Barbiturates, similar to other CNS depressants, have potent effects on cerebral metabolism. Several studies in the 1970s demonstrated the effect of barbiturates as a dose-related depression of the CMRO₂, which produces a progressive slowing of the EEG, a reduction in the rate of adenosine triphosphate consumption, and protection from incomplete or focal cerebral ischemia [53, 54]. When the results of the EEG became isoelectric, a point at which cerebral metabolic activity is approximately 50% of baseline, no further decrements in CMRO₂ occurred [55]. These findings support the hypothesis that metabolism and function are coupled. However, it must be noted that it is the portion of metabolic activity concerned with neuronal signaling and impulse traffic that is reduced by barbiturates, not that portion corresponding to basal metabolic function. The only way to suppress baseline metabolic activity concerned with cellular activity is through hypothermia. Thus, the effect of barbiturates on cerebral metabolism is maximized at a 50% depression of cerebral function in which less oxygen is required as CMRO₂ is diminished, leaving all metabolic energy to be used for the maintenance of cellular integrity [55]. This may be of importance to the elderly patient undergoing neurosurgery for aneurysm clipping or carotid endarterectomy in which focal ischemia may occur.

With the reduction in CMRO₂, there is a parallel reduction in cerebral perfusion, which is seen in decreased CBF and ICP (Table 17.1). With reduced CMRO₂, cerebral vascular resistance increases and CBF decreases [56]. However, for thiopental, the ratio of CBF to CMRO₂ is unchanged. Thus, the reduction in CBF after the administration of barbiturates causes a concurrent decrease in ICP. Furthermore, even though the MAP decreases, barbiturates do not compromise the overall CPP, because the CPP = MAP – ICP. In this relationship, ICP decreases more relative to the decrease in MAP after barbiturate use, thus preserving CPP. This is in contrast to propofol, which has a greater likelihood in the elderly patient of decreasing MAP to an extent that may compromise CPP, as noted above [57].

Barbiturates produce CNS effects when they cross the blood –brain barrier. There are several well-known factors that help to determine the rapidity with which a drug enters the cerebral spinal fluid (CSF) and brain tissue. These factors include the degree of lipid solubility, degree of ionization, level of protein binding, and the plasma drug concentration. Drugs with high lipid solubility and low degree of ionization rapidly cross the blood–brain barrier, producing a fast onset of action. Approximately 50% of thiopental is nonionized at physiologic pH, which accounts in part for the rapid accumulation of thiopental in the CSF after intravenous administration. Protein binding also affects the onset of action in the CNS. Barbiturates are highly bound to albumin and other plasma proteins. As only unbound drug (free drug) can cross the blood–brain barrier, an inverse relationship exists between the degree of plasma protein binding and the rapidity of drug passage across the blood–brain barrier.

The final factor governing the rapidity of drug penetration of the blood–brain barrier is the plasma drug concentration. Simply because of concentration gradient, higher levels of drug concentrations in the plasma produce greater amounts of drug that diffuses into the CSF and the brain. The two primary determinants of the plasma concentration are the dose administered and the rate (speed) of administration. The higher the dose and the more rapid its administration, the more rapid is the effect. This is of particular importance in the elderly patient who may have a reduced central volume of distribution and thus require a reduced dose of thiopental to reach the same plasma concentration as a younger adult [58].

Cardiovascular depression from barbiturates is a result of both central and peripheral (direct vascular and cardiac) effects [59]. The hemodynamic changes produced by barbiturates have been studied in healthy individuals and in patients with heart disease. The primary cardiovascular effect of barbiturate induction is peripheral vasodilation that results in a pooling of blood in the venous system. A decrease in contractility is another effect, which is related to reduced availability of calcium to the myofibrils. There is also an increase in heart rate. Mechanisms for the decrease in cardiac output include (1) direct negative inotropic action, (2) decreased ventricular filling because of increased capacitance, and (3) transiently decreased sympathetic outflow from the CNS. The increase in heart rate (10–36%) that accompanies thiopental administration probably results from the baroreceptor-mediated sympathetic reflex stimulation of the heart in response to the decrease in output and pressure. Thiopental produces dose-related negative inotropic effects, which seem to result from a decrease in calcium influx into the cells with a resultant diminished amount of calcium at sarcolemma sites. The cardiac index is unchanged or is reduced, and the MAP is maintained or is slightly reduced [59]. Thiopental infusions and lower doses tend to be accompanied by smaller hemodynamic changes than those noted with rapid bolus injections; however, their use in current practice is exceedingly limited if not gone altogether [25].

The increase in heart rate encountered in patients with coronary artery disease anesthetized with thiopental (1–4 mg/kg) is potentially deleterious because of the obligatory increase in myocardial oxygen consumption (MVO₂) that accompanies the increased heart rate. Patients who have normal coronary arteries have no difficulty in maintaining adequate coronary blood flow to meet the increased MVO₂ [60]. When thiopental is given to hypovolemic patients, there is a significant reduction in cardiac output (69%) as well as a substantial decrease in blood pressure [45–48]. Patients without adequate compensatory mechanisms, therefore, may have serious hemodynamic disturbance with thiopental induction. All of these concerns are of particular importance in geriatric patients, because they are more likely to have clinically significant coronary artery disease, are more likely to be intravascularly hypovolemic, and their compensatory mechanisms to maintain heart rate and blood pressure may be reduced because of age-related alterations and pharmacologic treatments such as beta-blockers or calcium channel blockers. Thus, it is of prime importance in the elderly patient to understand proper dose reduction (discussed further) and the effects of the rate of administration of an induction bolus. If these are not heeded, it becomes common to have significant hypotension in the geriatric patient with the need to administer vasopressors after induction , a practice that can be avoided if a proper understanding of the above principle is gained.

Barbiturates produce dose-related central respiratory depression. There is also a significant incidence of transient apnea after their administration for induction of anesthesia [15]. The evidence for central depression is a correlation between EEG suppression and minute ventilation [61]. With increased anesthetic effect, there is diminished minute ventilation. The time course of respiratory depression has not been fully studied, but it seems that peak respiratory depression (as measured by the slope of CO₂ concentration in the blood) and minute ventilation after delivery of thiopental 3.5 mg/kg occurs 1–1.5 min after administration. These parameters rapidly return to predrug levels, and within 15 min the drug effects are barely detectable [62]. Of note, respirations are lost sooner and return later than that seen with propofol. Patients with chronic lung disease are slightly more susceptible to the respiratory depression of thiopental. The usual ventilatory pattern with thiopental induction has been described as “double apnea.” The initial apnea that occurs during drug administration lasts a few seconds and is succeeded by a few breaths of reasonably adequate tidal volume, which is followed by a longer apneic period. During the induction of anesthesia with thiopental, ventilation must be assisted or controlled to provide adequate respiratory exchange. This is of particular concern in the elderly patient who will have an increased closing capacity, which will produce a shorter time to become hypoxemic, as compared with the young adult patient [21].

All benzodiazepines have hypnotic, sedative, anxiolytic, amnesic, anticonvulsant, and centrally produced muscle relaxant properties. The drugs differ in their potency and efficacy with regard to each of these PD actions. The binding of benzodiazepines to their respective receptors is of high affinity, stereospecific, and able to fully saturate the receptors; the order of receptor affinity (thus potency) of the three agonists is lorazepam > midazolam > diazepam. Midazolam is approximately three to six times as potent as diazepam [74].

The mechanism of action of benzodiazepines is reasonably well understood [75–77]. The interaction of ligands with the benzodiazepine receptor represents an example in which the complex systems of biochemistry, molecular pharmacology, genetic mutations, and clinical behavioral patterns are seen to interact. Through recent genetic studies, the GABA_A subtypes have been found to mediate the different effects (amnesic, anticonvulsant, anxiolytic, and sleep) [78]. Sedation, anterograde amnesia, and anticonvulsant properties are mediated via a1 receptors [78], and anxiolysis and muscle relaxation are mediated by the a2 GABA_A receptor [78]. The degree of effect exerted at these receptors is a function of plasma level. By using plasma concentration data and PK simulations, it has been estimated that a benzodiazepine receptor occupancy of less than 20% may be sufficient to produce the anxiolytic effect, whereas sedation is observed with 30–50% receptor occupancy and unconsciousness requires 60% or higher occupation of benzodiazepine agonist receptors [78].

Agonists and antagonists bind to a common (or at least overlapping) area of the benzodiazepine portion of the GABA_A receptor by forming differing reversible bonds with it [79, 80]. The effects of midazolam can be reversed by use of flumazenil, a benzodiazepine antagonist that occupies the benzodiazepine receptor, but produces no activity and therefore blocks the actions of midazolam. The duration of reversal is dependent on the dose of flumazenil and the residual concentration of midazolam.

The onset and duration of action of a bolus intravenous administration of midazolam largely depends on the dose given and time at which the dose is administered; the higher the dose given over a shorter time (bolus), the faster the onset. Midazolam has a rapid onset (usually within 30–60 s) of action. The time to establish equilibrium between plasma concentration and EEG effect of midazolam is approximately 2–3 min and is not affected by age [81]. Like onset, the duration of effect is related to lipid solubility and blood level [82]. Thus, termination of effect is relatively rapid after midazolam administration. But some physicians have a general sense that midazolam is associated with the production of confusion even after the termination of sedation. This has been reported in prior studies and case reports [83, 84]. However, a more recent study suggests that this might not be the case, particularly at lower doses [85]. Taken together, these data seem to suggest that single, lower doses of midazolam (0.03 mg/kg) will not cause confusion, whereas higher doses (0.05–0.07 mg/kg) along with an infusion of midazolam will have a greater association with confusion in the geriatric patient, as opposed to that seen with the use of a low-dose propofol infusion [83–85].

Midazolam, like most intravenous anesthetics and other benzodiazepines, produces dose-related central respiratory system depression. The peak decrease in minute ventilation after midazolam administration (0.15 mg/kg) is almost identical to that produced in healthy patients given diazepam (0.3 mg/kg) [86]. Respiratory depression is potentiated with opioids and must be carefully monitored in elderly patients getting both. The peak onset of ventilatory depression with midazolam (0.13–0.2 mg/kg) is rapid (about 3 min), and significant depression remains for about 60–120 min [63, 87]. The depression is dose related. The respiratory depression of midazolam is more pronounced and of longer duration in patients with chronic obstructive pulmonary disease, and the duration of ventilatory depression is longer with midazolam (0.19 mg/kg) than with thiopental (3.3 mg/kg) [63].

At sufficient doses, apnea occurs with midazolam as with other hypnotics. The incidence of apnea after thiopental or midazolam when these drugs are given for induction of anesthesia is similar. In clinical trials, apnea occurred in 20% of 1130 patients given midazolam for induction and 27% of 580 patients given thiopental [72]. Apnea is related to dose and is more likely to occur in the presence of opioids. Older age [88] debilitating disease and other respiratory depressant drugs probably also increase the incidence and degree of respiratory depression and apnea with midazolam.

Midazolam alone has modest hemodynamic effects. The predominant hemodynamic change is a slight reduction in arterial blood pressure, resulting from a decrease in systemic vascular resistance. The hypotensive effect is minimal and about the same as seen with thiopental [89]. Despite the hypotension, midazolam, in doses as high as 0.2 mg/kg, is safe and effective for induction of anesthesia even in patients with severe aortic stenosis. The hemodynamic effects of midazolam are dose related: the higher the plasma level, the greater the decrease in systemic blood pressure [90]; however, there is a plateau plasma drug effect above which little change in arterial blood pressure occurs. The plateau plasma level for midazolam is 100 ng/mL [90]. Heart rate, ventricular filling pressures, and cardiac output are maintained after induction of anesthesia with midazolam.

The stimulation of endotracheal intubation and surgery are not blocked by midazolam [91]. Thus, adjuvant anesthetics, usually opioids, are often combined with benzodiazepines. The combination of benzodiazepines with opioids and nitrous oxide has been investigated in patients with ischemic and valvular heart diseases [92–95]. Although the addition of nitrous oxide to midazolam (0.2 mg/kg) has trivial hemodynamic consequences, the combination of benzodiazepines with opioids does have a synergistic effect [96]. The combination of midazolam with fentanyl [93] or sufentanil [95] produces greater decreases in systemic blood pressure than does each drug alone.

Midazolam is used for sedation as preoperative premedication [110], intraoperatively during regional or local anesthesia, and postoperatively for sedation. The anxiolysis, amnesia, and elevation of the local anesthetic seizure threshold are desirable benzodiazepine actions for regional anesthesia. It should be given by titration for this use; endpoints of titration are adequate sedation or dysarthria and maintained ventilation. The onset of action is relatively rapid with midazolam, usually with peak effect reached within 2–3 min of administration. There is an excellent correlation within individuals of sedation score to blood level, but between individuals there is considerable variation in blood level and sedation [107]. The duration of action primarily depends on the dose used. There is often a disparity in the level of sedation compared with the presence of amnesia (patients can be seemingly conscious and coherent, yet they are amnesic for events and instructions). The degree of sedation and the reliable amnesia, as well as preservation of respiratory and hemodynamic function, are better overall with midazolam than with other sedative-hypnotic drugs used for conscious sedation with the possible exception of propofol [88]. However, in elderly patients, inadvertent overdose for sedation during endoscopy has been reported to cause a decline in cognitive function [111]. Also, in a small percentage of patients (1.4%), a paradoxical reaction successfully treated with flumazenil has been reported during endoscopy resulting in agitation rather than sedation [112]. When using midazolam for sedation, a sedation score such as the Ramsay Sedation Scale is commonly used whereas the BIS in elderly patients might result in less-reliable sedation monitoring [113]. When midazolam is compared with propofol for sedation, the two are generally similar except that emergence or wake-up is more rapid with propofol [88]. Midazolam and propofol require close medical supervision because of potential respiratory depression and hypotension [108, 109]. Despite the wide safety margin with midazolam, respiratory function must be monitored when it is used for sedation to prevent undesirable degrees of respiratory depression [63, 114, 115]. This is especially true in the geriatric patient and when opioids are also given [88, 116]. There may be a slight synergistic action between midazolam and spinal anesthesia with respect to ventilation [117]. Thus, the use of midazolam for sedation during regional and epidural anesthesia requires vigilance with regard to respiratory function, when these drugs are given with opioids. Sedation for longer periods, for example, in the ICU, is accomplished with benzodiazepines. Prolonged infusion will result in accumulation of drug and, in the case of midazolam, significant concentration of the active metabolite. The chief advantages are the amnesia and hemodynamic stability, and the disadvantage, compared with propofol, is the longer dissipation of effects when infusion is terminated.

With midazolam, induction of anesthesia is defined as unresponsiveness to command and loss of the eyelash reflex. When midazolam is used in appropriate doses, induction occurs less rapidly than with thiopental or propofol [72], but the amnesia is more reliable. Numerous factors influence the rapidity of action of midazolam. These factors are dose, speed of injection, degree of premedication, age, American Society of Anesthesiologists (ASA) physical status, and concurrent anesthetic drugs [72, 110]. In a well-premedicated, healthy patient, midazolam (0.2 mg/kg given in 5–15 s) will induce anesthesia in 28 s. The BIS may be used to monitor depth with midazolam and adjuvant drugs during anesthesia [118, 119]. Emergence time is related to the dose of midazolam as well as the dose of adjuvant anesthetic drugs [72]. Emergence is more prolonged with midazolam than with propofol [120, 121]. This difference accounts for some anesthesiologists’ preference for propofol induction for short operations. The best method of monitoring depth with midazolam is use of the BIS [122]. Over the years since its introduction into practice, midazolam has become much less used for induction and maintenance of anesthesia in part because of the delirium found with higher doses of the drug [123].

The amnesic period after an anesthetic dose is about 1–2 h. Infusions of midazolam have been used to ensure a constant and appropriate depth of anesthesia. Experience indicates that a plasma level of more than 50 ng/mL when used with adjuvant opioids (e.g., fentanyl) and/or inhalation anesthetics (e.g., nitrous oxideand volatile anesthetics) is achieved with a bolus loading dose of 0.05–0.15 mg/kg and a continuous infusion of 0.25–1 μg/ kg/min [124]. This is sufficient to keep the patient asleep and amnesic but arousable at the end of surgery. Midazolam when compared with dexmedetomidine MAC sedation during endoscopic nasal surgery produces more amnesia than dexmedetomidine [125]. Lower infusion doses almost certainly are required in elderly patients and with certain opioids.

Etomidate is a hypnotic drug that is structurally unrelated to all other induction medications. It contains a carboxylated imidazole ring that provides water solubility in an acidic milieu and lipid solubility at physiologic pH. It is dissolved in propylene glycol, which often causes pain on injection. Etomidate works by depressing the reticular activating system and enhances the inhibitory effects of GABA by binding to a subunit of the GABA_A receptor and thereby increasing its affinity for GABA. However, unlike the barbiturates, which have global depressant effects on the reticular activating system, etomidate has some disinhibitory effects, which accounts for the 30–60% rate of myoclonus with administration. Interestingly, one study has shown that this unwanted side effect can be reduced with pretreatment, similar to a defasciculating dose of neuromuscular blocking drugs (NMBDs) [130].

Etomidate induces changes in CBF, metabolic rate, and ICP to the same extent as thiopental and propofol. However, because this is not the result of a large reduction in arterial blood pressure, CPP is well maintained [130]. This is of particular importance in the elderly person who is at risk for ischemic stroke secondary to carotid occlusion. Etomidate has EEG changes similar to thiopental with a biphasic pattern of activation followed by depression. However, etomidate has been shown to activate somatosensory evoked potentials [57]. Additionally, etomidate causes myoclonic movements after induction, a disturbing effect of unknown significant, in approximately 75% of patients but this is not evidence of an insufficient induction dose [131]. Of note, etomidate does have a higher rate of postoperative nausea and vomiting associated with it than with the other intravenous induction drugs [132]. Finally, there are no PD changes with age with respect to etomidate as measured by EEG [130].

Unlike propofol, etomidate has minimal effects on the cardiovascular system. There is a slight decline in the arterial blood pressure secondary to a mild reduction in the systemic vascular resistance. Etomidate does not seem to have direct myocardial depressant effects, because myocardial contractility, heart rate, and cardiac output are usually unchanged [133]. Etomidate does not cause histamine release. These aspects of the PD of etomidate make it very useful in the patient with compromised intravascular volume, coronary artery disease, or reduced ventricular function, as is often encountered in the elderly patient.

Etomidate causes less respiratory depression than benzodiazepines, barbiturates, or propofol in induction doses. In fact, even an induction dose of etomidate often does not cause apnea [61]. This fact, combined with its minimal cardiovascular effects, makes etomidate a very useful drug in the setting of a hemodynamically brittle elderly patient with a possible difficult airway and little respiratory reserve.

Induction doses of etomidate temporarily inhibit the synthesis of cortisol and aldosterone, lasting approximately 12–18 h after a single bolus dose [134]. However, the clinical significance of this has been debated in septic patients. Some studies have shown an increased mortality with a single bolus dose in septic patients, whereas others have not [135]. Alternatively, long-term infusions or closely repeated exposures can lead to adrenocortical suppression, which may be associated with an increased susceptibility to infection and an increased mortality rate in the critically ill patient [136].