Duration of Effect

The elimination half-time is the time taken for the amount of drug in the body to decrease by 50%. This parameter is often quoted when describing the pharmacokinetic properties of a drug. However, elimination half-time only rarely reflects the duration of effect. When a drug is given intravenously, it is rapidly distributed to a central “virtual” compartment consisting of plasma, interstitial fluid, and organs with high blood flow (brain, heart, liver, kidneys). This central compartment is in equilibrium with the effect site and with the organs of elimination—the liver and the kidneys. Drugs with low lipid solubility, high ionization, and high protein binding tend to be confined to this central compartment and typically have a small steady-state volume of distribution (V_SS; Fig. 4-1). Such drugs can be described using a one-compartment model (V₁). For drugs that obey one-compartment kinetics, the duration of effect may be related to the elimination half-time. Examples include aminoglycosides and neuromuscular blocking drugs. However, highly lipid-soluble drugs, including most sedative-hypnotics and opioid analgesics, display multicompartment kinetics in which drugs are redistributed from the central compartment to one or two peripheral compartments (V₂, V₃; see Fig. 4-1).

Figure 4.1 Pharmacokinetic models. The top panel shows one- and three-compartment models. With a one-compartment model, drug is delivered to the central compartment, which has an apparent volume of distribution (V). Drug is eliminated from the body by the liver and the kidneys (or plasma enzymes). The clearance of a drug (CL) is the volume of plasma from which the drug is completely eliminated per unit time; it is usually reported in ml/min. For a one-compartment model, the volume of distribution and the clearance of the drug determine the elimination half-time (T1/2elim):T1/2elim = Ln(2)V/CL, where Ln(2) is the natural logarithm of 2(0.692). For drugs that obey one-compartment kinetics, the elimination half-time is an important determinant of the duration of effect of the drug. The central compartment is in equilibrium with the effect-site compartment (Ve). Keo is the rate constant for equilibrium between the central compartment and the effect site. The time taken for the effect-site concentration to increase to half the central compartment concentration is the effect-site half-time (T1/2keo); it is related to Keo as T1/2keo = Ln(2)/Keo. Effect-site half-time is an important determinant of the speed of onset of a drug.

For a three-compartment model, drug equilibrates between the central compartment (V₁) and two peripheral compartments (V₂ and V₃). The steady-state volume of distribution (V_SS) is given by: V_SS = V₁ + V₂ + V₃. Rate constants and distribution half-times between the central and peripheral compartments can be calculated. The bottom panel shows plasma-concentration-time curves for a drug that displays three-compartment kinetics. Following a single intravenous (IV) dose (curve A), the concentration initially falls very rapidly due to redistribution from the central compartment; the clinical effect is relatively short. Following multiple intravenous bolus doses (curve B), or prolonged infusion (curve C), the peripheral compartments become progressively saturated with drug. With each dose, the rate of decline of the plasma concentration decreases until the peripheral compartments are fully saturated. Once drug administration ceases, the plasma concentration slowly falls, and the clinical effect is prolonged.

In a three-compartment model, the drug is delivered to the central compartment and then distributed to two peripheral compartments: the second compartment is composed of tissues with intermediate blood flow (e.g., muscle); and the third compartment is composed of tissues with low blood flow, principally fat. For highly lipid-soluble drugs, this third compartment can provide a huge reservoir into which drug molecules slowly equilibrate. For example, the V_SS of propofol, a highly lipid-soluble drug, is more than 400 l, a volume much greater than the volume of a human being, implying that the drug is concentrated in the lipid-rich V₃. This compares with a V_SS of just 15 l for pancuronium, a poorly lipidsoluble drug that is confined mainly to the extracellular fluid. The offset of highly lipid-soluble drugs, such as fentanyl, midazolam, and propofol, is not due to their elimination from the body but to redistribution from the central compartment to peripheral compartments. However, following prolonged infusion, the peripheral compartment can become saturated with a drug. Once the infusion has stopped, as the concentration in the central compartment falls due to hepatic metabolism, the drug moves back into the central compartment from the peripheral compartments, prolonging the clinical effect.

For drugs that display multicompartment kinetics, distribution and context-sensitive half-times are more useful concepts than elimination half-time. The distribution half-time is the time taken for the concentration within the central compartment to fall by 50%. Following a single intravenous dose, the distribution half-time determines the duration of effect of the drug (see Fig. 4-1). The context-sensitive half-time is the time taken for the effect-site concentration to fall by 50% following discontinuation of an intravenous infusion.¹ Because a drug accumulates in the peripheral compartments over time, the context-sensitive half-time changes depending on the duration of infusion (Fig. 4-2). The context-sensitive half-time provides some indication of the duration of effect of the drug following both short- and long-term infusions (or repeated bolus doses). The percentage of decrease in concentration required for recovery from a drug’s effect is not necessarily 50%.

Figure 4.2 Context-sensitive half-times for various sedatives and opioids as a function of time.

(Redrawn from Hughes MA et al: Anesthesiology 76:334, 1992.)

Highly lipid-soluble drugs typically undergo extensive hepatic metabolism to produce water-soluble metabolites that are then excreted by the kidney. For a number of sedative and analgesic drugs, these metabolites are pharmacologically active and can prolong the clinical effect, particularly in the presence of renal dysfunction.

Multicompartment kinetics, prolonged infusion, the presence of active metabolites, and concomitant hepatic and renal dysfunction collectively explain why the duration of effect of many sedative and analgesic medications is greatly prolonged in critically ill patients.

Onset of Action

The speed of onset of a drug depends on multiple factors; two that are of clinical importance for intravenously administered sedatives and analgesics are (1) the speed with which the drug is distributed within the central compartment and (2) the half-time for equilibration between the central and effect-site compartments (T_1/2keo; see Fig. 4-1). Low cardiac output slows drug distribution within the central compartment and can greatly prolong the onset time. Thus, when administering potent sedative or analgesic medications to patients with low cardiac output, it is essential to give a small initial dose and wait a longer than normal time for the clinical effect to occur. Values for T_1/2keo vary among drugs. For instance, the T_1/2keo for morphine, fentanyl, and remifentanil are 17 minutes, 6.6 minutes, and 1.16 minutes, respectively. Therefore, morphine will have a slower onset of action than fentanyl and remifentanil.

Loading and Maintenance Doses

If a drug is administered by constant infusion or repeat doses it takes five (elimination) half-times to achieve 97% of the steady-state concentration (see Fig. 4-1). Thus, for a drug with a relatively long half-time, it is often useful to give a loading dose to rapidly increase the plasma concentration to within the therapeutic range. The loading dose (LD) depends on the volume of distribution (V) of the drug and the desired plasma concentration (C_P):

1

Note that loading dose does not depend on clearance; thus, the loading dose of some renally eliminated drugs, such as gentamicin, do not need to be reduced in patients with renal failure. However, for drugs that display multicompartment kinetics, it is important to be clear which loading dose is used in the calculation: V₁ or V_SS. A loading dose based on V_SS will initially produce very high plasma levels because the drug will be delivered only to the central compartment volume (V₁). To avoid this problem, multiple small loading doses based on V₁ may be required. This is the approach that is recommended for amiodarone in Chapter 3.

The dosing rate for maintenance therapy is equal to the rate of elimination of the drug, which at steady state is dependent on the clearance (CL) and plasma concentration:

2

Thus, maintenance dose is dependent on clearance, and the maintenance dose of renally eliminated drugs (or drugs with active metabolites) must be reduced in patients with renal failure.

Indications for Sedation

Ventilated patients require sedation to tolerate endotracheal intubation and mechanical ventilation, facilitate nursing care, minimize the stress response, reduce oxygen consumption, diminish recall of unpleasant experiences, and prevent the development of posttraumatic stress disorder.^2,3 Less commonly, sedation is indicated in extubated patients for the treatment of anxiety or delirium.

Mechanical ventilation, particularly using lungprotective strategies with long inspiratory times and permissive hypercapnia, is poorly tolerated by nonsedated patients and can result in ventilator dysynchrony (Chapter 29) and the sensation of dyspnea. Distressed patients may become tachycardic and hypertensive, which can exacerbate or provoke myocardial ischemia and bleeding. Such patients may also self-extubate or pull out their intravascular lines and surgical drains. A critically unwell patient commonly benefits from deep sedation, occasionally accompanied by neuromuscular blockade, during the acute phase of an illness. However, most patients do not require paralysis, only a level of sedation sufficient to allow tolerance of endotracheal intubation. Sedation of agitated patients should be commenced only after providing adequate analgesia and treating reversible physiologic causes.³

Adverse Effects of Sedation

Excessive sedation contributes to hypotension and delays awakening, needlessly prolonging the duration of mechanical ventilation.⁴ Sedation may also mask the development of intracranial, intrathoracic, or intraabdominal complications. The reduction in sympathetic tone that follows the administration of sedative and analgesic drugs can cause important hypotension. Hypotension is particularly marked in patients with high levels of endogenous catecholamines such as those that occur in the settings of hypovolemia and acute heart failure. Following prolonged administration of some sedatives (and opioid analgesics), tolerance may develop such that increased doses are required to elicit the same clinical effect. Abrupt discontinuation of certain sedatives, notably benzodiazepines, in a patient who has developed tolerance, may provoke a withdrawal syndrome (see discussion under subsequent heading Benzodiazepines). For these reasons, the need for sedation should be evaluated on an on-going basis and the depth of sedation regularly assessed.

Assessment of Sedation

If the clinical state allows, sedation should be stopped each day until the patient shows signs of awakening. Sedation can then be restarted if still indicated. A number of sedation scoring systems have been developed to quantify the depth of sedation and allow sedative drugs to be titrated to effect. Some commonly used scoring systems are shown in Table 4-1. Although primarily a system for monitoring neurologic function after trauma, the Glasgow Coma Scale (Table 4-2) may also be used to monitor sedation, although much information is lost in patients who are intubated and cannot respond verbally. One option in ventilated patients is to revise the Glasgow Coma Scale score to a maximum of 10, with the annotation that the patient is intubated.

Table 4-2 Glasgow Coma Scale

The Bispectral Index (BIS) monitor, a highly processed electroencephalogram, is widely used to assess depth of anesthesia during surgery and has been used on a limited basis to monitor depth of sedation in the ICU.^5,6 A potential problem with this monitor in the ICU environment is that BIS recordings are increased—implying a reduced depth of sedation—by electromyographic activity.⁷ Thus, BIS recordings tend to be lower in paralyzed patients than in nonparalyzed patients for an equivalent depth of sedation. This phenomenon could potentially lead to paralyzed patients receiving inadequate sedation, resulting in unpleasant awareness.

Sedative Drugs

The ideal sedative agent would have a rapid onset and offset of action (i.e., a short and constant context-sensitive half-time); cause minimal respiratory and cardiovascular depression; and be inexpensive. No single agent fulfills all three criteria.

Means of Action

“Typical” antipsychotics are thought to act primarily by inhibiting central dopaminergic neurotransmission; the term covers a diverse range of drugs, including the phenothiazines (e.g., chlorpromazine, thioridazine), the butyrophenones (e.g., haloperidol, droperidol), and the dibenzoxazepines (e.g., loxapine). To varying degrees the typical antipsychotics also produce antagonism at cholinergic, histamine, and α receptors. A newer class of “atypical” antipsychotics (e.g., risperidone) is thought to act by means of inhibition at central serotonin receptors. Many antipsychotics are also useful antiemetics (see Antiemetics in subsequent material).

Side Effects

The side effects of antipsychotics include sedation (central antihistamine and anticholinergic effects), peripheral anticholinergic effects (dry mouth, blurred vision, constipation, urinary retention), postural hypotension (α receptor blockade), and extrapyramidal motor effects. Extrapyramidal effects may be acute or chronic; acute effects include dystonic reactions (manifest as painful muscle spasms involving primarily the face and neck) and motor restlessness. Antipsychotics should be avoided in patients with Parkinson disease. The risks for respiratory depression and airway obstruction are lower than with the benzodiazepines and propofol, but they can still occur. A number of antipsychotic drugs (including thioridazine, haloperidol, and droperidol) are associated with QT prolongation and torsades de pointes ventricular tachycardia.¹¹ This has led to a “black-box warning” for droperidol that has greatly diminished its use. However, the risk for torsades de pointes with droperidol has probably been overstated.¹² Nevertheless, this class of drugs should be avoided in patients with QT prolongation and in those with other risk factors for torsades de pointes (Chapter 21). The QT interval should be monitored during treatment, and if it lengthens by >25%, treatment should be discontinued.

Antipsychotics

Pain Assessment

In awake patients, pain may be subjectively graded on a numeric rating scale of 0 to 10, where 0 means no pain and 10 means the worst pain imaginable. Pain scores should be recorded at rest and with coughing every few hours for the first two days following surgery. The goal is to maintain pain scores of 3/10 at rest and 5/10 with coughing. The degree of sedation and the presence of any nausea or vomiting should be assessed at the same time as the pain scores. In extubated patients receiving opioids, sedation may be rated on a 4-point scale where A = alert, V = rousable to voice, P = rousable to pain, and U = unrousable. The assessment of regional nerve blocks is discussed in Chapter 12.

In ventilated or confused patients who are unable to verbalize, pain is not easy to assess. It may be very difficult to distinguish inadequate analgesia from inadequate sedation. Doctors, and to a lesser extent nurses, tend to underestimate pain in ICU patients.²⁰ Various physiologic parameters, such as tachycardia, hypertension, and sweating, can indicate pain. Additionally, behavioral responses such as face grimacing, eye shutting, and fist clenching may be indicative of pain.²¹ If there is doubt, analgesics should be administered in preference to sedatives.

Analgesic Regimens

Most patients benefit from a multimodal analgesic regimen consisting of an opioid in combination with acetaminophen. In selected patients, a nonsteroidal antiinflammatory drug (NSAID) or tramadol is a useful adjunct.

While patients are sedated and intubated, an opioid is administered intravenously, usually by intermittent bolus and on the basis of the patient’s physiologic and behavioral responses. Once the patient is awake and extubated, intravenous opioid analgesia may be administered via a patient controlled analgesia (PCA) system (Table 4-4). PCA provides enhanced analgesia and increased patient satisfaction, and it may improve pulmonary function over that found with the use of conventional intravenous opioid techniques.^22,23 An alternative to PCA, once patients are able to take medications enterally, is the combination of a long- and a short-acting oral opioid (Table 4-5). The long-acting agent is taken morning and night for 2 to 3 days following surgery; the short-acting agent is taken for breakthrough pain or prior to planned activities such as physical therapy. A full agonist should not be combined with a weak agonist (see Strong and Weak Opioids in subsequent material). After the first few days following surgery, acetaminophen, either alone or in combination with a weak opioid, usually provides adequate analgesia. A simple measure that helps with pain following sternotomy is the use of a rolled towel or small pillow that the patient can clutch over the wound during repositioning and coughing.

In some centers, regional analgesia (Chapter 12), often involving epidural or intrathecal opioids, is used for the management of pain following cardiac surgery. The advantages of regional analgesia include improved postoperative pain relief and the potential for earlier extubation.^24,25 Countered against this is the fact that for most patients the pain following sternotomy is of moderate, rather than severe, intensity¹⁵ and is typically less than the pain experienced following a lateral thoracotomy or upper abdominal incision. Furthermore, despite the good safety record noted in reported series, many clinicians are concerned about the potential for the formation of epidural hematoma in patients in whom the epidural space has been instrumented and who then receive heparin shortly thereafter.

Analgesics: Opioids

Strong and Weak Opioids: Efficacy Versus Potency

Opioids function as competitive agonists, partial agonists, or antagonists at opioid receptors within the CNS. There are three basic types of opioid receptors—mu (μ), kappa (κ), and sigma (δ)—but the therapeutic and adverse effects of opioid drugs result mainly from their effect on μ receptors.

Efficacy describes the maximum clinical effect of a drug; potency describes the plasma concentration, and therefore the dose, required for a particular clinical effect—usually the EC50, the plasma concentration at which 50% of a drug’s maximal effect occurs (Fig. 4-3). Strong opioids, such as morphine and fentanyl, have high efficacy and are termed full agonists or simply agonists. However, fentanyl has a greater affinity for the μ receptor than does morphine, and it is able to produce a similar clinical effect at a lower dose; thus, fentanyl is more potent than morphine. A drug that produces a submaximal effect, irrespective of the dose, is called a partial agonist (see Fig. 4-3). Propoxyphene, oxycodone, and buprenorphine are examples of partial agonists at opioid receptors. Such drugs are also called weak opioids. Weak opioids may have high (e.g., buprenorphine) or low (e.g., codeine) potency.

Figure 4.3 Efficacy and potency. Three dose-response curves are shown. Drugs A and B are able to elicit a maximal effect (full agonist), whereas drug C produces a submaximal effect (partial agonist). EC50 is the plasma concentration at which 50% of a drug’s maximal effect occurs; EC50 defines potency. Drugs A and B have the same efficacy, but drug A is more potent than drug B. Drugs A and C have the same potency, which is greater than that of drug B.

When an opioid antagonist drug such as naloxone binds to an opioid receptor it exerts no pharmacologic effect (zero efficacy). An antagonist drug functions by preventing access to the receptor population by an agonist. Thus naloxone, which is highly potent, can antagonize the effect of morphine or fentanyl. Similarly, a potent partial agonist such as buprenorphine can potentially antagonize the effect of a less potent strong opioid such as morphine. For this reason, strong and weak opioids should not be used together. The characteristics of opioids that are commonly used in the ICU are shown in Table 4-5.

Analgesics: Acetaminophen (Paracetamol)

Acetaminophen is an analgesic and antipyretic drug that has almost no antiinflammatory effects. Its mechanism of action is unknown, but it is thought to act centrally through inhibition of the enzymes cyclooxygenase (COX) type 3 (constitutive) and COX-2b (inducible) (see NSAIDs below). At therapeutic doses it has almost no side effects, provides effective analgesia for mild to moderate pain, and has an opioid-sparing effect for severe pain.²⁹ Acetaminophen has been available for oral and rectal use for many years, and recently an intravenous formulation has become available. Intravenous acetaminophen is likely to replace the acetaminophen prodrug, propacetamol, which is currently available for intravenous use in some countries. The oral bioavailability of acetaminophen is 100%, and the dose, whether oral, rectal, or intravenous, is 500 to 1000 mg 6 hourly. Overdose with as little as 10 g of acetaminophen can cause fatal hepatotoxicity. Acetaminophen should be withheld in patients with hepatocellular dysfunction.

Analgesics: Nonsteroidal Antiinflammatory Drugs

NSAIDs inhibit COX, an enzyme necessary for the synthesis of prostaglandins and thromboxane. Two major forms of the enzyme are found in the periphery: COX-2, the inducible form, is involved in inflammation. COX-1, the constitutive form, is found in various sites throughout the body, including blood vessels, the stomach, and the kidney. Many of the prostaglandins formed under the influence of COX-1 have protective functions such as renal vasodilation (Chapter 1) and protection of the gastric mucosa. NSAIDs inhibit both forms of the enzyme.

NSAIDs have analgesic, antipyretic, antiplatelet, and antiinflammatory effects and may be used as supplements to acetaminophen and opioids for the treatment of pain following cardiac surgery.^30,31 However, NSAIDs are associated with a number of adverse effects, including renal dysfunction and gastrointestinal bleeding. NSAID-induced renal dysfunction is more likely in patients with preexisting renal dysfunction, impaired ventricular function, or hemodynamic instability. The risk for gastrointestinal bleeding is greatest in patients taking anticoagulant medications in those with histories of peptic ulcer disease. NSAIDs can provoke bronchospasm in patients with asthma and potentially can increase postoperative bleeding. For these reasons, the routine use of NSAIDs in cardiac surgery patients is controversial.^32,33

NSAIDs should be avoided in the following circumstances:

If a decision is made to use an NSAID, it may be prudent to wait until the first postoperative creatinine level has been confirmed as normal before administering the first dose. There is some evidence that the concomitant use of ibuprofen reduces the antiplatelet effects of aspirin,³⁴ suggesting that an alternative agent should be used following cardiac surgery. In a recent large-cohort study, the combined use of NSAIDs (including ibuprofen) and aspirin was not associated with an increased risk for myocardial infarction compared to the use of aspirin alone.³⁵

Two NSAIDs that are formulated for intravenous use are tenoxicam (20 to 40 mg once daily) and ketorolac (10 to 20 mg 6 hourly). Orally, diclofenac (75 mg twice daily) or ketoprofen (75 mg three times daily or 200 mg slow-release daily) are options.

Analgesics: COX-2 Inhibitors

Drugs with specific activity against COX-2 act as analgesics and cause less gastrointestinal side effects than the NSAIDs. COX-2 inhibitors have been associated with an increased risk for myocardial infarction, stroke, and sternal wound infection. A number have been withdrawn from the market. They should be avoided in cardiac surgery patients.^36,37

Mechanism of Action

Under the influence of a nerve action potential, acetylcholine is released from motor nerve fibers into the synaptic cleft of the neuromuscular junction of skeletal muscle. Acetylcholine binds to cholinergic nicotinic receptors on the surface of skeletal muscle, allowing the entry of sodium into the muscle cell which, depending on the number of receptors stimulated, may precipitate an action potential and muscular contraction. There are two types of neuromuscular-blocking drug that inhibit this cholinergic receptor in distinct ways.

Assessment of Neuromuscular Function

The onset of neuromuscular blockade is apparent in the presence of apnea, the loss of muscle tone, and the development of suitable conditions for endotracheal intubation. Recovery is evident by the resumption of spontaneous ventilation and the ability to perform certain motor functions, such as hand clench and head lift. These simple tests are adequate for most situations in the ICU, but they are not specific for neuromuscular function and do not rule out the presence of clinically important residual weakness. In certain circumstances, such as defining the cause of unexplained apnea or ensuring the absence of residual paralysis prior to extubation, assessment of neuromuscular function with a peripheral nerve stimulator is indicated.

One way of gauging neuromuscular function by using a nerve stimulator is to assess the strength of thumb adduction in response to a train-of-four stimulation of the ulnar nerve. The nerve stimulator is connected to surface electrodes on the distal ulnar side of the forearm. One electrode is placed 1 cm proximal to the wrist crease and the other electrode is placed 2 to 3 cm farther up on the forearm. The nerve stimulator is then set to deliver four brief impulses at 60 to 80 mA, 0.5 seconds apart. The normal response to train-of-four stimulation is four strong twitches of the thumb, each of equal intensity, whereas complete neuromuscular blockade fails to elicit any response. As recovery of function occurs, the first of the four stimuli elicits a weak twitch; then progressively, the second, third, and fourth twitches return (Fig. 4-4). Once four twitches have returned, residual neuromuscular blockade is identified by the presence of fade, where the intensity of the first twitch is greater than that of the last twitch. Patients should not be extubated until there are four strong twitches with minimal fade. Residual neuromuscular blockade resulting from nondepolarizing agents can be reversed with neostigmine, as described later.

Figure 4.4 Recovery from a nondepolarizing neuromuscular blocking drug. Twitch strength in response to train-of-four (TOF) electric stimulation over time is shown. The reversibility of neuromuscular blockade is also shown. See text for details.

Nondepolarizing Neuromuscular-blocking Drugs

Factors That Influence Nondepolarizing Muscle Relaxants

The effect of nondepolarizing neuromuscular-blocking drugs is augmented in a clinically important manner by magnesium, by certain antibiotics (e.g., clindamycin, colistin, tetracycline, aminoglycosides), and by some neuromuscular diseases, particularly myasthenia gravis.

Reversal of Nondepolarizing Muscle Relaxants

Residual neuromuscular blockade (see earlier material) caused by nondepolarizing neuromuscular blocking drugs can be reversed by intravenous neostigmine (50 μg/kg). Neostigmine inhibits the enzyme cholinesterase that is responsible for the breakdown of acetylcholine at the neuromuscular junction. Thus, acetylcholine levels rise and competitively antagonize the neuromuscular blocking drug. In addition, neostigmine inhibits cholinesterase at parasympathetic muscarinic cholinergic receptors and can therefore cause bradycardia, bronchospasm, and increased gastrointestinal activity. To avoid these side effects, neostigmine is typically coadministered with a muscarinic cholinergic receptor antagonist such as atropine (20 μg/kg) or glycopyrrolate (5 μg/kg).

Dopaminergic Antagonists

The antipsychotic medications described earlier all have antiemetic properties. Prochlorperazine (10 to 25 mg rectally, 8 hourly), droperidol (0.5 to 1 mg intravenously, 8 hourly), and metoclopramide (10 mg orally or intravenously, 6 hourly) are widely used as antiemetics. In addition to an antagonist effect at central dopamine receptors, metoclopramide is also a gastrointestinal prokinetic, which may contribute to its antiemetic effect. Metoclopramide, at the dose described, has a very low incidence of side effects but is less effective than droperidol and ondansetron.⁴²

Serotonin Antagonists

Antagonists of the 5-hydroxytryptamine type 3 receptor (5-HT₃), such as ondansetron and granisetron, are useful antiemetics and have very few side effects; headache and transient elevations in liver function tests occasionally occur.⁴³ For the treatment of established nausea and vomiting, 4 mg of ondansetron is as effective as 8 mg.⁴⁴ The dose may be repeated 8 hourly. Combined use of a 5-HT₃ receptor antagonist and tramadol should be avoided (see Tramadol, earlier).

Antihistamines

Centrally acting histamine (H₁) receptor antagonists are useful antiemetics, particularly in patients with histories of motion sickness. Examples include promethazine (25 to 50 mg), cyclizine (25 to 50 mg), and dimenhydrinate (50 to 100 mg), all of which are available for oral and intravenous use. Some antihistamines also have antidopaminergic effects; for instance, promethazine, which is a phenothiazine. Antihistamines can cause dry mouth, urinary retention, sedation, and confusion in the elderly.

Anticholinergics

Of the anticholinergic medications, only scopolamine is widely used clinically as an antiemetic because it has a greater ability to cross the blood-brain barrier than either atropine or glycopyrrolate. It is administered once every 3 days as a transdermal patch behind the ear. Dry mouth is the most important side effect when scopolamine is administered by this route, but sedation and confusion can also occur.