Clinical Anesthesia

Chapter 18

Intravenous Anesthetics

Paul F. White

Matthew R. Eng

Key Points

1. With the exception of ketamine, intravenous (IV) anesthetics lack intrinsic analgesic properties.

2. Dexmedetomidine is an α2-agonist with sedative and opioid-sparing effects that is used as an anesthetic adjuvant in the operating room and intensive care unit.

3. Low doses of IV anesthetics produce sedation, and high doses produce hypnosis (or unconsciousness).

4. All IV anesthetics are sedative-hypnotics and produce dose-dependent central nervous system (CNS) depression.

5. Compared to thiopental and propofol, methohexital produces less CNS depression.

6. Propofol possesses unique antiemetic and appetite-stimulating properties.

7. Etomidate produces less cardiovascular depression than the barbiturates and propofol.

8. Ketamine possesses analgesic and psychomimetic properties.

9. Midazolam possesses amnestic and anxiolytic properties.

10.     IV anesthetics in combination with potent opioid analgesics and/or local anesthetics can be used to produce total intravenous anesthesia.

The concept of intravenous (IV) anesthesia has evolved from primarily induction of general anesthesia to total IV anesthesia (TIVA). TIVA has assumed increasing importance for therapeutic, as well as diagnostic, procedures in both adults and children. IV anesthetic techniques are used for procedures in the operating room (OR) and remote from the OR. In many centers in Europe and South America, TIVA has become more popular for general anesthesia than classic “balanced anesthesia” or volatile anesthetic-based techniques. This change has been a result of (1) the development of rapid, short-acting IV hypnotic, analgesic, and muscle relaxant drugs; (2) the availability of pharmacokinetic and dynamic-based IV delivery systems; and (3) the development of the electroencephalogram (EEG)-based cerebral monitoring devices, which measure the hypnotic component of the anesthetic state. This chapter focuses on the pharmacologic properties and clinical uses of the currently available IV anesthetics.

Following its introduction into clinical practice, thiopental quickly became the gold standard of IV anesthetics against which all the newer IV drugs were compared. Many different hypnotic drugs are currently available for use during IV anesthesia (Fig. 18-1). However, it is clear that the “ideal” IV anesthetic is yet to be developed. The physical and pharmacologic properties that an ideal IV anesthetic would possess include the following:

1.   Drug compatibility (water-solubility) and stability in solution.

2.   Lack of pain on injection, veno-irritation, and local tissue damage following extravasation.

3.   Low potential to release histamine or precipitate hypersensitivity reactions.

4.   Rapid and smooth onset of hypnotic action without excitatory activity.

5.   Rapid metabolism to pharmacologically inactive metabolites.

6.   A steep dose-response relationship to enhance titratability and minimize tissue accumulation.

7.   Lack of acute cardiovascular and respiratory depression.

8.   Decreases in cerebral metabolism and intracranial pressure.

9.   Rapid and smooth return of consciousness and cognitive skills with residual analgesia.

10. Absence of postoperative nausea and vomiting, amnesia, psychomimetic reactions, dizziness, headache, or prolonged sedation (hangover effects).

Despite thiopental's proven clinical usefulness, safety, and widespread acceptance over many decades of use, it has been supplanted by a variety of agents from different drug groups. The sedative-hypnotic drugs that have been more recently introduced into clinical practice (e.g., midazolam, ketamine, etomidate, propofol) have proven to be extremely valuable in specific clinical situations. These newer compounds combine many of the characteristics of the ideal IV anesthetic, but fail in aspects where the other drugs succeed. For some of these IV


sedative-hypnotics, disadvantages have led to “restricted” indications (e.g., ketamine, etomidate). Because the optimal pharmacologic properties are not equally important in every clinical situation, the anesthesiologist must make the choice of the IV anesthetic drug that best fits the needs of the individual patient and the operative (or diagnostic) procedure.

Figure 18-1. Chemical structures of currently available nonopioid intravenous anesthetics.

General Pharmacology of Intravenous Hypnotics

Mechanism of Action

A widely accepted theory of anesthetic action is that both IV and inhalational anesthetics exert their primary sedative and hypnotic effects through an interaction with the inhibitory γ-aminobutyric acid (GABA) neurotransmitter system.1 GABA is the principal inhibitory neurotransmitter within the CNS. The GABA and adrenergic neurotransmitter systems counterbalance the action of excitatory neurotransmitters. The GABA type A (GABAA) receptor is a receptor complex consisting of up to five glycoprotein subunits. When the GABAAreceptor is activated, transmembrane chloride conductance increases, resulting in hyperpolarization of the postsynaptic cell membrane and functional inhibition of the postsynaptic neuron. Sedative-hypnotic drugs interact with different components of the GABA-receptor complex (Fig. 18-2). However, the allosteric (structural) requirements for activation of the receptor are different for IV and volatile anesthetics.

Benzodiazepines bind to specific receptor sites that are part of the GABAA-receptor complex. The binding of benzodiazepines to their receptor site increases the efficiency of the coupling between the GABA receptor and the chloride ion channel. The degree of modulation of the GABA-receptor function is limited, which explains the maximal “ceiling effect” produced by benzodiazepines with respect to CNS depression. The dose-dependent CNS depressant effect of benzodiazepines produce hypnosis, sedation, anxiolysis, amnesia, and anticonvulsant effects.2 These CNS effects are presumed to be associated with stimulation of different receptor subtypes and/or concentration-dependent receptor occupancy.1 For example, it has been suggested that benzodiazepine receptor occupancy of 20% provides anxiolysis, while 30 to 50% receptor occupancy is associated with amnesia to sedation, and 60% receptor occupancy is required for hypnosis (or unconsciousness).2

The interaction of barbiturates and propofol with specific membrane structures appears to decrease the rate of dissociation of GABA from its receptor, thereby increasing the duration of the GABA-activated opening of the chloride ion channel (Fig. 18-2). Barbiturates can also mimic the action of GABA by directly activating the chloride channels. The proposed mechanism of action of thiopental relates to its ability to function as a competitive inhibitor at the nicotinic acetylcholine receptors in the CNS.3 Etomidate augments GABA-gated chloride currents (i.e., indirect modulation) and at


higher concentrations evokes chloride currents in the absence of GABA (i.e., direct activation). Although the mechanism of action of propofol is similar to that of the barbiturates (i.e., enhancing the activity of the GABA-activated chloride channel), it also possesses ion channel-blocking effects in cerebral cortex tissue and nicotinic acetylcholine receptors, as well as an inhibitory effect on lysophosphatidate signaling in lipid mediator receptors.4

Figure 18-2. A. This model depicts the postsynaptic site of γ-aminobutyric acid (GABA) and glutamate within the CNS. GABA decreases the excitability of neurons by its action at the GABAA-receptor complex. When GABA occupies the binding site of this complex, it allows inward flux of chloride ion, resulting in hyperpolarizing of the cell and subsequent resistance of the neuron to stimulation by excitatory transmitters. Barbiturates, benzodiazepines, propofol, and etomidate decrease neuronal excitability by enhancing the effect of GABA at this complex, facilitating this inhibitory effect on the postsynaptic cell. Glutamate and its analog N-methyl-D-aspartate (NMDA) are excitatory amino acids. When glutamate occupies the binding site on the NMDA subtype of the glutamate receptor, the channel opens and allows Na+, K+, and Ca2+ to either enter or leave the cell. Flux of these ions leads to depolarization of the postsynaptic neuron and initiation of an action potential and activation of other pathways. Ketamine blocks this open channel and prevents further ion flux, thus inhibiting the excitatory response to glutamate. (Reprinted with permission from Van Hemelrijck J, Gonzales JM, White PF: Use of intravenous sedative agents, Principles and Practice of Anesthesiology. Edited by Rogers MC, Tinker JH, Covino BG, Longnecker DE. St. Louis, Mosby, 1992, p 1131.) B. Schematic model of the GABAA-receptor complex illustrating recognition sites for many of the substances that bind to the receptor. C. Model of the NMDA receptor showing sites for antagonist action. Ketamine binds to the site labeled PCP (phencyclidine). The pentameric structure of the receptor, composed of a combination of the subunits NR 1 and NR 2, is illustrated. (Altered with permission from Leeson TD, Iversen LL: The glycine site on the NMDA receptor: Structure-activity relationships and therapeutic potential. J Med Chem 1994; 37: 4053.)

Ketamine produces a functional dissociation between the thalamocortical and limbic systems, a state that has been termed dissociative anesthesia. Ketamine depresses neuronal function in the cerebral cortex and thalamus, while simultaneously activating the limbic system. Ketamine's effect on the medial medullary reticular formation may be involved in the affective component of its nociceptive activity. The CNS effects of ketamine appear to be primarily related to its antagonistic activity at the N-methyl-D-aspartate (NMDA) receptor (Fig. 18-2). Unlike the other IV anesthetics, ketamine does not interact with GABA receptors; however, it binds to non-NMDA glutamate receptors and nicotinic, muscarinic, monoaminergic, and opioid receptors. In addition, it also inhibits neuronal sodium channels (producing a modest local anesthetic action) and calcium channels (causing cerebral vasodilatation).

The centrally-active α2-adrenergic receptor agonists, clonidine and dexmedetomidine, have potent sedative and opioid analgesic-sparing properties. These drugs also have significant effects on the peripheral α2 receptors involved in regulating the cardiovascular system by inhibiting norepinephrine release. This class of anesthetic adjuvants can also reduce blood pressure and heart rate by decreasing the tonic levels of sympathetic outflow from the CNS and augmenting cardiac vagal activity, respectively.5,6 However, dexmedetomidine failed to block the acute hyperdynamic response to electroconvulsive therapy when administered as an adjuvant to methohexital anesthesia.7 Earlier studies with clonidine demonstrated that this α2 agonist-antagonist could


also reduce the IV8 and volatile9 anesthetic requirements, as well as the postoperative opioid analgesic requirement.

Pharmacokinetics and Metabolism

An understanding of basic pharmacokinetic principles is integral to the understanding the pharmacologic actions and interactions of IV anesthetic and adjunctive drugs, and will allow the anesthesiologist to develop more optimal dosing strategies when using IV techniques. Although lipid solubility facilitates diffusion of IV anesthetics across cellular membranes, including the blood–brain barrier, only the nonionized form is able to readily cross neuronal membranes. The ratio of the unionized-to-ionized fraction depends on the pKa of the drug and the pH of the body fluids.

The rapid onset of the CNS effect of most IV anesthetics can be explained by their high lipid solubility and the relatively high proportion of the cardiac output (20%) perfusing the brain. However, a variable degree of hysteresis exists between the blood concentration of the hypnotic drug and its onset of action on the CNS. The hysteresis is related in part to diffusion of these drugs into brain tissue and nonspecific CNS receptor binding. However, the number of CNS binding sites is usually saturable and only a small fraction of the available binding sites needs to be occupied to produce clinical effects. Although the total amount of drug in the blood is available for diffusion, the diffusion rate will be more limited for IV anesthetics with a high degree of plasma protein binding (90%) because only the “free” unbound drug can diffuse across membranes and exert central effects. When several drugs compete for the same binding sites, or when the protein concentration in the blood is decreased by preexisting disease (e.g., hepatic failure, malnutrition), a higher fraction of the unbound drug will be available to exert an effect on the CNS. Since only unbound drug is available for uptake and metabolism in the liver, highly protein-bound drugs may have a lower rate of hepatic metabolism as a result of their decreased hepatic extraction ratio (i.e., the fraction of the hepatic blood flow that is cleared of the drug).

The pharmacokinetics of IV hypnotics are characterized by rapid distribution and subsequent redistribution into several hypothetical compartments (determined by their effect on blood flow to various tissues), followed by elimination (Table 18-1). The initial pharmacologic effects are related to the activity of the drug in the central compartment. The primary mechanism for terminating the central effects of IV anesthetics administered for induction of anesthesia is redistribution from the central highly perfused compartment (brain) to the larger, but less well perfused “peripheral” compartments (muscle, fat). Even for drugs with a high hepatic extraction ratio, elimination does not usually play a major role in terminating the drug's CNS effects because elimination of the drug can occur only from the central compartment. The rate of elimination from the central compartment, the amount of drug present in the peripheral compartments, and the rate of redistribution from the peripheral compartments “back” into the central compartment determine the time necessary to eliminate the drug from the body and directly influence recovery times.

Most IV anesthetic agents are eliminated via hepatic metabolism followed by renal excretion of more water-soluble metabolites. Some metabolites have pharmacologic activity and can produce prolonged drug effects (e.g., oxazepam, desmethyldiazepam, norketamine). Moreover, there is considerable interpatient variability in the clearance rates for commonly used IV anesthetic drugs. The elimination clearance is the distribution volume cleared of drug over time and is a measure of the efficacy of the elimination process. The slow elimination of some anesthetics is partly due to their high degree of protein binding that reduces their hepatic extraction ratio. Other drugs may have a high hepatic extraction ratio and elimination clearance despite extensive plasma protein binding (e.g., propofol), indicating that protein binding is not always a rate-limiting factor.

For most drugs, the hepatic enzyme systems are not saturated at clinically relevant drug concentrations, and the rate of drug elimination will decrease as an exponential function of the drug's plasma concentration (first-order kinetics). However, when high steady-state plasma concentrations are achieved with prolonged infusions, hepatic enzyme systems can become saturated and the elimination rate becomes independent of the drug concentration (zero-order kinetics). The elimination half-life (t1/2β) is the time required for the anesthetic concentration to decrease by 50% during the terminal phase of the plasma decay curve. The t1/2β depends on the volume to be cleared (the distribution volume) and the efficiency of the metabolic clearance system. Because their volumes of distribution are similar, the wide variation in elimination half-life values for the IV anesthetics is a reflection of differences in their clearance values.

When a drug infusion is administered without a loading dose, at least 3 times the t1/2β value may be required to achieve a true “steady-state” plasma concentration. The steady-state concentration obtained during an anesthetic infusion depends on the rate of drug administration and its clearance rate. When an infusion is discontinued, the rate at which the plasma concentration


decreases largely depends on the clearance rate (as reflected by the terminal t1/2β value). For drugs with shorter elimination half-lives, plasma concentration will decrease at a rate that allows for a more rapid recovery (e.g., propofol). Drugs with long t1/2β values (e.g., thiopental and diazepam) are usually only administered by continuous IV infusion when the medical condition requires long-term treatment (e.g., elevated intracranial pressure [ICP] as a result of brain injury or prolonged sedation in the intensive care unit [ICU] because of respiratory failure).

Table 18-1 Pharmacokinetic Values for the Currently Available Intravenous Sedative-Hypnotic Drugs





CLEARANCE (mL/kg/min)


















































From White PF. Textbook of Intravenous Anesthesia. Baltimore, Williams & Wilkins, 1997, pp. 27 and 77.

Careful titration of an anesthetic drug to achieve the desired clinical effect is necessary to avoid drug accumulation and the resultant prolonged CNS effects after the infusion has been discontinued. Although the value of the t1/2β indicates how fast a drug is eliminated from the body, a more useful indicator of the acceptability of a hypnotic infusion for maintenance of anesthesia or sedation is the context-sensitive half-time, a value derived from computer simulations of drug infusions.10 The context-sensitive half-time is defined as the time necessary for the effect-compartment (i.e., effect site) concentration to decrease by 50% in relation to the duration of the infusion. The context-sensitive half-time becomes particularly important in determining recovery after prolonged infusions of sedative-hypnotic drugs. Drugs (e.g., propofol) may have a relatively short context-sensitive half-time despite the fact that a large amount of drug remains present in the “deep” (less well-perfused) compartment. The slow return of the anesthetic from the deep compartment contributes little to the concentration of drug in the central compartment from which it is rapidly cleared. Therefore, the concentration in the central compartment rapidly declines below the hypnotic threshold after discontinuation of the infusion, contributing to short emergence times despite the fact that a substantial quantity of anesthetic drug may remain in the body.

Marked interpatient variability exists in the pharmacokinetics of IV sedative-hypnotic drugs. Factors that can influence anesthetic drug disposition include the degree of protein binding, the efficiency of hepatic and renal elimination processes, physiologic changes with aging, pre-existing disease states, the operative site, body temperature, and drug interactions (e.g., coadministration of volatile anesthetics). For example, increased age, lean body (muscle) mass, and total body water decrease result in an increase in the steady-state volume of distribution of most IV anesthetics. The increased distribution volume and decreased hepatic clearance leads to a prolongation of their t1/2β values. Moreover, a decrease of the volume of the central compartment may result in higher initial drug concentrations and can at least partially explain the decreased induction requirement in the elderly. Additionally, the slower redistribution from the vessel-rich tissues to intermediate compartments (muscles) also contributes to the age-related decrease in the induction dose requirements.10 Although prolongation of the elimination half-time does not provide an explanation for the decreased induction dose requirement, it is responsible for producing higher steady-state plasma concentrations at any given infusion rate, contributing to a slower recovery from the subhypnotes (residual effects).

The hepatic clearance of IV anesthetics with a high (e.g., etomidate, propofol, ketamine) or intermediate (e.g., methohexital, midazolam) extraction ratio largely depends on hepatic blood flow, with most of the drug being removed from the blood as it flows through the liver (so-called perfusion-limited clearance). The elimination rate of drugs with low hepatic extraction ratios (e.g., thiopental, diazepam, lorazepam) depends on the enzymatic activity of the liver and is less dependent of hepatic blood flow (so-called capacity-limited clearance). Hepatic blood flow decreases during upper abdominal and laparoscopic surgery and, as a result, higher blood levels of drugs with perfusion-limited clearance are achieved at any given infusion rate. With aging, a decreased cardiac output and a redistribution of blood flow can partly explain the lower clearance rate for drugs with perfusion-limited clearance. Although concomitant administration of volatile anesthetics (which are known to decrease liver blood flow) has little influence on the elimination of thiopental, they can decrease the clearance of etomidate, ketamine, methohexital, and propofol. Other factors that decrease hepatic blood flow include hypocapnia, congestive heart failure, intravascular volume depletion, acute alcohol intoxication, circulatory collapse, increase intra-abdominal pressure, β-adrenergic blockade, and norepinephrine administration.

Hepatic disease can influence the pharmacokinetics of drugs by (1) altering the plasma protein content and changing the degree of protein binding, (2) decreasing hepatic blood flow and producing intrahepatic shunting, and (3) depressing the metabolic enzymatic activity of the liver. Therefore, the influence of hepatic disease on pharmacokinetics and dynamics of IV anesthetics is difficult to predict. Renal disease can also alter the concentration of plasma and tissue proteins, as well as the degree of protein binding, thereby producing changes in free drug concentrations. Because IV anesthetic agents are primarily metabolized by the liver, renal insufficiency has little influence on their rate of metabolic inactivation or elimination of the primary compound.

Pharmacodynamic Effects

The principal pharmacologic effect of IV anesthetics is to produce progressively increasing sedation and ultimately hypnosis as a result of dose-dependent CNS depression. However, all sedative-hypnotics also directly or indirectly affect other major organ systems. The relationship between the dose of a sedative-hypnotic and its CNS effects can be defined by dose-response curves. Although most IV anesthetics are characterized by steep dose-response curves, they are not always parallel (Fig. 18-3). However, the characteristics of a dose-response


curve can only be interpreted in relation to the specific response for which it was constructed.

Figure 18-3. Dose-response relationships for sedation with midazolam (•) and diazepam (□). The level of sedation (2 = awake and alert to 6 = asleep and unarousable) was assessed 5 minutes after bolus doses of midazolam (0.05, 0.1, or 0.15 mg/kg) or diazepam (0.1, 0.2, or 0.3 mg/kg). Values represent mean values ± SEM. (Reprinted with permission from White PF, Vascones LO, Mathes SA, et al: Comparison of midazolam and diazepam for sedation during plastic surgery. J Plast Reconstruct Surg 1988; 81: 703.)

Figure 18-4. The concentration of thiopental versus time and spectral edge in an elderly patient (top) and in a younger patient (bottom). Solid horizontal bars represent the length of thiopental infusion. Filled circles represent the measured thiopental concentration (linear scale), and the solid line next to them represents the fitted data from the pharmacokinetic model. The axis for spectral edge has been inverted for visual clarity. (Reprinted with permission from Homer TD, Stanski DR: The effect of increasing age on thiopental disposition and anesthetic requirement. Anesthesiology 1985; 62: 714.)

When steady-state plasma concentrations are achieved, one can assume that the plasma concentration is in quasiequilibrium with the effect-site concentration. Under these circumstances, it is possible to describe the relationship between drug and effect using a concentration-effect curve (Fig. 18-4). Because of the pharmacodynamic variability that exists among individuals, the plasma drug concentration necessary to obtain a particular effect is often described in terms of an effective concentration range, the so-called therapeutic window. Efficacy of an IV anesthetic relates to the maximum effect that can be achieved with respect to some measure of CNS function. Depending on the drug effect under consideration, the efficacy of sedative-hypnotics may appear to be <100%. For example, it is virtually impossible to produce a burst-suppressive EEG pattern with a benzodiazepine. Potency, on the other hand, relates to the quantity of drug necessary to obtain the maximum CNS effect. The relative potency of sedative-hypnotics also varies depending on the end point chosen. In the presence of an antagonist drug (e.g., flumazenil), the maximal response that can be obtained with a benzodiazepine agonist is further reduced because of competition for the same CNS receptor binding sites.

The influence of sedative-hypnotics on cerebral metabolism, cerebral hemodynamics, and ICP is of particular importance during neuroanesthesia. In patients with reduced cerebral compliance, a small increase in cerebral blood volume can cause a life-threatening increase in ICP. Most sedative-hypnotic drugs cause a proportional reduction in cerebral metabolism (CMRO2) and cerebral blood flow (CBF), resulting in a decrease in ICP. Although a decrease in CMRO2 probably provides only a modest degree of protection against CNS ischemia or hypoxia, some hypnotics appear to possess cerebroprotective potential (e.g., thiopental, propofol). Explanations for the alleged neuroprotective effects of these compounds include a biochemical role as free-radical scavengers and membrane stabilizers (barbiturates and propofol) or NMDA-receptor antagonists (ketamine). With the exception of ketamine, all sedative-hypnotics also lower intraocular pressure. The changes in intraocular pressure generally reflect the effects of the IV agent on systemic arterial pressure and intracranial hemodynamics. However, none of the available sedative-hypnotic drugs protect against the transient increase in intraocular pressure that occurs with laryngoscopy and tracheal intubation.

Most IV hypnotics have similar EEG effects. Activation of high-frequency EEG activity (15 to 30 Hz) is characteristic of low concentrations (so-called sedative doses) of IV anesthetics. At higher concentrations, an increase in the relative contribution of the lower frequency higher amplitude waves is observed. At high concentrations, a burst-suppressive pattern develops with an increase in the isoelectric periods. Most sedative-hypnotic drugs have been reported to cause occasional EEG seizurelike activity. Interestingly, these same drugs also possess anticonvulsant properties.11,12 When considering possible epileptogenic properties of CNS-depressant drugs, it is important to differentiate between true epileptogenic activity (e.g., methohexital) and myoclonic-like phenomena (e.g., etomidate, ketamine, propofol). Myoclonic activity is generally considered to be the result of an imbalance between excitatory and inhibitory subcortical centers, produced by an unequal degree of suppression of these brain centers by low concentrations of hypnotic drugs. Epileptic activity refers to a sudden alteration in CNS seizurelike activity resulting from a high-voltage electrical discharge at either cortical or subcortical sites, with subsequent spreading to the thalamic and brainstem centers. As a result of its vasoconstrictive effects on the cerebral vasculature, propofol may be useful for treatment of intractable migraine headaches.13

Although some induction drugs can increase airway sensitivity, coughing and airway irritation (e.g., bronchospasm) are usually a result of manipulation of the airway during “light” (inadequate) levels of IV anesthesia rather than to a direct drug effect. With the exception of ketamine (and to a lesser extent, etomidate), IV anesthetics produce dose-dependent respiratory depression, which is enhanced in patients with chronic obstructive pulmonary disease. The respiratory depression is characterized by a decrease in tidal volume and minute ventilation, as well as a transient rightward shift in the CO2 response curve. Following the rapid injection of a large bolus dose of an IV anesthetic, transient apnea lasting 30 to 90 seconds is usually produced. Ketamine causes minimal respiratory depression when administered in the usual induction doses, while etomidate is associated with less respiratory depressant effects than the barbiturate compounds or propofol. The α2-agonist dexmedetomidine has minimal depressant effects on respiratory function.14 The sympatholytic effects of dexmedetomidine when administered for premedication may increase the incidence of intraoperative hypotension and bradycardia.15

Many different factors contribute to the hemodynamic changes associated with IV induction of anesthesia, including the patient's pre-existing cardiovascular and fluid status, resting sympathetic nervous system tone, chronic cardiovascular drugs, preanesthetic medication, the speed of drug injection, and the onset of unconsciousness. In addition, cardiovascular changes can be attributed to the direct pharmacologic actions of anesthetic and analgesic drugs on the


heart and peripheral vasculature. IV anesthetics can depress the CNS and peripheral nervous system responses, blunt the compensatory baroreceptor reflex mechanisms, produce direct myocardial depression, and lower peripheral vascular resistance (and/or dilate venous capacitance vessels), thereby decreasing venous return. Profound hemodynamic effects occur at induction of anesthesia in the presence of hypovolemia because a higher than expected drug concentration is achieved in the central compartment. Not surprisingly, the acute cardiocirculatory depressant effects of all IV anesthetics are accentuated in the elderly, as well as in the presence of pre-existing cardiovascular disease (e.g., coronary artery disease, hypertension).

The effects of IV anesthetics on neuroendocrine function are also influenced by the surgical stimuli. Surgery-induced increases in stress hormones (e.g., vasopressin, catecholamines) can result in increased peripheral vascular resistance, and a reduction of urine output. Similarly, glucose tolerance appears to be decreased by surgical stress, resulting in elevations in the glucose concentration. Unlike ketamine and dexmedetomidine, most IV sedative-hypnotic drugs lack intrinsic analgesic activity. In fact, thiopental has been alleged to possess so-called antianalgesic activity (i.e., appearing to lower the pain threshold). Although propofol possesses dose-dependent effects on thalamocortical transfer of nociceptive information, pain-evoked cortical activity remains intact after loss of consciousness.16

Hypersensitivity (Allergic) Reactions

Allergic or hypersensitivity-type reactions to IV anesthetics are rare but can be severe and even life-threatening. IV drug administration bypasses the normal “protective barriers” against entrance of foreign molecules into the body. With the exception of etomidate, all IV induction agents have been alleged to cause some histamine release. However, the incidence of severe anaphylactic reactions is extremely low with the currently available IV induction agents. The high frequency of allergic reactions to the Cremophor EL–containing formulations led to the early withdrawal of IV anesthetics containing this solubilizing agent (e.g., propofol EL, propanidid, Alphadione [Althesin]). The possible mechanisms for immunologic reactions include (1) direct action on mast cells, (2) classic complement activation after previous exposure and antibody formation, (3) complement activation through the alternative pathway without previous antigen exposure, (4) antigen-antibody reactions, and (5) the “mixed type” of anaphylactoid reactions.

Severe anaphylactic reactions to IV anesthetics are extremely uncommon; however, profound hypotension attributed to nonimmunologically mediated histamine release has been reported with thiopental use. Although anaphylactic reactions to etomidate have been reported, it does not appear to release histamine, and is considered to be the most “immunologically safe” IV anesthetic. Propofol does not normally trigger histamine release, but life-threatening anaphylactoid reactions have been reported in patients with a previous history of multiple-drug allergies. Barbiturates can also precipitate episodes of acute intermittent porphyria and their use is contraindicated in patients who are predisposed to acute intermittent porphyria. Although benzodiazepines, ketamine, and etomidate are reported to be safe in humans, these drugs have been shown to be porphyrogenic in animal models. The most common cause of profound hypotension following IV induction of anesthesia is that of drug interactions and/or unrecognized hypovolemia.

Comparative Physicochemical and Clinical Pharmacologic Properties of Intravenous Agents


The most commonly used barbiturates are thiopental (5-ethyl-5-[1-methylbutyl]-2-thiobarbituric acid), methohexital (1-methyl-5-allyl-5-[1-methyl-2-pentanyl] barbituric acid), and thiamylal (5-allyl-5-[1-methylbutyl]-2-thiobarbituric acid). Thiopental (Pentothal) and thiamylal (Surital) are thiobarbiturates, while methohexital (Brevital) is an oxybarbiturate. Thiamylal is slightly more potent than thiopental but has a similar pharmacologic profile. Although the l-isomers of thiopental and thiamylal are twice as potent as the d-isomers, both hypnotics are commercially available as racemic mixtures. Because methohexital has two asymmetric centers, it has four stereoisomers. The β-l-isomer is 4 to 5 times more potent than the α-l-isomer, but it produces excessive motor responses. Therefore, methohexital is marketed as the racemic mixture of the two α-isomers.

All three barbiturates are available as sodium salts and must be dissolved in isotonic sodium chloride (0.9%) or water to prepare solutions of 2.5% thiopental, 1 to 2% methohexital, and 2% thiamylal. If refrigerated, solutions of the thiobarbiturates are stable for up to 2 weeks. Solutions of methohexital are stable for up to 6 weeks. When barbiturates are added to Ringer lactate or an acidic solution containing other water-soluble drugs, precipitation will occur and can occlude the IV catheter. Although the typical solution of thiopental (2.5%) is highly alkaline (pH 9) and can be irritating to the tissues if injected extravenously, it does not cause pain on injection and venoirritation is rare. In contrast, a 1% methohexital solution frequently causes discomfort when injected into small veins. Intra-arterial injection of thiobarbiturates is a serious complication as crystals can form in the arterioles and capillaries, causing intense vasoconstriction, thrombosis, and even tissue necrosis. Accidental intra-arterial injections should be treated promptly with intra-arterial administration of papaverine and lidocaine (or procaine), as well as a regional anesthesia-induced sympathectomy (stellate ganglion block, brachial plexus block) and heparinization.

Thiopental is metabolized in the liver to hydroxythiopental and the carboxylic acid derivative, which are more water soluble and have little CNS activity. When high doses of thiopental are administered, a desulfuration reaction can occur with the production of pentobarbital, which has long-lasting CNS-depressant activity. The low elimination clearance of thiopental (3.4 mL/kg/min) contributes to a long elimination half-life (t1/2β of 11 hours). Pre-existing hepatic and renal disease result in decreased plasma protein binding, thereby increasing the free fraction of thiopental and enhancing its CNS and cardiovascular-depressant properties. During prolonged continuous administration of thiopental, the concentration in the tissues approaches the concentration in the central compartment, with termination of its CNS effects becoming solely dependent on elimination by nonlinear hepatic metabolism. Methohexital is metabolized in the liver to inactive hydroxyderivatives. The clearance of methohexital (11 mL/kg/min) is higher and more dependent on hepatic blood flow than thiopental, resulting in a shorter elimination half-life (t1/2β 4 hours).

The usual induction dose of thiopental is 3 to 5 mg/kg in adults, 5 to 6 mg/kg in children, and 6 to 8 mg/kg in infants. Because methohexital is approximately 2.7 times more potent than thiopental, a dose of 1.5 mg/kg is equivalent to 4 mg/kg


of thiopental in adults. The dose of barbiturates necessary to induce anesthesia is reduced in premedicated patients, patients in early pregnancy (7 to 13 weeks' gestation), and those of more advanced American Society of Anesthesiologists physical status (III or IV). Geriatric patients require a 30 to 40% reduction in the usual adult dose because of a decrease of the volume of the central compartment and slowed redistribution of thiopental from the vessel-rich tissues to lean muscle.17 When the calculation of the induction dose is based on the lean body mass rather than total body weight, dosage adjustments for age, sex, or obesity are not necessary. Thiopental infusion is seldom used to maintain anesthesia because of the long context-sensitive half-time and prolonged recovery period. Plasma thiopental levels necessary to maintain a hypnotic state range between 10 and 20 mg/mL. A typical infusion rate necessary to treat intracranial hypertension or intractable convulsions is 2 to 4 mg/kg/hr. The plasma concentration of methohexital needed to maintain hypnosis during anesthesia ranges between 3 and 5 mg/mL and can be achieved with an infusion rate of 50 to 120 mg/kg/min.

Barbiturates produce a proportional decrease in CMRO2 and CBF, thereby lowering ICP. The maximal decrease in CMRO2 (55%) occurs when the EEG becomes isoelectric (burst-suppressive pattern). An isoelectric EEG can be maintained with a thiopental infusion rate of 4 to 6 mg/kg/hr (resulting in plasma concentrations of 30 to 50 µg/mL). Because the decrease in systemic arterial pressure is usually less than the reduction in ICP, thiopental should improve cerebral perfusion and compliance. Therefore, thiopental is widely used to improve brain relaxation during neurosurgery and to improve cerebral perfusion pressure (CPP) after acute brain injury. Although barbiturate therapy is widely used to control ICP after brain injury, the results of outcome studies are no better than with other aggressive forms of cerebral antihypertensive therapy.

It has been suggested that barbiturates also possess “neuroprotective” properties secondary to their ability to decrease oxygen demand. Alternative explanations have been suggested, including a reverse steal (“Robin Hood effect”) on CBF, free-radical scavenging, stabilization of liposomal membranes, as well as excitatory amino acid receptor blockade. Based on evidence from experimental studies and a large randomized prospective multi-institutional study,18 experts have concluded that barbiturates have no place in the therapy following resuscitation of a cardiac arrest patient. In contrast, barbiturates are frequently used for cerebroprotection during incomplete brain ischemia (e.g., carotid endarterectomy, temporary occlusion of cerebral arteries, profound hypotension, and cardiopulmonary bypass). By improving the brain's tolerance of incomplete ischemia in patients undergoing open heart surgery with cardiopulmonary bypass, barbiturates were alleged to decrease the incidence of postbypass neuropsychiatric disorders.19 However, during valvular open heart cardiac surgery, a protective effect of barbiturate loading could not be demonstrated.20 Given the lack a demonstrable neuroprotective effect, use of barbiturates during cardiac surgery is not recommended. Use of moderate degrees of hypothermia (33 to 34°C) might provide superior neuroprotection to the barbiturates without prolonging recovery.

Barbiturates cause predictable, dose-dependent EEG changes and possess potent anticonvulsant activity. Continuous infusions of thiopental have been used to treat refractory status epilepticus. However, low doses of thiopental may induce spike wave activity in epileptic patients. Methohexital has well-established epileptogenic effects in patients with psychomotor epilepsy. Low-dose methohexital infusions are frequently used to activate cortical EEG seizure discharges in patients with temporal lobe epilepsy. It is also the IV anesthetic of choice for electroconvulsive therapy.21 Since the frequency of epileptiform EEG activity during induction of anesthesia with methohexital is significantly less than that which occurs during normal periods of sleep in epileptic patients, this suggests that higher doses of methohexital produces anticonvulsant activity. Methohexital also causes myoclonic-like muscle tremors and other signs of excitatory activity (e.g., hiccoughing).

Barbiturates cause dose-dependent respiratory depression.22 However, bronchospasm or laryngospasm following induction with thiopental is usually the result of airway manipulation in “lightly” anesthetized patients. Laryngeal reflexes appear to be more active after induction with thiopental than with propofol. The cardiovascular effects of thiopental and methohexital include decreases in cardiac output, systemic arterial pressure, and peripheral vascular resistance. The depressant effects of thiopental on cardiac output are primarily a result of a decrease in venous return caused by peripheral pooling, as well as a result of a direct myocardial depressant effect, which assumes increasing importance in the presence of hypovolemia and myocardial disease.23 Use of appropriate doses can minimize the cardiodepressant effects of thiopental, even in infants. Bhutada et al.24 demonstrated that thiopental could be used for induction in infants without important changes in heart rate and blood pressure during the intubation period. An equipotent dose of methohexital produces even less hypotension than thiopental because of a greater tachycardic response to the blood pressure-lowering effects of the drug. If the blood pressure remains stable, the myocardial oxygen demand/supply ratio remains normal despite the increase in heart rate because of a concurrent decrease in coronary vascular resistance.


Propofol (2,6-disopropylphenol), an alkylphenol compound, is virtually insoluble in aqueous solution. The initial Cremophor EL formulation of propofol was withdrawn from clinical testing because of the high incidence of anaphylactic reactions. Subsequently, propofol (10 mg/mL) was reintroduced as an egg lecithin emulsion formulation (Diprivan), consisting of 10% soybean oil, 2.25% glycerol, and 1.2% egg phosphatide. Pain on injection occurs in 32 to 67% of patients when injected into small hand veins but can be minimized by injection into larger veins and by prior administration of either lidocaine or a potent opioid analgesic (e.g., fentanyl or remifentanil). A wide variety of drugs have been alleged to reduce pain on injection of propofol (e.g., metoprolol,25 granisetron,26 dolasetron,27 and even thiopental28). Diluting the formulation with additional solvent (Intralipid) or changing the lipid carrier (Lipofundin) also reduced propofol-induced injection pain, probably because of a decrease in the concentration of free propofol in the aqueous phase of the emulsion. A new propofol formulation with sodium metabisulphite (instead of disodium edentate) as an antimicrobial has been shown to be associated with less severe pain on injection.29 Although the presence of the metabisulphite has raised concerns regarding its use in sulphite-allergic patients, this does not appear to be a clinically important problem. Of interest, a 2% formulation is available for long-term sedation to decrease the fluid volume infused as well as the lipid load.

More recently, a lower-lipid formulation of propofol (Ampofol) has been introduced into clinical practice for both general anesthesia30 and sedation.31 The increased “free” fraction of propofol leads to increased pain when it is injected into small veins. Therefore, it is important to add lidocaine to the Ampofol formulation to minimize the pain on injection. A new water-soluble prodrug of propofol (Aquavan) is in clinical development. This prodrug is rapidly hydrolyzed by plasma


alkaline phosphatases in the circulation to release free propofol.32 It has a slower onset than propofol but a similar recovery profile.33 Although Aquavan does not produce injection site discomfort, a transient burning sensation has been reported in the perineal region following IV injection.

Propofol's pharmacokinetics has been studied using single-bolus dosing and continuous infusions.34 In studies using a two-compartment kinetic model, the initial distribution half-life is 2 to 4 minutes and the elimination half-life is 1 to 3 hours. Using a three-compartment model, the initial and slow distribution half-life values are 1 to 8 minutes and 30 to 70 minutes, respectively. The elimination half-life depends largely on the sampling time after discontinuing the administration of propofol and ranges from 2 to 24 hours. This long elimination half-life is indicative of the existence of a poorly perfused compartment from which propofol slowly diffuses back into the central compartment. Propofol is rapidly cleared from the central compartment by hepatic metabolism and the context-sensitive half-life for propofol infusions up to 8 hours is <40 minutes. Propofol is rapidly and extensively metabolized to inactive, water-soluble sulphate and glucuronic acid metabolites, which are eliminated by the kidneys. Propofol's clearance rate (20–30 mL/kg/min) exceeds hepatic blood flow, suggesting that an extrahepatic route of elimination (lungs) also contributes to its clearance. Nevertheless, changes in liver blood flow would be expected to produce marked alterations in propofol's clearance rate. Surprisingly, few changes in propofol's pharmacokinetics have been reported in the presence of hepatic or renal disease.

The induction dose of propofol in healthy adults is 1.5 to 2.5 mg/kg, with blood levels of 2 to 6 µg/mL producing unconsciousness depending on the concomitant medications (e.g., opioid analgesics), the patient's age and physical status, and the extent of the surgical stimulation.35 In one of the first reports describing the use of propofol for induction and maintenance of anesthesia with nitrous oxide, an average infusion rate of 120 µg/kg/min was required.36 The recommended maintenance infusion rate of propofol varies between 100 and 200 µg/kg/min for hypnosis and 25 to 75 µg/kg/min for sedation. Awakening typically occurs at plasma propofol concentrations of 1 to 1.5 µg/mL.37 Because a 50% decrease in the plasma propofol concentration is usually required for awakening, emergence following anesthesia is usually rapid even following prolonged infusions.

Analogous to the barbiturates, children require higher induction and maintenance doses of propofol on a milligram per kilogram basis as a result of their larger central distribution volume and higher clearance rate. Elderly patients and those in poor health require lower induction and maintenance doses of propofol as a result of their smaller central distribution volume and decreased clearance rate. Although subhypnotic doses of propofol produce sedation and amnesia,37 awareness has been reported even at higher infusion rates when propofol is used as the sole anesthetic.38 Propofol often produces a subjective feeling of well-being (and even euphoria) on emergence, and has been abused by health care professionals as a result of this CNS action.39

Propofol decreases CMRO2 and CBF, as well as ICP.40 However, when larger doses are administered, the marked depressant effect on systemic arterial pressure can significantly decrease CPP. Cerebrovascular autoregulation in response to changes in systemic arterial pressure and reactivity of the cerebral blood flow to changes in carbon dioxide tension are not affected by propofol. Evidence for a possible neuroprotective effect has been reported with in vitro preparations, and the use of propofol to produce EEG burst suppression has been proposed as a method for providing neuroprotection during aneurysm surgery. Its neuroprotective effect may at least partially be related to the antioxidant potential of propofol's phenol ring structure, which may act as a free-radical scavenger, decreasing free-radical–induced lipid peroxidation. A recent study reported that this antioxidant activity may offer many advantages in preventing the hypoperfusion-reperfusion phenomenon that can occur during major laparoscopic surgery.41 Although TIVA with propofol and an opioid analgesic is a safe and effective alternative to standard inhalation techniques (i.e., volatile anesthetic with nitrous oxide) for maintenance of anesthesia, concerns have been raised regarding the cost-effectiveness of this technique.42

Propofol produces cortical EEG changes that are similar to those of thiopental. However, sedative doses of propofol increase β-wave activity analogous to the benzodiazepines. Induction of anesthesia with propofol is occasionally accompanied by excitatory motor activity (so-called nonepileptic myoclonia). In a study involving patients without a history of seizure disorders, excitatory movements following propofol were not associated with EEG seizure activity.43 Propofol appears to possess profound anticonvulsant properties.44Propofol has been reported to decrease spike activity in patients with cortical electrodes implanted for resection of epileptogenic foci and has been used successfully to terminate status epilepticus. The duration of motor and EEG seizure activity following electroconvulsive therapy is significantly shorter with propofol than with other IV anesthetics. Propofol produces a decrease in the early components of somatosensory and motor-evoked potentials but does not influence the early components of the auditory-evoked potentials.

Propofol produces dose-dependent respiratory depression, with apnea occurring in 25 to 35% of patients after a typical induction dose. A maintenance infusion of propofol decreases tidal volume and increases respiratory rate. The ventilatory response to carbon dioxide and hypoxia is also significantly decreased by propofol. Propofol can produce bronchodilation in patients with chronic obstructive pulmonary disease and does not inhibit hypoxic pulmonary vasoconstriction.

Propofol's cardiovascular depressant effects are generally considered to be more profound than those of thiopental. Both direct myocardial depressant effects and decreased systemic vascular resistance have been implemented as important factors in producing cardiovascular depression. Direct myocardial depression and peripheral vasodilation are dose- and concentration-dependent. In addition to arterial vasodilation, propofol produces venodilation (caused both to a reduction in sympathetic activity and by a direct effect on the vascular smooth muscle), which further contributes to its hypotensive effect. The relaxation of the vascular smooth muscle may be because of an effect on intracellular calcium mobilization or because of an increase in the production of nitric oxide. Experiments in isolated myocardium suggest that the negative inotropic effect of propofol results from a decrease in intracellular calcium availability secondary to inhibition of transsarcolemmal calcium influx.

Propofol also alters the baroreflex mechanism, resulting in a smaller increase in heart rate for a given decrease in arterial pressure.45 The smaller increase in heart rate with propofol may account for the larger decrease in arterial pressure than with an equipotent dose of thiopental. Recent studies suggest that induction of anesthesia with propofol attenuates desflurane-mediated sympathetic activation.46 Age enhances the cardiodepressant response to propofol and a reduced dosage is required in the elderly. Patients with limited cardiac reserve seem to tolerate the cardiac depression and systemic vasodilation produced by carefully titrated doses of propofol, and maintenance infusions are increasingly used at the end of cardiac surgery when early extubation is desired.

Propofol appears to possess antiemetic properties that contribute to a lower incidence of emetic sequelae after general anesthesia.36 In fact, subanesthetic doses of propofol (10 to


20 mg) have also been successfully used to treat nausea and emesis in the early postoperative period.47 The postulated mechanisms include antidopaminergic activity, depressant effect on the chemoreceptor trigger zone and vagal nuclei, decreased release of glutamate and aspartate in the olfactory cortex, and reduction of serotonin concentrations in the area postrema. However, the ability of propofol to produce a sense of well-being may also contribute to its antiemetic action. Interestingly, propofol also decreases the pruritus produced by spinal opioids.

Propofol does not trigger malignant hyperthermia and may be considered the induction agent of choice in malignant hyperthermia-susceptible patients. The use of propofol infusions for sedation in the pediatric ICU has been linked to several deaths following prolonged administration because of lipid accumulation and hypotension. Although clinical doses of propofol do not affect cortisol synthesis or the response to adrenocorticotropic hormone stimulation, propofol has been reported to inhibit phagocytosis and killing of bacteria in vitro and to reduce proliferative responses when added to lymphocytes from critically ill patients.48 Because fat emulsions are known to support the growth of micro-organisms, contamination can occur as a result of dilution or fractionated use.49

In critically-ill children and adults receiving high-dose infusions of propofol, some patients have been reported to experience “propofol syndrome,” which is characterized by myocardial failure, metabolic acidosis, and rhabdomyolysis. The etiology of this syndrome may be related to the large lipid load associated with prolonged infusions of the current formulations of propofol.


The parenteral benzodiazepines include diazepam (Valium), lorazepam (Ativan), and midazolam (Versed), as well as the antagonist flumazenil (Romazicon). Diazepam and lorazepam are insoluble in water and their formulation contains propylene glycol, a tissue irritant that causes pain on injection and venous irritation. Diazepam is available in a lipid emulsion formulation, which does not cause pain or thrombophlebitis but is associated with a slightly lower bioavailability. Midazolam is a water-soluble benzodiazepine that is available in an acidified (pH 3.5) aqueous formulation that produces minimal local irritation after IV or intramuscular (IM) injection.50 At physiologic pH, an intramolecular rearrangement occurs that changes the physicochemical properties of midazolam such that it becomes more lipid soluble.

Benzodiazepines undergo hepatic metabolism via oxidation and glucuronide conjugation. Oxidation reactions are susceptible to hepatic dysfunction and coadministration of other anesthetic drugs. Diazepam is metabolized to active metabolites (desmethyldiazepam, 3-hydroxydiazepam), which can prolong diazepam's residual sedative effects because of their long t1/2 β values. These metabolites undergo secondary conjugation to form inactive water-soluble glucuronide conjugates. Drugs that inhibit the oxidative metabolism of diazepam include the H2-receptor blocking drug cimetidine. Severe liver disease reduces diazepam's protein-binding and hepatic-clearance rate, increases its volume of distribution, and thereby further prolongs the t1/2β value. Chronic renal disease decreases protein binding and increases the free drug fraction, resulting in enhanced hepatic metabolism and a shorter t1/2β value. In elderly patients, the clearance rate of diazepam is significantly decreased, prolonging its t1/2β to 75 to 150 hours.

Lorazepam is directly conjugated to glucuronic acid to form pharmacologically inactive metabolites. Age and renal disease have little influence on the kinetics of lorazepam; however, severe hepatic disease decreases its clearance rate. Midazolam undergoes extensive oxidation by hepatic enzymes to form water-soluble hydroxylated metabolites, which are excreted in the urine. However, the primary metabolite, 1-hydroxymethylmidazolam, has mild CNS-depressant activity. The hepatic clearance rate of midazolam is 5 times greater than lorazepam and 10 times greater than diazepam. Although changes in liver blood flow can affect the clearance of midazolam, age has relatively little influence on midazolam's elimination half-life.

The benzodiazepines used in anesthesia are classified as either short- (midazolam, flumazenil), intermediate- (diazepam), or long-acting (lorazepam). Because the distribution volumes are similar, the large difference in the elimination half-times is because of differences in their differing clearance rates (Table 18-1). The context-sensitive half-times for diazepam and lorazepam are very long; therefore, only midazolam should be used by continuous infusion to avoid excessive accumulation.

All benzodiazepines produce dose-dependent anxiolytic, anterograde amnestic, sedative, hypnotic, anticonvulsant, and spinally mediated muscle relaxant properties. Benzodiazepines differ in potency and efficacy with regard to their distinctive pharmacologic properties.50 The dose-dependent pharmacologic activity implies that the CNS effects of various benzodiazepine compounds depend on the affinity for receptor subtypes and their degree of receptor binding. Although benzodiazepines can be used as hypnotics, they are primarily used as premedicants and adjuvant drugs because of their anxiolytic, sedative, and amnestic properties. For example, midazolam (0.04 to 0.08 mg/kg IV/IM) is the most commonly used premedicant. In addition, midazolam, 0.4 to 0.8 mg/kg administered orally 10 to 15 minutes before parental separation, is an excellent premedicant in children. In contrast to lorazepam, both diazepam and midazolam can be used to induce anesthesia because they have a relatively short onset time after IV administration. The half-life of equilibration between the plasma concentration of midazolam and its maximal EEG effect is only 2 to 3 minutes. The therapeutic window to maintain unconsciousness with midazolam is reported to be 100 to 200 ng/mL, with awakening occurring at plasma concentrations below 50 ng/mL. However, significant hypnotic synergism occurs when midazolam and opioid analgesics are administered in combination.

The usual induction dose of midazolam in premedicated patients is 0.1 to 0.2 mg/kg IV, with infusion rates of 0.25 to 1 mg/kg/min required to maintain hypnosis and amnesia in combination with inhalational agents and/or opioid analgesics. Higher maintenance infusion rates and prolonged administration will result in accumulation and prolonged recovery times. Lower infusion rates are sufficient to provide sedation and amnesia during local and regional anesthesia.51 Patient-controlled administration of midazolam during procedures under local anesthesia is well accepted by patients and associated with few perioperative complications.52

Benzodiazepines decrease both CMRO2 and CBF analogous to the barbiturates and propofol. However, in contrast to these compounds, midazolam is unable to produce a burst-suppressive (isoelectric) pattern on the EEG. Accordingly, there is a “ceiling” effect with respect to the decrease in CMRO2 produced by increasing doses of midazolam. Midazolam produces a dose-related decrease in regional cerebral perfusion in the parts of the brain that subserve arousal, attention, and memory. Cerebral vasomotor responsiveness to carbon dioxide is preserved during midazolam anesthesia. In patients with severe head injury, a bolus dose of midazolam may decrease CPP with little effect on ICP. Although midazolam may improve neurologic outcome after incomplete ischemia in animal experiments, benzodiazepines have not


been shown to possess neuroprotective activity in humans. Like the other sedative-hypnotic drugs, the benzodiazepines are potent anticonvulsants that are commonly used to treat status epilepticus.

Benzodiazepines produce dose-dependent respiratory depression. In healthy patients, the respiratory depression associated with benzodiazepine premedication is insignificant. However, the depressant effect is enhanced in patients with chronic respiratory disease, and synergistic depressant effects occur when benzodiazepines are coadministered with opioid analgesics. Benzodiazepines also depress the swallowing reflex and decrease upper airway reflex activity.

Both midazolam and diazepam produce decreases in systemic vascular resistance and blood pressure when large doses are administered for induction of anesthesia. However, the cardiovascular depressant effects of benzodiazepines are frequently “masked” by the stimulus of laryngoscopy and intubation. The cardiovascular depressant effects are directly related to the plasma concentration; however, a plateau plasma concentration appears to exist above which little further change in arterial blood pressure occurs. In the presence of heart failure, the decrease in preload and afterload produced by benzodiazepines may be beneficial in improving cardiac output. However, the cardiodepressant effect of benzodiazepines may be more marked in hypovolemic patients.

A short-acting intravenous sedative, Ro 48-6791, is a water-soluble benzodiazepine that has full agonistic activity at CNS benzodiazepine receptors. Compared with midazolam, it is 2- to 2.5-fold more potent, has a higher plasma clearance rate, and has a similar onset and duration of action.53 In a study involving outpatients undergoing endoscopy procedures, the times to ambulation and to recovery from psychomotor impairment were decreased compared to midazolam, although the later recovery end points (e.g., “fitness-for-discharge”) were similar.54

In contrast to all other sedative-hypnotic drugs, there is a specific antagonist for benzodiazepines. Flumazenil, a 1,4-imidazobenzodiazepine derivative, has a high affinity for the benzodiazepine receptor but minimal intrinsic activity.55 Flumazenil's molecular structure is similar to other benzodiazepines except for the absence of a phenyl group, which is replaced by a carbonyl group. It is water soluble and possesses moderate lipid solubility at physiologic pH. Flumazenil is rapidly metabolized in the liver, and its metabolites are excreted in the urine as glucuronide conjugates. Flumazenil acts as a competitive antagonist in the presence of benzodiazepine agonist compounds. The residual activity of the benzodiazepines in the presence of flumazenil depends on the relative concentrations of the agonist and antagonist drugs. As a result, it is possible to reverse benzodiazepine-induced anesthesia (or deep sedation) either completely or partially, depending on the dose of flumazenil. Flumazenil is short acting, with an elimination half-life of ~1 hour.

Recurrence of the central effects of benzodiazepines (resedation) may occur after a single dose of flumazenil because of residual effects of the more slowly eliminated agonist drug.56If sustained antagonism is desired, it may be necessary to administer flumazenil as repeated bolus doses or a continuous infusion. In general, 45 to 90 minutes of antagonism can be expected following flumazenil 1 to 3 mg IV. However, the respiratory depression produced by benzodiazepines is not completely reversed by flumazenil.57 Reversal of benzodiazepine sedation with flumazenil is not associated with adverse cardiovascular effects or evidence of an acute stress response.58 Although flumazenil does not appear to change CBF or CMRO2following midazolam anesthesia for craniotomy, acute increases in ICP have been reported in head-injured patients receiving flumazenil.


Etomidate is a carboxylated imidazole-containing anesthetic compound (R-1-ethyl-1-[a-methylbenzyl] imidazole-5-carboxylate) that is structurally unrelated to any other IV anesthetic. Only the d-isomer of etomidate possesses anesthetic activity. Analogous to midazolam (which also contains an imidazole nucleus), etomidate undergoes an intramolecular rearrangement at physiologic pH, resulting in a closed-ring structure with enhanced lipid solubility. The aqueous solution of etomidate (Amidate) is unstable at physiologic pH and is formulated in a 0.2% solution with 35% propylene glycol (pH 6.9), contributing to a high incidence of pain on injection, venoirritation, and hemolysis. A new lipid emulsion formulation (Etomidate-Lipuro) has recently been introduced in Europe and appears to be associated with a lower incidence of side effects compared with the original propylene glycol formulation.

The standard induction dose of etomidate (0.2–0.3 mg/kg IV) produces a rapid onset of anesthesia. Involuntary myoclonic movements are common during the induction period as a result of subcortical disinhibition and are unrelated to cortical seizure activity. The frequency of this myocloniclike activity can be attenuated by prior administration of opioid analgesics, benzodiazepines, or small sedative doses (0.03 to 0.05 mg/kg) prior to induction of anesthesia.59 Recently, remifentanil reduced etomidate-induced myoclonic activity without increasing side effects like apnea, emesis, or pruritus.60 Emergence time after etomidate anesthesia is dose-dependent but remains short even after administration of repeated bolus doses or continuous infusions. For maintenance of hypnosis, the target concentration is 300 to 500 ng/mL and can be rapidly achieved by administering a two- or three-stage infusion (e.g., 100 mg/kg/min for 10 minutes followed by 10 mg/kg/min or 100 mg/kg/min for 3 to 5 minutes, followed by 20 mg/kg/min for 20 to 30 minutes, and then 10 mg/kg/min). The pharmacokinetics of etomidate are optimally described by a three-compartment open model.61 The high clearance rate of etomidate (18 to 25 mL/kg/min) is a result of extensive ester hydrolysis in the liver (forming inactive water-soluble metabolites). A significant decrease in plasma protein binding has been reported in the presence of uremia and hepatic cirrhosis. Severe hepatic disease causes a prolongation of the elimination half-life secondary to an increased volume of distribution and a decreased plasma clearance rate.

Analogous to the barbiturates, etomidate decreases CMRO2, CBF, and ICP. However, the hemodynamic stability associated with etomidate will maintain adequate CPP. Etomidate has been used successfully for both induction and maintenance of anesthesia for neurosurgery. Etomidate's well-known inhibitory effect on adrenocortical synthetic function62 limits its clinical usefulness for long-term treatment of elevated ICP. Although clear evidence for a neuroprotective effect in humans is lacking, etomidate is frequently used during temporary arterial occlusion and intraoperative angiography (for the treatment of cerebral aneurysms). Etomidate produces an EEG pattern that is similar to thiopental except for the absence of increased β activity at lower doses. Etomidate has been alleged to produce convulsionlike EEG potentials in epileptic patients without the appearance of myoclonic or convulsant-like motor activity, a property that has been proven useful for intraoperative mapping of seizure foci. Analogous to methohexital, etomidate possesses anticonvulsant properties and has been used to terminate status epilepticus. Etomidate also produces a significant increase of the amplitude of somatosensory-evoked potentials while only minimally increasing their latency. Consequently, etomidate can be used to facilitate the interpretation of somatosensory-evoked potentials when the signal quality is poor.


Etomidate causes minimal cardiorespiratory depression even in the presence of cardiovascular and pulmonary disease.63 The drug does not induce histamine release and can be safely used in patients with reactive airway disease. Consequently, etomidate is considered to be the induction agent of choice for poor-risk patients with cardiorespiratory disease, as well as in those situations in which preservation of a normal blood pressure is crucial (e.g., cerebrovascular disease). However, etomidate does not effectively blunt the sympathetic response to laryngoscopy and intubation unless combined with a potent opioid analgesic.

Etomidate is associated with a high incidence of postoperative nausea and emesis when used in combination with opioids for brief outpatient procedures. In addition, the increased mortality in critically ill patients sedated with an etomidate infusion has been attributed to its inhibitory effect on cortisol synthesis.64 Etomidate inhibits the activity of 11-β-hydroxylase, an enzyme necessary for the synthesis of cortisol, aldosterone, 17-hydroxyprogesterone, and corticosterone. Even after a single induction dose of etomidate,64 adrenal suppression persists for 5 to 8 hours. Although the clinical significance of short-term blockade of cortisol synthesis is not known, the use of etomidate for maintenance of anesthesia has been questioned. Recently, etomidate has been reported to inhibit platelet function, resulting in prolongation of the bleeding time.65 In spite of its side effect profile, etomidate remains a valuable induction drug for specific indications (e.g., in patients with severe cardiovascular and cerebrovascular disease).


Ketamine (Ketalar or Ketaject) is an arylcyclohexylamine that is structurally related to phencyclidine.66 Ketamine is a water-soluble compound with a pKa of 7.5 and is available in 1%, 5%, and 10% aqueous solutions. The ketamine molecule contains a chiral center producing two optical isomers. The S(+) isomer of ketamine possesses more potent anesthetic and analgesic properties despite having a similar pharmacokinetic and pharmacodynamic profile as the racemic mixture (or the R[–] isomer).67,68 Although the S(+)-ketamine is approved for clinical use in Europe, the commonly used solution is a racemic mixture of the two isomers. Ketamine is extensively metabolized by hepatic microsomal cytochrome P450 enzymes and its primary metabolite, norketamine, is one third to one fifth as potent as the parent compound. The metabolites of norketamine are excreted by the kidney as water-soluble hydroxylated and glucuronidated conjugates. Analogous to the barbiturates and propofol, ketamine has relatively short distribution and redistribution half-life values. Ketamine also has a high hepatic clearance rate (1 L/min) and a large distribution volume (3 L/kg), resulting in an elimination half-life of 2–4 hours. The high hepatic extraction ratio suggests that alterations in hepatic blood flow can significantly influence ketamine's clearance rate.

Ketamine produces dose-dependent CNS depression leading to a so-called dissociative anesthetic state characterized by profound analgesia and amnesia, even though patients may be conscious and maintain protective reflexes. The proposed mechanism for this cataleptic state includes electrophysiologic inhibition of thalamocortical pathways and stimulation of the limbic system. Although it is most commonly administered parenterally, oral and intranasal administration of ketamine (6 mg/kg) has been used for premedication of pediatric patients. Following benzodiazepine premedication, ketamine 1 to 2 mg/kg IV (or 4 to 8 mg/kg IM) can be used for induction of anesthesia. The duration of ketamine-induced anesthesia is in the range of 10 to 20 minutes after a single induction dose; however, recovery to full orientation may require an additional 60 to 90 minutes. Emergence times are even longer following repeated bolus injections or a continuous infusion. S(+)-ketamine has a shorter recovery time compared with the racemic mixture. The therapeutic window for maintenance of unconsciousness with ketamine is between 0.6 and 2 µg/mL in adults and between 0.8 and 4 µg/mL in children. Analgesic effects are evident at subanesthetic doses of 0.1 to 0.5 mg/kg IV and plasma concentrations of between 85 and 160 ng/mL. A low-dose infusion of 4 µg/kg/min IV was reported to result in equivalent postoperative analgesia as an IV morphine infusion at 2 mg/hr.

As a result of its NMDA-receptor blocking activity, ketamine should be highly effective for “pre-emptive” analgesia and opioid-resistant chronic pain states.69 Unfortunately, a well-controlled study failed to demonstrate a pre-emptive effect when ketamine was administered prior to the surgical incision (vs. intraoperatively).70 Nevertheless, other studies71,72described a beneficial opioid-sparing effect of small doses of ketamine (75 to 150 µg/kg IV) when administered as an adjuvant during surgery.

An important consideration in the use of ketamine anesthesia relates to the high incidence of psychomimetic reactions (namely, hallucinations, nightmares, altered short-term memory, and cognition) during the early recovery period. The incidence of these reactions is dose-dependent and can be reduced by coadministration of benzodiazepines, barbiturates, or propofol. Ketamine has been traditionally contraindicated for patients with increased ICP or reduced cerebral compliance because it increases CMRO2, CBF, and ICP. However, there is recent evidence that IV induction doses of ketamine actually decrease ICP in traumatic–brain-injury patients during controlled ventilation with propofol sedation.73Prior administration of thiopental or benzodiazepines can blunt ketamine-induced increases in CBF. Because ketamine has antagonistic activity at the NMDA receptor, it may possess some inherent protective effects against brain ischemia. However, ketamine can adversely affect neurologic outcome in the presence of brain ischemia despite its NMDA-receptor blocking activity. Cortical EEG recordings following ketamine induction are characterized by the appearance of fast β activity (30 to 40 Hz) followed by moderate-voltage θ activity, mixed with high-voltage δ waves recurring at 3- to 4-second intervals. At higher dosages, ketamine produces a unique EEG burst-suppression pattern (Fig. 18-5).

Although ketamine-induced myoclonic and seizurelike activity has been observed in normal (nonepileptic) patients, ketamine appears to possess anticonvulsant activity.11,12 Two studies have demonstrated the opioid-sparing effects of low-dose


ketamine (75 to 200 µg/kg) when administered as an adjuvant during anesthesia.71,72 Interestingly, small doses of ketamine have also been used in the treatment of severe depression in patients with chronic pain syndromes.74,75 However, ketamine can produce adverse effects when administered in the presence of tricyclic antidepressants because both drugs inhibit norepinephrine reuptake and could produce severe hypotension, heart failure, and/or myocardial ischemia.75,76

Figure 18-5. Progressive changes in the electroencephalogram (EEG) produced by ketamine. Stages I through III are achieved with racemic ketamine and its S(+)isomer. With R(-)ketamine, Stage II was the maximal EEG depression produced. (Reprinted with permission from Shüttler J, Stanski DR, White PF, et al: Pharmacodynamic modeling of the EEG effect of ketamine and its enantiomers in man. J Pharmacokinet Biopharm 1987; 15: 241.)

Ketamine has well-characterized bronchodilatory activity. In the presence of active bronchospasm, ketamine is considered to be the IV induction agent of choice. Ketamine has been used in subanesthetic dosages to treat persistent bronchospasm in the OR and ICU. It is also used in combination with midazolam to provide sedation and analgesia for asthmatic patients. In contrast to the other IV anesthetics, protective airway reflexes are more likely to be preserved with ketamine. However, it must be emphasized that the use of ketamine does not obviate the need for tracheal intubation in the patient with a full stomach (because tracheal soiling has been reported in this situation). Ketamine causes minimal respiratory depression in clinically relevant doses and can facilitate the transition from mechanical to spontaneous ventilation after anesthesia. However, its ability to increase oral secretions can lead to laryngospasm during “light” anesthesia.

Ketamine has prominent cardiovascular-stimulating effects secondary to direct stimulation of the sympathetic nervous system. Ketamine is the only anesthetic that actually increases peripheral arteriolar resistance. As a result of its vasoconstrictive properties, ketamine can reduce the magnitude of redistribution hypothermia.77 Induction of anesthesia with ketamine often produces significant increases in arterial blood pressure and heart rate. Although the mechanism of the cardiovascular stimulation is not entirely clear, it appears to be centrally mediated. There is evidence to suggest that ketamine attenuates baroreceptor activity via an effect on NMDA receptors in the nucleus tractus solitarius. Because of the increased cardiac work and myocardial oxygen consumption, ketamine negatively affects the balance between myocardial oxygen supply and demand. Consequently, its use is not recommended in patients with severe coronary artery disease. In contrast to the secondary cardiovascular stimulation, ketamine has intrinsic myocardial depressant properties that only become apparent in the seriously ill patient with depleted catecholamine reserves. Because ketamine can also increase pulmonary artery pressure, its use is contraindicated in adult patients with poor right ventricular reserve. Interestingly, the effect on the pulmonary vasculature seems to be attenuated in children.

The renewed interest in ketamine is related to the use of smaller doses (100 to 250 µg/kg) as an adjuvant during anesthesia.78 The anesthetic (sedative) and opioid analgesic-sparing effects of ketamine can reduce ventilatory depression during monitored anesthesia care.79,80,81 The availability of the stereoisomer of ketamine has increased the nonanesthetic adjunctive use of ketamine.82 The anesthetic and analgesic potency of S(+)-ketamine is 3 times greater than R(-)-ketamine and twice that of the racemic mixture (Fig. 18-6), reflecting its fourfold greater affinity at the phencyclidine binding site on the NMDA receptor compared with the R(-) isomer. The therapeutic index of S(+)-ketamine is 2.5 times greater than both the R(-) and the racemic forms. In addition, hepatic biotransformation of S(+)-ketamine occurs 20% faster than that of the R(-) enantiomer, contributing to shorter emergence times and faster return of cognitive function. Both isomers produce similar cardiovascular-stimulating effects and hormonal responses during surgery. Although the incidence of dreaming is similar with S(+)-ketamine and the racemic mixture, subjective mood and patient acceptance are higher with the S(+) isomer.67,68

Figure 18-6. Concentration-response relationship for racemic ketamine and S(+)ketamine in relation to specific clinical end points. The slowing of the median electroencephalogram frequency was used as the effect (end point) and was related to the arterial blood concentrations of ketamine. (Reprinted with permission from Schüttler J, Kloos S, Ihmsen H, et al: Pharmacokinetic-pharmacodynamic properties of S(+) ketamine versus racemic ketamine: A randomized double-blind study in volunteers. Anesthesiology 1992; 77: A330.)


Dexmedetomidine is a highly selective α2-adrenoceptor agonist that has been approved by the Food and Drug Administration for the short-term (<24 hours) sedation of mechanically ventilated patients in the ICU setting. In this setting it appears to offer some clinical advantages because it produces a unique type of sedation-analgesia with less ventilatory depression than the commonly used sedative-hypnotic and opioid analgesic drugs.14 Although dexmedetomidine is being used for sedating patients undergoing diagnostic and therapeutic procedures outside the operating and ICU environments, these represent “off label” uses of this drug.

When used for premedication prior to general anesthesia, dexmedetomidine produced preoperative sedation and anxiolysis comparable to midazolam.15 However, its use led to an increased incidence of intraoperative hypotension and bradycardia compared with the commonly used benzodiazepine compound. When used for premedication prior to regional anesthesia, dexmedetomidine reduced patient anxiety, sympathoadrenal (stress) responses, and perioperative opioid analgesic requirements.83 Hall et al.84 also demonstrated the sedative, amnestic, and analgesic effects of low-dose infusions of dexmedetomidine (0.2 to 0.6 µg/kg/hr).

As an IV adjuvant during induction and/or maintenance of general anesthesia, dexmedetomidine will blunt the acute hemodynamic response to laryngoscopy and intubation.85 It has also been used to facilitate awake fiberoptic intubation.86,87 When used as an anesthetic adjuvant during general anesthesia, dexmedetomidine has been reported to improve perioperative hemodynamic stability in neurosurgical patients,88 and improve postoperative pain control after major surgery.89,90 However, a recent study91 failed to demonstrate any clinically significant improvements in patient outcomes after bariatric surgery despite producing both anesthetic and analgesic-sparing effects.

In summary, dexmedetomidine appears to be a potentially useful adjuvant during local and regional anesthesia. It provides comparable sedation to midazolam92 but has a slower onset and offset of sedation than propofol.93 When administered as an adjuvant during IV regional anesthesia94 it improved the quality of both intra- and postoperative analgesia. Because of its high cost, dexmedetomidine's cost-benefit


ratio as an IV adjuvant during general anesthesia clearly requires further investigation.

Clinical Uses of Intravenous Anesthetics

Induction Agents

The induction characteristics and recommended dosages of the available IV anesthetic agents are summarized in Table 18-2. As a result of differences in pharmacokinetic (e.g., altered clearance and distribution volumes) and pharmacodynamic (altered brain sensitivity) variables, the induction dosages of all IV anesthetics need to be adjusted to meet the needs of individual patients. For example, advanced age, pre-existing diseases (e.g., hypothyroidism, hypovolemia), premedication (e.g., benzodiazepines), and coadministration of adjuvant drugs (e.g., opioids, α2-agonists) decrease the induction dose requirements. When there is concern regarding a possible abnormal response, assessing the effect of a small “test dose” (equal to 10 to 20% of the usual induction dose) will often identify those patients for whom a dosage adjustment is required. Before administering additional medication, adequate time should be allowed for the anesthetic to exert its effect, especially when using drugs with a slow onset of action (midazolam) or in the presence of a “slow” circulation time in elderly patients and those with congestive heart failure.

The clinical uses of propofol have expanded greatly since its introduction into clinical practice in 1989.95 IV administration of propofol results in a rapid loss of consciousness (usually within one arm-to-brain circulation) that is comparable to that of the barbiturates. Although an induction dose of 2.5 mg/kg was initially recommended, the use of smaller induction doses of propofol (1 to 2 mg/kg) has minimized its acute cardiovascular and respiratory depressant effects. Recovery from propofol's sedative-hypnotic effects is rapid with less residual sedation, fatigue (“hangover”), and cognitive impairment than with other available sedative-hypnotic drugs after short surgical procedures. Consequently, propofol has become the IV drug of choice for outpatients undergoing ambulatory surgery.

With benzodiazepines, there is wide variation in the dose-response relationships in unpremedicated elective surgery patients. Compared with midazolam, diazepam and lorazepam have slower onset times to achieve a peak effect and their dose-effect relationship is less predictable. As a result, diazepam and lorazepam are rarely used for induction of general anesthesia. In addition, the slow hepatic clearance of diazepam and lorazepam may contribute to prolonged residual effects (e.g., sedation, amnesia, fatigue) when they are used for premedication. Midazolam has a slightly more rapid onset and may be a useful induction agent for special indications (e.g., when nitrous oxide is contraindicated, or as part of a total IV anesthetic technique). However, when midazolam is used for induction and/or maintenance of anesthesia, return of consciousness takes substantially longer than with other sedative-hypnotic drugs. In spite of its extensive hepatic metabolism, recovery of cognitive function is still slower after midazolam compared with thiopental, methohexital, etomidate, or propofol.

In an effort to optimize the clinical use of midazolam during the induction period, it is used increasingly as a coinduction agent with other sedative-hypnotic drugs (propofol, ketamine). Midazolam 2 to 5 mg IV can provide for increased sedation, amnesia, and anxiolysis during the preinduction period. When midazolam is used in combination with propofol, 1.5 to 2 mg/kg IV,112 or ketamine, 0.75 to 1 mg/kg IV,96 it facilitates the onset of anesthesia and decreases the possibility of intraoperative recall without delaying emergence times. Midazolam also attenuates the cardiostimulatory response to ketamine, as well as its psychomimetic emergence reactions. Use of midazolam, 2 to 3 mg IV, with propofol reduces recall during the induction period; however, larger doses of midazolam (5 mg IV) will delay emergence after brief surgical procedures.

As a result of their side effect profiles, the clinical use of etomidate and ketamine for induction of anesthesia is restricted to specific situations in which their unique pharmacologic profiles offer advantages over other available IV anesthetics. For example, etomidate can facilitate maintenance of a stable blood pressure in high-risk patients with critical stenosis of the cerebral vasculature and in patients with severe cardiac impairment or unstable angina. Ketamine is a useful induction agent for patients with reactive airway disease, as well as for those situations where continued spontaneous ventilation is desirable during surgery.

Table 18-2 Induction Characteristics and Dosage Requirements for the Currently Available Sedative-Hypnotic Drugs



ONSET (sec)



































































0, none; +, minimal; ++, moderate; +++, severe; ↓, decrease; ↑, increase.
From White PF. Textbook of Intravenous Anesthesia. Baltimore, Williams & Wilkins, 1997, pp. 27–46 and 77–92.


Maintenance of Anesthesia

The continued popularity of volatile anesthetics for maintenance of anesthesia is primarily related to their rapid reversibility and ease of administration when using a conventional vaporizer delivery system. The availability of IV drugs with more rapid onset and shorter recovery profiles, as well as user-friendly infusion delivery systems, has facilitated the maintenance of anesthesia with continuous infusions of IV drugs, producing an anesthetic state (namely, TIVA) that compares favorably with the volatile anesthetics. In a comparison of the requirement of postoperative analgesics after inhalation and TIVA techniques, not surprisingly, the postoperative pain was reduced after TIVA.97 For example, in morbidly obese patients undergoing bariatric surgery, the use of TIVA technique was associated with a superior recovery profile compared with a sevoflurane-based inhalation technique.98However, TIVA techniques are more expensive than inhalation or “balanced” anesthetic techniques.42

The traditional intermittent bolus administration of IV drugs results in depth of anesthesia (and analgesia) that oscillates above and below the desired level.99 Because of rapid distribution and redistribution of the IV anesthetics, the high peak blood concentration after each bolus is followed by a rapid decrease, producing fluctuating drug levels in the blood and hence the brain. The magnitude of the drug level fluctuation depends on the size of the bolus dose and the frequency of its administration. Wide variation in the plasma drug concentrations can result in hemodynamic and respiratory instability as a result of changes in the depth of anesthesia or sedation. By providing more stable blood (and brain) concentrations with a continuous IV infusion, it might be possible to improve anesthetic conditions and hemodynamic stability, as well as decreasing side effects and recovery times with IV anesthetics.100 Administration of IV anesthetics by a variable-rate infusion is a logical extension of the incremental bolus method of drug titration, as a continuous infusion is equivalent to the sequential administration of infinitely small bolus doses.

Although an IV anesthetic can be titrated to achieve and maintain the desired clinical effect, a knowledge of basic pharmacokinetic principles is helpful in more accurately predicting the optimal dosage requirements. The required plasma concentration depends on the desired pharmacologic effect (hypnosis, sedation), the concomitant use of other adjunctive drugs (opioid analgesics, muscle relaxants, cardiovascular drugs), the type of operation (superficial, intraabdominal, intracranial), and the patient's sensitivity to the drug (age, drug history, preexisting diseases). Preexisting diseases (cirrhosis, congestive heart failure, renal failure) can markedly alter the pharmacokinetic variables of the highly protein-bound, lipophilic IV anesthetic drugs. In general, children have higher clearance rates, while the elderly have reduced clearance values. Various intraoperative interventions (e.g., laryngoscopy, tracheal intubation, skin incision, entry into body cavities) transiently increase the anesthetic and/or analgesic requirements. Therefore, the infusion scheme should be tailored to provide peak drug concentrations during the periods of most intense stimulation. For specific surgical interventions, the so-called therapeutic window of an IV anesthetic is defined as the blood concentration range required to produce a given effect (Table 18-3). It must be emphasized that the therapeutic window for sedative-hypnotics is markedly influenced by the presence of adjunctive drugs (e.g., opioids, α2-agonists, nitrous oxide).

The use of IV anesthetic techniques requires continuous titration of the drug infusion rate to the desired pharmacodynamic end-point.96 Most anesthesiologists rely on somatic and autonomic signs for assessing depth of IV anesthesia, analogous to the manner in which they titrate the volatile anesthetics. The most sensitive clinical signs of depth of anesthesia appear to be changes in muscle tone (i.e., electromyography [EMG]) and ventilatory rate and pattern.101 However, if the patient has been given muscle relaxants, the anesthesiologist must rely on signs of autonomic hyperactivity (e.g., tachycardia, hypertension, lacrimation, diaphoresis). Unfortunately, the anesthetic drugs (ketamine), as well as adjunctive agents (α2-agonists, beta-blockers, adenosine, calcium channel blockers), can directly influence the cardiovascular response to surgical stimulation. Although the cardiovascular signs of autonomic nervous system hyperactivity may be masked, other autonomic signs (e.g., diaphoresis) and purposeful movements may be more reliable indicators of depth of anesthesia than blood pressure because the latter depends on the ability of the heart to maintain the cardiac output in the face of acute changes in afterload. The heart rate response to surgical stimulation appears to be more useful than the blood pressure response in determining the need for additional analgesic medication. Moreover, it would appear that blood pressure and heart rate responses to surgical stimulation are a less useful guide with IV techniques than with volatile anesthetics. Interestingly, supplementation with a sedative-hypnotic (propofol) was as effective as a potent opioid analgesic in controlling acute autonomic responses during TIVA.102

The clinical assessment of anesthetic depth has become more challenging because IV anesthetic techniques involve a combination of hypnotics, opioids, muscle relaxants, and adjuvant drugs. The interactions between these drugs can result in additive, supra-additive, infra-additive, or even antagonistic effects. An ideal “depth of anesthesia” indicator would integrate the physiologic and neurologic information from all


aspects of the anesthetic state. In the absence of a global cerebral function monitor, the depth of anesthesia device should provide an indication of one or more of the key components of general anesthesia (e.g., hypnosis, analgesia, amnesia, suppression of the stress response, or muscle relaxation). A simple, noninvasive monitor of the depth of anesthesia, which would reliably predict a patient's response to surgical stimulation, would be extremely valuable when using IV anesthetic techniques.

Table 18-3 Therapeutic Blood Concentrations when Intravenous Anesthetics are Infused for Hypnosis or Sedation







10–20 µg/mL

10–20 µg/mL

4–8 µg/mL

4–8 µg/mL


6–15 µg/mL

5–10 µg/mL

1–3 µg/mL

1–3 µg/mL


4–6 µg/mL

2–4 µg/mL

1–2 µg/mL

1–1.5 µg/mL


100–200 ng/mL

50–200 ng/mL

40–100 ng/mL

50–150 ng/mL


500–1000 ng/mL

300–600 ng/mL

100–300 ng/mL

200–350 ng/mL


1–4 µg/mL

0.6–2 µg/mL

0.1–1 µg/mL


NA, not available.
From White PF. Textbook of Intravenous Anesthesia. Baltimore, Williams & Wilkins, 1997, pp. 27 and 77.

The EMG activity of the frontalis muscles increases significantly in patients who move in response to specific surgical stimuli.101 However, EMG changes occur late and their interpretation is obscured by muscle relaxant drugs. The EEG changes depend largely on the type of anesthetic drugs used. Although a common EEG pattern can be recognized with increasing depression of CNS function by sedative-hypnotics and opioid analgesics, there is no characteristic EEG pattern associated with unconscious and amnestic states.103Univariate descriptors of EEG activity appear to be of limited clinical usefulness, and no meaningful correlation could be found between EEG spectral edge frequency and hemodynamic response to surgical stimuli during propofol anesthesia.104 Although EEG variables (spectral edge frequency, median frequency) appear to be useful indicators of the CNS effects of anesthetic and analgesic drugs in the experimental setting, their usefulness in clinical practice is limited because the many confounding factors during the operation (changing drug levels and surgical stimulation). The EEG-based bispectral index (BIS), patient state index, state entropy and response entropy, and cerebral state index represent monitoring approaches that reply on sophisticated computerized algorithms to analyze the spontaneous EEG. All of these cerebral monitoring devices have proved to be a useful indicator of anesthetic (hypnotic) depth. Several recent studies have demonstrated that the use of these indices can improve titration of both IV and volatile anesthetics during surgery, thereby facilitating the recovery process.105 Using EEG-based monitoring can reduce the time required to achieve fast-track eligibility and facilitate earlier discharge home after ambulatory surgery.106,107

An alternative to the spontaneous EEG involves the use of the evoked response of the EEG to sensory stimuli (e.g., auditory-evoked potential monitors). The ability to quantitatively assess the response of the body to varying levels of stimulation (sensory- or auditory-evoked responses) may be useful in improving the assessment of depth of anesthesia.108 Although all sedative-hypnotic drugs affect the brainstem evoked potentials, uncertainty still exists regarding the most useful evoked response(s) to measure. The complexity associated with recording evoked responses is much greater than recording the spontaneous EEG because the value is critically dependent on technical factors (e.g., stimulus intensity, stimulus rate, electrode position), body temperature, as well as the anesthetic drugs. Although most IV anesthetics produce dose-dependent changes in the somatosensory-evoked potentials, the correlation between the acute hemodynamic changes to surgical stimuli and the early auditory-evoked responses is poor. However, the early cortical (midlatency) auditory-evoked response might be useful in detecting awareness under anesthesia. Furthermore, the auditory-evoked potential index may be more discriminating than the spontaneous EEG-based devices in characterizing the transition from wakefulness to unresponsiveness.109

Figure 18-7. Simulated drug level curves when a constant infusion is administered following a “full” loading dose equal to [Cp] × Vdss (Curve A), a smaller loading dose equal to [Cp] × Vc (Curve B), or in the absence of a loading (Curve C). See text for details. (Reprinted with permission from White PF: Clinical uses of intravenous anesthetic and analgesic infusions. Anesth Analg 1989; 68: 161.)

As a result of the availability of more rapid and shorter acting sedative-hypnotics, sophisticated computer technology, and new insights into pharmacokinetic-dynamic interactions, use of TIVA techniques has been steadily increasing throughout the world during the last decade. When using constant rate IV infusions, 4 to 5 half-lives may be required to achieve a steady-state anesthetic concentration (Fig. 18-7). To more rapidly achieve a therapeutic blood concentration, it is necessary to administer a loading (priming) dose and to maintain the desired drug concentration using a maintenance infusion. The loading dose (LD) and initial maintenance infusion rate (MIR) can be calculated from previously determined population kinetic values using the following equations:

LD = Cp (mg/mL) · Vd (mL/kg)

MIR = Cp (mL/kg) · Cl (mL/kg/min)

where Cp = plasma drug concentration, Vd = distribution volume, and Cl = drug clearance.

The use of the smaller central volume of distribution (Vc) for the Vd component of the LD equation will underestimate the LD, whereas use of the larger steady-state volume of distribution (Vdss) will result in drug levels that transiently exceed those that are desired. If a smaller LD is administered, a higher initial MIR will be required to compensate for the drug that is removed from the brain by both redistribution and elimination processes. As the redistribution phase assumes less importance, the MIR will decrease because it becomes solely dependent on the drug's elimination and the desired plasma concentration.

An alternative approach is to begin with a rapid loading infusion with a bolus-elimination transfer scheme that combines three functions, as shown in the following equation:

Input = V1 · Css + Cl · Css + V1 · Css (k21 · e-k21t)

where V1 = distribution volume of the central compartment, Css = steady-state plasma concentration, Cl = drug clearance; k21 = redistribution constant from the central to the peripheral compartment, and k21 = redistribution constant from the peripheral to the central compartment. Implementation of the bolus-elimination transfer infusion scheme requires the use of a microprocessor-controlled pump. If a continuous infusion is to be used in an optimal manner to suppress responses to surgical stimuli, the MIR should be varied according to the individual patient responses (Fig. 18-8). Using an MIR large enough to suppress responses to the most intense surgical stimuli will lead to excessive drug accumulation, postoperative side effects, and delayed recovery. More gradual signs of inadequate or excessive anesthesia can be treated by making 50 to 100% changes in the MIR. Abrupt increases in autonomic activity can be treated by giving a small bolus dose equal to 10 to 25% of the initial loading dose and increasing the MIR.


Figure 18-8. The “landscape” of surgical anesthesia. The surgical stimuli are not constant during an operation; therefore, the plasma concentration of an intravenous anesthetic should be titrated to match the needs of the individual patient. ICU, intensive care unit. (Reprinted with permission from Glass PSA, Shafer SL, Jacobs JR, et al: Intravenous drug delivery systems, Miller's Anesthesia, 4th ed. New York, Churchill Livingstone, 1994, p. 391.)

Despite the marked pharmacokinetic and pharmacodynamic variability that exists among surgical patients, computer programs have been developed that allow reasonable predictions of concentration-time profiles for IV anesthetics and analgesics. This new technology has led to the development of target-controlled infusions (TCI), whereby the anesthesiologist chooses a “target” blood or brain (effective site) drug concentration and the micropressor-controlled infusion pump infuses the drug at the rate needed to rapidly achieve and maintain the desired concentration based on population pharmacokinetic-dynamic data.109 It is obvious that the target concentration must be altered depending on the observed pharmacodynamic effect and the anticipated changes in surgical stimulation.

Closed-loop control based on plasma drug concentrations is not possible because there is no available method to obtain frequent measurements of drug concentrations in real time. A more advanced form of TCI uses a feedback signal generated by simulating a mathematical model of the control process. Clearly, the precision of control achievable with a model-based system is only as accurate as the model. An example of a model-based drug delivery system is the computer-assisted continuous infusion system. An ideal automatic anesthesia delivery device would titrate anesthetic to meet the needs of the individual patient using an acquired feedback signal that accurately reflects the effect site concentration of the drug. The most successful efforts at feedback control of anesthesia have used the BIS and cortical auditory-evoked responses to assess the pharmacodynamic end point.108

The rapid, short-acting sedative-hypnotics (e.g., methohexital, propofol) and opioids (e.g., alfentanil, remifentanil) are better suited for continuous administration techniques than the more traditional anesthetic and analgesic agents because they can be more precisely titrated to meet the unique and changing needs of the individual patient. Traditionally, the elimination half-life of a particular drug has been used in attempting to predict the duration of drug action and the time to awakening after discontinuation of the anesthetic infusion. Using conceptual modeling techniques, it has been shown that the concept of context-sensitive half-time is more appropriate in choosing drugs for continuous IV administration (Fig. 18-9). Because none of the currently available IV drugs can provide for a complete anesthetic state without producing prolonged recovery times and undesirable side effects, it is necessary to administer a combination of IV drugs that provide for hypnosis, amnesia, hemodynamic stability, analgesia, and muscle relaxation. Selecting a combination of drugs with similar pharmacokinetics and compatible pharmacodynamic profiles should improve the anesthetic and surgical conditions. Sedative-hypnotics, opioids, sympatholytics, and muscle relaxants can be successfully administered using continuous infusion TIVA techniques as alternatives to the volatile anesthetics and nitrous oxide.

Sedation in the Operating Room and Intensive Care Unit

The use of sedative-hypnotic drugs as part of a monitored anesthesia care technique in combination with local anesthetics is becoming increasingly popular.110,111,112 During local or regional anesthesia, subhypnotic dosages of IV anesthetics can be infused to produce sedation, anxiolysis, and amnesia and enhance patient comfort. The optimum sedation technique achieves the desired clinical end points without producing perioperative side effects (e.g., respiratory depression, nausea, and vomiting).113 In addition, it should provide for ease of titration to the desired level of sedation while providing for a rapid return to a “clear-headed” state on completion of the surgical procedure.

Sedation also constitutes an essential element in the management of patients in the ICU. The ideal sedative agent for critically ill patients would have minimal depressant effects on the respiratory and cardiovascular systems, would not influence biodegradation of other drugs, and would be independent of renal and hepatic function for its elimination. Recently, the BIS monitor has been used to monitor the depth of sedation in the ICU. For patients undergoing cardiac surgery, rapid reversibility of the sedative state may result in earlier extubation and lead to a shorter stay in the ICU. Although intermittent bolus injections of sedative-hypnotic drugs (e.g., diazepam 2.5 to 5 mg, lorazepam 0.5 to 1 mg, midazolam 1.25


to 2.5 mg) have been administered during local anesthesia, continuous infusion techniques with propofol are becoming increasingly popular for maintaining a stable level of sedation in the OR and ICU settings.

Figure 18-9. Context-sensitive half-time values as a function of infusion duration for intravenous anesthetics, including thiopental, midazolam, diazepam, ketamine, etomidate, and propofol. The context-sensitive half-time for thiopental and diazepam is significantly longer compared with etomidate, propofol, and midazolam, with an increasing infusion duration increase. (Reprinted with permission from Hughes MA, Glass PSA, Jacobs JR: Context-sensitive half-time in multicompartment pharmacokinetic models for intravenous anesthesia. Anesthesiology 1992; 76: 334.)

Benzodiazepines, particularly midazolam, are still the most widely used for sedation in the ICU and for relief of acute situational anxiety during local and regional anesthesia. Midazolam has a steeper dose-response curve than diazepam (Fig. 18-3),111 and therefore careful titration is necessary to avoid oversedation and respiratory depression. Midazolam infusion, 0.05 to 5 mg/kg/min, can be highly effective in providing sedation for hemodynamically unstable patients in the ICU.114 Use of a midazolam infusion has been shown to control agitation and decrease analgesic requirements without producing cardiovascular or respiratory instability. However, marked variability exists for midazolam in the individual patient dose-effect relationships. In addition, marked tolerance may develop to the CNS effects of midazolam with prolonged administration.

Propofol sedation offers advantages over the other sedative-hypnotics (including midazolam) because of its rapid recovery and favorable side effect profile. In addition, the degree of sedation is readily changeable from “light” to “deep” levels by varying the MIR. Following a propofol loading dose of 0.25 to 0.5 mg/kg, a carefully titrated subhypnotic infusion of 25 to 75 µg/kg/min produces a stable level of sedation with minimal cardiorespiratory depression and a short recovery period. Because even low concentrations of propofol can depress the ventilatory response to hypoxia, supplemental oxygen should always be provided. Sedative infusions of propofol produce less perioperative amnesia than midazolam, and propofol-induced amnesia appears to be directly related to the infusion rate.

A small dose of midazolam (2 mg IV) administered immediately before a variable-rate infusion of propofol has also been shown to significantly decrease intraoperative anxiety and recall of uncomfortable events without compromising the rapid recovery from propofol sedation.112 Propofol sedation can also be supplemented with potent opioid and nonopioid analgesics to provide sedation analgesia. In comparing propofol and midazolam for patient-controlled sedation,115 midazolam was associated with less intraoperative recall and pain on injection than propofol, while propofol was associated with less residual impairment of cognitive function. Compared with anesthesiologist-controlled sedation, patient-controlled sedation was associated with fewer propofol dosages, “lighter” levels of sedation, and reduced patient comfort.116 Computer “target-controlled” sedation was also associated with more frequent “oversedation.”117 Finally, music can reduce the propofol dosage requirement during local and regional anesthesia.118

Compared with midazolam in the ICU setting, use of propofol sedation allowed for more rapid weaning of critically ill patients from artificial ventilation.119 It has been suggested that the more rapid weaning after propofol sedation may be cost-saving compared with midazolam when only a limited period of sedation (<48 hours) is required.120 Although a pharmacokinetic study yielded no evidence of a change in receptor sensitivity or drug accumulation over a 4-day study period, preliminary data suggest that tolerance to the CNS effects of propofol may develop with more prolonged administration (>1 week). Increasingly, dexmedetomidine infusions are being used in critically ill patients who require both sedation and analgesia.

Concerns have been raised about elevated lipid plasma levels in patients sedated with standard formulations of propofol over a period of several days, especially when high infusion rates (>6 mg/kg/hr) are used. However, the availability of a propofol formulation with reduced lipid content (Ampofol) should decrease the risk of this problem in the future. Because of conflicting evidence regarding increased mortality as a result of myocardial failure when propofol was used for sedation in the neonatal ICU,121,122,123,124 more safety data are needed to define the indications for the use of prolonged propofol infusions, especially in this patient population. Low-dose ketamine infusions (5 to 25 mg/kg/min) can also be used for sedation and analgesia during local or regional anesthetic procedures, as well as in the ICU setting.67 Midazolam, 0.07 to 0.15 mg/kg infused over 3 to 5 minutes, followed by ketamine, 0.25 to 0.5 mg/kg IV over 1 to 3 minutes, produced excellent sedation, amnesia, and analgesia without significant cardiorespiratory depression.

Another alternative to propofol for sedation outside the OR is dexmedetomidine. The α2-agonist can be infused at rates of 0.25 to 0.75 µg/kg/hr to produce sedation during gynecologic procedures90 and in the ICU. Although the onset of sedation is slower than that of propofol, its opioid-sparing effects reduce the risk of ventilatory depression during procedures outside the OR and may facilitate weaning from mechanical ventilation in the ICU. In the ambulatory setting, recovery from dexmedetomidine's sedative effects is slower than with propofol.


Despite the introduction of new anesthetic agents, it is obvious that many of the goals desirable in an ideal IV anesthetic have not been achieved with any of the currently available drugs. Nevertheless, each of these sedative-hypnotic drugs possesses characteristics that may be useful in specific clinical situations and when combined with an appropriate multimodal analgesic technique (e.g., opioids, nonsteroidal anti-inflammatory drugs, local anesthetics) can provide excellent anesthetic conditions. In situations in which a rapid recovery is not essential (e.g., inpatient procedures), the barbiturates thiopental and methohexital may be the most cost-effective IV anesthetics. Although recovery from anesthesia with methohexital is more rapid than with thiopental (and compares favorably with propofol), excitatory side effects (e.g., myoclonus, hiccoughing) are more prominent than with thiopental or propofol. Methohexital remains the anesthetic of choice for electroconvulsive therapy procedures.

Propofol is the IV drug of choice when a rapid and smooth recovery is essential (e.g., outpatient [ambulatory] anesthesia); increasingly, propofol has been used for all inpatient procedures because of the availability of less costly generic formulations. Recovery from propofol anesthesia is characterized by the absence of a “hangover effect” and reduced postoperative nausea and vomiting symptoms. The cardiovascular-depressant effects produced by propofol appear to be more pronounced than those of thiopental, but can be minimized by careful titration and the use of a variable-rate infusion during the maintenance period. The ability to combine propofol with potent, rapid, and short-acting opioid analgesics (e.g., remifentanil) has facilitated the use of TIVA techniques. Improvements in the TCI delivery systems for IV anesthetics (propofol) and analgesics (remifentanil) will lead to an ever greater acceptance of TIVA techniques in the future.125

When administered alone for induction of anesthesia, benzodiazepines are associated with a slower onset and more prolonged recovery profile. In the usual induction doses, benzodiazepines are associated with minimal cardiorespiratory depression and the reliable amnestic effect may be valuable during TIVA (e.g., for acute sedation prior to induction of anesthesia, for maintenance in the absence of nitrous oxide). When administered in smaller doses, midazolam can also be a valuable adjunct as part of a coinduction and/or maintenance


technique. Other shorter-acting benzodiazepines may be developed in the future (e.g., Ro 48-6791).

Etomidate has minimal cardiovascular and respiratory depressant effects and is therefore an extremely useful induction agent in high-risk patients. It is also occasionally used as an alternative to methohexital for electroconvulsive therapy procedures. The occurrence of pain on injection, excitatory phenomena, adrenocortical suppression, and a high incidence of postoperative nausea and vomiting have limited the use of etomidate to special situations in which its cardiovascular profile offers significant advantages over other available IV anesthetics. A new lipid formulation of etomidate is apparently associated with its fewer side effects and may allow this IV anesthetic to gain wider clinical acceptance in the future.

Ketamine is a unique IV anesthetic that produces a wide spectrum of pharmacologic effects including sedation, hypnosis, somatic analgesia, bronchodilation, and sympathetic nervous system stimulation. Induction of anesthesia can be rapidly achieved following IM injection, making ketamine a valuable alternative to an inhalation induction when IV access is difficult to establish. Ketamine is also indicated for induction of anesthesia in the presence of severe hypovolemic shock, acute bronchospastic states, right-to-left intracardiac shunts, and cardiac tamponade. The adverse hyperdynamic cardiovascular, cerebrodynamic, and psychomimetic effects of ketamine can be minimized by prior administration of a benzodiazepine (e.g., midazolam) or a sedative-hypnotic drug (e.g., thiopental, propofol). Ketamine is also useful as part of coinduction and maintenance anesthetic techniques when avoiding opioid analgesics is desirable. The introduction of the more potent S(+)-ketamine may increase use of ketamine in small doses or by continuous infusion as an IV adjuvant during general anesthesia because of its anesthetic and analgesic-sparing activity.

In summary, IV anesthesia has evolved from being used mainly for induction of anesthesia to providing unconsciousness and amnesia for surgical procedures performed under local, regional, and general anesthesia. New insights into the pharmacokinetics and dynamics of IV anesthetics, as well as the development of computer technology to facilitate IV drug delivery (e.g., TCIs), have greatly enhanced the use of TIVA techniques. The shorter context-sensitive half-life values of the newer sedative-hypnotic drugs make these compounds more useful as continuous infusions for maintenance of anesthesia and sedation. While the search for the ideal IV anesthetic continues, the major challenge for anesthesiologists is to choose the sedative-hypnotic drug that most closely matches the patient's needs in specific clinical situations.


1. Franks NP, Lieb WR: Molecular and cellular mechanisms of general anaesthesia. Nature 1994; 367: 607

2. White PF: Textbook of Intravenous Anesthesia. Baltimore, Williams & Wilkins, 1997, pp 27 and 77

3. Coates KM, Mather LE, Johnson R, et al: Thiopental is a competitive inhibitor at the human alpha-7 nicotinic acet ylcholine receptor. Anesth Anal 2001; 92: 930

4. Rossi MA, Chan CK, Christensen JD, et al: Interactions between propofol and lipid mediator receptors: inhibition of lysophosphatidate signaling. Anesth Analg 1996; 83: 1090

5. Shelly MP: Dexmedetomidine: A real innovation or more of the same? Br J Anaesth 2001; 87: 677

6. Thomas JE, Judith E, Hall, MA et al: The effects of increasing plasma concentrations of Dexmedetomidine in humans. Anesthesiology 2000; 93: 382

7. Fu W, White PF: Dexmedetomidine failed to block the acute hyperdynamic response to electroconvulsive therapy. Anesthesiology 1999; 90: 422

8. Higuchi H, Adachi Y, Dahan A, et al: The interaction between propofol and clonidine for loss of consciousness. Anesth Analg 2002; 94: 886

9. Segal IS, Jarvis DJ, Duncan SR, et al: Clinical efficacy of oral-transdermal clonidine combinations during the perioperative period. Anesthesiology 1991; 74: 220

10. Hughes MA, Jacobs JR, Glass PSA: Context-sensitive half-time in multicompartment pharmacokinetic models for intravenous anesthesia. Anesthesiology 1992; 76: 334

11. Modica PA, Tempelhoff R, White PF: Pro- and anticonvulsant effects of anesthetics (Part I). Anesth Analg. 1990; 70: 303

12. Modica PA, Tempelhoff R, White PF: Pro- and anticonvulsant effects of anesthetics (Part II). Anesth Analg 1990; 70: 433

13. Drummond-Lewis J, Scher C: Propofol: A new treatment strategy for refractory migraine headache. Pain Med 2002; 3: 366

14. Hsu YW, Cortinez LI, Robertson KM, et al: Dexmedetomidine pharmacodynamics: part I: crossover comparison of the respiratory effects of dexmedetomidine and remifentanil in healthy volunteers. Anesthesiology 2004; 101: 1066

15. Scheinin H, Jaakola ML, Sjovall S, et al: Intramuscular dexmedetomidine as premedication for general anesthesia. A comparative multicenter study. Anesthesiology 1993; 78: 1065

16. Hofbauer RK, Fiset P, Plourde G, et al: Dose-dependent effects of propofol on the central processing of thermal pain. Anesthesiology 2004; 100: 386

17. Avram J, Krejcie TC, Henthorn TK: The relationship of age to pharmacokinetics of early drug distribution: The concurrent disposition of thiopental and indocyanine green. Anesthesiology 1990; 72: 403

18. (No authors listed): Randomized clinical study of thiopental loading in comatose survivors of cardiac arrest. Am J Emerg Med 1986; 4: 72

19. Gunaydin B, Babacan A: Cerebral hypoperfusion after cardiac surgery and anesthetic strategies: A comparative study with high-dose fentanyl and barbiturate anesthesia. Ann Thorac Cardiovasc Surg 1998; 4: 12

20. Newman MF, Croughwell ND, White WD, et al: Pharmacologic electroencephalograhic suppression during cardiopulmonary bypass: A comparison of thiopental and isoflurane. Anesth Analg 1998; 86: 246

21. Ding Z, White PF: Anesthesia for electroconvulsive therapy. Anesth Analg 2002; 94: 1351

22. Blouin RT, Conard PF, Gross JB: Time course of ventilatory depression following induction doses of propofol and thiopental. Anesthesiology 1991; 75: 940

23. Vohra A, Thomas AN, Harper NJN, et al: Non-invasive measurement of cardiac output during induction of anaesthesia and tracheal intubation: Thiopentone and propofol compared. Br J Anaesth 1991; 67: 64

24. Bhutada A, Shani R, Rastogi S, et al: Randomised controlled trial of thiopental for intubation in neonates. Arch Dis Child Fetal Neonatal Ed 2000; 82: F34

25. Asik I, Yorukoglu D, Gulay I, et al: Pain on injection of propofol: Comparison of metoprolol with lidocaine. Eur J Anaesthesiol 2003; 20: 487

26. Dubey PK, Prasad SS: Pain on injection of propofol: The effect of granisetron pretreatment. Clini J Pain 2003; 19: 121

27. Piper SN, Rohm KD, Papsdorf M, et al: Dolasetron reduces pain on injection of propofol. Anaesthesiol Intensivmed Notfallmed Schmerzther 2002; 37: 528

28. Agarwal A, Ansari MF, Gupta D, et al: Pretreatment with thiopental for prevention of pain associated with propofol injection. Anaesth Analg 2004; 98: 683

29. Shao X, Li H, White PF, et al: Bisulfite-containing propofol: is it a cost-effective alternative to Diprivan for induction of anesthesia? Anesth Analg 2000; 91: 871

30. Song D, Hamza M, White PF, et al: The pharmacodynamic effects of a lower-lipid emulsion of propofol: A comparison with the standard propofol emulsion. Anesth Analg 2004; 98: 687

31. Song D, Hamza M, White PF, et al: Comparison of a lower-lipid propofol emulsion with the standard emulsion for sedation during monitored anesthesia care. Anesthesiology, 2004; 100: 1072

32. Gibiansky E, Struys MM, Gibiansky L, et al: AQUAVAN injection, a water-soluble prodrug of propofol, as a bolus injection: a phase I dose-escalation comparison with DIPRIVAN (part 1): pharmacokinetics. Anesthesiology 2005; 103: 718

33. Struys MM, Vanluchene AL, Gibiansky E, et al: AQUAVAN injection, a water-soluble prodrug of propofol, as a bolus injection: a phase I dose-escalation comparison with DIPRIVAN (part 2): pharmacodynamics and safety. Anesthesiology 2005; 103: 730

34. Shafer A, Doze VA, Shafer SL, et al: Pharmacokinetics and pharmacodynamics of propofol infusions during general anesthesia. Anesthesiology 1988; 69: 348

35. Sebel PS, Lowdon JD: Propofol: A new intravenous anesthetic. Anesthesiology 1989; 71: 260

36. Doze VA, Westphal LM, White PF: Comparison of propofol with methohexital for outpatient anesthesia. Anesth Analg 1986; 65: 1189

37. Smith I, White PF, Nathanson M, et al: Propofol: An update on its clinical use. Anesthesiology 1994; 81: 1005

38. Glass PSA: Prevention of awareness during total intravenous anesthesia. Anesthesiology 1993; 78: 399

39. Oxorn D, Orser B, Ferris LE, et al: Propofol and thiopental anesthesia: A comparison of the incidence of dreams and perioperative mood alterations. Anesth Analg 1994; 79: 553

40. Pinaud M, Lelausque JN, Chetanneau A, et al: Effects of propofol on cerebral hemodynamics and metabolism in patients with brain trauma. Anesthesiology 1990; 73: 404


41. Yagmurdur H, Cakan T, Bayrak A, et al: The effects of etomidate, thiopental, and propofol in induction on hypoperfusion-reperfusion phenomenon during laparoscopic cholecystectomy. Acta Anaesthesiol Scand 2004; 48: 772

42. Dolk A, Cannerfelt R, Anderson RE, et al: Inhalation anaesthesia is cost-effective for ambulatory surgery clinical comparison with propofol during elective knee arthroscopy. Eur J Anaesthesiol 2002; 19: 88

43. Reddy RV, Moorthy SS, Dierdorf SF, et al: Excitatory effects and electroencephalographic correlation of etomidate, thiopental, methohexital, and propofol. Anesth Analg 1993; 77: 1008

44. Ebrahim ZY, Schubert A, Van Ness P, et al: The effect of propofol on the electroencephalogram of patients with epilepsy. Anesth Analg 1994; 78: 275

45. Sellgren J, Ejnell H, Elam M, et al: Sympathetic muscle nerve activity, peripheral blood flows, and baroreceptor reflexes in humans during propofol anesthesia and surgery. Anesthesiology 1994; 80: 534

46. Lopatka CW, Muzi M, Ebert TJ: Propofol, but not etomidate, reduces desflurane-mediated sympathetic activation in humans. Can J Anaesth 1999; 46: 342

47. Gan TJ, Glass PSA, Howell ST, et al: Determination of plasma concentrations associated with 50% reduction in postoperative nausea. Anesthesiology 1997; 87: 779

48. Krumholz W, Endrass J, Hempelmann G: Propofol inhibits phagocytosis and killing of Staphylococcus aureus and Escherichia coli by polymorphonuclear leukocytes in vitro. Can J Anaesth 1994; 41: 446

49. Crowther J, Hrazdil J, Jolly DT, et al: Growth of microorganisms in propofol, thiopental, and a 1: 1 mixture of propofol and thiopental. Anesth Analg 1996; 82: 475

50. Reves JG, Fragen RJ, Vinik HR, et al: Midazolam—Pharmacology and uses. Anesthesiology 1985; 62: 310

51. Urquhart ML, White PF: Comparison of sedative infusions during regional anesthesia: Methohexital, etomidate, and midazolam. Anesth Analg 1988; 68: 249

52. Ghouri A, Taylor E, White PF: Patient-controlled drug administration during local anesthesia: A comparison of midazolam, propofol, and alfentanil. J Clin Anesth 1992; 4: 476

53. Dingemanse J, van Gerven JMA, Schoemaker RC, et al: Integrated pharmacokinetics and pharmacodynamics of Ro 48-6791, a new benzodiazepine, in comparison with midazolam during first administration to healthy male subjects. Br J Clin Pharmacol 1997; 44: 477

54. Tang J, Wang B, White PF, et al: Comparison of the sedation and recovery profiles of Ro 48-6791, a new benzodiazepine, and midazolam in combination with meperidine for outpatient endoscopic procedures. Anesth Analg 1999; 89: 893

55. Brodgen RN, Goa KL: Flumazenil. Drugs 1991; 42: 1061

56. Ghouri AF, Ramirez Ruiz MA, et al: Effect of flumazenil on recovery after midazolam and propofol sedation. Anesthesiology 1994; 81: 333

57. Flogel CM, Ward DS, Wada DR, et al: The effects of large-dose flumazenil on midazolam-induced ventilatory depression. Anesth Analg 1993; 77: 1207

58. White PF, Shafer A, Boyle WA, et al: Benzodiazepine antagonism does not provoke a stress response. Anesthesiology 1989; 70: 636

59. Doenicke AW, Roizen MF, Kugler J, et al: Reducing myoclonus after etomidate. Anesthesiology 1999; 90: 113

60. Kelsaka E, Karakaya D, Sarihasan B, et al: Remifentanil pretreatment reduces myoclonus after etomidate. J Clin Anesth 2006; 18: 83

61. Van Hamme MJ, Ghoneim MM, Amber JJ: Pharmacokinetics of etomidate, a new intravenous anesthetic. Anesthesiology 1978; 49: 274

62. Wagner RL, White PF, Kan PB, et al: Inhibition of adrenal steroidogenesis by the anesthetic etomidate. N Engl J Med 1984; 310: 1415

63. Gooding JM, Weng JT, Smith RA, et al: Cardiovascular and pulmonary response following etomidate induction of anesthesia in patients with demonstrated cardiac disease. Anesth Analg 1979; 50: 40

64. Wagner RL, White PF: Etomidate inhibits adrenocortical function in surgical patients. Anesthesiology 1984; 61: 647

65. Gries A, Weis S, Herr A, et al: Etomidate and thiopental inhibit platelet function in patients undergoing infrainguinal vascular surgery. Acta Anaesthesiol Scand 2001; 45: 449

66. White PF, Way WL, Trevor AJ: Ketamine—Its pharmacology and therapeutic uses. Anesthesiology 1982; 56: 119

67. White PF, Ham J, Way WL, et al: Pharmacology of ketamine isomers in surgical patients. Anesthesiology 1980; 52: 231

68. White PF, Schuttler J, Shafer A, et al: Comparative pharmacology of the ketamine isomers. Studies in volunteers. Br J Anaesth 1985; 57: 197

69. Rabben T, Skjelbred P, Oye I: Prolonged analgesia effect of ketamine, an N-methyl-D-aspartate receptor inhibitor, in patients with chronic pain. J Pharmocol Ther 1999; 289: 1060

70. Dahl V, Ernoe PE, Steen T, et al: Does ketamine have preemptive effects in women undergoing abdominal hysterectomy procedures? Anesth Analg 2000; 90: 1419

71. Susuki M, Tsueda K, Lansing PS, et al: Small-dose ketamine enhances morphine-induced analgesia after outpatient surgery. Anesth Analg 1999; 89: 98

72. Menigaux C, Fletcher D, Dupont X, et al: The benefits of intraoperative small-dose ketamine on postoperative pain after anterior cruciate ligament repair. Anesth Analg 2000; 90: 129

73. Albanese J, Arnaud S, Rey M, et al: Ketamine decreases intracranial pressure and electroencephalographic activity in traumatic brain injury patients during propofol sedation. Anesthesiology 1997; 87: 1328

74. Berman RM, Capiello A, Anand A, et al: Antidepressant effects of ketamine in depressed patients. Biol Psychiatry 2000; 47: 351

75. Kudoh A, Takahira Y, Katagai H, et al: Small-dose ketamine improves the postoperative state of depressed patients. Anesth Analg 2002; 95: 114

76. Mortero RF, Clark LD, Tolan MM, et al: The effects of small-dose ketamine on propofol sedation: Respiration, postoperative mood, perception, cognition and pain. Anesth Analg 2001; 92: 1465

77. Ikeda T, Kazama T, Sessler DI, et al: Induction of anesthesia with ketamine reduces the magnitude of redistribution hypothermia. Anesth Analg 2001; 93: 934

78. Khos R, Duriex ME: Ketamine: Teaching an old drug new tricks. Anesth Analg 1998; 87: 1186

79. Badrinath S, Avramov MN, Shadrick M, et al: The use of a ketamine-propofol combination during monitored anesthesia care. Anesth Analg 2000; 90: 858

80. Deng XM, Xiao WJ, Luo MP, et al: The use of midazolam and small-dose ketamine for sedation and analgesia during local anesthesia. Anesth Analg 2001; 93: 1174

81. Mortero RF, Clark LD, Tolan MM, et al: The effects of small-dose ketamine on propofol sedation: respiration, postoperative mood, perception, cognition, and pain. Anesth Analg 2001; 92: 1465

82. Sneyd JR: Recent advances in intravenous anaesthesia. Br J Anaesth 2004; 93: 725

83. Jaakola ML: Dexmedetomidine premedication before intravenous regional anesthesia in minor outpatient hand surgery. J Clin Anesth 1994; 6: 204

84. Hall JE, Uhrich TD, Barney JA, et al: Sedative, amnestic, and analgesic properties of small-dose dexmedetomidine infusions. Anesth Analg 1995; 90: 699

85. Yildiz M, Tavlan A, Tuncer S, et al: Effect of dexmedetomidine on haemodynamic responses to laryngoscopy and intubation: perioperative haemodynamics and anaesthetic requirements. Drugs R D 2006; 7: 43

86. Scher CS, Gitlin MC: Dexmedetomidine and low-dose ketamine provide adequate sedation for awake fibreoptic intubation. Can J Anaesth 2003; 50: 607

87. Bergese SD, Khabiri B, Roberts WD, et al: Dexmedetomidine for conscious sedation in difficult awake fiberoptic intubation cases. J Clin Anesth 2007; 19: 141

88. Tanskanen PE, Kyttä JV, Randell TT, et al: Dexmedetomidine as an anaesthetic adjuvant in patients undergoing intracranial tumour surgery: a double-blind, randomized and placebo-controlled study. Br J Anaesth. 2006; 97: 658

89. Arain SR, Ruehlow RM, Uhrich TD, et al: The efficacy of dexmedetomidine versus morphine for postoperative analgesia after major inpatient surgery. Anesth Analg 2004; 98: 153

90. Gurbet A, Basagan-Mogol E, Turker G, et al: Intraoperative infusion of dexmedetomidine reduces perioperative analgesic requirements. Can J Anaesth 2006; 53: 646

91. Tufanogullari B, White PF, Peixoto MP, et al: Dexmedetomidine infusion during laparoscopic bariatric surgery: Effect on recovery outcome variables. Anesth Analg 2008 (in press)

92. Alhashemi JA: Dexmedetomidine vs midazolam for monitored anaesthesia care during cataract surgery. Br J Anaesth 2006; 96: 722

93. Arain SR, Ebert TJ: The efficacy, side effects, and recovery characteristics of dexmedetomidine versus propofol when used for intraoperative sedation. Anesth Analg 2002; 95: 461

94. Memis D, Turan A, Karamanlioglu B, et al: Adding dexmedetomidine to lidocaine for intravenous regional anesthesia. Anesth Analg 2004; 98: 835

95. Smith I, White PF, Nathanson M, et al: Propofol: An update on its clinical use. Anesthesiology 1994; 81: 1005

96. White PF: Comparative evaluation of intravenous agents for rapid sequence induction: Thiopental, ketamine, and midazolam. Anesthesiology 1982; 57: 279

97. Kamata K, Nagata O, Iwakiri H, et al: Comparison of requirement for postoperative analgesics after inhalation and total intravenous anesthesia. Masui 2003; 52: 1200

98. Salihoglu Z, Karaca S, Kose Y, et al: Total intravenous anesthesia versus single breath technique and anesthesia maintenance with sevoflurane for bariatric operations. Obes Surg 2001; 11: 496

99. White PF: Use of continuous infusion versus intermittent bolus administration of fentanyl or ketamine during outpatient anesthesia. Anesthesiology 1983; 59: 294

100. White PF: Clinical uses of intravenous anesthetic and analgesic infusions. Anesth Analg 1989; 68: 161

101. Chang T, Dworsky WA, White PF: Continuous electromyography for monitoring depth of anesthesia. Anesth Analg 1980; 53: 315

102. Monk TG, Ding Y, White PF: Total intravenous anesthesia: effects of opioid versus hypnotic supplementation on autonomic responses and recovery. Anesth Analg 1992; 75: 798

103. Plourde G: Depth of anaesthesia. Can J Anaesth 1991; 31: 270

104. White PF, Boyle WA: Relationship between hemodynamic and electroencephalographic changes during general anesthesia. Anesth Analg 1989; 68: 177

105. White PF: Use of cerebral monitoring during anesthesia: Effect on recovery profile. Best Prac Res Clin Anaesth 2006; 20: 181

106. Song D, van Vlymen J, White PF: Is the bispectral index useful in predicting fast-track eligibility after ambulatory anesthesia with propofol and desflurane? Anesth Analg 1998; 87: 1245


107. White PF, Ma H, Tang J, et al: Does the use of electroencephalographic bispectral index or auditory evoked potential index monitoring facilitate recovery after desflurane anesthesia in the ambulatory setting? Anesthesiology 2004; 100: 811

108. Struys M, Versichelen L, Mortier E, et al: Comparison of spontaneous frontal EMG, EEG power spectrum and bispectral index to monitor propofol drug effect and emergence. Acta Anaesthesiol Scand 1998; 42: 628

109. Schraag S, Bothner U, Gajraj R, et al: The performance of electroencephalogram bispectral index and auditory evoked potential index to predict loss of consciousness during propofol infusion. Anesth Analg 1999; 89: 1311

110. Milne SE, Kenny GN: Future applications for TCI systems. Anaesthesia 1998; 53: 56

111. White PF, Vascones LO, Mathes SA, et al: Comparison of midazolam and diazepam for sedation during plastic surgery. J Plast Reconstruct Surg 1998; 81: 703

112. Taylor E, Ghouri AF, White PF: Midazolam in combination with propofol for sedation during local anesthesia. J Clin Anesth 1992; 4: 213

113. Sä Règo MM, Watcha, MF, White PF: The changing role of monitored anesthesia care in the ambulatory setting. Anesth Analg 1997; 85: 1020

114. Shafer A, Doze VA, White PF: Pharmacokinetic variability of midazolam infusions in critically ill patients. Crit Care Med 1990; 18: 1039

115. Ghouri AF, Taylor E, White PF: Patient-controlled drug administration during local anesthesia: a comparison of midazolam, propofol, and alfentanil. J Clin Anesth 1992; 4: 476

116. Alhashemi JA, Kaki AM: Anesthesiologist-controlled versus patient-controlled propofol sedation for shockwave lithotripsy. Can J Anaesth 2006; 53: 449

117. Burns R, McCrae AF, Tiplady B: A comparison of target-controlled therapy with patient-controlled administration of propofol combined with midazolam for sedation during dental surgery. Anaesthesia 2003; 58: 170

118. Ayoub CM, Rizk LB, Yaacoub CI, et al: Music and ambient operating room noise in patients undergoing spinal anesthesia. Anesth Analg 2005; 100: 1316

119. White PF, Negus JB: Sedative infusions during local or regional anesthesia: A comparison of midazolam and propofol. J Clin Anesth 1991; 3: 32

120. Aitkenhead AR, Pepperman ML, Willatts SM, et al: Comparison of propofol and midazolam for long-term sedation in critically ill patients. Lancet 1989; 2: 704

121. Carrasco G, Molina R, Costa J, et al: Propofol vs. midazolam in short-, medium-, and long-term sedation of critically ill patients: A cost-benefit analysis. Chest 1993; 103: 557

122. McFarlan CS, Anderson BJ, Short TG: The use of propofol infusions in paediatric anaesthesia: A practical guide. Paediatr Anaesth 1999; 9: 209

123. Parke TJ, Steven JE, Rice ASC, et al: Metabolic acidosis and fatal myocardial failure after propofol infusion in children: five case reports. BMJ 1992; 305: 613

124. Martin PH, Murphy BVS, Petros AJ: Metabolic, biochemical and haemodynamic effects of infusion of propofol for long-term sedation of children undergoing intensive care. Br J Anaesth 1997; 79: 276

125. Egan TD, Shafer SL: Target-controlled infusions for intravenous anesthetics. Anesthesiology 2003; 99: 1039

Editors: Barash, Paul G.; Cullen, Bruce F.; Stoelting, Robert K.; Cahalan, Michael K.; Stock, M. Christine