Chapter 41. The Critically Ill Poisoned Patient

The 2001 publication by the American Heart Association, TOX-ACLS Toxicologic-Oriented Advanced Life Support,1 marked wide recognition that critical illness resulting from poisoning may require very different management from similar illness occurring in the nonpoisoned patient. That publication made specific suggestions for management of dysrhythmias and other toxicity caused by cocaine, calcium channel blockers and β-blockers, opioids, tricyclic antidepressants, and drug-induced cardiovascular shock. In 2010, the updated American Heart Association ACLS guidelines contained specific evaluation and recommendations regarding poisoning with these same aforementioned toxins, as well as cyanide, digoxin, and antidotal therapy with flumazenil and lipid emulsion.2 The relevance of those publications is recognition that proper management of many clinical problems caused by toxins differs or deviates from management of the same clinical problem occurring in the nonpoisoned patient.

The main focus of this chapter is on the most commonly encountered problems in the clinical management of poisoned patients. We present a general approach to management and discuss the unique management issues involving poisoning. This chapter cannot cover all poisoning circumstances that require unique management, but it does cover the most common problems relevant to intensivists.

Although the severe consequences of poisoning and toxicity may cause a wide array of clinical problems, the majority of critical illnesses resulting from poisoning involve the following problems: (1) airway or respiratory compromise, (2) cardiovascular depression manifested as hypotension and/or bradycardia, (3) cardiovascular stimulation manifested as hypertension, tachycardia, and/or tachydysrhythmia, (4) hyperthermia, and (5) seizure and status epilepticus (Table 41-1).

As with other clinical circumstances, airway management is nearly universally the primary concern in management of the poisoned patient. Airway compromise, respiratory depression, and/or respiratory compromise must be immediately addressed, and, when possible, their underlying etiology corrected or addressed.

Endotracheal Intubation and Unique Considerations

The decision to perform endotracheal intubation in the poisoned patient is based on expectation of lack of continuing airway patency and prognostic factors. Although Glasgow Coma Score (GCS) was not developed to be used as a guide to airway and respiratory status, in many poisonings it can predict the need for endotracheal intubation and/or mechanical ventilation. Patients with a GCS of 6 or less as a result of poisoning often require endotracheal intubation.3

There are numerous exceptions to the principle of GCS <6 being predictive of need for intubation, including toxicity from dissociative agents that cause depressed mental status or coma but do not typically compromise airway and respiration.4 Such agents include ketamine, phencyclidine (PCP), and dextromethorphan; agents that cause waxing and waning respiratory comprise primarily clonidine; γ-hydroxybutyrate (GHB) and its congeners may also fit into this category. This latter group may result in apnea and/or respiratory depression that is reversible with stimulation of the patient.

An exceptional circumstance warranting rapid securing of the airway is airway compromise secondary to caustic exposure. Due to the potential for rapid deterioration and loss of ability to secure an airway later, it is advisable to endotracheally intubate patients with caustic exposure and stridor, aphonia, or dysphonia in a manner similar to that conducted for an airway burn.5

Certain toxins require specific considerations when performing endotracheal intubation. Poisoning with organophosphate pesticides deactivates pseudocholinesterase, resulting in an extremely prolonged half-life of paralytic agents used for neuromuscular blockade.6 This should be taken into consideration when administering these medications and when selecting agents based on their half-life and duration of action.

Salicylate toxicity results in metabolic acidemia with respiratory alkalemia due to salicylate stimulation of central respiratory drive and compensatory increase in respiratory drive.7 In salicylate-poisoned patients, increased minute ventilation is achieved through tachypnea and/or hyperpnea. Interrupting this for even the brief time required to perform endotracheal intubation may result in rapid or immediate seizure or cardiovascular collapse.8 Endotracheal intubation in a patient with salicylate toxicity should be undertaken only when necessary, and performed by a clinician with the highest capability to complete the procedure rapidly. Immediately after endotracheal intubation, the patient should be ventilated at both a rate and volume 150% of the rate and volume typically used. Serial blood gas analysis will allow rate and volume settings to be adjusted. Failure to maintain hyperventilation may result in rapid or immediate death.

A unique exception to the “airway, breathing, circulation” paradigm familiar to emergency medicine and critical care clinicians is physical decontamination of patients. Allowing a contaminated patient to enter clinical areas without proper decontamination is a well-documented risk to caregivers and other patients, and is capable of rendering an emergency department's critical care unit incapable of providing effective care. Even the presence of patients with detectable odors due to minimally toxic substances has well-recognized potential to be disruptive and generate unease, panic, and hysteria, causing vague, nonspecific but incapacitating symptoms in a manner best described as hysteria or mass hysteria.9 Accordingly, any patient contaminated with highly toxic substances capable of cross-contaminating staff or clinical areas, such as organophosphate pesticides and certain hydrocarbons, absolutely must be decontaminated in a “hot” or “warm” zone outside the “cool” zone in which only decontaminated patients are managed and in which clinical care is provided. It is an appropriate practice for a contaminated patient to be kept outside of the emergency department, even if unstable, apneic, or in a state of cardiovascular collapse, before being brought inside a clinical care area without decontamination.

When securing the airway in a setting with clinicians wearing personal protective equipment, consider use of a laryngeal mask airway rather than endotracheal intubation, as LMA placement is less impacted by wearing cumbersome PPE suits.10

Poisons Affecting Respiratory Drive and Ventilation

Numerous drugs and medications may cause hypoventilation by blunting central respiratory drive, most notably sedative–hypnotic medications such as benzodiazepines, barbiturates, and alcohol, and opioids such as morphine, heroin, and fentanyl.

In addition to central blunting of respiratory drive, hypoventilation may occur as a result of impaired chest wall movement. This may be due to weakness or paralysis, as occurs from exposure to botulinum toxin, pesticides and other organophosphates, and neuromuscular blockers, or from hypokalemia and hypermagnesemia. Chest wall rigidity can cause hypoventilation, and may result from tetanus, strychnine, or fentanyl exposure. For fentanyl chest wall rigidity, also termed “wooden chest,” use of naloxone in standard or high doses may be attempted. Chest wall rigidity from tetanus or strychnine may be relieved with neuromuscular blockers (Table 41-2).

By far the most common effect, if any, toxin exposure will have on respiratory status is to induce respiratory depression. Certain unique toxins, however, result in increase in central respiratory drive. Caffeine and theophylline, which are used therapeutically to increase respiratory drive in neonatal apnea syndrome,11 salicylates, and cocaine result in increased respiratory rate by stimulation of central respiratory drive. Salicylates additionally have metabolic effects that cause peripheral effects that increase respiratory drive.

Poisons Affecting Cellular Respiration

Certain poisons affect respiration on the cellular or molecular level. This may commonly result from alteration of hemoglobin to form methemoglobin or carboxyhemoglobin, both of which are incapable of normal oxygen delivery. Cellular or molecular impairment of respiration may also occur by interference with oxidative phosphorylation, such as by poisoning with cyanide, carbon monoxide, or hydrogen sulfide.

Methemoglobin

Methemoglobinemia results from oxidative stress on hemoglobin that results in iron oxidation to the ferric (Fe3+) rather than standard ferrous (Fe2+) state.12 This derivative of hemoglobin binds H2O rather than oxygen, and does not deliver oxygen to tissues. Methemoglobinemia presents clinically with expected results of hypoxia: tachypnea, dyspnea, and severe cyanosis. Pulse oximetry readings in this condition are inaccurate due to inability of standard pulse oximeters to interpret methemoglobin light absorption, as standard pulse oximeters are only intended to quantify oxyhemoglobin and deoxyhemoglobin. Methemoglobinemia typically results in pulse oximetry readings that range from 75% to 85% on standard pulse oximeters. Co-oximeters are capable of precisely measuring oxyhemoglobin, deoxyhemoglobin, methemoglobin, and carboxyhemoglobin, and can accurately quantify the precise methemoglobin level.13 They are laboratory devices that interpret blood gas samples, not to be confused with pulse oximeters, which are bedside devices that estimate hemoglobin saturation using light absorption. Normal methemoglobin levels are 0.5–3%. Methemoglobin levels >10% may be associated with symptomatic illness, and levels >50% may result in rapid death. Treatment of methemoglobinemia involves administration of high-flow oxygen and chemical reduction of the methemoglobin back to hemoglobin using methylene blue.

In the normal state, small quantities of methemoglobin may be converted back to hemoglobin through an NADH-dependent reaction catalyzed by cytochrome b5 reductase. This is the mechanism by which nontoxic quantities of methemoglobin that form during day-to-day exposure to oxidants correct methemoglobinemia. During severe methemoglobinemia, this reaction is insufficient, and an alternative metabolic pathway that does not function without the aid of an exogenously administered reducing agent is needed. The reducing agent used therapeutically is methylene blue, and it works by action through the hexose monophosphate shunt to reduce methemoglobin, yielding normal, functional hemoglobin.

Optimal methylene blue dosing is unknown, and numerous varied recommended doses exist. Methylene blue may result in hemolysis in G6PD-deficient persons, and should be avoided or used with extreme caution in them. Dosing of methylene blue is 1–2 mg/kg given IV over 5 minutes.14 Response is usually very rapid, but if methemoglobin levels remain elevated, 1 hour later methylene blue may be readministered at the same dose. Methylene blue interferes with pulse oximetry reading, and continued use of co-oximetry by measurement of venous blood gas samples is required to monitor methemoglobin levels. Dapsone toxicity may cause prolonged methemoglobinemia due to the long half-life of dapsone. Repeated treatments with methylene blue, not to exceed 5 mg/kg/24 hours, may be required in symptomatic patients.

Carbon Monoxide

Carbon monoxide is a by-product of combustion. Carbon monoxide poisoning typically results from closed-space fires or by exposure to exhaust from combustion engines. Carbon monoxide binds hemoglobin with an affinity approximately 250 times greater than that of oxygen.15 Carbon monoxide binding of hemoglobin results in carboxyhemoglobin, which is a nonfunctional form of hemoglobin that does not transport oxygen.

Clinical problems resulting from acute carbon monoxide poisoning include headache, nausea, vomiting, disorientation, altered mental status or coma, syncope, seizure, and cardiac arrest. Chronic carbon monoxide poisoning presents differently, often with headache and malaise that may be misdiagnosed as a viral syndrome.

Carboxyhemoglobin is not detected by standard bedside pulse oximetry. Pulse oximeters misinterpret carboxyhemoglobin as oxyhemoglobin, and therefore give a falsely normal pulse oximetry reading in patients with carbon monoxide poisoning. Co-oximetry measurement of carboxyhemoglobin level in venous or arterial blood gas is required to quantify degree of carbon monoxide binding of hemoglobin. There is no “normal” carboxyhemoglobin level, but average persons have levels <3% that may result from exposure to automobile exhaust and other sources. Smokers have notably higher carbon monoxide levels16 based on how heavily they smoke, and they may have baseline carboxyhemoglobin levels as high as 10%.

Symptomatic illness from carbon monoxide poisoning may occur acutely with any level of carboxyhemoglobin, but is typical at levels >10%. Depending on health of the patient, severe illness and injury may occur with carboxyhemoglobin levels as low as 10%, although healthy individuals usually tolerate higher levels. Significant acute carbon monoxide poisoning results from levels >25%, and levels >45% are immediately life threatening.17

Treatment of carbon monoxide toxicity involves administering supplemental oxygen. This may be normobaric, or if possible hyperbaric oxygen. Hyperbaric oxygen does not have notable immediate benefit to the patient, but is administered to prevent the potentially devastating neurologic sequelae of poisoning.18 These sequelae include a Parkinson-like syndrome and extreme neuropsychiatric disability that may render the patient incapable to work, study, or pursue the usual activities of daily living.

Cyanide

Cyanide toxicity results from cyanide binding to cytochrome α3, which interferes with oxidative phosphorylation. This prevents cellular respiration, in effect suffocating the tissues on a cellular level. Cyanide exposure may result from inhalation of smoke in closed-space fires.19,20 Fires have been recognized to often produce copious quantities of cyanide that result from burning of plastic, polyurethane, rubber, silk, wool, and many other materials typically contained in homes and offices. Cyanide poisoning may result iatrogenically from prolonged use of nitroprusside, which contains cyanide and the antiquated antineoplastic agent laetrile. Certain plants, such as pits or seed from peaches, apricots, plums, pears, apples, and bitter almond, are cyanogenic. Cyanide is used commonly in the jewelry industry and in some fields such as photography. Homicidal and suicidal use of cyanide is well reported.

Cyanide toxicity presents clinically with acute onset of severe illness after exposure. This is typically syncope, coma, seizure, cardiac dysrhythmia, or cardiac arrest. Due to inability to use oxygen in oxidative phosphorylation, fair-skinned patients typically have a flushed, pink appearance. Lab corroboration of cyanide toxicity may be obtained by comparing arterial and venous blood gas samples drawn simultaneously and by noting a lack of oxygen extraction across the capillary bed.21 Metabolic acidemia is always present and usually severe. In the setting of closed-space fires, lactate concentration >10 mmol/L is pathognomonic or cyanide toxicity.19 This appears to be true regardless of the presence of carbon monoxide poisoning or the extent of body surface burn.19

Serum cyanide levels are rarely clinically available, but are of use if they can be obtained rapidly. Cyanide levels <1.0 mg/L correlate with tachycardia and flushing, 1.0–2.5 mg/L with altered mental status, seizure, and hypotension, and levels >3.0 mg/L are typically rapidly fatal.

Treatment for cyanide toxicity involves use of hydroxocobalamin, which is the optimal antidote, or use of all or part of the cyanide antidote kit, which is a combination of amyl nitrite pearls, sodium nitrite, and sodium thiosulfate.22,23

Cyanide antidote therapy should be given to any patient with known or suspected exposure to cyanide or from a closed-space fire if he or she has metabolic acidemia, elevation of lactate concentration, loss of consciousness or altered mental status, shock, cardiac dysrhythmia, or cardiac arrest.

Hydroxocobalamin, a vitamin B12 precursor, directly binds cyanide to form vitamin B12, which is harmless and excreted in the urine. Dosing of hydroxocobalamin is 70 mg/kg, to a maximum dose of 5 g, given IV over 30 minutes. In cases of cardiac arrest, it can be given as an IV push. The dose can be repeated to a maximum total of 15 g. Use of hydroxocobalamin may subsequently interfere with pulse oximetry readings and co-oximetry readings, rendering it difficult or impossible to know oxygen saturation, leaving only measurements of Po2 as a guide.24 These interferences may last as long as several days.

Use of the cyanide antidote kit involves three parts: amyl nitrite inhalation pearls, sodium nitrite for IV administration, and sodium thiosulfate for IV administration. Nitrates are used to induce methemoglobinemia. These are only used in cyanide toxicity not resulting from closed-space fires and smoke inhalation. After closed-space fire, carbon monoxide poisoning may concomitantly be present, and decrease oxygen-carrying capacity by causing formation of methemoglobin is contraindicated. For patients who possibly have carbon monoxide poisoning, only the sodium thiosulfate portion of the kit is administered.22 This works by enhancing formation of cyanomethemoglobin through the enzyme rhodanese.

Use of the various portions of the kit are as follows: amyl nitrite pearl is crushed and inhaled for 1 minute until IV access is obtained. The sodium nitrite dose is 10 mL of the 3% solution included in the kit; pediatric dosing is 0.33 mL/kg. Hypotension may result from nitrite use. Nitrites are intended to cause methemoglobinemia; if methemoglobin level >10–15% is not induced by the initial dosing of sodium nitrite, half of the original dose may be administered 30–60 minutes after the first dose.

Sodium thiosulfate is administered as 12.5 g IV, which is the full 50-mL bottle of 25% sodium thiosulfate solution included in the kit. The pediatric dose is 1.65 mL/kg of the same 25% solution. Repeat dosing of sodium thiosulfate may be given as one half of the original dose 30–60 minutes after the initial dose.

Cardiovascular depression in the form of hypotension and/or bradycardia may result from exposure to cardioselective medications, such as digoxin, β-adrenergic antagonists, calcium channel antagonists, and clonidine, and also occurs from numerous other toxins. As a preterminal event, cardiovascular depression may result secondary to toxicity from any poison, including cardiovascular stimulants. Treatment of cardiovascular depression from certain toxins, particularly cardioselective medications, may require very specific, unique therapy.

Asymptomatic hypotension and/or bradycardia, particularly in the absence of end-organ manifestations, do not necessarily indicate treatment. The lower limit at which clinicians are comfortable allowing heart rate or blood pressure to remain varies. Generally, maintaining the heart rate at 45 beats/min or greater, systolic blood pressure >90 mm Hg, and diastolic blood pressure >40 mm Hg are limits above which vital signs should be maintained.

Intravenous fluid bolus, atropine, and pressors may be used to treat cardiovascular depression with exceptions noted here. Bradycardia resulting from most cardioactive medications infrequently responds to atropine because homeostatic mechanisms will have already decreased or removed vagal tone in an attempt to compensate. Use of atropine is not contraindicated for medications such as digoxin, β-adrenergic antagonists, calcium channel antagonists, and clonidine, but more effective and definitive therapy should not be delayed due to atropine administration.

Pressors similarly may lack their typical effectiveness when used in poisoned patients. As with all patients, care should be exercised in balancing attempts to maintain central arterial or venous pressure with end-organ or extremity capillary perfusion pressure to prevent paradoxical hypoperfusion of organs, digits, and extremities with high-dose pressor infusions (Tables 41-3 and 41-4).

Tricyclic Antidepressant Hypotension

If an indirect or mixed-acting pressor is used without success, particularly in cases of hypotension from tricyclic antidepressants, norepinephrine should be initiated. The pathophysiology of cyclic antidepressant poisoning may result in catecholamine depletion to a degree that no quantity of an indirect-acting agent such as dopamine will be effective, and norepinephrine or epinephrine may be necessary.

Digoxin Cardiovascular Depression

Specific unique therapies for cardioactive medication toxicity may be indicated. Cardiotoxicity from digoxin, digitoxin, or other cardioactive steroids such as bufotoxin may be treated with digoxin-specific Fab.25 Although this antidote is produced for digoxin, cross-specificity with digitoxin, bufotoxin from toad species, oleandrin from oleander, and other botanical cardiac glycosides typically respond to digoxin-specific Fab. Empirical dosing for acute digitoxin toxicity is 10–15 vials in adults or children. In cases of oleander, toad, or other related poisoning, the typical dose for acute digoxin toxicity may be used, and additional doses may be needed.

β-Blocker Cardiovascular Depression

β-Blocker toxicity typically does not respond significantly to IV fluid, atropine, or pressors. Glucagon is often effective as a result of its activity, since it is independent and unaffected by blockade of the β-adrenergic receptors.26 An empirical dose of glucagon is 5 mg SC or IV in adults, 1 mg in children <20 kg, or 2 mg in children >20 kg. If effective, glucagon may be administered again as necessary. If glucagon therapy proves to be ineffective, more than two consecutive attempts to restore cardiovascular function should not be attempted using this therapy.

Calcium Channel Blocker Cardiovascular Depression

Calcium channel antagonists are unique cardiotoxins. Due to calcium channel blockade, patients with calcium channel antagonist overdose often maintain normal mentation even in the state of extremely low systolic and mean arterial blood pressure. This phenomenon is so characteristic and notable that normal mental status in the presence of extreme hypotension/bradycardia should be considered highly suggestive of calcium channel antagonist toxicity.

Therapy for calcium channel antagonists includes administration of high-dose calcium. Calcium gluconate or calcium glubionate is effective and its concentration is safe to be used in peripheral veins. Calcium chloride contains three times the elemental calcium of calcium gluconate and therefore has some advantage in use. Extreme caution must be exercised to avoid extravasation of calcium chloride. An antidotal therapy used regularly for the past decade that is highly effective in treating calcium channel antagonist-induced cardiovascular depression is insulin–euglycemia.27

Pathophysiologically, the healthy myocardium uses free fatty acids for energy. Myocardium that is unhealthy, stressed, or in a shock state will utilize glucose for energy, and it is believed that this is the mechanism by which insulin–glucose infusion aids calcium channel blocker–induced cardiovascular depression.

Hyperinsulinemia/ Euglycemia Therapy

Insulin–euglycemia therapy involves initiating a 1-U/kg bolus of regular insulin with 0.5 g/kg of dextrose. If the blood glucose is >400 mg/dL prior to the insulin bolus, no glucose bolus is necessary. After the initial bolus, an insulin infusion of 0.5–1.0 U/kg/h with a continuous dextrose infusion beginning at 0.5 g/kg/h should be started. This dextrose is best infused at D25 or D50 given by central venous access to limit free water administration. The dextrose infusion may be titrated to give more or less dextrose as needed to maintain acceptable serum glucose.

Cardiac function should be reassessed every 20–30 minutes. In cases of persistent cardiovascular depression, the insulin infusion may be increased in increments of 0.5 U/kg/h every 30 minutes if needed to a maximum dose of 2.5 U/kg/h. Increasing the insulin infusion dose will require increase in the amount of dextrose infused.

The initial response to insulin–euglycemia therapy is typically not seen immediately, and it may take 20–40 minutes after initiation of therapy to detect clinical response in cases in which the therapy is successful.

It is critical to monitor the blood glucose frequently during this therapy, every 30 minutes at minimum until stable, and then every 1 hour after the insulin infusion, glucose infusion, and blood glucose are stable. It is also necessary to monitor the serum potassium, since some degree of hypokalemia is expected. Stable hypokalemia at levels down to 2.5 mEq/L does not require supplemental potassium administration.

Lipid Emulsion Therapy

Lipid emulsion therapy may be used in treatment of lipophilic cardiotoxins of all types: calcium channel blocker, β-blocker, local anesthetic, tricyclic antidepressants, and others.28 Lipid emulsion is administered as an initial bolus of intralipid or other lipid emulsion in a 20% concentration in a 1.5 mL/kg followed by 0.25 mL/kg/min or 15 mL/kg/h to run for 30 or 60 minutes. Occasionally a prolonged infusion in a dose of 1–2 g/kg per day or 5–10 mL/kg per day is necessary. Although propofol contains lipid, it should never be used as the agent to provide the lipid for this antidotal therapy. The quantity of propofol to lipid is such that an extremely toxic overdose of propofol would be necessary to deliver an adequate dose of lipid for lipid emulsion therapy.

Clonidine Cardiovascular Depression

Clonidine is an α-agonist with opioidlike effects, and toxicity often mimics the opioid toxidrome of miosis, coma, and apnea.29 In known or suspected clonidine toxicity, high-dose naloxone may be used if the patient is not opioid tolerant. Opioid-tolerant patients are expected to have severe opioid withdrawal as a result of naloxone administration. This may result in severe vomiting and aspiration if depressed mental status is present. If high-dose naloxone is effective at improving respiratory effort or cardiovascular depression resulting from clonidine, an infusion of two-thirds of the dose that achieved clinical response given hourly is recommended. This may be titrated as needed.

Caffeine and Theophylline Hypotension

Therapy of hypotension resulting from severe theophylline or caffeine toxicity is best carried out by typical use of IV fluid and pressors. If this is unsuccessful, administration of a short-acting β-adrenergic antagonist, such as esmolol, may be highly effective in reducing or eliminating hypotension.30

It is useful to discuss the underlying pathophysiology by which a β-blocker is used to treat hypotension: Hypotension from theophylline or caffeine results from excess β-stimulation, including β2, which causes hypotension. This hypotension may involve a characteristically widened pulse pressure, in which the difference between systolic and diastolic pressures may be 150% or greater than the diastolic pressure. Blockade of β2 agonism in such cases may result in rapid and complete resolution of hypotension as well as typical metabolic effects of theophylline toxicity such as hyperglycemia and hypokalemia.29 Esmolol is preferred due to a titratable effect; long-acting β-blockers are not recommended.

Cardiovascular stimulation in the form of hypertension, tachycardia, and/or tachydysrhythmia commonly results from a variety of poisonings. Cocaine toxicity is the commonly encountered situation requiring unique therapy to control cardiovascular stimulation.

Cocaine causes release of large quantities of catecholamines in a dose-dependent manner. Therefore, the primary pharmacologic treatment of cocaine toxicity, including cardiovascular stimulation and cocaine chest pain, is benzodiazepines, often required in very large doses. Benzodiazepines counteract this effect and typically result in abatement of hypertension, tachycardia, as well as psychomotor agitation.

After liberal administration of benzodiazepines, use of antihypertensives may be required. During the initial years of the introduction of crack cocaine into the United States, numerous case reports followed by large case series documented a paradoxical rise in blood pressure in cocaine-intoxicated patients treated with β-blockers. Some of these resulted in catastrophic or fatal intracranial hemorrhage and other sequelae of severe hypertension. The mechanism by which β-blockers may cause increased rather than decreased blood pressure in cocaine-intoxicated patients is by removal of β-adrenergic tone, leaving unopposed α-adrenergic tone and extreme vasoconstriction. For this reason, β-blockers should generally be avoided in management of cocaine-induced cardiovascular stimulation. If benzodiazepine therapy is inadequate to control cardiovascular stimulation, use of phentolamine is preferred to treat hypertension. Other therapies may include use of nitrites such as nitroglycerine or nitroprusside, calcium channel blockers, and both conduction-modulating agents such as verapamil or diltiazem and newer agents such as nifedipine, nicardipine, and others.

Some centers do not adhere to the admonishment to avoid β-blocker use. In such settings, labetalol is often used because it has some degree of α- as well as β-blockade. Use of other β-blockers should only follow or accompany concomitant use of an α-blocking agent such as phentolamine.

Sympathetic nervous system stimulation by cholinergic poisons such as pesticides is a less commonly encountered situation requiring unique care. Acetylcholine excess may cause muscarinic excess, secondary bradycardia/hypotension, or nicotinic excess with secondary stimulation of the sympathetic chain ganglia and tachycardia/hypertension. It may also cause alternation between cardiovascular stimulation and depression. For this reason, management of cardiovascular stimulation secondary to cholinergic poisons such as organophosphates should be managed by use of short-acting, titratable medications such as esmolol or nitroprusside. Such use allows rapid cessation of the medication if the patient transitions from a state of cardiovascular stimulation to cardiovascular depression (Table 41-5).

Pathophysiologically, there are several ways in which poison exposure and toxicity may result in hyperthermia. By alteration of normal mental state, cognition, and psychomotor agitation, patients may be unaware of own temperatures or that of their surroundings. They may fail to avoid exercise or activity in hot ambient environments, become unable to leave the environment, or continue to exert themselves while restrained. Examples include becoming comatose in a closed automobile during daytime, and comatose on a hot surface such as asphalt, where heat gain by conduction may occur quickly. Commonly, this may occur with drugs of abuse such as ethanol, cocaine, opioids, and PCP. Psychomotor agitation associated with drugs of abuse may also result in significant heat production.

Such cases are understandably more common in warmer months, summer in the Northern hemisphere and winter in the Southern hemisphere. A clear relationship between deaths from cocaine hyperthermia and ambient temperature exists: Deaths from cocaine hyperthermia in New York City, for example, dramatically peak during the warmest summer months and are rare during other times.31 This is likely true for other drugs that cause psychomotor agitation and hyperthermia, as well.

Other pathophysiologic mechanisms for hyperthermia include uncoupling of oxidative phosphorylation, such as with salicylate or dinitrophenol toxicity; increased metabolism, such as from thyroid hormone or thyroid extract toxicity; impaired sweating, such as from antihistamines and anticholinergics; vasoconstriction due to α-adrenergic agonism, such as from amphetamines, cocaine, pseudoephedrine, and other sympathomimetics. Malignant hyperthermia (MH), resulting from ryanodine receptor dysfunction, as well as serotonin syndrome (SS) and neuroleptic malignant syndrome (NMS), is discussed later in this chapter.

Although different pathophysiology underlies the manner in which hyperthermia is reached, the initial acute treatments are similar. The temperature at which permanent neurologic injury will occur in any patient cannot be known, but a core temperature of 107°F or 42°C warrants active cooling, preferably by immersion in ice or an ice bath. Tepid sponging, mist spray and fans, or other less effective measures should only be used in the extraordinary instance when true ice immersion cannot be achieved.

A simple method of ice immersion is placing the patient in a partially closed body bag enclosed with ice. Covering the patient in ice and wrapping in a sheet or blanket is also suitable, understanding that this will quickly result in water pooling on the floor surrounding the patient as the ice melts. If a cholera bed is available, this aids in collection of melting ice water and is preferable from a nursing and housekeeping perspective. Using immersion makes cardiopulmonary monitoring more difficult.

Typically patients with hyperthermia resulting from psychomotor agitation, hypermetabolism, or uncoupled oxidative phosphorylation do not typically feel uncomfortable in an ice pack or ice bath. After some period of time when the temperature decreases, they may communicate that they feel cold or uncomfortable, and this often correlates with reaching a goal temperature of 100–102°F. Care should be taken to carefully monitor patients to avoid overcooling below normal body temperature.

Treatment of psychomotor agitation should include chemical restraint by benzodiazepine administration. Use of haloperidol is contraindicated for this purpose as it lowers seizure threshold, results in increased incidence of cardiac dysrhythmia, and impairs heat dissipation. In cases of true MH, dantrolene is indicated. Dantrolene is often errantly used for hyperthermia from causes other than MH, in which there is no potential benefit and thus the small risk of use is not justified. Use of COX inhibitors such as aspirin, acetaminophen, ibuprofen, ketorolac, naproxen, or others plays no role whatsoever in toxin-induced hyperthermia.

Hyperthermic Syndromes: Serotonin Syndrome, Neuroleptic Malignant Syndrome, and Malignant Hyperthermia

SS, NMS, and MH are toxin-induced illnesses that result in hyperthermia by different mechanisms (Table 41-6). These syndromes have significant overlap in clinical presentation, but careful evaluation can clearly differentiate between them. Hotline numbers for the Neuroleptic Malignant Syndrome Information Service are 1-888-667-8367 and 1-315-464-4001. Another hotline to aid in management of MH is the Malignant Hyperthermia Association of the United States, which can be reached at 1-800-644-9737 or 1-315-434-7079. These services are intended to provide advice about diagnosis and management of NMS and MH, respectively. They are supported by medical toxicologists and are able to assist in differentiating NMS, MH, and SS, and to make treatment recommendations.

Onset of illness and progression of illness are distinct for SS, NMS, and MH. SS develops over hours and universally <24 hours after exposure to the serotonin agonist. SS is rapidly progressive, and can quickly transition from mild illness to critical instability or death in hours. This helps differentiate from NMS, which develops over days, and for which both the progression and resolution occur over a more prolonged time period. MH develops more acutely than both SS and NMS, within minutes to hours, and nearly universally within 12 hours of exposure to the causal medication(s). It may rapidly progress and rapidly dissipate. As a result, initial diagnosis and treatment of MH infrequently involves emergency medicine or critical care physicians, usually occurs in operating rooms or postoperative recovery rooms, and is overseen by anesthesiologists managing the patient at that time.

Serotonin Syndrome

SS results from excess serotonergic agonism, typically as a result of exposure to two or more serotonin agonists or massive exposure to a single serotonin agonist. It characteristically causes mental status changes, autonomic hyperactivity, and neuromuscular abnormalities. Alteration of mental status usually does not involve coma or impaired consciousness. Typically it is anxiety, disorientation, psychomotor agitation, and hyperalertness, with patients startling easily. The neuromuscular findings may be hyperreflexia, clonus, tremor, muscle rigidity, myoclonus, hyperreflexia, and a unique form of shivering that is sometimes rhythmic and progressive along the torso, similar to that of a dog shaking water from its coat. Autonomic manifestations are tachycardia and hypertension, with hyperthermia.

To fulfill Hunter criteria to diagnose SS, the patient must have been exposed to a serotonergic medication or drug, and have any of the following: (1) spontaneous clonus; (2) inducible clonus plus agitation or diaphoresis; (3) ocular clonus plus agitation or diaphoresis; (4) tremor and hyperreflexia; (5) hypertonia; (6) temperature >38°C plus ocular clonus or inducible clonus.

Laboratory abnormalities include myoglobinuria, elevated creatine phosphokinase (CPK), and hyperkalemia. In suspected cases, obtain blood gas analysis with lactate concentration, serum electrolytes, liver function tests, and CBC.

Treatment of SS includes maintaining vital signs within acceptable limits, including cooling to <39°C (102.2°F), and use of benzodiazepines, as well as possibly cyproheptadine. Benzodiazepines treat agitation, serve as muscle relaxants, and are useful because the CNS side effects of benzodiazepines do not overlap with CNS changes that result from SS. Lorazepam 0.05–0.1 mg/kg IV is given every 20–30 minutes until clinical effect is reached, followed by repeat administration at the appropriate dose in 2- to 6-hour periods. Diazepam 0.1–0.5 mg/kg may also be used, with initial doses repeated every 10–15 minutes and repeat dosing every 1–2 hours as needed. If benzodiazepines fail to sedate thoroughly, cyproheptadine is given empirically in adults as a 12-mg initial dose followed by 2 mg every 2 hours until symptoms resolve. Dosing would be modified on a weight basis in children.

Cyproheptadine is only available in a PO formulation, but can be crushed and injected down a nasogastric tube to patients with altered mental status. Medications such as chlorpromazine and olanzapine should not be used, since they lower seizure threshold and increase risk of developing NMS.

Neuroleptic Malignant Syndrome

NMS is an extrapyramidal syndrome associated with hyperthermia, muscle rigidity, autonomic instability, and altered mental status. This occurs predominantly with use of antipsychotics and less commonly when anti-Parkinson dopamine agonists are withdrawn.

NMS occurs in particular with potent antipsychotics such as haloperidol and fluphenazine, and with depot formulations such as long-acting depot haloperidol injection, but it has been reported to occur from all classes of neuroleptic drugs, as well as new atypical antipsychotic agents.

Epidemiologically, NMS occurs more commonly in males and in younger patients, but can occur in any gender or age patient. NMS is an idiosyncratic reaction, meaning it is not dose dependent. It may occur with the first dose of medication or may occur in a patient who has been receiving the medication for years without any adverse side effects. NMS occurs more frequently within the first 2 weeks of initiating a neuroleptic or antipsychotic treatment, with depot injection use, and with rapid dose escalation. The pathophysiology and etiology of NMS are unknown, but are widely believed to be mediated by central dopamine antagonism.

As mentioned, NMS may be differentiated from SS by timing of onset. NMS symptoms typically develop over several days, whereas SS develops over hours. Because NMS occurs in patients with psychiatric illness and is slower in onset, there is more likely to be delayed or missed diagnosis. Hyperthermia, altered mental status, autonomic instability, and muscle rigidity are universally present in patients with NMS.

Hyperthermia is usually not as extreme as with SS or NMS, with temperatures typically in the range of 38–39°C (100.4–102.2°F) and uncommonly greater than 40°C (104°F). Muscular rigidity with NMS is more catatonic, lead-pipe rigidity, whereas SS is associated with fasciculation, twitching, shivering, and hyperreflexia. Mental status changes are also more similar to those of catatonic states, and patients may be mute, stuporous, or comatose.

Lab testing is critical in management of NMS. Elevations of CPK may be severe. Rhabdomyolysis and myoglobinuric renal failure may result. Expectedly, associated electrolyte abnormalities may include hyperkalemia and mild elevations of lactate. Low serum iron concentration has >95% sensitivity in detecting NMS.32

Management of NMS includes immediate discontinuation of the offending drug and supportive care with correction of dehydration and electrolyte imbalance. Cooling to decrease temperature to acceptable range may be carried out by the physical methods mentioned earlier in this section. Pharmacologic treatment for NMS should be administered. Hyperthermia from NMS is usually less severe than SS, and typically will not require the aggressive measures to lower temperature that are commonly needed for SS.

Pharmacologic therapy for NMS includes bromocriptine, which agonizes dopamine receptors. This is only available as a PO formulation, and may be crushed and given by nasogastric tube. Dosing is 2.5 mg PO Q 6–8 hourly. It is recommended to continue this for 10–14 days after the symptoms of NMS have resolved. Amantadine may be used instead of bromocriptine. Benzodiazepines may also be used for muscle relaxation and to relieve psychomotor agitation. Lorazepam 2 mg IV/PO Q 6 h is typically effective, but this dose may be increased as necessary.

NMS resolution typically takes days to weeks, on average 5–15 days, to resolve. This is in contrast to SS, for which onset and resolution are often within hours.

Malignant Hyperthermia

MH is a hypermetabolic crisis typically encountered in the setting of anesthesia administration, and may be seen in genetically susceptible patients who receive inhalational anesthetics and/or succinylcholine. As previously mentioned, MH is rarely encountered in the emergency department or ICU setting, and is typically managed in the operative and postoperative setting. Because MH is the toxin-induced hyperthermic crisis that develops most rapidly, within minutes to hours, and is most likely to result in severe morbidity or mortality, clinicians who administer succinylcholine or manage postoperative patients should be aware of this entity and its management.

MH involves excessive release of calcium from the sarcoplasmic reticulum of myocytes, and an ensuing hypermetabolism that results in hypercarbia, mixed respiratory and metabolic acidosis, rhabdomyolysis, and hyperthermia that is sometimes severe, with temperatures rapidly rising up to 113°F. It is widely misunderstood and misstated that hyperthermia rapidly develops in patients with MH: in fact, hyperthermia and rhabdomyolysis may be the last of the clinical symptoms to become apparent, occurring after muscle rigidity, hypercarbia, and mixed respiratory and metabolic acidemia.

The earliest clinical indication of MH is often hypercarbia. There is no absolute Pco2 that is diagnostic, but a Paco2 >60–65 or end-tidal CO2 >55–60 in the absence of other obvious cause should be considered suggestive in postoperative patients. This hypercapnia may be managed by increasing minute ventilation, although the increases required by mechanical ventilation are often greater than would normally be expected. If the patient is not already being mechanically ventilated, he or she should be endotracheally intubated and mechanically ventilated with 100% FiO2, with a minute ventilation that corrects the Pco2 as reasonably as possible.

It is appropriate to correct hypercarbia, but investigation for other evidence of MH should be initiated. This includes physical exam to detect increased muscle tone; evaluation of arterial blood gas; urinary myoglobin, serum CPK, and serum potassium that may be associated with rhabdomyolysis; PT/PTT, INR, and fibrin split products to detect disseminated intravascular coagulation; and rectal or core temperature monitoring. Although hyperthermia is often not present when MH is initially suspected, when elevation of temperature does begin, it may be rapid, with temperatures rising as much as 2°F every 5 minutes.

At any time when MH is strongly suspected, any possible inciting medications should be discontinued and dantrolene administration should begin. Despite any other supportive care given, without dantrolene patients are extremely unlikely to survive MH.33 Prior to the development of dantrolene, 70% of cases of MH were fatal. With current supportive care and dantrolene administration in the United States, approximately 10% of cases are fatal.

Supportive care includes correcting hypercarbia, providing 100% oxygen to support the hypermetabolic state, correcting hyperkalemia and treating rhabdomyolysis, managing disseminated intravascular coagulation if it occurs, and, last, management of hyperthermia.

Use of active cooling as described in the beginning of this section is recommended in conjunction with dantrolene administration.

Seizure may be the consequence of numerous pathophysiologic events, most commonly metabolic and neurochemical. Management of seizure or status epilepticus resulting from poisoning varies significantly from management of epilepsy or trauma-associated seizure.34 The initial management of toxin-induced seizures should include rapid bedside assessment of blood glucose and assessment for hypoxia.

Management of toxin-induced seizures differs from epileptic or traumatic seizures in that phenytoin use is contraindicated.33 Specifically, phenytoin results in increases in severe seizure activity and increased incidence of cardiac dysrhythmia and death. Although the sodium channel blocking activity of phenytoin is effective in decreasing activity of epileptogenic foci or focal activity of traumatized brain, toxin-induced seizure activity is the culmination of diffuse and global cerebral dysfunction, for which sodium channel blockade both is unhelpful and may worsen seizure activity. It is ill-advised to treat toxin-induced seizure with phenytoin, as doing so is expected to be ineffective and, more important, to increase morbidity and mortality.

A management protocol for toxin-induced seizures can be summarized as follows: administration of a benzodiazepine, such as lorazepam 0.05–0.1 mg/kg every 10–15 minutes or diazepam 0.1–0.2 mg/kg every 5–10 minutes to a maximum of three doses. Further dosing is not detrimental, but use of benzodiazepines should be considered inadequate if three doses have not successfully terminated seizures, and escalation to more intensive therapy is warranted.

Benzodiazepine use should be followed by empirical dosing of pyridoxine (vitamin B6) to treat potential effect of isoniazid or other hydralazine-induced seizure. Dosing is 1 g pyridoxine for every 1 g of isoniazid or other hydrazine ingested. Empirical dosing is 2–4 g in adults or 70 mg/kg in children. Because pyridoxine is not often immediately accessible at the dose required, this may be ordered from the pharmacy and progression down the treatment algorithm continues, with pyridoxine administered as soon as it is available.

The traditional algorithm for management of seizures or status epilepticus has been three doses of a benzodiazepine, after which use of a barbiturate begins (Figure 41-1). Due to the advent of highly effective medications such as propofol and valproic acid, these newer medications may be used instead of barbiturates after initial use of benzodiazepines fails. Decision to use barbiturates or newer medications such as propofol or valproic acid is based on clinician preference and comfort.

Propofol

Propofol is extraordinarily rapid in its ability to terminate seizure activity, and its activity as both a GABA agonist and NMDA antagonist makes it the most useful and preferred agent for toxin-induced status epilepticus. We recommend propofol for use after failure of benzodiazepine therapy by administration of 1 mg/kg IV followed by repeated boluses of titration of propofol infusion at 0.1–0.3 mg/kg/min titrated to clinical effect.

Barbiturate

Barbiturate loading would be with pentobarbital 5 mg/kg IV or phenobarbital 10–20 mg IV, typically loaded over 20 minutes.

Valproic Acid or Midazolam

Valproic acid works by GABA agonism; dosing is 25 mg/kg infused over 5–10 minutes. Midazolam is also a GABA agonist; loading and infusion dosing are: loading dose 0.15 mg/kg IV followed by infusion of 1 mcg/kg/min. Every 5 minutes of seizure activity, double infusion rate to maximum rate of 16 mcg/kg/min. If continuous seizure, this would take 20 minutes to reach the maximum infusion dose rate.

Respiratory depression and need for endotracheal intubation should be anticipated when administering a barbiturate subsequent to a benzodiazepine or with any therapeutic propofol use. For this reason, administration of propofol with planned endotracheal intubation is reasonable. Unless continuous EEG monitoring is available, use of any long-acting paralytic agent is not recommended, since it may mask seizure activity and prevent appropriate escalation of pharmacologic therapy if seizures continue. Due to brief duration of activity, succinylcholine may be used as a paralytic if there are no contraindications such as significant hyperkalemia.

Third-Line Anticonvulsants

Other medications that may be employed are third-line agents levetiracetam35 and ketamine.36

Levetiracetam mechanism of action is unknown, and therefore less desirable for use in toxin-induced seizures, as there is potential, such as with phenytoin, to exacerbate seizures or to increase morbidity and mortality. Levetiracetam dosing is 20–40 mg/kg IV diluted in 100 mL of saline and infused over 15 minutes.

Ketamine is an NMDA antagonist with potential to dissipate seizures by a different mechanism of action, NMDA antagonism, than other medications reviewed here, most of which work by GABA agonism. Little data on ketamine therapy for status epilepticus exist. Dosing is initial 1 mg/kg IV bolus, followed by 0.05–0.1 mg/kg/min infusion. Further treatment may involve general anesthetic agents administered by an anesthesiologist.

Management of seizures from other toxin-induced phenomena such as hyponatremia or hypocalcemia may be carried out in the typical fashion.