sign in sign up
<<
Home , Lithium toxicity, Poison control center, Protamine, Spider bite

critical care review Adult in Critical Care* Part I: General Approach to the Intoxicated Patient

Babak Mokhlesi, MD; Jerrold B. Leiken, MD; Patrick Murray, MD; and Thomas C. Corbridge, MD, FCCP

Intensivists are confronted with poisoned patients on a routine basis, with clinical scenarios ranging from known or toxic exposure, illicit drug use, suicide attempt, or accidental exposure. In addition, drug can also manifest in hospitalized patients from inappropriate dosing and drug interactions. In this review article, we describe the epidemiology of in the United States, review physical examination findings and laboratory data that may aid the intensivist in recognizing a toxidrome (symptom complex of specific poisoning) or specific poisoning, and describe a rational and systematic approach to the poisoned patient. It is important to recognize that there is a paucity of evidence-based information on the management of poisoned patient. However, the most current recommendations by the American Academy of Clinical Toxicology and European Association of Centers and Clinical Toxicologists will be reviewed. Specific will be reviewed in the second section of these review articles. (CHEST 2003; 123:577–592)

Key words: critical care; ICU; poisoning; toxicology; toxidromes

Abbreviations: GL ϭ ; pKa ϭ negative logarithm of the acid ionization equilibrium constant

high index of suspicion for intoxication is war- unintentional, provide an approach to the diagnosis A ranted in the practice of critical care medicine. of the poisoned patient, and discuss strategies for The protean manifestations of intoxication challenge general supportive care. In part II, we will review the even the most astute clinicians, particularly when assessment and management of specific intoxications. patients present with altered mental status or when there is no history of intoxication. Recognition of a specific toxic syndrome (or toxidrome) helps (Table Epidemiology 1), but symptoms are often nonspecific (as in early acetaminophen poisoning) or masked by other con- Since 1983, the American Association of ditions (eg, myocardial ischemia in the setting of Control Centers has compiled data from the Toxic carbon monoxide poisoning). Exposure Surveillance System. In their 2000 annual In the first of this two-part series, we will review report, 63 poison centers reported a total of 2,168,248 the epidemiology of poisonings, both intentional and human toxic exposure cases. Adults accounted for approximately one third of exposures. Most exposures *From the Division of Pulmonary and Critical Care Medicine were unintentional (71% of cases) and involved a single (Dr. Mokhlesi), Cook County /Rush Medical College, toxic substance (92%). Fewer than 5% of cases involved Chicago; Evanston Northwestern Healthcare-OMEGA (Dr. Leiken), Chicago; Section of Nephrology (Dr. Murray), Univer- an adverse reaction to a or food. Oral sity of Chicago, Chicago; and Medical Intensive Care Unit ingestion was the commonest route of exposure (Fig 1). (Dr. Corbridge), Northwestern University Medical School, Most exposures occurred at the patient’s own resi- Chicago, IL. Manuscript received March 19, 2002; revision accepted July 12, dence, and most patients (75%) were managed on-site 2002. with assistance from a poison information center and Correspondence to: Babak Mokhlesi, MD, Division of Pulmonary did not require an emergency department visit. Only and Critical Care Medicine, Cook County Hospital/Rush Medical College, 1900 West Polk St, Chicago, IL 60612; e-mail: 3% of patients required critical care. [email protected] The categories of substances/toxins with the larg- www.chestjournal.org CHEST / 123/2/FEBRUARY, 2003 577 Table 1—Common Toxidromes

Toxidrome Features Drugs/ Drug Treatment Anticholinergic Mydriasis (for life-threatening “Hot as a hare, dry as a Blurred vision events, do not use in cyclic bone, red as a beet, mad Fever Baclofen antidepressant overdose because of as a hatter” Dry skin Benztropine potential worsening of conduction Flushing Tricyclic antidepressants disturbances) Ileus Phenothiazines Urinary retention Propantheline Tachycardia Scopolamine Hypertension Coma Seizures Myoclonus Cholinergic “SLUDGE” Salivation Carbamate Atropine Lacrimation Organophosphates for organophosphates Urination Physostigmine Pilocarpine GI cramps Emesis Wheezing Diaphoresis Bronchorrhea Bradycardia Miosis ␤-Adrenergic Tachycardia Albuterol ␤-Blockade (caution in asthmatics) Hypotension Caffeine Potassium replacement Terbutaline Theophylline ␣-Adrenergic Hypertension Phenylephrine Treat hypertension with phentolamine Bradycardia Phenylpropanolamine or nitroprusside, not with ␤- Mydriasis blockers alone ␤- and ␣-Adrenergic Hypertension Amphetamines Benzodiazepines Tachycardia Cocaine Mydriasis Ephedrine Diaphoresis Phencyclidine Dry mucus membranes Pseudoephedrine Sedative/hypnotic Stupor and coma Confusion Slurred speech Barbiturates Urinary alkalinization for Apnea Benzodiazepines phenobarbital Ethanol Meprobamate Opiates Hallucinogenic Amphetamines Benzodiazepines Psychosis Cannabinoids Panic Cocaine Fever Lysergic acid diethylamide Mydriasis Phencyclidine (may Hyperthermia present with miosis) Synesthesia Extrapyramidal Rigidity/tremor Haloperidol Diphenhydramine Opisthotonos Phenothiazines Benztropine Trismus Risperidone Hyperreflexia Olanzapine Choreoathetosis Narcotic Altered mental status Dextromethorphan Naloxone Slow shallow breaths Opiates Miosis Pentazocine Bradycardia Propoxyphene Hypotension Hypothermia Decreased bowel sounds

578 Critical Care Review Table 1—Continued

Toxidrome Features Drugs/Toxins Drug Treatment

Serotonin Irritability Benzodiazepine Hyperreflexia Meperidine Withdrawal of drug Flushing Paroxetine Diarrhea Diaphoresis Trazodone Fever Clomipramine Trismus Tremor Myoclonus Epileptogenic Hyperthermia Strychnine Antiseizure Hyperreflexia Nicotine Pyridoxine for isoniazid Lindane Extracorporeal removal of drug (lindane, May mimic stimulant Lidocaine camphor, xanthines) patterns Cocaine Physostigmine for anticholinergic agents Xanthines Avoid phenytoin for theophylline Isoniazid induced seizures Chlorinated hydrocarbons Anticholinergics Camphor Phencyclidine Solvent Lethargy Hydrocarbons Avoid catecholamines Confusion Acetone Withdrawal of Headache Toluene Restlessness Naphthalene Incoordination Trichloroethane Derealization Chlorinated hydrocarbons Depersonalization Uncoupling of oxidative Hyperthermia Aluminum phosphide Sodium bicarbonate for metabolic acidosis phosphorylation Tachycardia Salicylates Patient cooling Metabolic acidosis 2,4-Dichlorophenol Avoid atropine and salicylates Dinitrophenol in refractory acidosis Glyphosate Pentachlorophenol Zinc phosphide

est number of deaths were , antidepres- on vital signs, ocular findings, mental status, and muscle sants, sedative/hypnotics/antipsychotics, stimulants, tone can help determine drug or toxin class.3 “street” drugs, cardiovascular drugs, and (Table 2). Of all deaths, 920 fatalities, a 5% increase Vital Signs compared to 1999, 88% occurred in 20- to 99-year- old individuals. The mortality rate was higher in Anticholinergic and sympathomimetic substances increase heart rate, BP, and temperature. In contrast, intentional rather than unintentional exposures (79% ␤ vs 10.5%, respectively).1 organophosphates, opiates, barbiturates, -blockers, benzodiazepines, , and clonidine cause hypo- thermia, bradycardia, and respiratory . Table Diagnosis of Toxic Ingestion 4 lists various toxins altering temperature. Drugs/toxins History and Physical Examination causing tachycardia or bradycardia are listed in Table 5. Table 3 includes clinical features mandating con- Ocular Findings sideration of toxic ingestion. Although the history is important, it may be unreliable or incomplete.2 Con- Anticholinergics and sympathomimetics cause my- sider that family members, friends, and pharmacists driasis. In contrast to anticholingeric overdose, the may have additional information. In the absence of a pupils remain somewhat light responsive in cocaine classic presentation or toxidrome, separating patients intoxication. Table 6 lists drugs that affect pupil size. with suspected poisoning into broad categories based Horizontal is common in alcohol intoxi- www.chestjournal.org CHEST / 123/2/FEBRUARY, 2003 579 Figure 1. Route of exposure for human poisoning. Data from the 2000 Toxic Exposure Surveillance System of the American Association of Poison Control Centers.1 cation. Other drugs causing nystagmus are , cology screening confirms (or not) toxin exposure but , solvents, meprobamate, quinine, rarely alters management (see below). and primidone. Phencyclidine and phenytoin cause horizontal, vertical, and rotary nystagmus. Anion Gap The normal range of anion gap may vary from 3 Mental Status, Behavior, and Muscle Tone to 12 mEq/L in some laboratories.4 An increase in It is important to determine whether the patient is anion gap (Ͼ 20 mEq/L) suggests lactic acidemia, comatose, stuporous, lethargic, delirious, confused, uremia, ketoacidemia, or selected intoxications or alert (Table 7). Some toxins cause seizures (Table 8); others alter muscle tone (Table 9). Table 3—Clinical Features Mandating Consideration Laboratory Evaluation of Toxic Ingestion*

Three gaps are important in toxicology: the anion History of drug overdose or substance abuse gap, osmolal gap, and oxygen saturation gap. Toxi- or prior suicide attempt History of other psychiatric illness Agitation and hallucinations Table 2—Most Lethal Human Toxin Exposures Stupor or coma * Reported to Poison Control Centers in 2000 Rotary nystagmus Total Deaths or confusion per Category Seizures Muscle rigidity Adult Exposures, Including Dystonia Substance/Toxin No. (% of All Children and Cardiopulmonary arrest Category Adult Exposures) Adults, No.† Unexplained cardiac arrhythmia Analgesics 92,245 (13.3) 405 Hyper/hypotension Alcohols (ethanol 37,451 (5.4) 103 Ventilatory failure and nonethanol) Aspiration Antidepressants 55,429 (8) 242 Bronchospasm Cardiovascular 28,941 (4.2) 108 Liver failure drugs Renal failure Sedatives/hypnotics/ 67,946 (9.8) 225 Hyper/hypothermia antipsychotics Rhabdomyolysis Stimulants and 17,423 (2.5) 187 Osmolal gap street drugs Anion gap acidosis Hyper/hypoglycemia *Data obtained from cases reported by 63 poison control centers Hyper/hyponatremia during 2000. Not all poisonings and intoxications are reported to Hyper/hypokalemia poison control centers.1 Polypharmacy †No. of deaths are based on an unlimited number of substances coded per exposure. *Modified with permission from Corbridge and Murray.68

580 Critical Care Review Table 4—Drugs Affecting Temperature Table 6—Selected Drugs Affecting Pupil Size*

Hypothermia Hyperthermia Miosis Mydriasis Alcohols Amphetamines Barbiturates Amphetamines Barbiturates Anticholinergics Carbamates Anticholinergics Cyclic antidepressants Antihistamines Clonidine Antihistamines Hypoglycemic agents Cocaine Ethanol Cocaine Cyclic antidepressants Isopropyl alcohol Cyclic antidepressants Phenothiazines Drug withdrawal Organophosphates Dopamine Colchicine Lysergic acid diethylamide Opioids (meperidine may Drug withdrawal Akee fruit poisoning Monoamine oxidase inhibitors cause mydriasis) Glutethimide Lithium Malignant hyperthermia Phencyclidine Lysergic acid diethylamide Neuroleptic malignant syndrome Phenothiazines Monamine oxidase inhibitors Phencyclidine Physostigmine Phencylidine Phenothiazines Pilocarpine Salicylates *Modified with permission from Corbridge and Murray.68 syndrome

Osmolal Gap (Tables 10, 11). A normal anion gap does not Low-molecular-weight drugs and toxins increase preclude intoxication because most toxins do not the discrepancy between measured and calculated elevate the anion gap or there may be a coexisting plasma osmolality (Table 12). Normal plasma osmo- condition that lowers the gap (Table 10). Common lality is 285 to 295 mOsm. The calculated value is among these conditions is hypoalbuminemia: for determined as follows: every 1 g/L decrease in the plasma albumin, the anion gap falls by 2.5 mEq/L.5 Intensivists should pay special attention to this correction factor to avoid missing a clinically significant anion gap. Table 7—Selected Drugs Altering Mental Status Also, in or polyethylene glycol poison- Depressed Physiologic Agitated Physiologic Delirium and ing, concurrent ethanol use delays the develop- State State Confusion ment of an elevated anion gap metabolic acidosis. Sympatholytics Sympathomimetics Alcohol/drug In this case, an elevated osmolal gap may be the Adrenergic blockers Adrenergic agonists withdrawal only early clue to the diagnosis.6 Antiarrhythmics Amphetamines Anticholinergics Antihypertensives Caffeine Antihistamines Antipsychotics Cocaine Carbon monoxide Cyclic antidepressants Ergot alkaloids Cimetidine Cholinergics Monoamine oxidase Heavy Table 5—Selected Drugs/Toxins Causing Tachycardia Bethanechol inhibitors Lithium and Bradycardia* Carbamates Theophylline Salicylates Nicotine Anticholinergics Tachycardia Bradycardia Organophosphates Antihistamines Amphetamines Antiarrhythmics (types 1a and 1c) Physostigmine Antiparkinsonian Anticholinergics ␤-Blockers Pilocarpine drugs Antihistamines Calcium-channel blockers Sedative/hypnotics Antipsychotics Caffeine Carbamates Alcohols Antispasmotics Carbon monoxide Clonidine Barbiturates Cyclic antidepressants Clonidine Cyclic antidepressants Benzodiazepines Cyclobenzaprine Cocaine Digoxin Gamma Drug withdrawal Cyanide Lithium hydroxybutyrate ␤-blockers Cyclic antidepressants Metoclopramide Ethchlorvynol Clonidine Drug withdrawal Opioids Narcotics Ethanol Ephedrine Organophosphates Analgesics Opioids Hydralazine Phenylpropanolamine Antidiarrheal agents Sedative/hypnotics Hydrogen sulfide Physostigmine Other Hallucinogens Methemoglobinemia Propoxyphene Cyanide Lysergic acid Phencyclidine Quinidine Hydrogen sulfide diethylamide Phenothiazines Hypoglycemic agents Marijuana Pseudoephedrine Lithium Mescaline Theophylline Salicylates Phencyclidine hormone overdose Other Thyroid hormones *Modified with permission from Corbridge and Murray.68 www.chestjournal.org CHEST / 123/2/FEBRUARY, 2003 581 Table 8—Common Drugs and Toxins Causing Table 10—Common Causes of Abnormal Anion Gap Generalized Seizures* Elevated Anion Gap Decreased Anion Gap Amphetamines Lactic acidosis (type A) Increased unmeasured cation Antihistamines/anticholinergic agents Uremia Hyperkalemia Antipsychotics Sepsis Hypercalcemia Caffeine/theophylline Rhabdomyolysis Carbamates Ketoacidosis Acute lithium intoxication Carbon monoxide Diabetic Elevated IgG (myeloma; Cocaine Alcoholic cationic paraprotein) Cyclic antidepressants Starvation Decreased unmeasured anion Ethylene glycol Toxic ingestions* Hypoalbuminemia Isoniazid Ethylene glycol Drugs Lead Methanol Bromide Lidocaine Paraldehyde Iodide Lithium Salicylate Lithium Methanol Metabolic alkalosis with Polymyxin B Organophosphates volume depletion Tromethamine Phencyclidine Analytical artifact Hypoglycemic agent (focal) Hypernatremia (Ͼ 170 mEq/L) Chlorambucil (focal) Hyperlipidemia Propranolol Salicylates *See Table 11. Withdrawal from alcohol or sedative/hypnotics *Modified with permission from Corbridge and Murray.68 ered the upper limit of normal.9 However, an osmo- lal gap of 10 mOsm in a patient who started at Ϫ 9 ϩ calculated osmolality ϭ 1.86[Na ] ϩ BUN/2.8 mOsm may be significantly elevated.10–12 ϩ ϩ glucose/18 ethanol/4.6 Oxygen Saturation Gap ϩ in which Na (in millimoles per liter) is multiplied An oxygen saturation gap is present when there is by nearly two to account for accompanying anions more than a 5% difference between the saturation (chloride and bicarbonate), and measured BUN, calculated from an arterial blood gas and the satura- glucose, and ethanol are converted from milligrams tion measured by co-oximetry. Co-oximetry deter- per deciliter to mmol/L by the appropriate denomi- mines oxygen saturation by detecting the absorption nator. of four different wavelengths, enabling it to directly The osmolal gap must be interpreted with caution. measure levels of four types of hemoglobin species: Measurement of osmolality by vapor pressure os- oxyhemoglobin, reduced hemoglobin, carboxyhemo- mometry does not detect volatile alcohols such as globin, and methemoglobin. However, arterial blood ethanol and methanol; however, it does detect eth- gas analysis calculates oxygen saturation from the ylene glycol. Freezing point depression osmometry, measured oxygen tension using an assumed standard the most frequently used method, measures all of these solutes.7,8 Therefore, it is important for clini- cians to know the method used by their institution to avoid missing methanol poisoning. By using the Table 11—Selected Drugs Associated With an Elevated Anion Gap Metabolic Acidosis standard formula, the normal osmolal gap may range from Ϫ 9 mOsm to ϩ 5 mOsm; 10 mOsm is consid- Acetaminophen (Ͼ 75 g) Ketamine Amiloride Metformin Ascorbic acid Methanol Carbon monoxide Niacin Table 9—Selected Drugs Affecting Muscle Tone* Chloramphenicol Nitroprusside Colchicine Nonsteroidal anti-inflammatory drugs Dystonic Reactions Dyskinesias Rigidity Nitroprusside Papaverine Haloperidol Anticholinergics Black widow bite Paraldehyde (hippuric acid) Metoclopramide Cocaine Malignant hyperthermia Epinephrine Phenformin Olanzapine Phenylcyclidine Neuroleptic malignant Ethanol Propofol syndrome Ethylene glycol Salicylates Phenothiazines Risperidone Phenylcyclidine Formaldehyde Terbutaline Risperidone Strychnine Hydrogen sulfide Tetracycline (outdated) Fentanyl Iron Toluene (hippuric acid) Isoniazid Verapamil *Modified with permission from Corbridge and Murray.68

582 Critical Care Review Table 12—Drugs/Toxins Associated With an Elevated Table 13—Drugs Commonly Included in Urine Osmolal Gap* Substances-of-Abuse Screens (Available in 30 min)*

Ethanol (if not included in the formula) Amphetamines Ethylene glycol/glycolaldehyde Barbiturates Glycerol Benzodiazepines Glycine Cannabinoids IV immunoglobulin (maltose) Cocaine Isopropanol/acetone Opioids Mannitol Phencyclidine Methanol/formaldehyde *Immunoassay technique; modified with permission from Corbridge Propylene glycol and Murray.68 Radiocontrast media Hypermagnesemia (Ͼ 9.5 mEq/L) Sorbitol 18,19 *Modified with permission from Corbridge and Murray.68 cases. Toxicology screening can identify a spe- cific toxin for which an is available and in some instances quantify a toxin allowing for titrated oxygen-hemoglobin dissociation curve. Toxins that therapy. are associated with an elevated oxygen saturation gap Most institutions offer urine testing for six or include carbon monoxide, methemoglobinemia, cy- seven of the most commonly abused drugs (Table anide, and hydrogen sulfide (sulfhemoglobin is not 13). Results are generally available in 30 min. More routinely measured by co-oximetry). Pulse oximetry comprehensive urine screening (usually performed estimates oxygen saturation by emitting a red light off-site) may take up to 2 to 3 h. Testing of blood or (wavelength of 660 nm) absorbed mainly by reduced gastric contents is rarely indicated.20 However, blood hemoglobin and a near-infrared light (wavelength of quantification of certain toxins is useful, particularly 940 nm) absorbed by oxyhemoglobin.13 Methemo- in cases of alcohol (ethanol and nonethanol), acet- globin absorbs almost equally at both these wave- aminophen, salicylate, phenobarbital, theophylline, lengths. At high methemoglobin levels (35%), the digoxin, iron, and lithium intoxication. A strong oxygen saturation by pulse oximetry tends to regress argument can be made for checking acetaminophen toward 85% and plateaus at that level despite further levels in all cases of suspected intoxication given the increments in methemoglobin levels. Thus, if the subtle manifestations of early acetaminophen poi- actual oxygen saturation by co-oximetry is Ͼ 85%, soning and importance of targeted therapy. the pulse oximetry would be underestimating it; if it is Ͻ 85% by co-oximetry, it would be overestimating Consultation oxygen saturation.14 Therefore, pulse oximetry may become unreliable in the setting of methemoglobin- Regional poison control center consultation is emia registering falsely high in patients with severe highly recommended in cases of suspected poisoning 21 methemoglobinemia and falsely low with mild met- and to help guide management in confirmed cases. hemoglobinemia. Since many laboratories do not These centers provide 24-h emergency and up-to- routinely use co-oximetry, a more commonly seen date technical information. They are staffed by gap may be the disparity between measured oxygen nurses, pharmacists, pharmacologists, and physicians saturation by blood gas and that measured by pulse trained and certified in toxicology. The national oximetry.15 Oxygen saturation measured by pulse toll-free number for poison control centers is 800- oximetry may be falsely elevated in methemoglobin- 222-1222. emia and should be utilized with caution in deter- mining the oxygen saturation gap.16,17 Carbon mon- oxide has a wavelength absorption coefficient similar Initial Supportive Measures to that of oxyhemoglobin; therefore, it is registered Airway, Breathing, Circulation as oxyhemoglobin by pulse oximetry leading to over- Supportive measures including the “ABCs” (air- estimation of oxygen saturation when compared to way, breathing, circulation) are often required be- co-oximetry. An abnormally high venous oxygen fore confirmation of intoxication. With cervical spine content (arteriolization of venous blood) is charac- precautions in place (unless trauma has been exclud- teristic of cyanide and hydrogen sulfide poisoning. ed), airway patency must be ensured in all cases. Endotracheal intubation is not always necessary Toxicology Screening when cough and gag reflexes are present and there is In spite of providing direct evidence of intoxica- adequate spontaneous ventilation, but when there is tion, screening tests alter management in Ͻ 5% of concern regarding airway protection and clinical www.chestjournal.org CHEST / 123/2/FEBRUARY, 2003 583 deterioration it is better to secure the airway. Intu- bation is indicated in acute respiratory failure (Table Table 14—Selected Causes of Hypoxemia in Drug Overdose and Toxic Ingestion* 14 for causes of hypoxemia in intoxicated patients). Other specific indications include the need for high Cause Drugs/Toxins levels of supplemental oxygen in carbon monoxide Hypoventilation Alcohols poisoning and the need to protect the airway for Barbiturates gastric emptying. Endotracheal intubation decreases Benzodiazepines (but does not eliminate) the risk of aspiration (which Botulinum toxin is approximately 11% in the comatose patient with Cyclic antidepressants 22–24 Neuromuscular blockade drug overdose). Opioids Depending on the intoxication, patients may Sedative/hypnotics present with hypotension or hypertension, brady- Snake bite arrhythmias or tachyarrhythmias. The pathogenesis Strychnine of hypotension varies and may include , Aspiration Drugs/toxins depressing mental status myocardial depression, cardiac arrhythmias, and sys- temic vasodilation. Treatment should be individual- Pneumonia Drugs resulting in aspiration; IV drug ized, but an initial strategy of rapid IV normal abuse with pulmonary seeding of solution infusion is indicated in most instances. infectious agents; inhalation interfering with lung protective Vasopressors may be required for refractory hypo- mechanisms tension. The vasopressor of choice depends on the Cardiogenic Antiarrhythmics type of intoxication (see below). Hypertension occurs pulmonary edema ␤-Blockers in the setting of sympathomimetic drugs, anticholin- Cyclic antidepressants ergics, ergot derivatives, phenylpropanolamine over- Verapamil Inert gases Carbon dioxide dose, and withdrawal from nicotine, alcohol, and Methane sedatives. Treatment of the hypertension depends on its chronicity and severity and the inciting agent (see Propane below). Hypertension-induced (reflex) bradycardia Noncardiogenic Cocaine generally should not be treated. pulmonary edema Ethylene glycol Hydrocarbons Inhalation injury Coma Cocktail Opioids Phosgene Immediately after establishing IV access, a “cock- Paraquat tail” of , dextrose, and naloxone should be Salicylates administered to patients with depressed mental sta- Bronchospasm ␤-Blockers Cocaine tus. This cocktail can be both therapeutic and diag- 25 Heroin nostic. Thiamine (100 mg by vein) is administered Organophosphates to treat and/or avoid Wernicke-Korsakoff syndrome Drugs resulting in aspiration in comatose patients. This strategy is not well sup- Drugs associated with myocardial ported by the literature, and few patients regain depression (cardiac asthma) consciousness following thiamine infusion. Still, rou- Alveolar hemorrhage Cocaine tine use of thiamine is safe, inexpensive, and Thrombolytics prevents the possibility of delayed deterioration sec- Amiodarone ondary to nutritional deficiency.25 Thiamine is par- Paraldehyde ticularly important in the nutritionally deplete alco- Nitrofurantoin holic. There is no evidence that dextrose should be Penicillamine 26 Toluene withheld until thiamine is administered. Comatose Pneumothorax Cocaine patients should receive dextrose, 50 g IV. A normal IV drug abuse with aberrant value by blood dipstick does not necessarily exclude venipuncture or bullous lung low serum glucose. A high value on dipstick testing disease should lead to rapid confirmation by blood draw, Kerosene Cellular hypoxia Carbon monoxide thus avoiding unnecessary dextrose (although admin- Cyanide istration of dextrose to hyperglycemic patients is Hydrogen sulfide unlikely to cause harm).25 Naloxone rapidly reverses Methemoglobinemia coma, respiratory depression, and hypotension in- Sulfhemoglobinemia duced by opioids. An initial dose of 0.2 to 0.4 mg is *Modified with permission from Corbridge and Murray.68 administered IV (or endotracheally). If there is no

584 Critical Care Review response after 2 to 3 min, an additional 1 to 2 mg can exposure and must be removed with caution and be administered and repeated up to 10 mg as placed in plastic bags or other containers that are required. Using a higher dose up front may pre- impervious to the toxin. This will limit exposure to cipitate large cardiovascular changes in medical personnel and patient. Ocular decontamina- dependent patients. Several opioids such as me- tion may require prolonged periods of irrigation with peridine, propoxyphene, diphenoxylate, metha- normal saline solution using a Morgan lens (MorTan; done, and pentazocine require large doses of Missoula, MT). Inhalational exposure presents a naloxone,27 but lack of response to 10 mg of greater challenge since the toxin cannot be accessed naloxone generally excludes opioid toxicity. Opioid and removed. Inhalational lung injury is beyond the antagonism with naloxone lasts 1 to 4 h requiring scope of this review. The majority of toxin exposures repeat doses or continuous infusion in significant and poisonings managed by intensivists occur intoxication.28 Acute pulmonary edema,29,30 opioid through the GI tract. There are four methods of GI withdrawal,31 and seizures32 have been reported decontamination including three mechanical ap- with naloxone administration. proaches (emesis, gastric emptying or gastric lavage Flumazenil should be considered in cases were [GL], and whole-bowel irrigation) and the use of is suspected or reversal of activated charcoal combined with a cathartic. therapeutic conscious sedation is desired.25,33 Case reports have cautioned clinicians of the risk of Emesis precipitating seizures with flumazenil when there is a Ipecac-induced emesis should be considered only suspicion of benzodiazepine plus cyclic antidepres- in fully alert patients, and is virtually never indicated sant overdose.34,35 Nonetheless, data suggest that after hospital admission. Ipecac is generally less flumazenil is safe as part of the coma cocktail even traumatic than GL, and is therefore the preferred with coma induced by the combination of benzodi- method of gastric emptying in pediatric patients. azepines and cyclic antidepressants.36 In a large Ipecac may be helpful at home if administered prospective trial of unconscious patients suspected of immediately after ingestion. In the best of circum- benzodiazepine overdose, Weinbroum et al36 ran- stances, a 30 to 40% removal rate can be achieved domized patients to receive either placebo or fluma- within 1 h after ingestion.39 Because of questionable zenil in addition to usual care. Seventy-one percent efficacy hours after ingestion, in-hospital use is de- of the patients had concomitant cyclic antidepressant creasing.40,41 Contraindications to its use include ingestion. These investigators did not observe any poisoning with corrosives, petroleum products, or significant side effects with flumazenil, even in pa- antiemetics. The potential for aspiration precludes tients with coma caused by a mixed overdose of its use in situations where there is a high risk of benzodiazepine and cyclic antidepressants. We typ- seizures (ingestion of a rapidly acting convulsant ically administer an initial 0.2 mg of IV flumazenil such as strychnine) or altered consciousness.42 The over 30 s followed by another 0.3-mg dose if neces- usual dose of ipecac syrup in adults is 30 mL sary. Doses beyond 3 mg generally do not provide followed by 16 oz of water. This dose usually additional benefit. Repeat sedation may occur in the induces within 20 to 30 min. The dose setting of high-dose or long-term use of benzodiaz- can be repeated once after 30 min if vomiting does epines. Although flumazenil is successful in improv- not occur. There is little evidence that ipecac ing the Glasgow coma scale score, it does not appear prevents drug absorption or systemic toxicity,43 to alter cost or major diagnostic/therapeutic inter- and there are no convincing data that it signifi- ventions in patients presenting with decreased level cantly alters the clinical outcome of patients who of consciousness due to an intentional unknown drug are awake and alert on presentation to the emer- overdose.37 Therefore, the cost-effectiveness of rou- gency department. Ipecac is rarely used (approxi- tine use of flumazenil as part of the coma cocktail mately 1% of all overdoses reported to the poison remains controversial,38 except in cases of acute centers),1 and its use may soon be confined to the benzodiazepine overdose.36 medical history books.

Gastric Emptying Prevention of Absorption GL through a 28F to 40F Ewald tube is similarly The route of entry for toxic substances can be aimed at physically removing a toxin. Prior to insert- dermal, ocular, GI, inhalational, or parenteral (Fig ing the Ewald tube, the mouth should be inspected 1). Skin decontamination requires removal of the for foreign material and equipment should be ready toxin with nonabrasive soap and water. Contami- for suctioning. Large gastric tubes (37F to 40F) are nated clothing may serve as a reservoir for continued less likely to enter the trachea than smaller nasogas- www.chestjournal.org CHEST / 123/2/FEBRUARY, 2003 585 tric tubes, and are necessary to facilitate removal of ment of poisoning unless a patient has ingested a gastric debris. After insertion, proper position needs potentially life-threatening amount of a poison and to be confirmed by aspirating acidic stomach con- the procedure can be undertaken within 60 min of tents and auscultating the left upper abdominal ingestion.50 Although controversial, some experts quadrant during insufflation of air. Experienced suggest that the time limit may be extended to 12 h personnel should perform GL in a facility where in cases of poisoning with agents that delay gastric resources are available to manage complications. emptying such as tricyclic antidepressants, opioids, Nonintubated patients must be alert (and be ex- or salicylates. In addition, gastric emptying may be pected to remain alert) and have adequate pharyn- beneficial if the ingested drug is not adsorbed by geal and laryngeal protective reflexes. In semicoma- activated charcoal (eg, ferrous sulfate, lithium). In tose patients, GL should be performed only after a cases of ingestion of a caustic liquid such as kerosene cuffed endotracheal tube has been inserted. Intuba- or its derivatives, GL should be avoided because of tion for the sole purpose of gastric emptying is the risk of aspiration-induced lung injury. Clinical reasonable only if there is a high likelihood that a studies evaluating the efficacy of GL are limited by highly lethal agent remains in the stomach. small study size, heterogeneity of toxins studied, and GL is performed by instilling 200-mL aliquots of different methodologies. There is also a concern that warmed tap water until there is clearing of aspirated GL may propel material into the duodenum increas- fluid. Stomach contents should be retained for anal- ing the chance of drug absorption.51 ysis. Tap water may avoid unnecessary salt loading compared to normal saline solution. Neither irrigant Activated Charcoal has been shown to significantly alter blood cell or electrolyte concentrations.44 After clearing, the Charcoal is a by-product of the combustion of Ewald tube may be replaced by a nasogastric tube various organic compounds such as wood, coconut for subsequent intermittent suctioning and/or ad- parts, bone, sucrose, rice, and starch. Its adsorptive ministration of activated charcoal. capacity is increased or activated by removing mate- GL has been advocated in the initial management rials previously adsorbed by a process that involves of many orally ingested agents. The risks associated steam heating and chemical treatment, thereby in- with this procedure include aspiration, arrhythmias, creasing the surface area available for adsorption to and stomach perforation.45 Because of these risks, between 1,000 m2/g and 3,000 m2/g. This results in a GL should not be performed in patients who have powerful, inert, nontoxic, and nonspecific adsorbent ingested a nontoxic substance, a nontoxic amount of that irreversibly binds intraluminal drugs and inter- a toxic substance, or when the toxin is no longer feres with their absorption. It is particularly effective expected to be present in the stomach. Examples in binding high-molecular-weight compounds. Acti- include patients who have vomited extensively prior vated charcoal decreases serum drug levels in some to hospital admission, patients who present several cases by creating a favorable diffusion gradient be- hours after ingesting an agent that does not decrease tween blood and gut, referred to as GI dialysis (see gut motility, and patients who have received agents below).52 The efficacy of activated charcoal has lead that are readily absorbed from the GI tract. Although to a resurgence of its use over the past few years. GL has been common in the management of patients Charcoal can be administered after both GL or with toxic ingestion, its use remains controversial.46 ipecac-induced emesis, but it is usually administered In obtunded patients, GL results in a more satisfac- as the sole GI decontaminating agent. Airway pro- tory clinical outcome only if performed within 1 h47 tection is imperative in stuporous, comatose, or or2hofingestion.48 Kulig et al47 compared the convulsing patients. Prior gastric stapling is an addi- utility of GL plus activated charcoal in 72 obtunded tional risk factor for emesis and aspiration with single patients with 44 obtunded patients who received or repeated doses.53 Charcoal aspiration has been only activated charcoal by nasogastric tube and associated with pneumonia54 (including fungal pneu- supportive care. They reported an improved clinical monia55), bronchiolitis obliterans,56 ARDS,57 and course if lavage was performed within1hofinges- death.58 tion. In contrast, Pond et al49 performed a prospec- Despite the mentioned complications, activated tive, randomized, controlled trial of 347 obtunded charcoal is generally effective and well tolerated. patients receiving GL plus activated charcoal or Complications are infrequent. The ideal dose should activated charcoal alone. There was no significant give a charcoal-to-drug ratio of 10:1. However, since difference in outcome even when patients presented the quantity of poison ingested is usually unknown to within 60 min of ingestion. Because of limited data, the clinician, the dose is based on actual patient the American Academy of Clinical Toxicology does weight (1 g/kg). It is commonly co-administered with not recommend routine use of GL in the manage- a cathartic (see below) to facilitate evacuation of the

586 Critical Care Review Table 15—Toxins and Drugs Not Adsorbed by administered, it should be limited to a single dose in Activated Charcoal order to minimize adverse effects.63 Alcohols Hydrocarbons Whole-Bowel Irrigation Organophosphates Carbamates Whole-bowel irrigation with a polyethylene glycol, Acids electrolyte solution (Colyte; Schwarz Pharma; Mil- Potassium waukee, WI) or potassium chloride (Golytely; Brain- Dichloro diphenyl trichloroethane (DDT) Alkali tree Laboratories; Braintree, MA), 1 to 2 L/h, in Iron adults is used to push tablets or packages through the Lithium GI track. The optimal regimen in regards to volume infused per hour, duration of use, and dosage of activated charcoal prior to whole-bowel irrigation has not been well established.46 It may take 3 to 5 h for toxic substance and avoid constipation. Commonly complete bowel irrigation to clear the rectal effluent. used agents include magnesium sulfate, magnesium These isotonic solutions are not absorbed and do not citrate, or sorbitol. Mixing the solution with juice cause major electrolyte shift or imbalance.64 The may increase acceptance of this black and gritty technique is time consuming and requires a cooper- adsorbent in children and adults. Single-dose acti- ative patient. Most studies supporting this approach vated charcoal is effective against most toxins and are limited to case reports, and there are no estab- drugs. Table 15 lists selected toxins for which char- lished indications for its use. However, whole-bowel coal is not particularly effective. Based on volunteer irrigation may have a role in intoxications where studies, the effectiveness of activated charcoal de- activated charcoal is not effective, such as ingestion creases with time; the greatest benefit is within1hof 59 of iron and sustained-release tablets, lithium, or in ingestion. cases of “body packing” with packages of illicit drugs.65 Contraindications to whole-bowel irrigation Catharsis include ileus, GI hemorrhage, and bowel perforation. The use of cathartics with activated charcoal may reduce the transit time of drugs and toxins in the GI tract and decrease the constipating effects of char- Enhancement of Elimination coal. Sorbitol is the cathartic of choice. It is generally Forced Diuresis and Urinary pH Manipulation administered only with the first dose of activated charcoal. The usual dosage is 1 to 2 mL/kg of a 70% Routine use of volume-loading to promote diuresis solution of sorbitol titrated to several loose stools has not been well studied or supported in the over the first day of treatment (total dose, 1 g/kg). literature and cannot be recommended. Its goal is to Magnesium-based cathartics (2 to 3 mL/kg po of a augment elimination of renally excreted toxins 10% solution of magnesium sulfate) may lead to through inhibition of tubular reabsorption. Thus, in magnesium accumulation in the setting of renal order to be effective, the toxin needs to undergo failure, and sodium-based products carry the risk of extensive tubular reabsorption that can be inhibited exacerbating hypertension or congestive heart fail- by forced diuresis. However, forced diuresis has the ure. Oil-based cathartics, if aspirated, may produce potential to cause electrolyte imbalance, pulmonary lipoid pneumonia. edema, and raised intracranial pressure.66 The tech- Cathartics have never been shown to decrease nique consists of achieving a urine flow rate from 3 morbidity and mortality or to decrease hospital to 6 mL/kg/h with a combination of isotonic fluids stay.60 In a cross-over study, Keller et al61 demon- and/or diuretics.67 When tubular reabsorption of a strated that activated charcoal with sorbitol led to a toxin is pH sensitive, then increased urine flow does 28% decrease in the absorption of salicylates when not significantly increase urinary drug elimination compared to activated charcoal alone. However, when added to alkaline or acid diuresis. McNamara and colleagues62 were unable to demon- Manipulation of urinary pH can be used therapeu- strate enhanced efficacy of activated charcoal with tically to enhance elimination of some intoxicants sorbitol catharsis in a simulated acetaminophen over- (Table 16). The limits of urinary pH are 4.5 to 7.5 dose. Based on available data, the routine use of a under conditions of enhanced acidification and alka- cathartic in combination with activated charcoal is linization. Thus, elimination of very strong (negative not endorsed by the American Academy of Clinical logarithm of the acid ionization equilibrium constant Toxicology and the European Association of Poisons [pKa] Ͻ 3) or very weak (pKa Ͼ 8) acids is unaltered Centers and Clinical Toxicologists. If a cathartic is by urinary pH manipulation. Other acidic or basic www.chestjournal.org CHEST / 123/2/FEBRUARY, 2003 587 Table 16—Toxins Eliminated by Urinary have been absorbed.69 The mechanism by which this Alkalinization modality accomplishes enhancement of elimination 2,4 Dichlorophenoxy-acetic acid is either by interrupting the enterohepatic/entero- Fluoride gastric circulation of drugs or through the binding of Isoniazid any drug that diffuses from the circulation into the Mephobarbital gut lumen (called GI dialysis). However, it has Methotrexate Phenobarbital limited application because the toxin must have a low Primidone volume of distribution, low binding, pro- Quinolone longed elimination half-life, and low pKa, which maximizes transport across mucosal membranes into Uranium the GI tract.67 Although optimal dosage and fre- quency of administration following the initial dose of drugs do not undergo renal tubular absorption, activated charcoal is not well established, most ex- irrespective of urinary pH, since they are polar in perts recommend a dose not Ͻ 12.5 g/h.70 After the their nonionized form. initial dose of 1 g/kg, activated charcoal may be Ͼ Urinary alkalinization (pH 7) is most often used administered at 0.5 g/kg every 2 to 4 h for at least to eliminate salicylates and phenobarbital. It can be three doses. Cathartics are generally not adminis- achieved by administration of IV sodium bicarbonate tered to avoid hypernatremia, hypokalemia, and (1 to 2 mEq/kg every 3 to 4 h); this may be hypermagnesemia. Multiple dosing should be used administered as two 50-mL ampules of 8.4% sodium with caution in patients with decreased bowel bicarbonate (each containing 50 mEq of NaHCO3) sounds, abdominal distension, and persistent emesis. per liter of 5% dextrose in water infused at Unless a patient has an intact or protected airway, 250 mL/h. Complications of this therapy include the administration of multidose charcoal is contrain- alkalemia (particularly in the presence of concurrent dicated. In a review of all the relevant scientific respiratory alkalosis), volume overload, hypernatre- literature, the American Academy of Clinical Toxi- mia, and hypokalemia. It is particularly important to avoid hypokalemia, which prevents excretion of al- cologists reported that although multidose charcoal kaline urine by promoting distal tubular potassium enhances drug elimination significantly, it has not yet reabsorption in exchange for hydrogen ion. Accord- been evaluated in a controlled trial of poisoned ingly, bicarbonate administration in the presence of patients with the objective of demonstrating a reduc- significant hypokalemia will not alkalinize the urine, tion in morbidity and mortality.71 Table 17 provides yet will increase the risk of alkalemia. Since urinary a list of drugs and toxins where there may be a role alkalinization therapy can cause hypokalemia (due to for multiple dosing of activated charcoal.67 However, alkalemia-induced intracellular potassium shift and based on experimental and clinical studies, it should increased urinary potassium loss with alkaline diure- be considered only in patients with a life-threatening sis), addition of potassium chloride to the bicarbon- ingestion of carbamazepine, dapsone, phenobarbital, ate infusion is commonly required. Acetazolamide quinine, or theophylline.71 should not be used to alkalinize urine. Resultant metabolic acidosis can increase toxicity of certain poisonings (particularly in the case of ).68 Table 17—Toxins and Drugs Eliminated by Multiple Urinary acidification (pH Ͻ 5.5) increases renal Dosing of Activated Charcoal* clearance of some nonpolar weak bases with pKa Amitriptyline Meprobamate values between 6 and 12. or lysine hydro- Amoxapine Methyprylon chloride or ammonium chloride have been used for Baclofen (?) Nadolol urinary acidification. However, due to the potential Benzodiazepines (?) Nortriptyline of urinary acidification to exacerbate myoglobinuric (?) Phencyclidine renal tubular injury, this therapy is virtually never Carbamazepine Phenobarbital Chlordecone Phenylbutazone used. Also, systemic acidosis must be avoided in Dapsone Phenytoin (?) order to avoid potential additive effects with toxin- Piroxicam induced metabolic or respiratory acidosis.68 Digitoxin Propoxyphene Digoxin Quinine Multiple-Dose Activated Charcoal Disopyramide Salicylates (?) Glutethimide Sotalol Multiple-dose activated charcoal can be an effec- Maprotiline Theophylline tive way to enhance the elimination of toxins that *? Represents equivocal data.

588 Critical Care Review Table 18—* Table 19—Criteria for Admission of the Poisoned Patient to the ICU* Drug/poison Antidotes Respiratory depression (Paco Ͼ 45 mm Hg) Acetaminophen N- 2 Emergency intubation Anticholinergics Physostigmine Seizures Anticholinesterases Atropine Cardiac arrhythmia (second- or third-degree atrioventricular block) Benzodiazepines Flumazenil Systolic BP Ͻ 80 mm Hg Black widow Equine-derived antivenin Unresponsiveness to verbal stimuli Carbon monoxide Oxygen Glasgow coma scale score Ͻ 12 Coral snake (Eastern and Equine-derived antivenin Need for emergency dialysis, hemoperfusion, or ECMO Texas) bite Increasing metabolic acidosis Cyanide Amyl , , sodium Pulmonary edema induced by toxins (including inhalation) or drugs thiosulfate, hydroxycobalamin Hypothermia or hyperthermia including neuroleptic malignant Digoxin Digoxin-specific antibodies syndrome Ethylene glycol Ethanol/, thiamine, and Tricyclic or phenothiazine overdose manifesting anticholinergic pyridoxine signs, neurologic abnormalities, QRS duration Ͼ 0.12 s, or QT Heavy metals (arsenic, (BAL), EDTA, Ͼ 0.5 s copper, gold, lead, penicillamine Body packers and stuffers mercury) Concretions caused by drugs Hypoglycemic agents Dextrose, , octreotide Emergency surgical intervention Iron Deferoxamine mesylate Administration of pralidoxime in organophosphate toxicity Isoniazid Pyridoxine administration in Crotalidae, coral snake, or Methanol Ethanol or fomepizole, folic acid Methemoglobinemia Need for continuous infusion of naloxone Opioids Naloxone Hypokalemia secondary to digitalis overdose (or need for digoxin- Organophosphate Atropine, pralidoxamine immune antibody Fab fragments) Rattlesnake bite Equine-derived antivenin *ECMO ϭ extracorporeal membrane oxygenation. *EDTA ϭ ethylenediamine tetra-acetic acid.

Extracorporeal Removal of Toxins hemodialysis may be useful should have a low mo- lecular weight (Ͻ 500 d), be water soluble, have low In situations where previously mentioned support- protein binding (Ͻ 70 to 80%), and have a small ive measures fail to improve a patient’s condition, volume of distribution (Ͻ 1 L/kg). It can especially extracorporeal removal of toxins can be lifesav- be effective in correcting concomitant electrolyte ing.72,73 Although clear proof that extracorporeal abnormality and metabolic acidosis. Toxins in which toxin removal favorably alters the course of any hemodialysis may be required in an early stage of intoxication is generally lacking,74 it should be con- intoxication include methanol, ethylene glycol, boric sidered when the intoxication is projected to un- acid, salicylates, and lithium. Hemodialysis can also dergo delayed or insufficient clearance because of be used for heavy chelation in patients with other organ dysfunction, the intoxicating agent pro- renal failure. duces toxic metabolites, or delayed toxicity is char- acteristic of the intoxication. In addition to physico- Hemoperfusion chemical properties of the intoxicant, serum toxin Hemoperfusion is defined as direct contact of levels or certain clinical features may mandate extra- 75 corporeal removal techniques. Three methods for blood with an adsorbent system. Charcoal he- extracorporeal removal of toxins are generally avail- moperfusion involves pumping blood through a able: (1) dialysis (usually hemodialysis rather than charcoal canister. Unlike hemodialysis, drug clear- peritoneal dialysis), (2) hemoperfusion; and (3) he- ance is not limited by low water solubility, high mofiltration. Plasmapheresis and exchange transfu- molecular weight, or increased protein binding, but sion are rarely used and will not be further discussed on the ability of the adsorbent to bind to the in this review. A complete list of drugs and toxins drug/toxin. However, the toxin needs to be present that may be removed by different extracorporeal in the central compartment for hemoperfusion to be removal techniques is beyond the scope of this effective. Hemoperfusion is essentially the paren- review.67,68 teral analog of oral activated charcoal. Complications of hemoperfusion include the following: (1) cartridge saturation; (2) thrombocytopenia that commonly oc- Hemodialysis curs due to platelet adsorption, inducing up to 30% Hemodialysis is the primary extracorporeal decrement in platelet count; (3) hypoglycemia and method to remove toxins or drugs. Toxins for which hypocalcemia; (4) access complications; (5) hypo- www.chestjournal.org CHEST / 123/2/FEBRUARY, 2003 589 thermia, since hemoperfusion pumps do not warm Indications for ICU Admission blood as hemodialysis does; and (6) charcoal embo- lization (prevented by a filter in the line returning In the current health-care climate, the practice of routinely admitting the poisoned patient to the ICU effluent blood to the patient). is being questioned. Brett et al81 identified eight Most drugs are extractable by hemoperfusion, clinical risk factors that can predict ICU interven- which is particularly suitable for extracorporeal re- Ͼ tions: (1) Paco2 45 mm Hg, (2) need for endotra- moval of toxins that are of high molecular weight, cheal intubation, (3) toxin-induced seizures, (4) car- highly protein bound, or lipid soluble. It has been diac arrhythmias, (5) QRS duration Ն 0.12 s, effectively used to enhance elimination of theophyl- (6) systolic BP Ͻ 80 mm Hg, (7) second- or third- line, phenobarbital, phenytoin, carbamazepine, para- degree atrioventricular block, and (8) unresponsive- quat, and glutethimide. Drugs poorly extracted by ness to verbal stimuli. In this retrospective study, if a hemoperfusion include the following: heavy metals poisoned patient did not exhibit any of the eight (lithium, bromide), some alcohols (ethanol, metha- characteristics, no ICU interventions (intubation, nol), carbon monoxide, and some illicit drugs (co- vasopressors or antiarrhythmics, and dialysis or he- caine, phencyclidine, and others). Efficacy of intox- moperfusion) were required. Other indications for icant removal is diminished for substances with a ICU admission include a Glasgow coma scale score large volume of distribution that are highly lipid Ͻ 12,82 need for emergency dialysis or hemoperfu- soluble and/or extensively tissue bound. These intox- sion, progressive metabolic acidosis, and a cyclic icants may be more effectively removed by hemofil- antidepressant or phenothiazine overdose with signs tration. of anticholinergic cardiac toxicity.83,84 Severe hyper- kalemia, wide alterations in body temperature, and need for continuous infusion of naloxone are also Hemofiltration reasons to admit a patient to an ICU. In addition, Hemofiltration achieves drug and toxin removal by staffing issues such as the availability of a “sitter” in convection. It transports solutes through a highly cases of attempted suicide may impact patient dis- porous membrane that is permeable to substances position. Table 19 provides a list of criteria for ICU with weights of up to 6,000 d, including virtually all admission.67 drugs. In some cases, hemofiltration membranes are permeable to substances weighing up to 20,000 d.76,77 Although the application of this technique has References not been vigorously studied in poisoned patients, 1 Litovitz TL, Klein-Schwartz W, White S, et al. 2000 annual report of the American Association of Poison Control Centers there are increasing numbers of case reports of toxic exposure surveillance system. Am J Emerg Med 2001; extracorporeal intoxicant removal by either the con- 19:337–395 tinuous arteriovenous or venovenous hemofiltration 2 Wright N. An assessment of the unreliability of the history methods.78–80 Hemofiltration is potentially useful for given by self-poisoned patients. Clin Toxicol 1980; 16:381– 384 removal of substances with a large volume of distri- 3 Olson KR, Pentel PR, Kelley MT. Physical assessment and bution, slow intercompartmental transfer, or exten- differential diagnosis of the poisoned patient. Med Toxicol sive tissue binding. Specific highly porous hemofil- 1987; 2:52–81 tration cartridges are also particularly useful for 4 Winter SD, Pearson R, Gabow PA, et al. The fall of the serum anion gap. Arch Intern Med 1990; 150:311–313 removal of large-molecular-weight solutes or com- 5 Gabow PA. Disorders associated with an altered anion gap. plexes, such as combined digoxin-Fab fragment com- Int 1985; 27:472–483 plexes, or desferoxamine complexes with iron or with 6 Ammar KA, Heckerling PS. Ethylene glycol poisoning with a aluminum. normal anion gap caused by concurrent ethanol ingestion: importance of the osmolal gap. Am J Kidney Dis 1996; 27:130–133 7 Walker JA, Schwartzbard A, Krauss EA, et al. The missing gap: a pitfall in the diagnosis of by Antidotes osmometry. Arch Intern Med 1986; 146:1843–1844 8 Sweeney TE, Beuchat CA. Limitations of methods of osmom- An antidote is a substance that increases the mean etry: measuring the osmolality of biological fluids. Am J lethal dose of a toxin, or that can favorably affect the Physiol 1993; 264:R469–R480 toxic effects of a poison. Some are toxic themselves 9 Glasser L, Sternglanz PD, Combie J, et al. Serum osmolality and therefore should be used only when indicated. and its applicability to drug overdose. Am J Clin Pathol 1973; 60:695–699 Table 18 lists antidotes for specific drugs/poisons. 10 Glaser DS. Utility of the serum osmolal gap in the diagnosis These will be discussed further next month in part II of methanol or ethylene glycol ingestion. Ann Emerg Med of this article. 1996; 27:343–346

590 Critical Care Review 11 Trummel J, Ford M, Autin P. Ingestion of an unknown 37 Barnett R, Grace M, Boothe P, et al. Flumazenil in drug alcohol. Ann Emerg Med 1996; 27:368–374 overdose: randomized, placebo-controlled study to assess cost 12 Hoffman RS, Smilkstein MJ, Howland MA, et al. Osmolal effectiveness. Crit Care Med 1999; 27:78–81 gaps revisited: normal values and limitations. J Toxicol Clin 38 Paloucek F. Antidotal flumazenil use: the protamine of the Toxicol 1993; 31:81–93 90s. Crit Care Med 1999; 27:10–11 13 Tremper KK, Barker SJ. Pulse oximetry. Anesthesiology 39 Bond GR, Requa RK, Krenzelok EP, et al. Influence of time 1989; 70:98–108 until emesis on the efficacy of decontamination using acet- 14 Barker SJ, Tremper KK, Hyatt J. Effects of methemoglobin- aminophen as a marker in a pediatric population. Ann Emerg emia on pulse oximetry and mixed venous oximetry. Anesthe- Med 1993; 22:1403–1407 siology 1989; 70:112–117 40 Merigan KS, Woodard M, Hedges JR, et al. Prospective 15 LeGrand TS, Peters JI. Pulse oximetry: advantages and evaluation of gastric emptying in the self-poisoned patient. pitfalls. J Crit Illn 1999; 20:195–206 Am J Emerg Med 1990; 8:479–483 16 Hampson NB. Pulse oximetry in severe carbon monoxide 41 Krenzelok EP, McGuigan M, Lheur P. Position statement: poisoning. Chest 1998; 114:1036–1041 ipecac syrup. American Academy of Clinical Toxicology, 17 Bozeman WP, Myers RA, Barish RA. Confirmation of the European Association of Poisons Centres and Clinical Toxi- pulse oximetry gap in carbon monoxide poisoning. Ann cologists. J Toxicol Clin Toxicol 1997; 35:699–709 Emerg Med 1997; 30:608–611 42 Shannon M. Ingestion of toxic substances by children. N Engl 18 Brett AS. Implications of discordance between clinical im- J Med 2000; 342:186–191 pression and toxicology analysis in drug overdose. Arch Intern 43 Vale JA, Meredith TJ, Proudfoot AT. Syrup of ipecacuanha: is Med 1988; 148:437–441 it really useful? BMJ 1986; 293:1321–1322 19 Kellerman AL, Fihn SD, LoGerfo JP, et al. Impact of drug 44 Rudolph JP. Automated gastric lavage and a comparison of screening in suspected overdose. Ann Emerg Med 1987; 0.9% normal saline solution and tap water irrigant. Ann 16:1209–1216 Emerg Med 1985; 14:1156–1159 20 Mahoney JD, Gross PL, Stern TA, et al. Quantitative serum 45 Thompson AM, Robins JB, Prescott LF, et al. Changes in toxic screening in the management of suspected drug over- cardiorespiratory function during gastric lavage for drug dose. Am J Emerg Med 1990; 8:16–22 overdose. Hum Toxicol 1987; 6:215–218 21 Wigder WN, Erickson T, Morse T, et al. Emergency depart- 46 Zimmerman JL. Management issues in toxicology. Semin ment poison advice telephone calls. Ann Emerg Med 1995; Respir Crit Care Med 2001; 22:23–28 25:349–352 47 Kulig K, Bar-Or D, Cantrill SV, et al. Management of acutely 22 Marik PE. Aspiration pneumonitis and aspiration pneumonia. poisoned patients without gastric emptying. Ann Emerg Med N Engl J Med 2001; 344:665–671 1985; 14:562–567 23 Roy TM, Ossorio MA, Cipolla LM, et al. Pulmonary compli- 48 Comstock EG, Faulkner TP, Boisaubin EV, et al. Studies on cations after tricyclic antidepressant overdose. Chest 1989; the efficacy of gastric lavage as practiced in a large metropol- 96:852–856 itan hospital. Clin Toxicol 1981; 18:581–597 24 Aldrich T, Morrsion J, Cesario T. Aspiration after overdosage 49 Pond SM, Lewis-Driver DJ, Williams GM, et al. Gastric of sedative or hypnotic drugs. South Med J 1980; 73:456–458 emptying in acute overdose: a prospective randomised con- 25 Hoffman RS, Goldfrank LR. The poisoned patient with trolled trial. Med J Aust 1995; 163:345–349 altered consciousness: controversies in the use of a “coma 50 Vale JA. Position statement: gastric lavage. American Acad- cocktail.” JAMA 1995; 274:562–569 emy of Clinical Toxicology, European Association of Poison 26 Reuler JB, Girard DE, Cooney TG. Wernicke’s encephalop- Centres and Clinical Toxicologists. J Toxicol Clin Toxicol athy. N Engl J Med 1985; 312:1035–1039 1997; 35:711–719 27 Goldfrank LR. The several uses of naloxone. Emerg Med 51 Saetta JP, March S, Gaunt ME, et al. Gastric emptying proce- 1984; 16:110–116 dures in the self-poisoned patient: are we forcing gastric content 28 Goldfrank LR, Weisman RS, Errick JK, et al. A dosing beyond the pylorus? J R Soc Med 1991; 84:274–276 nomogram for continuous infusion intravenous naloxone. Ann 52 Levy G. Gastrointestinal clearance of drugs with activated Emerg Med 1986; 15:566–570 charcoal. N Engl J Med 1982; 307:676–678 29 Ward CF. Pulmonary edema and naloxone [letter]. J Clin 53 Varga DW, Roy TM, Ossorio MA. Gastric stapling: a signifi- Anesth 1996; 8:690 cant risk factor in the treatment of tricyclic antidepressant 30 Schwartz JA, Koenigsberg MD. Naloxone-induced pulmo- overdose. J Ky Med Assoc 1989; 87:501–503 nary edema. Ann Emerg Med 1987; 16:1294–1296 54 Givens T, Holloway M, Wason S. Pulmonary aspiration of 31 Chiang WK, Goldfrank LR. Substance withdrawal. Emerg activated charcoal: a of its misuse in overdose Med Clin North Am 1990; 8:616–632 management. Pediatr Emerg Care 1992; 8:137–140 32 Mariani PJ. Seizure associated with low-dose naloxone. Am J 55 George DL, McLeod R, Weinstein RA. Contaminated com- Emerg Med 1989; 7:127–129 mercial charcoal as a source of fungi in the respiratory tract. 33 Doyon S, Roberts JR. Reappraisal of the coma cocktail: Infect Control Hosp Epidemiol 1991; 12:732–734 dextrose, flumazenil, naloxone, and thiamine. Emerg Med 56 Elliott CG, Colby TV, Kelly TM. Charcoal lung: bronchiolitis Clin North Am 1994; 12:301–316 obliterans after aspiration of activated charcoal. Chest 1989; 34 Longmire AW, Seger DL. Topics in clinical pharmacology: 96:672–674 flumazenil, a benzodiazepine antagonist. Am J Med Sci 1993; 57 Harris CR, Filandrinos D. Accidental administration of acti- 306:49–52 vated charcoal into the lung: aspiration by proxy. Ann Emerg 35 Mordel A, Winkler E, Almog S, et al. Seizures after fluma- Med 1993; 22:1470–1473 zenil administration in a case of combined benzodiazepine 58 Menzies DG, Busuttil A, Prescott LF. Fatal pulmonary and tricyclic antidepressant overdose. Crit Care Med 1992; aspiration of oral activated charcoal. BMJ 1988; 297:459–460 20:1733–1734 59 Chyka PA, Seger D. Position statement: single-dose activated 36 Weinbroum A, Rudick V, Sorkine P, et al. Use of flumazenil in charcoal. American Academy of Clinical Toxicology, Euro- the treatment of drug overdose: a double-blind and open clinical pean Association of Poisons Centres and Clinical Toxicolo- study in 110 patients. Crit Care Med 1996; 24:199–206 gists. J Toxicol Clin Toxicol 1997; 35:721–741 www.chestjournal.org CHEST / 123/2/FEBRUARY, 2003 591 60 Krenzelok EP, Keller R, Stewart RD. Gastrointestinal transit ican Academy of Clinical Toxicology, European Association of times of cathartics combined with charcoal. Ann Emerg Med Poisons Centres and Clinical Toxicologists. J Toxicol Clin 1985; 14:1152–1155 Toxicol 1999; 37:731–751 61 Keller RE, Schwab RA, Krenzelok EP. Contribution of 72 Cutler RE, Forland SC, St. John Hammond PG, et al. Extra- sorbitol combined with activated charcoal in prevention of corporeal removal of drugs and poisons by hemodialysis and salicylate absorption. Ann Emerg Med 1990; 19:654–656 hemoperfusion. Ann Rev Pharmacol Toxicol 1987; 27:169–191 62 McNamara RM, Aaron CK, Gembroys M, et al. Sorbitol 73 Pond SM. Extracorporeal techniques in the treatment of catharsis does not enhance efficacy of charcoal in a simulated poisoned patients. Med J Aust 1991; 154:617–622 acetaminophen overdose. Ann Emerg Med 1988; 17:243–246 74 Garella S. Extracorporeal techniques in the treatment of 63 Barceloux D, McGuigan M, Hartigan-Go K. Position state- exogenous intoxications. Kidney Int 1988; 33:735–754 ment: cathartics. American Academy of Clinical Toxicology, 75 Rommes JH. Haemoperfusion, indications and side-effects. European Association of Poisons Centres and Clinical Toxi- Arch Toxicol 1992; 15:S40–S49 cologists. J Toxicol Clin Toxicol 1997; 35:743–752 64 Davis GR, Santa Ana CA, Morawski SG, et al. Development 76 Golper TA, Wedel SK, Kaplan AA, et al. Drug removal by of a lavage solution associated with minimal water and continuous arteriovenous hemofiltration: theory and clinical electrolyte absorption or secretion. Gastroenterology 1980; observations. Int J Artif Organs 1985; 8:307–312 78:991–995 77 Golper TA. Drug removal during continuous renal replace- 65 Tenenbein M. Position statement: . ment therapies. Dial Transpl 1993; 22:185–212 American Academy of Clinical Toxicology, European Associ- 78 Bellomo R, Kearly Y, Parkin G, et al. Treatment of life- ation of Poison Centres and Clinical Toxicologists. J Toxicol threatening with continuous arterio-venous Clin Toxicol 1997; 35:753–762 hemodiafiltration. Crit Care Med 1991; 19:836–837 66 Pond SM. Principles of techniques applied to enhance elim- 79 Leblanc M, Raymond M, Bonnardeaux A, et al. Lithium ination of toxic compounds. In: Goldfrank LR, ed. Gold- poisoning treated by high-performance continuous arterio- frank’s toxicologic emergencies. Stamford, CT: Appelton and venous and venovenous hemodiafiltration. Am J Kidney Dis Lange, 1998; 53–62 1996; 27:365–372 67 Krenzelok EP, Leikin JB. Approach to a poisoned patient. Dis 80 Menghini VV, Albright RC. Treatment of lithium intoxication Mon 1996; 42:513–608 with continuous venovenous hemodiafiltration. Am J Kidney 68 Corbridge TC, Murray P. Toxicology in adults In. Hall J, Dis 2000; 3:E21 Schmidt GS, Wood LDH, eds. Principles of critical care. New 81 Brett AS, Rothschild N, Gray R, et al. Predicting the clinical York, NY: McGraw Hill, 1998; 1473–1525 course in intentional drug overdose: implications for the use of 69 Bradberry SM, Vale JA. Mutiple-dose activated charcoal: a the intensive care unit. Arch Intern Med 1987; 147:133–137 review of relevant clinical studies. J Toxicol Clin Toxicol 1995; 82 Hamad AE, Al-Ghadban A, Carvounis CP, et al. Predicting 33:407–416 the need for medical intensive care monitoring in drug- 70 Ilkhanipour K, Yealy DM, Krenzelok EP. The comparative overdosed patients. J Intensive Care Med 2000; 15:321–328 efficacy of various multiple-dose activated charcoal regimens. 83 Kulling P, Persson H. The role of the intensive care unit in Am J Emerg Med 1992; 10:298–300 the management of the poisoned patient. Med Toxicol 1986; 71 Vale JA, Krenzelok EP, Barceloux GD, et al. Position state- 1:375–386 ment and practice guidelines on the use of multi-dose 84 Callaham M. Admission criteria for tricyclic antidepressant activated charcoal in the treatment of acute poisoning. Amer- ingestion. West J Med 1982; 137:425–429

592 Critical Care Review