed-loop controlled administration of propofol using bispectral analysis. Anaesthesia 1998;53:749–54
21. Held CM, Roy RJ. Multiple drug hemodynamic control by means of a supervisory-fuzzy rule-based adaptive control system: validation on a model. IEEE Trans Biomed Eng 1995;42:371–85
22. Hemmerling TM, Charabati S, Zaouter C, Minardi C, Mathieu PA. A randomized controlled trial demonstrates that a novel closed-loop propofol system performs better hypnosis control than manual administration. Can J Anaesth 2010;57:725–35
23. Bullinaria JA, Riddell PM. Neural network control systems that learn to perform appropriately. Int J Neural Syst 2001;11:79–88
24. Kashihara K, Kawada T, Uemura K, Sugimachi M, Sunagawa K. Adaptive predictive control of arterial blood pressure based on a neural network during acute hypotension. Ann Biomed Eng 2004;32:1365–83
25. Kamangar FA, Behbehani K. An artificial neural network-based controller for the control of induced paralysis using vecuronium bromide. Ann Biomed Eng 1997;25:1040–51
26. Haddad WM, Bailey JM, Hayakawa T, Hovakimyan N. Neural network adaptive output feedback control for intensive care unit sedation and intraoperative anesthesia. IEEE Trans Neural Netw 2007;18:1049–66
27. Tosun M, Ferikoglu A, Gunturkun R, Unal C. Control of sevoflurane anesthetic agent via neural network using electroencephalogram signals during anesthesia. J Med Syst 2010 . [Epub ahead of print]
28. Moore BL, Doufas AG, Pyeatt LD. Reinforcement learning: a novel method for optimal control of propofol-induced hypnosis. Anesth Analg 2011;112:360–7
29. John RI, Innocent PR. Modeling uncertainty in clinical diagnosis using fuzzy logic. IEEE Trans Syst Man Cybern B Cybern 2005;35:1340–50
30. Griffiths F, Green E, Tsouroufli M. The nature of medical evidence and its inherent uncertainty for the clinical consultation: qualitative study. BMJ 2005;330:511
31. Zadeh LA. Fuzzy sets. Inform Control 1965:338–53
32. Torres A, Nieto JJ. Fuzzy logic in medicine and bioinformatics. J Biomed Biotechnol 2006;2006:91908
33. Pauldine R, Beck G, Salinas J, Kaczka DW. Closed-loop strategies for patient care systems. J Trauma 2008;64:S289–94
34. Anderson P. Building on success: academic emergency department leverages clinical automation to reduce overcrowding, streamline workflow and enhance revenue. Health Manag Technol 2005;26:32–4
35. Weinger MB, Herndon OW, Gaba DM. The effect of electronic record keeping and transesophageal echocardiography on task distribution, workload, and vigilance during cardiac anesthesia. Anesthesiology 1997;87:144–55
36. Westenskow DR. Fundamentals of feedback control: PID, fuzzy logic, and neural networks. J Clin Anesth 1997;9:33S–5S
37. Adams BD, Carr B, Raez A, Hunter CJ. Cardiopulmonary resuscitation in the combat hospital and forward operating base: use of automated external defibrillators. Mil Med 2009;174:584–7
38. Terry P, Rushton CH. Allocation of scarce resources: ethical challenges, clinical realities. Am J Crit Care 1996;5:326–30
39. Cass NM, Lampard DG, Brown WA, Coles JR. Computer controlled muscle relaxation: a comparison of four muscle relaxants in the sheep. Anaesth Intensive Care 1976;4:16–22
40. Bradlow HS, Rametti LB, Uys PC, Coetzee WP. Microcomputer-based muscle relaxation monitor and controller for clinical use. Med Biol Eng Comput 1985;23:547–55
41. Asbury AJ, Linkens DA. Clinical automatic control of neuromuscular blockade. Anaesthesia 1986;41:316–20
42. Bradlow HS, Uys PC, Rametti LB. On-line control of atracurium induced muscle relaxation. J Biomed Eng 1986;8:72–5
43. de Vries JW, Ros HH, Booij LH. Infusion of vecuronium controlled by a closed-loop system. Br J Anaesth 1986;58:1100–3
44. Jaklitsch RR, Westenskow DR, Pace NL, Streisand JB, East KA. A comparison of computer-controlled versus manual administration of vecuronium in humans. J Clin Monit 1987;3:269–76
45. Webster NR, Cohen AT. Closed-loop administration of atracurium: steady-state neuromuscular blockade during surgery using a computer controlled closed-loop atracurium infusion. Anaesthesia 1987;42:1085–91
46. Wait CM, Goat VA. Atracurium infusion during paediatric craniofacial surgery: closed loop control of neuromuscular block. Anaesthesia 1989;44:567–70
47. Olkkola KT, Schwilden H. Adaptive closed-loop feedback control of vecuronium-induced neuromuscular relaxation. Eur J Anaesthesiol 1991;8:7–12
48. Assef SJ, Lennon RL, Jones KA, Burke MJ, Behrens TL. A versatile, computer-controlled, closed-loop system for continuous infusion of muscle relaxants. Mayo Clin Proc 1993;68:1074–80
49. Rowaan CJ, Vandenbrom RH, Wierda JM. The Relaxometer: a complete and comprehensive computer-controlled neuromuscular transmission measurement system developed for clinical research on muscle relaxants. J Clin Monit 1993;9:38–44
50. Mason DG, Linkens DA, Edwards ND, Reilly CS. Development of a portable closed-loop atracurium infusion system: systems methodology and safety issues. Int J Clin Monit Comput 1996;13:243–52
51. Sorgenfrei IF, Norrild K, Larsen PB, Stensballe J, Ostergaard D, Prins ME, Viby-Mogensen J. Reversal of rocuronium-induced neuromuscular block by the selective relaxant binding agent sugammadex: a dose-finding and safety study. Anesthesiology 2006;104:667–74
52. Chisholm DJ, Kraegen EW, Bell DJ, Chipps DR. A semi-closed loop computer-assisted insulin infusion system: hospital use for control of diabetes in patients. Med J Aust 1984;141:784–9
53. Chanoch LH, Jovanovic L, Peterson CM. The evaluation of a pocket computer as an aid to insulin dose determination by patients. Diabetes Care 1985;8:172–6
54. Leiberman E, Katz M, Weitzman S, Mazor M, Kalay A, Tiran J. Control of a pregnant diabetic with daily insulin changes by insulin dosage computer (IDC). Life Support Syst 1985;3 Suppl 1:573–7
55. Schiffrin A, Mihic M, Leibel BS, Albisser AM. Computer-assisted insulin dosage adjustment. Diabetes Care 1985;8:545–52
56. Strack T, Bergeler J, Beyer J, Hutten H. Computer assisted conventional insulin therapy. Life Support Syst 1985;3 Suppl 1:568–72
57. Albisser AM, Schiffrin A, Schulz M, Tiran J, Leibel BS. Insulin dosage adjustment using manual methods and computer algorithms: a comparative study. Med Biol Eng Comput 1986;24:577–84
58. Pernick NL, Rodbard D. Personal computer programs to assist with self-monitoring of blood glucose and self-adjustment of insulin dosage. Diabetes Care 1986;9:61–9
59. Shimoda S, Nishida K, Sakakida M, Konno Y, Ichinose K, Uehara M, Nowak T, Shichiri M. Closed-loop subcutaneous insulin infusion algorithm with a short-acting insulin analog for long-term clinical application of a wearable artificial endocrine pancreas. Front Med Biol Eng 1997;8:197–211
60. Shichiri M, Kawamori R, Goriya Y, Yamasaki Y, Nomura M, Asakawa N, Kubota M, Kamada T. Long-term application of wearable artificial endocrine pancreas: closed-loop intravenous vs subcutaneous insulin infusion. Life Support Syst 1985;3 Suppl 1:583–7
61. Mehanovic S, Mujic M. The automatic regulation of the basal dose on the insulin pump for the treatment of patients that have diabetes type 1. Bosn J Basic Med Sci 2010;10:100–6
62. Van Herpe T, De Moor B, Van den Berghe G. Towards closed-loop glycaemic control. Best Pract Res Clin Anaesthesiol 2009;23:69–80
63. El-Khatib FH, Jiang J, Damiano ER. A feasibility study of bihormonal closed-loop blood glucose control using dual subcutaneous infusion of insulin and glucagon in ambulatory diabetic swine. J Diabetes Sci Technol 2009;3:789–803
64. Wilinska ME, Chassin LJ, Hovorka R. In silico testing: impact on the progress of the closed loop insulin infusion for critically ill patients project. J Diabetes Sci Technol 2008;2:417–23
65. El-Khatib FH, Jiang J, Damiano ER. Adaptive closed-loop control provides blood-glucose regulation using dual subcutaneous insulin and glucagon infusion in diabetic swine. J Diabetes Sci Technol 2007;1:181–92
66. Okabayashi T, Maeda H, Sun ZL, Montgomery RA, Nishimori I, Hanazaki K. Perioperative insulin therapy using a closed-loop artificial endocrine pancreas after hepatic resection. World J Gastroenterol 2009;15:4116–21
67. Fort A, Narsinghani U, Bowyer F. Evaluating the safety and efficacy of Glucommander, a computer-based insulin infusion method, in management of diabetic ketoacidosis in children, and comparing its clinical performance with manually titrated insulin infusion. J Pediatr Endocrinol Metab 2009;22:119–25
68. Cavalcanti AB, Silva E, Pereira AJ, Caldeira-Filho M, Almeida FP, Westphal GA, Beims R, Fernandes CC, Correa TD, Gouvea MR, Eluf-Neto J. A randomized controlled trial comparing a computer-assisted insulin infusion protocol with a strict and a conventional protocol for glucose control in critically ill patients. J Crit Care 2009;24:371–8
69. Hovorka R, Allen JM, Elleri D, Chassin LJ, Harris J, Xing D, Kollman C, Hovorka T, Larsen AM, Nodale M, De Palma A, Wilinska ME, Acerini CL, Dunger DB. Manual closed-loop insulin delivery in children and adolescents with type 1 diabetes: a phase 2 randomised crossover trial. Lancet 2010;375:743–51
70. Caduff A, Talary MS, Mueller M, Dewarrat F, Klisic J, Donath M, Heinemann L, Stahel WA. Non-invasive glucose monitoring in patients with type 1 diabetes: a Multisensor system combining sensors for dielectric and optical characterisation of skin. Biosens Bioelectron 2009;24:2778–84
71. Ferrante do Amaral CE, Wolf B. Current development in non-invasive glucose monitoring. Med Eng Phys 2008;30:541–9
72. De Block C, Vertommen J, Manuel-y-Keenoy B, Van Gaal L. Minimally-invasive and non-invasive continuous glucose monitoring systems: indications, advantages, limitations and clinical aspects. Curr Diabetes Rev 2008;4:159–68
73. Renard E, Place J, Cantwell M, Chevassus H, Palerm CC. Closed-loop insulin delivery using a subcutaneous glucose sensor and intraperitoneal insulin delivery: feasibility study testing a new model for the artificial pancreas. Diabetes Care 2010;33:121–7
74. Renard E. Closed-loop insulin delivery: is the Holy Grail near? Lancet 2010;375:702–3
75. Mayo CW, Bickford RG, Faulconer A Jr. Electroencephalographically controlled anesthesia in abdominal surgery. J Am Med Assoc 1950;144:1081–3
76. Westenskow DR, Jordan WS. The Utah system: computer-controlled anesthetic delivery. Contemp Anesth Pract 1984;8:221–33
77. Hayes JK, Westenskow DR, East TD, Jordan WS. Computer-controlled anesthesia delivery system. Med Instrum 1984;18:224–31
78. Westenskow DR, Wallroth CF. Closed-loop control for anesthesia breathing systems. J Clin Monit 1990;6:249–56
79. DeBey RK, Westenskow DR, Jordan WS, McJames SW. A urine based control system for fluid infusion. Biomed Sci Instrum 1987;23:195–8
80. Westenskow DR, Zbinden AM, Thomson DA, Kohler B. Control of end-tidal halothane concentration. Part A. Anaesthesia breathing system and feedback control of gas delivery. Br J Anaesth 1986;58:555–62
81. Westenskow DR. Fundamentals of feedback control applied to microcomputer instrumentation design. Int J Clin Monit Comput 1986;3:239–44
82. Meline LJ, Westenskow DR, Somerville A, Wernick RT, Jacobs J, Pace NL. Evaluation of two adaptive sodium nitroprusside control algorithms. J Clin Monit 1986;2:79–86
83. Meline LJ, Westenskow DR, Pace NL, Bodily MN. Computer-controlled regulation of sodium nitroprusside infusion. Anesth Analg 1985;64:38–42
84. Ohlson KB, Westenskow DR, Jordan WS. A microprocessor based feedback controller for mechanical ventilation. Ann Biomed Eng 1982;10:35–48
85. Westenskow DR, Tucker SM. Evaluation of a ventilation monitor. Crit Care Med 1981;9:64–6
86. Westenskow DR. Control of PaCO2 during mechanical ventilation: monitoring and feedback techniques. Ann Biomed Eng 1981;9:659–67
87. Ritchie RG, Ernst EA, Pate BL, Pearson JD, Sheppard LC. Closed-loop control of an anesthesia delivery system: development and animal testing. IEEE Trans Biomed Eng 1987;34:437–43
88. Kenny GN, Mantzaridis H. Closed-loop control of propofol anaesthesia. Br J Anaesth 1999;83:223–8
89. Liu N, Chazot T, Trillat B, Michel-Cherqui M, Marandon JY, Law-Koune JD, Rives B, Fischler M. Closed-loop control of consciousness during lung transplantation: an observational study. J Cardiothorac Vasc Anesth 2008;22:611–5
90. Liu N, Chazot T, Trillat B, Pirracchio R, Law-Koune JD, Barvais L, Fischler M. Feasibility of closed-loop titration of propofol guided by the Bispectral Index for general anaesthesia induction: a prospective randomized study. Eur J Anaesthesiol 2006;23:465–9
91. Agarwal J, Puri GD, Mathew PJ. Comparison of closed loop vs. manual administration of propofol using the Bispectral Index in cardiac surgery. Acta Anaesthesiol Scand 2009;53:390–7
92. Hegde HV, Puri GD, Kumar B, Behera A. Bi-spectral index guided closed-loop anaesthesia delivery system (CLADS) in pheochromocytoma. J Clin Monit Comput 2009;23:189–96
93. Liu N, Chazot T, Hamada S, Landais A, Boichut N, Dussaussoy C, Trillat B, Beydon L, Samain E, Sessler DI, Fischler M. Closed-loop coadministration of propofol and remifentanil guided by bispectral index: a randomized multicenter study. Anesth Analg 2011;112:546–57
94. Morley A, Derrick J, Mainland P, Lee BB, Short TG. Closed loop control of anaesthesia: an assessment of the bispectral index as the target of control. Anaesthesia 2000;55:953–9
95. Locher S, Stadler KS, Boehlen T, Bouillon T, Leibundgut D, Schumacher PM, Wymann R, Zbinden AM. A new closed-loop control system for isoflurane using bispectral index outperforms manual control. Anesthesiology 2004;101:591–602
96. Bibian S, Dumont GA, Zikov T. Dynamic behavior of BIS, M-entropy and neuroSENSE brain function monitors. J Clin Monit Comput 2011;25:81–7
97. Palanca BJ, Mashour GA, Avidan MS. Processed electroencephalogram in depth of anesthesia monitoring. Curr Opin Anaesthesiol 2009;22:553–9
98. Kent CD, Domino KB. Depth of anesthesia. Curr Opin Anaesthesiol 2009;22:782–7
99. Luginbuhl M, Bieniok C, Leibundgut D, Wymann R, Gentilini A, Schnider TW. Closed-loop control of mean arterial blood pressure during surgery with alfentanil: clinical evaluation of a novel model-based predictive controller. Anesthesiology 2006;105:462–70
100. Hemmerling TM. Automated anesthesia. Curr Opin Anaesthesiol 2009;22:757–63
101. Schwilden H, Stoeckel H. Closed-loop feedback controlled administration of alfentanil during alfentanil-nitrous oxide anaesthesia. Br J Anaesth 1993;70:389–93
102. Mathews DM, Cirullo PM, Struys MM, De Smet T, Malik RJ, Chang CL, Neuman GG. Feasibility study for the administration of remifentanil based on the difference between response entropy and state entropy. Br J Anaesth 2007;98:785–91
103. Pambianco DJ, Vargo JJ, Pruitt RE, Hardi R, Martin JF. Computer-assisted personalized sedation for upper endoscopy and colonoscopy: a comparative, multicenter randomized study. Gastrointest Endosc 2011;73:765–72
104. Saxton GA Jr, Myers G. A servomechanism for automatic regulation of pulmonary ventilation. J Appl Physiol 1957;11:326–8
105. Hewlett AM, Platt AS, Terry VG. Mandatory minute volume: a new concept in weaning from mechanical ventilation. Anaesthesia 1977;32:163–9
106. Dojat M, Pachet F, Guessoum Z, Touchard D, Harf A, Brochard L. NeoGanesh: a working system for the automated control of assisted ventilation in ICUs. Artif Intell Med 1997;11:97–117
107. Jandre FC, Pino AV, Lacorte I, Neves JH, Giannella-Neto A. A closed-loop mechanical ventilation controller with explicit objective functions. IEEE Trans Biomed Eng 2004;51:823–31
108. Spahija J, Beck J, de Marchie M, Comtois A, Sinderby C. Closed-loop control of respiratory drive using pressure-support ventilation: target drive ventilation. Am J Respir Crit Care Med 2005;171:1009–14
109. Tehrani F, Rogers M, Lo T, Malinowski T, Afuwape S, Lum M, Grundl B, Terry M. Closed-loop control of the inspired fraction of oxygen in mechanical ventilation. J Clin Monit Comput 2002;17:367–76
110. Tehrani FT, Roum JH. Flex: a new computerized system for mechanical ventilation. J Clin Monit Comput 2008;22:121–30
111. Younes M. Proportional assist ventilation, a new approach to ventilatory support. Theory. Am Rev Respir Dis 1992;145:114–20
112. Gruber PC, Gomersall CD, Leung P, Joynt GM, Ng SK, Ho KM, Underwood MJ. Randomized controlled trial comparing adaptive-support ventilation with pressure-regulated volume-controlled ventilation with automode in weaning patients after cardiac surgery. Anesthesiology 2008;109:81–7
113. Lellouche F, Mancebo J, Jolliet P, Roeseler J, Schortgen F, Dojat M, Cabello B, Bouadma L, Rodriguez P, Maggiore S, Reynaert M, Mersmann S, Brochard L. A multicenter randomized trial of computer-driven protocolized weaning from mechanical ventilation. Am J Respir Crit Care Med 2006;174:894–900
114. Wysocki M, Brunner JX. Closed-loop ventilation: an emerging standard of care? Crit Care Clin 2007;23:223–40,ix
115. Lellouche F, Brochard L. Advanced closed loops during mechanical ventilation (PAV, NAVA, ASV, SmartCare). Best Pract Res Clin Anaesthesiol 2009;23:81–93
116. Steer PJ, Carter MC, Choong K, Hanson M, Gordon AJ, Pradhan P. A multicentre prospective randomized controlled trial of induction of labour with an automatic closed-loop feedback controlled oxytocin infusion system. Br J Obstet Gynaecol 1985;92:1127–33
117. Arulkumaran S, Ingemarsson I, Ratnam SS. Closed loop automatic infusion system for induction of labour based on external tocography. Asia Oceania J Obstet Gynaecol 1986;12:221–6
118. Hammond JJ, Kirkendall WM, Calfee RV. Hypertensive crisis managed by computer controlled infusion of sodium nitroprusside: a model for the closed loop administration of short acting vasoactive agents. Comput Biomed Res 1979;12:97–108
119. Potter DR, Moyle JT, Lester RJ, Ware RJ. Closed loop control of vasoactive drug infusion: a preliminary report. Anaesthesia 1984;39:670–7
120. Ngan Kee WD, Tam YH, Khaw KS, Ng FF, Critchley LA, Karmakar MK. Closed-loop feedback computer-controlled infusion of phenylephrine for maintaining blood pressure during spinal anaesthesia for caesarean section: a preliminary descriptive study. Anaesthesia 2007;62:1251–6
121. Blankenship HB, Wallace FD, Pacifico AD. Clinical application of closed-loop postoperative autotransfusion. Med Prog Technol 1990;16:89–93
122. Zelano JA, Li JK, Welkowitz W. A closed-loop control scheme for intraaortic balloon pumping. IEEE Trans Biomed Eng 1990;37:182–92
123. Guyton AC, Hall JE. Heart muscle: the heart as a pump and function of the heart valves. In: Guyton AC, Hall JE eds. Textbook of Medical Physiology. 11th ed. Philadelphia: Elsevier/Saunders, 2006:103–15
124. Guyton AC, Hall JE. Overview of the circulation: medical physics of pressure, flow, and resistance. In: Guyton AC, Hall JE eds. Textbook of Medical Physiology. 11th ed. Philadelphia: Elsevier/Saunders, 2006:161–70
125. Gan TJ, Soppitt A, Maroof M, el-Moalem H, Robertson KM, Moretti E, Dwane P, Glass PS. Goal-directed intraoperative fluid administration reduces length of hospital stay after major surgery. Anesthesiology 2002;97:820–6
126. Pearse R, Dawson D, Fawcett J, Rhodes A, Grounds RM, Bennet ED. Early goal-directed therapy after major surgery reduces complications and duration of hospital stay: a randomised, controlled trial [ISRCTN38797445]. Crit Care 2005;9:R687–93
127. Wakeling HG, McFall MR, Jenkins CS, Woods WG, Miles WF, Barclay GR, Fleming SC. Intraoperative oesophageal Doppler guided fluid management shortens postoperative hospital stay after major bowel surgery. Br J Anaesth 2005;95:634–42
128. Mayer J, Boldt J, Mengistu AM, Rohm KD, Suttner S. Goal-directed intraoperative therapy based on autocalibrated arterial pressure waveform analysis reduces hospital stay in high-risk surgical patients: a randomized, controlled trial. Crit Care 2010;14:R18
129. Rhodes A, Cecconi M, Hamilton M, Poloniecki J, Woods J, Boyd O, Bennett D, Grounds RM. Goal-directed therapy in high-risk surgical patients: a 15-year follow-up study. Intensive Care Med 2010;36:1327–32
130. Hamilton MA, Cecconi M, Rhodes A. A systematic review and meta-analysis on the use of preemptive hemodynamic intervention to improve postoperative outcomes in moderate and high-risk surgical patients. Anesth Analg 2011;112:1392–402
131. Rosenbaum A, Kirby C, Breen PH. Measurement of oxygen uptake and carbon dioxide elimination using the bymixer: validation in a metabolic lung simulator. Anesthesiology 2004;100:1427–37
132. Perel A, Pizov R, Cotev S. Systolic blood pressure variation is a sensitive indicator of hypovolemia in ventilated dogs subjected to graded hemorrhage. Anesthesiology 1987;67:498–502
133. Marik PE, Baram M, Vahid B. Does central venous pressure predict fluid responsiveness? A systematic review of the literature and the tale of seven mares. Chest 2008;134:172–8
134. Marik PE, Cavallazzi R, Vasu T, Hirani A. Dynamic changes in arterial waveform derived variables and fluid responsiveness in mechanically ventilated patients: a systematic review of the literature. Crit Care Med 2009;37:2642–7
135. Michard F. Volume management using dynamic parameters: the good, the bad, and the ugly. Chest 2005;128:1902–3
136. Marik PE. Techniques for assessment of intravascular volume in critically ill patients. J Intensive Care Med 2009;24:329–37
137. Bendjelid K, Romand JA. Fluid responsiveness in mechanically ventilated patients: a review of indices used in intensive care. Intensive Care Med 2003;29:352–60
138. Harvey S, Harrison DA, Singer M, Ashcroft J, Jones CM, Elbourne D, Brampton W, Williams D, Young D, Rowan K. Assessment of the clinical effectiveness of pulmonary artery catheters in management of patients in intensive care (PAC-Man): a randomised controlled trial. Lancet 2005;366:472–7
139. Yu DT, Platt R, Lanken PN, Black E, Sands KE, Schwartz JS, Hibberd PL, Graman PS, Kahn KL, Snydman DR, Parsonnet J, Moore R, Bates DW. Relationship of pulmonary artery catheter use to mortality and resource utilization in patients with severe sepsis. Crit Care Med 2003;31:2734–41
140. Richard C, Warszawski J, Anguel N, Deye N, Combes A, Barnoud D, Boulain T, Lefort Y, Fartoukh M, Baud F, Boyer A, Brochard L, Teboul JL. Early use of the pulmonary artery catheter and outcomes in patients with shock and acute respiratory distress syndrome: a randomized controlled trial. JAMA 2003;290:2713–20
141. Sandham JD, Hull RD, Brant RF, Knox L, Pineo GF, Doig CJ, Laporta DP, Viner S, Passerini L, Devitt H, Kirby A, Jacka M. A randomized, controlled trial of the use of pulmonary-artery catheters in high-risk surgical patients. N Engl J Med 2003;348:5–14
142. Bonazzi M, Gentile F, Biasi GM, Migliavacca S, Esposti D, Cipolla M, Marsicano M, Prampolini F, Ornaghi M, Sternjakob S, Tshomba Y. Impact of perioperative haemodynamic monitoring on cardiac morbidity after major vascular surgery in low risk patients: a randomised pilot trial. Eur J Vasc Endovasc Surg 2002;23:445–51
143. Connors AF Jr, Speroff T, Dawson NV, Thomas C, Harrell FE Jr, Wagner D, Desbiens N, Goldman L, Wu AW, Califf RM, Fulkerson WJ Jr, Vidaillet H, Broste S, Bellamy P, Lynn J, Knaus WA. The effectiveness of right heart catheterization in the initial care of critically ill patients. SUPPORT Investigators. JAMA 1996;276:889–97
144. Chittock DR, Dhingra VK, Ronco JJ, Russell JA, Forrest DM, Tweeddale M, Fenwick JC. Severity of illness and risk of death associated with pulmonary artery catheter use. Crit Care Med 2004;32:911–5
145. Hadian M, Pinsky MR. Functional hemodynamic monitoring. Curr Opin Crit Care 2007;13:318–23
146. Roche AM, Miller TE, Gan TJ. Goal-directed fluid management with trans-oesophageal Doppler. Best Pract Res Clin Anaesthesiol 2009;23:327–34
147. Jorgensen CC, Bundgaard-Nielsen M, Skovgaard LT, Secher NH, Kehlet H. Stroke volume averaging for individualized goal-directed fluid therapy with oesophageal Doppler. Acta Anaesthesiol Scand 2009;53:34–8
148. Forrester JS, Ganz W, Diamond G, McHugh T, Chonette DW, Swan HJ. Thermodilution cardiac output determination with a single flow-directed catheter. Am Heart J 1972;83:306–11
149. Ganz W, Donoso R, Marcus HS, Forrester JS, Swan HJ. A new technique for measurement of cardiac output by thermodilution in man. Am J Cardiol 1971;27:392–6
150. Ganz W, Swan HJ. Measurement of blood flow by thermodilution. Am J Cardiol 1972;29:241–6
151. Swan HJ, Ganz W. Variability between measurements of cardiac output. Crit Care Med 1976;4:279–80
152. Rick JJ, Burke SS. Respirator paradox. South Med J 1978;71:1376–8, 1382
153. Coyle JP, Teplick RS, Long MC, Davison JK. Respiratory variations in systemic arterial pressure as an indicator of volume status. Anesthesiology 1983;59:A53
154. Coriat P, Vrillon M, Perel A, Baron JF, Le Bret F, Saada M, Viars P. A comparison of systolic blood pressure variations and echocardiographic estimates of end-diastolic left ventricular size in patients after aortic surgery. Anesth Analg 1994;78:46–53
155. Tavernier B, Makhotine O, Lebuffe G, Dupont J, Scherpereel P. Systolic pressure variation as a guide to fluid therapy in patients with sepsis-induced hypotension. Anesthesiology 1998;89:1313–21
156. Michard F, Boussat S, Chemla D, Anguel N, Mercat A, Lecarpentier Y, Richard C, Pinsky MR, Teboul JL. Relation between respiratory changes in arterial pulse pressure and fluid responsiveness in septic patients with acute circulatory failure. Am J Respir Crit Care Med 2000;162:134–8
157. Partridge BL. Use of pulse oximetry as a noninvasive indicator of intravascular volume status. J Clin Monit 1987;3:263–8
158. Pizov R, Ya'ari Y, Perel A. The arterial pressure waveform during acute ventricular failure and synchronized external chest compression. Anesth Analg 1989;68:150–6
159. Cannesson M, Attof Y, Rosamel P, Desebbe O, Joseph P, Metton O, Bastien O, Lehot JJ. Respiratory variations in pulse oximetry plethysmographic waveform amplitude to predict fluid responsiveness in the operating room. Anesthesiology 2007;106:1105–11
160. Natalini G, Rosano A, Franschetti ME, Fachetti P, Bernardini A. Variations in arterial blood pressure and photoplethysmography during mechanical ventilation. Anesth Analg 2006;103:1182–8
161. Natalini G, Rosano A, Taranto M, Faggian B, Vittorielli E, Bernardini A. Arterial versus plethysmographic dynamic indices to test responsiveness for testing fluid administration in hypotensive patients: a clinical trial. Anesth Analg 2006;103:1478–84
162. Wyffels PA, Durnez PJ, Helderweirt J, Stockman WM, De Kegel D. Ventilation-induced plethysmographic variations predict fluid responsiveness in ventilated postoperative cardiac surgery patients. Anesth Analg 2007;105:448–52
163. Feissel M, Teboul JL, Merlani P, Badie J, Faller JP, Bendjelid K. Plethysmographic dynamic indices predict fluid responsiveness in septic ventilated patients. Intensive Care Med 2007;33:993–9
164. Solus-Biguenet H, Fleyfel M, Tavernier B, Kipnis E, Onimus J, Robin E, Lebuffe G, Decoene C, Pruvot FR, Vallet B. Non-invasive prediction of fluid responsiveness during major hepatic surgery. Br J Anaesth 2006;97:808–16
165. Cannesson M, Besnard C, Durand PG, Bohe J, Jacques D. Relation between respiratory variations in pulse oximetry plethysmographic waveform amplitude and arterial pulse pressure in ventilated patients. Crit Care 2005;9:R562–8
166. De Backer D, Pinsky MR. Can one predict fluid responsiveness in spontaneously breathing patients? Intensive Care Med 2007;33:1111–3
167. Lamia B, Ochagavia A, Monnet X, Chemla D, Richard C, Teboul JL. Echocardiographic prediction of volume responsiveness in critically ill patients with spontaneously breathing activity. Intensive Care Med 2007;33:1125–32
168. Coudray A, Romand JA, Treggiari M, Bendjelid K. Fluid responsiveness in spontaneously breathing patients: a review of indexes used in intensive care. Crit Care Med 2005;33:2757–62
169. Charron C, Fessenmeyer C, Cosson C, Mazoit JX, Hebert JL, Benhamou D, Edouard AR. The influence of tidal volume on the dynamic variables of fluid responsiveness in critically ill patients. Anesth Analg 2006;102:1511–7
170. De Backer D, Heenen S, Piagnerelli M, Koch M, Vincent JL. Pulse pressure variations to predict fluid responsiveness: influence of tidal volume. Intensive Care Med 2005;31:517–23
171. de Waal EE, Rex S, Kruitwagen CL, Kalkman CJ, Buhre WF. Dynamic preload indicators fail to predict fluid responsiveness in open-chest conditions. Crit Care Med 2009;37:510–5
172. Duperret S, Lhuillier F, Piriou V, Vivier E, Metton O, Branche P, Annat G, Bendjelid K, Viale JP. Increased intra-abdominal pressure affects respiratory variations in arterial pressure in normovolaemic and hypovolaemic mechanically ventilated pigs. Intensive Care Med 2007;33:163–71
173. Landsverk SA, Hoiseth LO, Kvandal P, Hisdal J, Skare O, Kirkeboen KA. Poor agreement between respiratory variations in pulse oximetry photoplethysmographic waveform amplitude and pulse pressure in intensive care unit patients. Anesthesiology 2008;109:849–55
174. Shelley KH, Murray WB, Chang D. Arterial-pulse oximetry loops: a new method of monitoring vascular tone. J Clin Monit 1997;13:223–8
175. Cannesson M, Desebbe O, Lehot JJ. Fluid responsiveness assessment using the pulse oxymeter waveform: not yet ready for prime time. Anesth Analg 2007;104:1598–9
176. Cannesson M, Desebbe O, Lehot JJ. Comment on “Plethysmographic dynamic indices predict fluid responsiveness in septic ventilated patients” by Feissel et al. Intensive Care Med 2007;33:1853
177. Landsverk SA, Kvandal P, Bernjak A, Stefanovska A, Kirkeboen KA. The effects of general anesthesia on human skin microcirculation evaluated by wavelet transform. Anesth Analg 2007;105:1012–9
178. Desebbe O, Cannesson M. Using ventilation induced plethysmographic variations to optimize patient fluid status. Curr Opin Anaesthesiol 2008;21:772–8
179. Aboy M, McNames J, Thong T, Phillips CR, Ellenby MS, Goldstein B. A novel algorithm to estimate the pulse pressure variation index deltaPP. IEEE Trans Biomed Eng 2004;51:2198–203
180. Cannesson M, Slieker J, Desebbe O, Bauer C, Chiari P, Hénaine R, Lehot JJ. The ability of a novel algorithm for automatic estimation of the respiratory variations in arterial pulse pressure to monitor fluid responsiveness in the operating room. Anesth Analg 2008;106:1195–2000
181. Cannesson M, Delannoy B, Morand A, Rosamel P, Attof Y, Bastien O, Lehot JJ. Does the pleth variability index indicate the respiratory induced variation in the plethysmogram and arterial pressure waveforms? Anesth Analg 2008;106:1189–94
182. Cannesson M, Musard H, Desebbe O, Boucau C, Simon R, Hénaine R, Lehot JJ. The ability of stroke volume variations obtained with Vigileo/FloTrac system to monitor fluid responsiveness in mechanically ventilated patients. Anesth Analg 2009;108:513–7
183. Biais M, Nouette-Gaulain K, Roullet S, Quinart A, Revel P, Sztark F. A comparison of stroke volume variation measured by Vigileo/FloTrac system and aortic Doppler echocardiography. Anesth Analg 2009;109:466–9
184. Reuter DA, Bayerlein J, Goepfert MS, Weis FC, Kilger E, Lamm P, Goetz AE. Influence of tidal volume on left ventricular stroke volume variation measured by pulse contour analysis in mechanically ventilated patients. Intensive Care Med 2003;29:476–80
185. Forget P, Lois F, de Kock M. Goal-directed fluid management based on the pulse oximeter-derived pleth variability index reduces lactate levels and improves fluid management. Anesth Analg 2010;111:910–4
186. Benes J, Chytra I, Altmann P, Hluchy M, Kasal E, Svitak R, Pradl R, Stepan M. Intraoperative fluid optimization using stroke volume variation in high risk surgical patients: results of prospective randomized study. Crit Care 2010;14:R118
187. Buettner M, Schummer W, Huettemann E, Schenke S, van Hout N, Sakka SG. Influence of systolic-pressure-variation-guided intraoperative fluid management on organ function and oxygen transport. Br J Anaesth 2008;101:194–9
188. Lopes MR, Oliveira MA, Pereira VO, Lemos IP, Auler JO Jr, Michard F. Goal-directed fluid management based on pulse pressure variation monitoring during high-risk surgery: a pilot randomized controlled trial. Crit Care 2007;11:R100
189. Parkin G, Wright C, Bellomo R, Boyce N. Use of a mean systemic filling pressure analogue during the closed-loop control of fluid replacement in continuous hemodiafiltration. J Crit Care 1994;9:124–33
190. Kramer GC, Kinsky MP, Prough DS, Salinas J, Sondeen JL, Hazel-Scerbo ML, Mitchell CE. Closed-loop control of fluid therapy for treatment of hypovolemia. J Trauma 2008;64:S333–41
191. Vaid SU, Shah A, Michell MW, Rafie AD, Deyo DJ, Prough DS, Kramer GC. Normotensive and hypotensive closed-loop resuscitation using 3.0% NaCl to treat multiple hemorrhages in sheep. Crit Care Med 2006;34:1185–92
192. Salinas J, Drew G, Gallagher J, Cancio LC, Wolf SE, Wade CE, Holcomb JB, Herndon DN, Kramer GC. Closed-loop and decision-assist resuscitation of burn patients. J Trauma 2008;64:S321–32
193. Hoskins SL, Elgjo GI, Lu J, Ying H, Grady JJ, Herndon DN, Kramer GC. Closed-loop resuscitation of burn shock. J Burn Care Res 2006;27:377–85
194. Cannesson M. Arterial pressure variation and goal-directed fluid therapy. J Cardiothorac Vasc Anesth 2010;24:487–97
195. Puskarich MA, Marchick MR, Kline JA, Steuerwald MT, Jones AE. One year mortality of patients treated with an emergency department based early goal directed therapy protocol for severe sepsis and septic shock: a before and after study. Crit Care 2009;13:R167
196. Goepfert MS, Reuter DA, Akyol D, Lamm P, Kilger E, Goetz AE. Goal-directed fluid management reduces vasopressor and catecholamine use in cardiac surgery patients. Intensive Care Med 2007;33:96–103
197. Maguire S, Rinehart J, Vakharia S, Cannesson M. Technical communication: respiratory variation in pulse pressure and plethysmographic waveforms: intraoperative applicability in a North American academic center. Anesth Analg 2011;112:94–6
198. Rinehart J, Alexander B, Meng L, Cannesson M. Closed-loop algorithm (LIR™) vs. anesthesiologist management of simulated massive hemorrhage. Society for Technology in Anesthesia Annual Meeting, Las Vegas, NV, 2011
199. Leslie K, Absalom A, Kenny GN. Closed loop control of sedation for colonoscopy using the Bispectral Index. Anaesthesia 2002;57:693–7
200. Absalom AR, Kenny GN. Closed-loop control of propofol anaesthesia using bispectral index: performance assessment in patients receiving computer-controlled propofol and manually controlled remifentanil infusions for minor surgery. Br J Anaesth 2003;90:737–41
201. Méndez JA, Torres S, Reboso JA, Reboso H. Adaptive computer control of anesthesia in humans. Comput Methods Biomech Biomed Engin 2009;12:727–34
202. Liu N, Chazot T, Hamada S, Landais A, Boichut N, Dussaussoy C, Trillat B, Beydon L, Samain E, Sessler DI, Fischler M. Closed-loop coadministration of propofol and remifentanil guided by bispectral index: a randomized multicenter study. Anesth Analg 2011;112:546–57

Supplemental Digital Content

© 2012 International Anesthesia Research Society