- Email: [email protected]
Assessment and Diagnosis of Poisoned Patients
ASSESSMENT AND MANAGEMENT
Assessment and Diagnosis of Poisoned Patients
Assessment and diagnosis of poisoned patients Assessment of level of consciousness, ventilation and circulation • What is the Glasgow Coma Score?
Allister Vale
• Are laryngeal reflexes present?
Sally Bradberry
• Is ventilatory insufficiency present?
Alex Proudfoot
• What are the pulse and blood pressure? History • Toxicological, medical, psychiatric and social
Assessment of an acutely poisoned patient follows the established clinical method (Figure 1): • immediate assessment of consciousness level, ventilation and circulation • history • examination • appropriate toxicological and non-toxicological investigations.
Circumstantial evidence • Suicide note • Circumstances in which patient was found Examination See Figure 2 Diagnostic trials of antidotes • Naloxone
Assessment of consciousness level, ventilation and circulation
• Flumazenil
Consciousness level, ventilation and circulation should be assessed in all patients.
Toxicological investigations • Specific
Consciousness level: the Glasgow Coma Scale (GCS) is the most commonly used method to assess the degree of impairment of consciousness, though it has never been validated in poisoned patients. A GCS score of less than 8 (not obeying commands, not speaking, not eye opening) should prompt careful respiratory assessment, particularly if the laryngeal (gag) reflex is lost.
• Screening
Ventilation: pulse oximetry can be used to measure oxygen saturation. The displayed reading may be inaccurate when the saturation is below 70%, when peripheral perfusion is poor, and in the presence of carboxyhaemoglobin and methaemoglobin. Only measurement of arterial blood gases indicates the presence both of hypercapnia and hypoxia. The presence of ventilatory insufficiency (as determined by arterial partial pressure of oxygen < 9 kPa on air and/or arterial partial pressure of carbon dioxide
• Radiology
Non-toxicological investigations • Haematology • Biochemistry • ECG
1
> 6 kPa) should lead to immediate intubation and mechanical ventilation, if the central respiratory depression cannot be reversed by administration of a specific antidote such as naloxone (see page 6). Circulation: pulse rate, blood pressure and temperature should be measured in all patients at presentation. Although hypotension (systolic blood pressure < 80 mm Hg) is a recognized feature of acute poisoning, the classical features of shock (tachycardia and pale, cold skin) are seen only rarely because a minority of patients are severely poisoned.
Allister Vale is Director of the National Poisons Information Service (Birmingham Centre) and the West Midlands Poisons Unit at City Hospital, Birmingham, UK.
History Adults: about 80% of adults who have ingested an overdose are conscious on arrival at hospital and the diagnosis of self-poisoning can usually be made from the history. In unconscious patients, a history from friends or relatives is helpful, and the diagnosis can often be inferred from tablet bottles or a ‘suicide note’ brought by the paramedics, or made by exclusion of other causes. Self-
Sally Bradberry is Assistant Director of the National Poisons Information Service (Birmingham Centre) at City Hospital, Birmingham, UK. Alex Proudfoot is Consulting Clinical Toxicologist at the National Poisons Information Service (Birmingham Centre) at City Hospital, Birmingham, UK.
MEDICINE
4
© 2003 The Medicine Publishing Company Ltd
ASSESSMENT AND MANAGEMENT
poisoning must always be considered in the differential diagnosis in any patient with an altered consciousness level. Acutely poisoned patients may be emotionally and psychiatrically distressed, and require competent, sympathetic assessment if essential information is not to be missed. It is pertinent to try to establish the nature of the substance taken, the amount involved, the route (ingestion, injection, inhalation or dermal) and the time of exposure, so that the clinical course can be anticipated and the risk assessed. Statements about the nature and amount of what has been taken should be regarded with clinical suspicion, however, because these are often inconsistent with laboratory analysis of blood or urine. Patients may not use generic drug names, and it is important to clarify the specific preparation involved because the composition of formulations with similar names can differ. Furthermore, selfpoisoning is often an impulsive act involving swallowing the contents of the first bottle or blister pack that comes to hand; sometimes, the drugs used may have been prescribed for another individual. Few patients count the number of tablets taken; the amount is often estimated in unquantifiable terms such as ‘handfuls’ or ‘mouthfuls’, though the patient may be able to recall the number of strips or packets. When the time of exposure is important (e.g. paracetamol poisoning), the accuracy can be improved by relating events to activities of daily life (e.g. the time of a television programme). Assessment of the psychological aspects of self-poisoning is discussed on page 16.
Children: a clear history is unlikely to be obtained from the patient or older witnesses. Statements about amounts taken are usually unreliable because the quantities present in containers before such incidents are often unknown. Circumstantial evidence is important in establishing a diagnosis of acute poisoning when the patient is very young, has hearing difficulties, or is demented or unconscious. Circumstances in which patient was found – children may be found eating potential poisons, or with tablets or other materials around their mouth or on their clothing. Similar evidence may be found on unconscious adults, or there may be empty drug containers, tablets or capsules nearby. A lack of personal effects to indicate the identity of an unconscious adult should arouse suspicion of a drug overdose. Protestations of relatives that an individual would never take an overdose are usually incorrect. Suicide notes are reliable indicators of drug overdose in the absence of evidence of physical violence as a cause of coma.
Examination Physical signs are particularly important when trying to elucidate the cause of unexplained coma. A diagnosis of acute poisoning can never be made on the basis of a single physical sign, but there are typical clusters of signs that make a diagnosis of poisoning with specific drugs very likely (Figure 2). Head injury should be excluded as a contributing or causative factor in comatose patients.
Common clusters of features that may be diagnostic Feature cluster • Coma, hypertonia, hyper-reflexia, extensor plantar responses, myoclonus, strabismus, mydriasis, sinus tachycardia
Likely poisons Tricyclic antidepressants Less commonly, orphenadrine, thioridazine, antihistamines
• Coma, hypotonia, hyporeflexia, plantar responses flexor or non-elicitable, hypotension
Barbiturates, benzodiazepines and alcohol combinations Severe tricyclic antidepressant poisoning
• Coma, miosis, reduced respiratory rate
Opioid analgesics
• Nausea, vomiting, tinnitus, deafness, sweating, hyperventilation, vasodilatation, metabolic acidosis
Salicylates
• Restlessness, agitation, mydriasis, anxiety, tremor, tachycardia, convulsions, arrhythmias
Sympathomimetics
• Hyperthermia, tachycardia, delirium, agitation, mydriasis
Ecstasy
• Blindness (usually with other features)
Quinine, methanol
• Miosis and hypersalivation
Organophosphorus and carbamate insecticides, nerve agents
2
MEDICINE
5
© 2003 The Medicine Publishing Company Ltd
ASSESSMENT AND MANAGEMENT
General observations may reveal useful information. For example, solvents or alcohol may be smelled on the breath, track marks may reveal undisclosed illicit substance abuse, atypical bruising may warn of domestic or other violence, and the stigmata of alcoholic liver disease may be revealed.
In acute poisoning, however, such a conclusion is not justified. Overdose with carbamazepine, phenytoin and tricyclic antidepressants can be associated with loss of these reflexes, but patients recover completely. Skin blisters may be found in poisoned patients who are, or have been, unconscious. Such lesions are not diagnostic of specific poisons, but are sufficiently common in poisoned patients (and sufficiently uncommon in patients unconscious from other causes) to be of diagnostic value.
Lateralizing neurological signs: with the exception of transient inequality of pupil size, lateralizing neurological signs effectively exclude a diagnosis of acute poisoning unless they can be explained by a pre-existing illness. Pyramidal tract signs: the usual features of pyramidal tract involvement (hypertonia, hyper-reflexia and extensor plantar responses) are commonly found in tricyclic antidepressant poisoning, and with other drugs with marked anticholinergic actions (e.g. the older antihistamines). However, all of these signs may be abolished in deep coma.
Trials of antidotes Administration of intravenous naloxone and flumazenil has been used to make rapid clinical diagnoses of opioid and benzodiazepine poisoning, respectively; this approach depends on noticeable improvement in the patient’s clinical condition within 1–2 minutes. Naloxone and flumazenil have a low incidence of adverse reactions and achieve considerable clinical improvement in some patients; their use is therefore justified in those whose clinical presentation suggests inadequate ventilation caused by an opiate and/or a benzodiazepine. These antagonists compete with the agonist agent for receptor sites. In severe poisoning, a large number of agonist molecules are present and an adequate dose of antagonist must be given to reverse toxicity to a clinically detectable extent; this is typically at least 1.2 mg of naloxone and 0.5 mg of flumazenil. Caution should be exercised with flumazenil in patients who might have co-ingested a tricyclic antidepressant, because seizures could be precipitated.
Abnormal movements: decerebrate and decorticate movements of the limbs often occur in unconscious poisoned patients, but in most cases there is no irreversible brain damage and the patient recovers fully. Acute dystonic movements (including acute torticollis, orolingual dyskinesias and oculogyric crises) are also produced; these are usually caused by metoclopramide, or less commonly by haloperidol, droperidol, prochlorperazine or trifluoperazine. Choreo-athetosis has been reported as a rare presenting feature of poisoning with organophosphorus insecticides. Pupillary changes in poisoning: inequality of the pupils is not uncommon in poisoned patients. Widely dilated pupils that react poorly to light may be caused by poisons with anticholinergic actions (e.g. tricyclic antidepressants) or sympathomimetic effects, or which cause blindness (e.g. quinine, methanol). Miosis is usually caused by opioid analgesics or poisons with cholinergic or anticholinesterase actions (e.g. organophosphorus insecticides, nerve agents). The degree and speed of reaction of the pupils to light is of no clinical value.
Toxicological investigations Measurement of the concentration of a specific toxin in the blood or toxicological screening of blood or urine can be used to establish a diagnosis of poisoning.
Ocular movements: strabismus, internuclear ophthalmoplegia and total external ophthalmoplegia have been described in acute poisoning, with impairment of consciousness caused by various drugs that act on the brain. Transient and variable strabismus (usually with the optic axes divergent in the horizontal plane) has been attributed to phenytoin, carbamazepine and tricyclic antidepressants. Dysconjugate, roving eye movements may also be seen if both eyes are observed for a period of time, and occasionally there may be total external ophthalmoplegia, even in patients in whom consciousness is no more than minimally impaired. Dysconjugate eye movements may become apparent only when the oculovestibular reflexes are examined using caloric stimuli, and have been reported in poisoning with tricyclic antidepressants, phenothiazines, benzodiazepines, barbiturates and ethanol. Instillation of ice-cold water into an ear should make both eyes turn to the irrigated side; failure of one eye to deviate suggests internuclear ophthalmoplegia and disturbed function of the medial longitudinal fasciculus.
Poisons for which emergency measurement of plasma or serum concentration is essential • Carboxyhaemoglobin • Ethanol (when monitoring treatment in ethylene glycol and methanol poisoning) • Ethylene glycol • Iron • Lithium • Methanol • Paracetamol • Salicylate • Theophylline
Loss of oculocephalic and oculovestibular reflexes is usually regarded as evidence of severe brain stem damage and brain death.
MEDICINE
3
6
© 2003 The Medicine Publishing Company Ltd
ASSESSMENT AND MANAGEMENT
Known toxins: it is neither possible nor necessary to measure the concentration of every toxin in biological fluids. Emergency measurement of plasma concentration is essential for the toxins shown in Figure 3; the concentration of these toxins is vital to ensure appropriate clinical management. Other agents for which assays should be available on a 24-hour basis at specialist laboratories include carbamazepine, digoxin, paraquat and phenobarbital. The concentration of a known toxin is usually measured in plasma or serum rather than whole blood. (Carboxyhaemoglobin is an exception – the sample should be taken in a heparinized blood gas syringe, taking care to exclude air bubbles.) If lithium concentrations are required, serum (or clotted blood) rather than plasma (or anticoagulated blood) should be sent to the laboratory. Many of the commercially available tubes that are used to prevent the sample from clotting already contain lithium as part of the anticoagulant, and their use would lead to falsely high concentrations.
Non-toxicological investigations • Serum sodium (e.g. hyponatraemia in Ecstasy poisoning) • Serum potassium (e.g. hypokalaemia in theophylline poisoning, hyperkalaemia in digoxin poisoning, rhabdomyolysis, haemolysis) • Plasma creatinine (e.g. renal failure in ethylene glycol poisoning) • Blood sugar (e.g. hypoglycaemia in insulin poisoning, hypoglycaemia and hyperglycaemia in salicylate poisoning) • Serum calcium (e.g. hypocalcaemia in ethylene glycol poisoning) • Serum alanine aminotransferase/aspartate aminotransferase activities (e.g. increased in paracetamol poisoning) • Serum phosphate (e.g. hypophosphataemia in severe paracetamol-induced renal tubular damage) • Acid–base disturbances, including metabolic acidosis
Screening for unknown toxins: plasma paracetamol concentration should be measured in all unconscious poisoned patients – multiple drug overdoses are common and paracetamol is one of the drugs most often involved, particularly in Europe. It is important to diagnose an overdose at a stage when administration of antidotes might prevent liver damage and death. In contrast, routine requests for salicylate concentrations cannot be justified. Salicylate poisoning (see MEDICINE 31:10) seldom causes coma and it is unlikely that clinically significant concentrations will be present in patients without the typical signs of salicylism. Laboratory screening for poisons in an unconscious patient is usually requested when the cause of coma is unknown. Although identification of a toxin may reassure the clinician, this is not a good reason for the request. Before proceeding, the clinician should consider how the result of a screen will alter management. The pattern of drugs involved in poisoning in most developed countries is such that specific treatment (e.g. antidotes, techniques to enhance elimination of the poison) is unlikely to be available, and management will therefore be supportive. Screening is labourintensive, time-consuming and expensive, and in most cases cannot be justified on an emergency basis because it will not alter the management of the patient. Furthermore, drugs given therapeutically may confuse interpretation of the result of a qualitative drug screen. If a laboratory screen for poisons is considered essential, it is important to discuss it with the laboratory staff in advance, to inform them of substances that might cause analytical confusion and to determine the nature of the samples to be assayed. Urine (20 ml) is usually a more appropriate fluid for detecting unknown poisons. Once identified, their concentration in blood or plasma can be determined by specific methods, if necessary.
• RBC cholinesterase activity (e.g. organophosphorus insecticide and nerve agent poisoning) • Whole blood methaemoglobin concentration (e.g. in nitrite poisoning)
4
Inspection of blood • Chocolate-coloured blood usually indicates methaemoglobinaemia caused by drugs such as dapsone, or ingestion of inorganic or organic nitrates or nitrites. • Pink plasma suggests haemolytic poisons (e.g. sodium chlorate). • Brown plasma suggests the presence of circulating myoglobin secondary to rhabdomyolysis. Inspection of urine • Brown discoloration of the urine may be caused by the presence of haemoglobin (if there is intravascular haemolysis), myoglobin secondary to rhabdomyolysis or metabolites of paracetamol. • Crystals may be prominent after an overdose with primidone or ingestion of ethylene glycol. Haematology is of little diagnostic value, though prolongation of prothrombin time may be secondary to ingestion of anticoagulants or may be caused by liver necrosis in paracetamol or hepatotoxic mushroom poisoning. Biochemistry: routine biochemistry and arterial blood gas analysis may provide useful and important information. Electrolyte, glucose, acid–base and osmolality disorders are discussed on page 8. In developed countries, undisclosed paracetamol overdose is a strong diagnostic possibility in any patient who presents with, or without, jaundice associated with plasma alanine aminotransferase activity of more than 5000 IU/litre.
Non-toxicological investigations Information about the nature of poisons ingested can occasionally be obtained from standard haematological and biochemical investigations, and from arterial blood gas analysis (Figure 4).
MEDICINE
7
© 2003 The Medicine Publishing Company Ltd
ASSESSMENT AND MANAGEMENT
Radiology: routine radiology is of little diagnostic value. It can be used to confirm ingestion of metallic objects (e.g. coins, button batteries) or injection of globules of metallic mercury. Rarely, hydrocarbon solvents (e.g. carbon tetrachloride) may be seen as a slightly opaque layer floating on the top of the gastric contents with the patient upright, or outlining the small bowel. Some enteric-coated or sustained-release drug formulations may be seen on plain abdominal radiographs, but, with the exception of iron salts, ordinary formulations are seldom seen. Radiography may have a limited role in confirming iron overdose in children, but more widespread use in acute poisoning is restricted by the fact that most adult patients are women of child-bearing age. Ingested packets of illicit substances may be discernible on a plain radiograph, but CT or MRI is more reliably able to detect such objects.
Metabolic Effects of Poisoning Alan Jones
Biochemical and metabolic abnormalities are common in critically ill patients, and in poisoning may occur as a: • primary effect of the poison on a particular metabolic pathway • secondary effect of the dysfunction of one or more organs which have been damaged by the poison. Management of biochemical abnormalities is an essential part of the care of severely poisoned patients, and particular biochemical features may be of diagnostic or prognostic value.
ECG: routine ECG is of limited diagnostic value. Sinus tachycardia with prolongation of the PR and QRS intervals in an unconscious patient should prompt consideration of tricyclic antidepressant overdose. With increasing cardiotoxicity, it may be impossible to detect P waves, and the pattern then resembles ventricular tachycardia. Overdose with cardiac glycosides or potassium salts also induces characteristic ECG changes. Q–T interval prolongation is a recognized adverse effect of several drugs in overdose and predisposes to ventricular arrhythmias, notably torsade de pointes.
Acid–base disturbances Acid–base disturbances are common in critically ill patients and may progress rapidly. Major abnormalities are likely to impair cell function and must therefore be recognized and treated effectively. Ready access to a blood gas analyser and an understanding of the results are essential.
Clinical features and complications of poisoning The toxicity of a substance, and therefore the features of poisoning, can generally be predicted from: • its physicochemical properties • its pharmacological/toxicological actions • the route of exposure • the dose. These features are classified as either direct or systemic. Those caused by the poisons that are commonly encountered in clinical practice are discussed from page 30 onwards and in MEDICINE 31:10.
Metabolic acidosis Acidaemia (pH < 7.35 or H+ > 45 nmol/litre) in a patient with low standard bicarbonate (< 21 mmol/litre) or a base deficit (< –2.3 mmol/litre) indicates metabolic acidosis. pCO2 is also usually low (< 4.6 kPa, 35 mm Hg) as a result of compensatory hyperventilation (Kussmaul’s breathing). Metabolic acidosis arises through accumulation of non-volatile acids or, less commonly, loss of bicarbonate. These two causes may be differentiated by measurement of the plasma anion gap. The total concentration of plasma cations must be equal to the concentration of anions, but conventional laboratory profiles do not measure all of the anions. Hence, there is an apparent anion gap that represents these unmeasured plasma anions (Figure 1). • The anion gap is increased when metabolic acidosis is caused by accumulation of acids (the conjugate anions of which contribute to the gap). • The anion gap is normal when metabolic acidosis is caused by loss of base (bicarbonate anion).
Direct (local) toxicity is confined to the site of contact of the substance with body surfaces. The route of exposure (eye, skin, respiratory or gastrointestinal tract) determines the anatomical location of the interaction; the physicochemical characteristics of the substance (solubility, volatility, pH) define its nature and extent. Systemic toxicity depends on the fraction of the dose of the poison that is absorbed into the circulation, and may be organ-specific or may involve several organs. Systemic toxicity is generally doserelated. The pharmacological/toxicological effects of the poison are proportional to the amount that has been absorbed, but are modulated by variations between individuals. Some individuals react in a non-dose-dependent, idiosyncratic manner to some agents (notably metoclopramide, and some phenothiazines and butyrophenones). The speed with which features appear depends partly on the route of exposure. It is greater with inhalation and injection than with dermal exposure and ingestion.
MEDICINE
Alan Jones is Consultant Chemical Pathologist and Physician at Birmingham Heartlands Hospital, Birmingham, UK, and Consulting Clinical Toxicologist at the National Poisons Information Service (Birmingham Centre). He qualified from the University of Cambridge, and trained in Leicester and Birmingham. His research interest is metabolic medicine.
8
© 2003 The Medicine Publishing Company Ltd
Assessment and Diagnosis of Poisoned Patients
Assessment and diagnosis of poisoned patients Assessment of level of consciousness, ventilation and circulation • What is the Glasgow Coma Score?
Allister Vale
• Are laryngeal reflexes present?
Sally Bradberry
• Is ventilatory insufficiency present?
Alex Proudfoot
• What are the pulse and blood pressure? History • Toxicological, medical, psychiatric and social
Assessment of an acutely poisoned patient follows the established clinical method (Figure 1): • immediate assessment of consciousness level, ventilation and circulation • history • examination • appropriate toxicological and non-toxicological investigations.
Circumstantial evidence • Suicide note • Circumstances in which patient was found Examination See Figure 2 Diagnostic trials of antidotes • Naloxone
Assessment of consciousness level, ventilation and circulation
• Flumazenil
Consciousness level, ventilation and circulation should be assessed in all patients.
Toxicological investigations • Specific
Consciousness level: the Glasgow Coma Scale (GCS) is the most commonly used method to assess the degree of impairment of consciousness, though it has never been validated in poisoned patients. A GCS score of less than 8 (not obeying commands, not speaking, not eye opening) should prompt careful respiratory assessment, particularly if the laryngeal (gag) reflex is lost.
• Screening
Ventilation: pulse oximetry can be used to measure oxygen saturation. The displayed reading may be inaccurate when the saturation is below 70%, when peripheral perfusion is poor, and in the presence of carboxyhaemoglobin and methaemoglobin. Only measurement of arterial blood gases indicates the presence both of hypercapnia and hypoxia. The presence of ventilatory insufficiency (as determined by arterial partial pressure of oxygen < 9 kPa on air and/or arterial partial pressure of carbon dioxide
• Radiology
Non-toxicological investigations • Haematology • Biochemistry • ECG
1
> 6 kPa) should lead to immediate intubation and mechanical ventilation, if the central respiratory depression cannot be reversed by administration of a specific antidote such as naloxone (see page 6). Circulation: pulse rate, blood pressure and temperature should be measured in all patients at presentation. Although hypotension (systolic blood pressure < 80 mm Hg) is a recognized feature of acute poisoning, the classical features of shock (tachycardia and pale, cold skin) are seen only rarely because a minority of patients are severely poisoned.
Allister Vale is Director of the National Poisons Information Service (Birmingham Centre) and the West Midlands Poisons Unit at City Hospital, Birmingham, UK.
History Adults: about 80% of adults who have ingested an overdose are conscious on arrival at hospital and the diagnosis of self-poisoning can usually be made from the history. In unconscious patients, a history from friends or relatives is helpful, and the diagnosis can often be inferred from tablet bottles or a ‘suicide note’ brought by the paramedics, or made by exclusion of other causes. Self-
Sally Bradberry is Assistant Director of the National Poisons Information Service (Birmingham Centre) at City Hospital, Birmingham, UK. Alex Proudfoot is Consulting Clinical Toxicologist at the National Poisons Information Service (Birmingham Centre) at City Hospital, Birmingham, UK.
MEDICINE
4
© 2003 The Medicine Publishing Company Ltd
ASSESSMENT AND MANAGEMENT
poisoning must always be considered in the differential diagnosis in any patient with an altered consciousness level. Acutely poisoned patients may be emotionally and psychiatrically distressed, and require competent, sympathetic assessment if essential information is not to be missed. It is pertinent to try to establish the nature of the substance taken, the amount involved, the route (ingestion, injection, inhalation or dermal) and the time of exposure, so that the clinical course can be anticipated and the risk assessed. Statements about the nature and amount of what has been taken should be regarded with clinical suspicion, however, because these are often inconsistent with laboratory analysis of blood or urine. Patients may not use generic drug names, and it is important to clarify the specific preparation involved because the composition of formulations with similar names can differ. Furthermore, selfpoisoning is often an impulsive act involving swallowing the contents of the first bottle or blister pack that comes to hand; sometimes, the drugs used may have been prescribed for another individual. Few patients count the number of tablets taken; the amount is often estimated in unquantifiable terms such as ‘handfuls’ or ‘mouthfuls’, though the patient may be able to recall the number of strips or packets. When the time of exposure is important (e.g. paracetamol poisoning), the accuracy can be improved by relating events to activities of daily life (e.g. the time of a television programme). Assessment of the psychological aspects of self-poisoning is discussed on page 16.
Children: a clear history is unlikely to be obtained from the patient or older witnesses. Statements about amounts taken are usually unreliable because the quantities present in containers before such incidents are often unknown. Circumstantial evidence is important in establishing a diagnosis of acute poisoning when the patient is very young, has hearing difficulties, or is demented or unconscious. Circumstances in which patient was found – children may be found eating potential poisons, or with tablets or other materials around their mouth or on their clothing. Similar evidence may be found on unconscious adults, or there may be empty drug containers, tablets or capsules nearby. A lack of personal effects to indicate the identity of an unconscious adult should arouse suspicion of a drug overdose. Protestations of relatives that an individual would never take an overdose are usually incorrect. Suicide notes are reliable indicators of drug overdose in the absence of evidence of physical violence as a cause of coma.
Examination Physical signs are particularly important when trying to elucidate the cause of unexplained coma. A diagnosis of acute poisoning can never be made on the basis of a single physical sign, but there are typical clusters of signs that make a diagnosis of poisoning with specific drugs very likely (Figure 2). Head injury should be excluded as a contributing or causative factor in comatose patients.
Common clusters of features that may be diagnostic Feature cluster • Coma, hypertonia, hyper-reflexia, extensor plantar responses, myoclonus, strabismus, mydriasis, sinus tachycardia
Likely poisons Tricyclic antidepressants Less commonly, orphenadrine, thioridazine, antihistamines
• Coma, hypotonia, hyporeflexia, plantar responses flexor or non-elicitable, hypotension
Barbiturates, benzodiazepines and alcohol combinations Severe tricyclic antidepressant poisoning
• Coma, miosis, reduced respiratory rate
Opioid analgesics
• Nausea, vomiting, tinnitus, deafness, sweating, hyperventilation, vasodilatation, metabolic acidosis
Salicylates
• Restlessness, agitation, mydriasis, anxiety, tremor, tachycardia, convulsions, arrhythmias
Sympathomimetics
• Hyperthermia, tachycardia, delirium, agitation, mydriasis
Ecstasy
• Blindness (usually with other features)
Quinine, methanol
• Miosis and hypersalivation
Organophosphorus and carbamate insecticides, nerve agents
2
MEDICINE
5
© 2003 The Medicine Publishing Company Ltd
ASSESSMENT AND MANAGEMENT
General observations may reveal useful information. For example, solvents or alcohol may be smelled on the breath, track marks may reveal undisclosed illicit substance abuse, atypical bruising may warn of domestic or other violence, and the stigmata of alcoholic liver disease may be revealed.
In acute poisoning, however, such a conclusion is not justified. Overdose with carbamazepine, phenytoin and tricyclic antidepressants can be associated with loss of these reflexes, but patients recover completely. Skin blisters may be found in poisoned patients who are, or have been, unconscious. Such lesions are not diagnostic of specific poisons, but are sufficiently common in poisoned patients (and sufficiently uncommon in patients unconscious from other causes) to be of diagnostic value.
Lateralizing neurological signs: with the exception of transient inequality of pupil size, lateralizing neurological signs effectively exclude a diagnosis of acute poisoning unless they can be explained by a pre-existing illness. Pyramidal tract signs: the usual features of pyramidal tract involvement (hypertonia, hyper-reflexia and extensor plantar responses) are commonly found in tricyclic antidepressant poisoning, and with other drugs with marked anticholinergic actions (e.g. the older antihistamines). However, all of these signs may be abolished in deep coma.
Trials of antidotes Administration of intravenous naloxone and flumazenil has been used to make rapid clinical diagnoses of opioid and benzodiazepine poisoning, respectively; this approach depends on noticeable improvement in the patient’s clinical condition within 1–2 minutes. Naloxone and flumazenil have a low incidence of adverse reactions and achieve considerable clinical improvement in some patients; their use is therefore justified in those whose clinical presentation suggests inadequate ventilation caused by an opiate and/or a benzodiazepine. These antagonists compete with the agonist agent for receptor sites. In severe poisoning, a large number of agonist molecules are present and an adequate dose of antagonist must be given to reverse toxicity to a clinically detectable extent; this is typically at least 1.2 mg of naloxone and 0.5 mg of flumazenil. Caution should be exercised with flumazenil in patients who might have co-ingested a tricyclic antidepressant, because seizures could be precipitated.
Abnormal movements: decerebrate and decorticate movements of the limbs often occur in unconscious poisoned patients, but in most cases there is no irreversible brain damage and the patient recovers fully. Acute dystonic movements (including acute torticollis, orolingual dyskinesias and oculogyric crises) are also produced; these are usually caused by metoclopramide, or less commonly by haloperidol, droperidol, prochlorperazine or trifluoperazine. Choreo-athetosis has been reported as a rare presenting feature of poisoning with organophosphorus insecticides. Pupillary changes in poisoning: inequality of the pupils is not uncommon in poisoned patients. Widely dilated pupils that react poorly to light may be caused by poisons with anticholinergic actions (e.g. tricyclic antidepressants) or sympathomimetic effects, or which cause blindness (e.g. quinine, methanol). Miosis is usually caused by opioid analgesics or poisons with cholinergic or anticholinesterase actions (e.g. organophosphorus insecticides, nerve agents). The degree and speed of reaction of the pupils to light is of no clinical value.
Toxicological investigations Measurement of the concentration of a specific toxin in the blood or toxicological screening of blood or urine can be used to establish a diagnosis of poisoning.
Ocular movements: strabismus, internuclear ophthalmoplegia and total external ophthalmoplegia have been described in acute poisoning, with impairment of consciousness caused by various drugs that act on the brain. Transient and variable strabismus (usually with the optic axes divergent in the horizontal plane) has been attributed to phenytoin, carbamazepine and tricyclic antidepressants. Dysconjugate, roving eye movements may also be seen if both eyes are observed for a period of time, and occasionally there may be total external ophthalmoplegia, even in patients in whom consciousness is no more than minimally impaired. Dysconjugate eye movements may become apparent only when the oculovestibular reflexes are examined using caloric stimuli, and have been reported in poisoning with tricyclic antidepressants, phenothiazines, benzodiazepines, barbiturates and ethanol. Instillation of ice-cold water into an ear should make both eyes turn to the irrigated side; failure of one eye to deviate suggests internuclear ophthalmoplegia and disturbed function of the medial longitudinal fasciculus.
Poisons for which emergency measurement of plasma or serum concentration is essential • Carboxyhaemoglobin • Ethanol (when monitoring treatment in ethylene glycol and methanol poisoning) • Ethylene glycol • Iron • Lithium • Methanol • Paracetamol • Salicylate • Theophylline
Loss of oculocephalic and oculovestibular reflexes is usually regarded as evidence of severe brain stem damage and brain death.
MEDICINE
3
6
© 2003 The Medicine Publishing Company Ltd
ASSESSMENT AND MANAGEMENT
Known toxins: it is neither possible nor necessary to measure the concentration of every toxin in biological fluids. Emergency measurement of plasma concentration is essential for the toxins shown in Figure 3; the concentration of these toxins is vital to ensure appropriate clinical management. Other agents for which assays should be available on a 24-hour basis at specialist laboratories include carbamazepine, digoxin, paraquat and phenobarbital. The concentration of a known toxin is usually measured in plasma or serum rather than whole blood. (Carboxyhaemoglobin is an exception – the sample should be taken in a heparinized blood gas syringe, taking care to exclude air bubbles.) If lithium concentrations are required, serum (or clotted blood) rather than plasma (or anticoagulated blood) should be sent to the laboratory. Many of the commercially available tubes that are used to prevent the sample from clotting already contain lithium as part of the anticoagulant, and their use would lead to falsely high concentrations.
Non-toxicological investigations • Serum sodium (e.g. hyponatraemia in Ecstasy poisoning) • Serum potassium (e.g. hypokalaemia in theophylline poisoning, hyperkalaemia in digoxin poisoning, rhabdomyolysis, haemolysis) • Plasma creatinine (e.g. renal failure in ethylene glycol poisoning) • Blood sugar (e.g. hypoglycaemia in insulin poisoning, hypoglycaemia and hyperglycaemia in salicylate poisoning) • Serum calcium (e.g. hypocalcaemia in ethylene glycol poisoning) • Serum alanine aminotransferase/aspartate aminotransferase activities (e.g. increased in paracetamol poisoning) • Serum phosphate (e.g. hypophosphataemia in severe paracetamol-induced renal tubular damage) • Acid–base disturbances, including metabolic acidosis
Screening for unknown toxins: plasma paracetamol concentration should be measured in all unconscious poisoned patients – multiple drug overdoses are common and paracetamol is one of the drugs most often involved, particularly in Europe. It is important to diagnose an overdose at a stage when administration of antidotes might prevent liver damage and death. In contrast, routine requests for salicylate concentrations cannot be justified. Salicylate poisoning (see MEDICINE 31:10) seldom causes coma and it is unlikely that clinically significant concentrations will be present in patients without the typical signs of salicylism. Laboratory screening for poisons in an unconscious patient is usually requested when the cause of coma is unknown. Although identification of a toxin may reassure the clinician, this is not a good reason for the request. Before proceeding, the clinician should consider how the result of a screen will alter management. The pattern of drugs involved in poisoning in most developed countries is such that specific treatment (e.g. antidotes, techniques to enhance elimination of the poison) is unlikely to be available, and management will therefore be supportive. Screening is labourintensive, time-consuming and expensive, and in most cases cannot be justified on an emergency basis because it will not alter the management of the patient. Furthermore, drugs given therapeutically may confuse interpretation of the result of a qualitative drug screen. If a laboratory screen for poisons is considered essential, it is important to discuss it with the laboratory staff in advance, to inform them of substances that might cause analytical confusion and to determine the nature of the samples to be assayed. Urine (20 ml) is usually a more appropriate fluid for detecting unknown poisons. Once identified, their concentration in blood or plasma can be determined by specific methods, if necessary.
• RBC cholinesterase activity (e.g. organophosphorus insecticide and nerve agent poisoning) • Whole blood methaemoglobin concentration (e.g. in nitrite poisoning)
4
Inspection of blood • Chocolate-coloured blood usually indicates methaemoglobinaemia caused by drugs such as dapsone, or ingestion of inorganic or organic nitrates or nitrites. • Pink plasma suggests haemolytic poisons (e.g. sodium chlorate). • Brown plasma suggests the presence of circulating myoglobin secondary to rhabdomyolysis. Inspection of urine • Brown discoloration of the urine may be caused by the presence of haemoglobin (if there is intravascular haemolysis), myoglobin secondary to rhabdomyolysis or metabolites of paracetamol. • Crystals may be prominent after an overdose with primidone or ingestion of ethylene glycol. Haematology is of little diagnostic value, though prolongation of prothrombin time may be secondary to ingestion of anticoagulants or may be caused by liver necrosis in paracetamol or hepatotoxic mushroom poisoning. Biochemistry: routine biochemistry and arterial blood gas analysis may provide useful and important information. Electrolyte, glucose, acid–base and osmolality disorders are discussed on page 8. In developed countries, undisclosed paracetamol overdose is a strong diagnostic possibility in any patient who presents with, or without, jaundice associated with plasma alanine aminotransferase activity of more than 5000 IU/litre.
Non-toxicological investigations Information about the nature of poisons ingested can occasionally be obtained from standard haematological and biochemical investigations, and from arterial blood gas analysis (Figure 4).
MEDICINE
7
© 2003 The Medicine Publishing Company Ltd
ASSESSMENT AND MANAGEMENT
Radiology: routine radiology is of little diagnostic value. It can be used to confirm ingestion of metallic objects (e.g. coins, button batteries) or injection of globules of metallic mercury. Rarely, hydrocarbon solvents (e.g. carbon tetrachloride) may be seen as a slightly opaque layer floating on the top of the gastric contents with the patient upright, or outlining the small bowel. Some enteric-coated or sustained-release drug formulations may be seen on plain abdominal radiographs, but, with the exception of iron salts, ordinary formulations are seldom seen. Radiography may have a limited role in confirming iron overdose in children, but more widespread use in acute poisoning is restricted by the fact that most adult patients are women of child-bearing age. Ingested packets of illicit substances may be discernible on a plain radiograph, but CT or MRI is more reliably able to detect such objects.
Metabolic Effects of Poisoning Alan Jones
Biochemical and metabolic abnormalities are common in critically ill patients, and in poisoning may occur as a: • primary effect of the poison on a particular metabolic pathway • secondary effect of the dysfunction of one or more organs which have been damaged by the poison. Management of biochemical abnormalities is an essential part of the care of severely poisoned patients, and particular biochemical features may be of diagnostic or prognostic value.
ECG: routine ECG is of limited diagnostic value. Sinus tachycardia with prolongation of the PR and QRS intervals in an unconscious patient should prompt consideration of tricyclic antidepressant overdose. With increasing cardiotoxicity, it may be impossible to detect P waves, and the pattern then resembles ventricular tachycardia. Overdose with cardiac glycosides or potassium salts also induces characteristic ECG changes. Q–T interval prolongation is a recognized adverse effect of several drugs in overdose and predisposes to ventricular arrhythmias, notably torsade de pointes.
Acid–base disturbances Acid–base disturbances are common in critically ill patients and may progress rapidly. Major abnormalities are likely to impair cell function and must therefore be recognized and treated effectively. Ready access to a blood gas analyser and an understanding of the results are essential.
Clinical features and complications of poisoning The toxicity of a substance, and therefore the features of poisoning, can generally be predicted from: • its physicochemical properties • its pharmacological/toxicological actions • the route of exposure • the dose. These features are classified as either direct or systemic. Those caused by the poisons that are commonly encountered in clinical practice are discussed from page 30 onwards and in MEDICINE 31:10.
Metabolic acidosis Acidaemia (pH < 7.35 or H+ > 45 nmol/litre) in a patient with low standard bicarbonate (< 21 mmol/litre) or a base deficit (< –2.3 mmol/litre) indicates metabolic acidosis. pCO2 is also usually low (< 4.6 kPa, 35 mm Hg) as a result of compensatory hyperventilation (Kussmaul’s breathing). Metabolic acidosis arises through accumulation of non-volatile acids or, less commonly, loss of bicarbonate. These two causes may be differentiated by measurement of the plasma anion gap. The total concentration of plasma cations must be equal to the concentration of anions, but conventional laboratory profiles do not measure all of the anions. Hence, there is an apparent anion gap that represents these unmeasured plasma anions (Figure 1). • The anion gap is increased when metabolic acidosis is caused by accumulation of acids (the conjugate anions of which contribute to the gap). • The anion gap is normal when metabolic acidosis is caused by loss of base (bicarbonate anion).
Direct (local) toxicity is confined to the site of contact of the substance with body surfaces. The route of exposure (eye, skin, respiratory or gastrointestinal tract) determines the anatomical location of the interaction; the physicochemical characteristics of the substance (solubility, volatility, pH) define its nature and extent. Systemic toxicity depends on the fraction of the dose of the poison that is absorbed into the circulation, and may be organ-specific or may involve several organs. Systemic toxicity is generally doserelated. The pharmacological/toxicological effects of the poison are proportional to the amount that has been absorbed, but are modulated by variations between individuals. Some individuals react in a non-dose-dependent, idiosyncratic manner to some agents (notably metoclopramide, and some phenothiazines and butyrophenones). The speed with which features appear depends partly on the route of exposure. It is greater with inhalation and injection than with dermal exposure and ingestion.
MEDICINE
Alan Jones is Consultant Chemical Pathologist and Physician at Birmingham Heartlands Hospital, Birmingham, UK, and Consulting Clinical Toxicologist at the National Poisons Information Service (Birmingham Centre). He qualified from the University of Cambridge, and trained in Leicester and Birmingham. His research interest is metabolic medicine.
8
© 2003 The Medicine Publishing Company Ltd