Oxford Handbook of Clinical and Laboratory Investigation

Many poisoned patients recover without specific management other than good supportive care. A minority have life-threatening toxicity. In assessing the poisoned patient, it is important to ensure adequate airway, breathing, and circulation, take a thorough history, and undertake a full clinical examination. Tablets, bottles, syringes, aerosol containers, and other items found with or near the patient should be retained and any corroborative history obtained. It is usually best to analyse biological specimens (usually blood and/or urine) if analytical confirmation of toxin exposures is required.

The role of blood and urine tests in toxicology

Close collaboration between analytical staff and clinicians is required if anything other than the simplest toxicological analysis is to be useful.

Toxicological analysis using blood or urine is used to confirm:

•The diagnosis of poisoning, when this is in doubt or for medicolegal purposes.

•To help in the management, or in the diagnosis of brain death.

•To work out the time to restart chronic drug therapy e.g. anticonvulsants.

Few centres have full analytical toxicology services, and a ‘toxicology screen’ rarely influences acute inpatient management, with the exception of paracetamol, salicylate, lithium, digoxin, and iron poisoning, and on occasions a drugs of abuse screen. Toxicological analysis of blood plasma or serum is also of value if an extracorporeal method of elimination, such as haemodialysis or MARS^® (Molecular Absorbance Recirculating Systems),¹ is being contemplated. Any toxicology analysis should be tailored to that patient’s circumstances and the poisons commonly encountered in that country. In Western Europe and North America, most patients will have taken pharmaceutical agents (often in combination), but pesticide poisoning, for example, is common in less well-developed countries.

Plasma paracetamol, salicylate, lithium, digoxin, and iron measurements in blood are usually available on an urgent basis. For other patients, particularly those who present a complex clinical picture or who are unconscious, a 50mL sample of urine and a 10mL sample of heparinized blood should be collected on admission and stored at 4°C (refrigerated). This can be analysed later if it is felt the result will influence your management or if needed for medicolegal purposes ( Samples of medicolegal importance, p. 702). Urine is useful for screening, especially for drugs of abuse, as it is often available in large volumes and often contains higher concentrations of poisons and their metabolites than blood samples. The samples should be obtained as soon as possible after admission, ideally before any therapeutic drugs are administered. Urine samples usually provide qualitative results, e.g. detect the presence of amphetamines or benzodiazepines. Quantitative measurements in urine are of little use because some compounds, such as benzodiazepines, are extensively metabolized prior to excretion in urine.

Sample requirements

Plasma or serum is normally used for quantitative assays for drugs and drug metabolites, and in general there are no marked differences in concentration between these fluids. Evacuated blood tubes and containers containing gel separators or soft rubber stoppers are not recommended if a toxicological analysis is to be performed, as the plasticizers (phosphates and phthalates) used in many such tubes may interfere with chromatographic methods, and volatile compounds, such as CO or ethanol, may be lost.

EDTA tubes are preferred for COHb assays and for measurement of lead and some other metals, as these are concentrated in RBCs. A fluoride/oxalate tube should be used if ethanol, cocaine, or benzodiazepines are being assayed, although special tubes containing 1% (w/v) fluoride are needed if enzymic hydrolysis of these and other compounds is to be completely prevented.

The use of disinfectant swabs containing alcohols should be avoided, as should heparin, which contains phenolic preservatives (chlorbutol, cresol), and preservatives containing mercury salts (see Table 11.1).

Table 11.1 Sample requirements for metals/trace elements analysis

Metal	Sample needed
Aluminium	10mL whole blood in plastic (not glass) tube—no anticoagulant/beads*
Antimony	5mL heparinized whole blood; 20mL urine
Arsenic	5mL heparinized whole blood; 20mL urine
Bismuth	5mL heparinized whole blood
Cadmium	2mL EDTA whole blood; 10mL urine
Chromium	2mL heparinized whole blood; 20mL urine
Copper	2mL heparinized or clotted whole blood, or 1mL plasma; 10mL urine
Iron	5mL clotted blood or 2mL serum (avoid haemolysis)
Lead	2mL EDTA whole blood
Lithium	5mL clotted blood or 2mL serum (NOT lithium heparin tube!)
Manganese	1mL heparinized whole blood or 0.5mL plasma*
Mercury	5mL heparinized whole blood; 20mL urine
Selenium	2mL heparinized whole blood or 1mL plasma/serum
Thallium	5mL heparinized whole blood; 20mL urine
Zinc	2mL whole blood (not EDTA) or 1mL plasma/serum

* Send unused container from the same batch to check for possible contamination.

Samples of medicolegal importance

A toxicology screen is helpful if murder, assault, or child abuse is suspected. Samples collected in such cases are often so important that they should be kept securely at −20°C or below, until investigation of the incident is concluded. Legal requirements mean that all specimens should be clearly labelled with the patient’s family or last name and any forenames, the date and time of collection, and the nature of the specimen, if this is not obvious. Strict chain of custody procedures should be implemented, and the doctor or nurse taking the sample should seal the bag with a tamper-proof device and sign and date the seal. A chain of custody form must accompany the sample and should be signed and dated by every person taking possession of the sample. The sample should be secured in a locked container or refrigerator if left unattended before arrival at the laboratory.

Post-mortem concentrations of some drugs may not necessarily represent antemortem levels and should therefore be interpreted with caution, e.g. in one case where a patient died ~2h after admission to hospital, the post-mortem serum olanzapine concentration was elevated 5-fold over the antemortem level.² With passage of time after death, some drugs (especially basic drugs with a high volume of distribution) move along a concentration gradient from areas of high concentration in solid organs into the local blood, a phenomenon called ‘post-mortem redistribution’. Therefore, the highest blood drug concentrations are found in central vessels such as the pulmonary artery or vein and the lowest levels in femoral veins e.g. morphine, amphetamines. Thus, the most appropriate post-mortem blood sample is a femoral blood sample.

The stability of the drug in post-mortem blood samples that are stored also needs to be considered. After a specimen has been collected, enzymes may remain active and continue to metabolize the drug in vitro. In general, drug instability occurs because functional groups are susceptible to transformation (e.g. 6-acetylmorphine) or readily oxidized or reduced. Olanzapine blood concentrations ↓ with time by 23–84% during storage.³ However, conjugated drugs, such as the glucuronides, may deconjugate, which ↑ the free drug concentration. In general, drug stability should be evaluated with consideration of long-term storage, the effect of freeze–thaw cycles, short-term stability (e.g. refrigerated samples), and bench-top (room temperature) stability. Thus, femoral blood sample drug concentrations should be measured as soon as possible after death and samples must be stored appropriately.

Further reading

Drummer OH. Requirements for bioanalytical procedures in post-mortem toxicology. Anal Bioanal Chem 2007; 388: 1495–503.

Kerrigan S. Sampling, storage and stability. In: A Negrusz, G Cooper (eds.). Clarke’s Analytical Forensic Toxicology, 2nd edn. London: Pharmaceutical Press, 2013; pp. 335–50.

Methods used in analytical toxicology

Older and newer methods are discussed here to reflect global use. A range of chromatographic and other methods, such as radioligand immunoassays, are available for toxicological analyses. Plasma concentrations associated with serious toxicity range from µg/L in the case of drugs such as digoxin to g/L in the case of ethanol.

Specialized laboratories use a combination of solvent extraction and thin-layer chromatography (TLC) together with gas chromatography-mass spectrometry (GC-MS) or liquid chromatography tandem-mass spectrometry (LC-MS/MS) as the basis for a ‘toxicology screen’. It is unwise to use TLC without corroboration of results by another method, e.g. LC-MS/MS, because the resolution power of TLC is limited and interpretation of chromatograms is subjective. A commercial kit for TLC (Toxi-lab, Marion Laboratories) is supplied with a compendium of colour plates, but even so problems can arise in the differentiation of compounds with similar mobility and colour reactions. The kit is aimed at the US market, and some common UK drugs are not included. Spectrophotometry is commonly used to measure salicylates, iron, and COHb. However, UV spectrophotometry and spectrophotofluorimetry are often used as detectors for high-performance liquid chromatography (HPLC) and in immunoassays. Spectrophotometric methods and immunoassays often suffer from interference from metabolites or other drugs. Immunoassays have the advantage of long shelf-life and simplicity, but all require confirmation with a chromatographic method if the results are to stand scrutiny. This is because immunoassays for small molecules are often not specific, e.g. some urine amphetamine immunoassays give +ve results with proguanil, isoxsuprine, labetalol, tranylcypromine, and phenylethylamine. The Syva Emit antidepressant assay cross-reacts with phenothiazines after overdose. Chromatographic methods have the advantage of selectivity and sensitivity and the ability to perform quantitative measurements, but they are more expensive. Generally, gas–liquid chromatography (GLC) or LC-MS/MS are used to measure basic drugs. Acidic or neutral moieties can also be analysed by LC-MS/MS. For screening very large numbers of compounds simultaneously or sequentially, then the Applied Biosystems Q-Trap technology is one of the most versatile—this is an LC-MS/MS where one of the quadruples of the triple quadrupole is used as an ion trap. Most laboratories that aim to be as versatile as possible have a fast GLC, in addition to the LC-MS/MS.

Modern methods of assay for heavy metals vary enormously. Inductively coupled plasma mass spectroscopy (ICPMS) is the most commonly used method in the UK. Here, ICPMS is used, instead of the flame or electrothermal furnace for atomization of the elements and mass is the detector. Atomic absorption spectrophotometry, with either flame or electrothermal atomization is the older method. In the case of iron, reliable kits based on the formation of a coloured complex are available.

There is wide variation in the units that various laboratories use to report results. This has caused confusion and errors in treatment, and great care is needed to ensure that clinical interpretation is undertaken in full knowledge of the units used by the reporting laboratory. Particular care is also required in interpretation and application of analytical techniques in post-mortem toxicology, due to changes in concentrations of blood in storage and differential post-mortem redistribution of drug after death ( Samples of medicolegal importance, p. 702).

Further reading

Baselt RC, Cravey RH.Disposition of Toxic Drugs and Chemicals in Man, 9th edn. San Francisco, Chemical Toxicology Institute, 2011.

Drummer OH. Requirements for bioanalytical procedures in post-mortem toxicology. Anal Bioanal Chem 2007; 388: 1495–503.

Brainstem death testing and organ donation

Brain death cannot be diagnosed in the presence of drugs that mask CNS activity, e.g. baclofen. The rule of thumb, based on the pharmacological principle that most drugs need five half-lives to be effectively eliminated from the circulation, is to allow four half-lives of any drug to elapse before declaring death, or to allow at least 2–3 days for drug effects to wear off. Whether this is satisfactory for patients with organ failure, and hence impaired drug elimination, is unclear. Often in patients being assessed for organ donation, measurement of plasma concentrations of residual drugs, with expert interpretation, is required to determine whether brainstem death tests are valid or whether a drug could be interfering with the results.

Selected donor organs from those who have died from poisoning by tricyclic antidepressants, benzodiazepines, barbiturates, insulin, CO, cocaine, methanol, and paracetamol have been used in transplantation. It is important to identify which organs act as reservoirs for drugs and either not consider such organs, e.g. a liver from a paracetamol-poisoned patient, or take prophylactic precautions like acetylcysteine administration in the case of donation of a heart from a paracetamol-poisoned patient.

Further reading

Baselt RC, Cravey RH.Disposition of Toxic Drugs and Chemicals in Man, 9th edn. San Francisco, Chemical Toxicology Institute, 2011.

Hanson P. Organ donation after fatal poisoning: an update with recent literature data. Adv Exp Med Biol 2004; 550: 207.

Wood DM, Chan WL, Dargan P. Using drug-intoxicated deaths as potential organ donors: impression of attendees at the American College of Medical Toxicology 2014 annual scientific meeting. J Med Toxicol 2014; 10: 360–3.

Interpretation of arterial blood gases in poisoned patients

Interpretation of blood gas values may be found in OHCM 10e, p. 162, pp. 188–9. Essentially four patterns emerge, which may occur together.

Respiratory acidosis

Hypoventilation results in retention of CO₂. This can occur after an overdose with any drugs that depress the CNS, e.g. tricyclic antidepressants, opioids, and ketamine.

Respiratory alkalosis

Hyperventilation with respiratory alkalosis is classically caused by aspirin (salicylate). It can also occur in response to hypoxia, drugs, and CNS injury.

Metabolic alkalosis

Metabolic alkalosis is very uncommon in poisoning. Rarely, it may result from excess administration of alkali, e.g. deliberate alkali ingestion.

Metabolic acidosis

This is the commonest metabolic abnormality in poisoning. If acidosis is particularly severe (e.g. pH <7.2), this should raise the question of poisoning by ethanol, methanol, or ethylene glycol. Measuring the anion gap and osmolar gaps are helpful in further differentiation of the medical or toxicological cause ( Ethylene glycol, ethanol, and methanol poisoning, pp. 720–722).

Amphetamines and derivatives

For example, methylene dioxymethamphetamine (MDMA) (ecstasy), MDEA, crystal methamphetamine (‘ICE’), paramethoxymethamphetamine (PMA).

Acute amphetamine overdose may cause sympathetic hyperstimulation, cardiovascular collapse, rhabdomyolysis, ventricular tachyarrhythmias, and death (often trauma-related). The following investigations should be considered in patients presenting to hospital with acute amphetamine(s) intoxication. In man, the half-life of methamphetamine ranges from 10 to 20h, depending on urine pH and the dose taken.

Plasma urea and electrolytes and glucose

It is critical that at least one set of U&E and creatinine are checked in every patient. Most are profoundly dehydrated and require vigorous rehydration. Some patients develop hyponatraemia, often after drinking excess water, and ADH secretion may also be responsible for this ( OHCM 10e, p. 672). Hypoglycaemia may occur.

Dipstick test of urine for myoglobin and subsequent serum creatine kinase

Hyperthermia can develop after amphetamine exposure and may cause rhabdomyolysis. Hyperthermia may be one of the features of serotonin toxicity (accompanied by autonomic instability, ocular and peripheral clonus. and hyper-reflexia). Dipstick testing of urine is +ve for blood in rhabdomyolysis, as myoglobin is detected by the Hb assay. This is an indication that serum CK should then be measured. If found to be elevated, adequate rehydration is needed to reduce deposition of myoglobin in renal tubules and reduce the risk of ARF as a consequence. Caution in use of alkalinization as for certain drugs, e.g. amphetamines; this potentially delays drug excretion.

Full blood count

Rarely, aplastic anaemia ( OHCM 10e, p. 364) has been reported after ecstasy (MDMA) ingestion.

Clotting studies

DIC ( OHCM 10e, p. 353) can occur, often in the context of hyperthermia. Once liver damage ensues, INR/PT ( OHCM 10e, p. 351) will rise.

Temperature

Hyperpyrexia can lead to rhabdomyolysis, DIC, and hepatocellular necrosis. Risks relate to the time in hours spent above 39°C. A rectal thermometer is the most accurate measure of temperature.

Blood pressure

A BP >180/120mmHg requires urgent medical care to reduce the risk of stroke.

Liver function tests

Acute liver injury can occur with a rise in AST or ALT, often of several thousands.

►►Note: do not miss a hidden paracetamol overdose—check paracetamol levels in blood, from the earliest sample you have on that patient! Checking an INR or prothrombin ratio (PTR) is essential in suspected late paracetamol poisoning and is of good prognostic value.

ECG

An ECG should be carried out in patients with tachycardia, chest pain, or reduced GCS. Patients should undergo cardiac monitoring. Cardiac arrhythmias are common and deaths, which occur soon after ingestion, may be due to these. Arrhythmias are often supraventricular, although ventricular arrhythmias also occur. Serial troponin levels are required if there has been/is chest pain.

Urine tests

Urine tests, e.g. EMIT dipstick system or by immunoassay in the laboratory, are sensitive and group-specific for amphetamines and can confirm an amphetamine has been ingested if that is in doubt, e.g. agitated patient in the emergency department. Note: amphetamine, MDMA, and MDA concentrations in blood are of no value in determining clinical management. Newer synthetic amphetamines (cathinones) may not be detected by urine dipstick tests but can be detected by chromatographic techniques in the laboratory.

Imaging

A CT brain scan should be performed for patients with altered conscious state, focal neurological signs, or severe headache.

Patients who are suspected body packers or body stuffers should undergo abdominal imaging, e.g. CT scan.

Further reading

Jones AL. Amphetamine overdose. BMJ Best Practice (online), Sept 2015. us.bestpractice.bmj.com.

Volkow ND, Fowler JS, Wang GJ, et al. Distribution and pharmacokinetics of methamphetamine in the human body: clinical implications. PLoS One 2010; 5: e152369.

Anticonvulsants

For most anticonvulsants, LC-MS/MS will be the analytical method of choice. However, some (e.g. valproic acid) are difficult to assay by LC-MS/MS, and GLC may be more appropriate.

Carbamazepine toxicity

Plasma concentrations of carbamazepine and its active metabolite the 10,11-epoxide can be measured by HPLC or LC-MS/MS but do not correlate well with the degree of toxicity. They are seldom performed, unless the diagnosis is in doubt or there is concern about a therapeutic excess. The therapeutic range in plasma is between 8 and 12mg/L. Toxicity has been seen with carbamazepine concentrations above 20mg/L (85mmol/L). Coma, fits, respiratory failure, and conduction abnormalities have been seen with concentrations in excess of 40mg/L (170mmol/L). In seven fatalities due to carbamazepine overdose, femoral blood concentrations taken post-mortem averaged 45mg/L (range 35–70).⁴

An FBC should be performed. U&E and creatinine should also be checked, as hyponatraemia and SIADH ( OHCM 10e, p. 241) have been reported. Hypoglycaemia has also been reported. LFTs and PTR/INR should be performed to assess hepatotoxicity. An ECG should be performed in all but the most trivial carbamazepine overdosage to assess if any conduction abnormalities or interval prolongations (e.g. QRS widening, sinus tachycardia, AV block, or QTc prolongation) are present.

Lamotrigine toxicity

Overall, most patients exposed to lamotrigine in overdose experienced no clinical effects.⁵ Plasma concentrations of lamotrigine can be measured for compliance purposes (therapeutic range 1–4mg/L; upper limit may be as high as 10mg/L) but are not of value in the overdose situation. In a fatal case, the femoral blood concentration was 54mg/kg.⁴

Valproate toxicity

Plasma concentrations of sodium valproate can be measured by HPLC but do not correlate well with either the depth of coma or risk of seizures after overdose. The therapeutic plasma range is 40–100mg/L. U&E, creatinine, and glucose should be measured, as hypernatraemia, hypoglycaemia, and hypocalcaemia have been reported after overdosage. LFTs and PTR/INR should be taken to assess hepatotoxicity. Depending on the clinical need, a CT head scan and/or ECG may be required. Post-mortem blood concentration in a 34-year-old man was 720mg/L.⁴

Phenytoin toxicity

Most patients with acute phenytoin poisoning do not require measurement of their plasma phenytoin concentration and are treated with multi-dose activated charcoal. Blood glucose, U&E, creatinine, LFTs, and PTR/INR should be checked. An ECG should be obtained. Most cardiac complications have occurred with rapid (>50mg/min) IV administration. A CT head scan is needed if there is focal neurology (excluding ataxia or nystagmus) or any history of traumatic injury.

Most phenytoin toxicity is managed by repeat multi-dose activated charcoal. Rarely, an urgent phenytoin measurement may help in severe phenytoin poisoning where charcoal haemoperfusion or some other extracorporeal elimination method, such as MARS, is being considered,⁶ e.g. if the plasma phenytoin concentration is above 40mg/L and rapidly rising or is close to, or exceeding, the potential lethal level of 100mg/L. The overall clinical value of such elimination methods remains to be established. Post-mortem phenytoin concentration in femoral blood of a 4.5-year-old child who ingested 2g was 45mg/L.⁴ Phenytoin undergoes post-mortem redistribution.

Patients with suspected chronic phenytoin toxicity as a result of therapeutic dosing should have their plasma phenytoin concentration measured. The ‘therapeutic range’ is 10–20mg/L. Routine measurements may be useful to monitor anticonvulsant therapy or to time re-institution of chronic therapy after overdose.

Benzodiazepines

Most patients who have taken an overdose with benzodiazepines just sleep off the drug without sequelae within 24h. However, more severe effects can occur when benzodiazepines are mixed with other drugs, especially in patients with pre-existing cardiovascular or respiratory disease. Pulse oximetry is useful for monitoring the adequacy of ventilation if significant CNS depression is present.

Generally, measuring benzodiazepine concentrations in blood or urine is not of value in the management of benzodiazepine overdose patients. Benzodiazepines can be detected in routine urine drugs screens. LC-MS/MS is the method of choice for quantitative analysis of diazepam and its metabolites. Liquid chromatography simultaneously assays diazepam and its polar metabolites, and post-mortem blood concentrations of 5 and 19mg/L have been found in fatalities.

Cannabis (marijuana)

Cannabis use may be detected in routine urine drug screening. They may be detected for up to several days after use, depending on the dose and frequency of use.

Synthetic cannabinoids may be consumed in herbal mixtures for recreational use, as an alternative to cannabis. Blood glucose, U&E, creatinine, and ECGs should be undertaken. Note that synthetic cannabinoids, e.g. JWH018, are often not detected by standard bedside urine drug screens (and thus are often the drugs of abuse selected by workers who undergo regular employment screening). However, they are detected by GC-MS, mass spectrometry, or LC-MS/MS if the sample is sent to the laboratory.

Further reading

Baselt RC, Cravey RH.Disposition of Toxic Drugs and Chemicals in Man, 9th edn, San Francisco, Chemical Toxicology Institute, 2011.

Hermanns-Clausen M, Kneisel S, Szabo B, Auwärter V. Acute toxicity due to the confirmed consumption of synthetic cannabinoids: clinical and laboratory findings. Addiction 2013; 108: 534–44.

Carbon monoxide

CO non-fire poisoning deaths have started to reduce in the UK, e.g. in 2012, compared with previous years. Incidence peaks in winter months, due to ↑ use of indoor heating and fossil fuel-powered generators and reduced external ventilation. It is highly likely that CO poisoning remains significantly underdiagnosed with patients presenting with ‘flu-like’ illness.

Survival rates for those resuscitated post-cardiac arrest with CO poisoning is very low, especially if there was pre-existing cardiovascular or respiratory disease. Pregnant women and their fetus are particularly susceptible to CO poisoning. Patients surviving significant CO poisoning may develop acute or delayed neurological sequelae leading to loss of consciousness, coma, and death. In the recovery phase, neurological sequelae, such as poor concentration and memory problems, may be seen. Symptoms include cognitive or personality changes, incontinence, psychosis, and Parkinsonian features such as rigidity. ~50–75% of people recover from their neurological features within 1 year.

A COHb concentration in blood confirms recent exposure to CO and should be measured urgently in all patients with suspected CO poisoning, including those with smoke inhalation. The space above the blood in the sample tube (headspace) should be minimized prior to blood gas analysis. Expected ‘normal’ values for COHb are up to 5% in non-smokers and up to 10% in smokers. However, people with pre-existing cardiovascular or respiratory disease can present with symptoms of toxicity at COHb concentrations in the low or normal range. A COHb concentration, however, does not measure severity of poisoning because COHb begins to dissociate from the moment of removal from the source of CO, and the rate of dissociation is also dependent on factors such as exogenous O₂ administration. Thus, any attempt to ‘back-extrapolate’ to find the initial highest COHb, e.g. by using nomograms, is flawed.

Management of the CO-poisoned patient is determined by the patient’s clinical condition and also on circumstantial evidence such as the intensity and duration of exposure, rather than a COHb concentration per se, although a level of >40% has been used as one criterion to guide the use of hyperbaric O₂ therapy (which is of controversial clinical value). The CO-exposed patient should be administered high-flow O₂ (e.g. 12L/min through a tight-fitting non-rebreather mask) until the COHb is <5% and clinical signs of CO poisoning, such as impaired heel–toe walking and finger–nose incoordination, have resolved. If in doubt, O₂ is administered for a minimum of 6h. It is currently unknown if hyperbaric O₂ is more effective than normobaric 100% O₂ in prevention of neurological complications in patients with CO poisoning.

Arterial blood gases

Any patient with suspected poisoning by CO requires ABG analysis. O₂ saturation monitors are misleading, as they read COHb as oxyhaemoglobin (HbO), and the true O₂ saturation of the patient can only be determined by ABG analysis.

Electrocardiogram

An ECG should be performed in anyone severely poisoned (e.g. drowsiness or any neurological abnormality, chest pain, or breathlessness) or with pre-existing ischaemic cardiomyopathy. ECG changes, such as ST-segment depression, T wave abnormalities, ventricular tachycardia or fibrillation, and cardiac arrest can occur. If ischaemia/infarction is seen on the ECG or suspected clinically, then serial troponins should be requested.

Further reading

Fisher DS, Leonardi G, Flanagan RJ. Fatal unintentional non-fire-related carbon monoxide poisoning: England and Wales, 1979–2012. Clin Toxicol 2014; 52: 166–70.

Smollin C, Olson K. Carbon monoxide poisoning (acute). BMJ Clin Evid. 2010 Oct 12; 2010–103.

Cocaine

Cocaine is snorted into the nose or injected IV.

Blood pressure monitoring

Patients with cocaine intoxication should have frequent measurements of their BP, as hypertension is a significant risk, and strokes and chest pain have been widely reported. A continuous monitoring device for repeated measurements is suitable.

ECG and cardiac markers

Cocaine-induced angina and MI are common. ECG monitoring is advised for all but the most trivial exposure to cocaine.

The predominant coronary pathology of cocaine is vasoconstriction. Thrombosis is less common, and troponin is the most sensitive indicator of myocardial damage (MI) due to cocaine. Cardiac troponins should also be performed in any patient with chest pain or ECG abnormalities. They are the most sensitive and specific markers. If elevated, then coronary angiography is usually indicated.

The ECG is of less diagnostic value. Up to 84% of patients with cocaine-related chest pain have abnormal ECGs, even in the absence of MI. In another study, 42% of patients with ‘MI-like’ ST elevation on their ECGs had MI excluded by cardiac markers. In a further study, total CK was elevated in 75% of patients, including 65% without MI.

Urine or blood testing

The use of cocaine can be determined by self-reports from patients or urine analysis of metabolites of cocaine (benzoylecgonine), e.g. EMIT test. This remains +ve for 24–48h after cocaine use. In assessing patients who may be a body packer, a +ve urine test may indicate pre-existing cocaine use or leaking cocaine from the packets.

GC-MS is more specific and can be carried out on blood or urine. Cocaine is unstable in blood, and samples are best taken into 1% w/v fluoride oxalate tubes if medicolegal sequelae of cocaine use are a possibility. ‘Nasal insufflation of 106mg of the drug to six volunteers produced mean peak plasma concentrations of 0.22mg/L at 0.5h and 0.61mg/L for the metabolite benzoylecgonine at 3h’.⁷ Smoking 50mg in six volunteers produced mean peak plasma concentrations of 0.2mg/L at 0.08h and 0.15mg/L for benzoylecgonine at 1.5h. Patients have survived plasma concentrations of 5.2mg/L, but usually fatalities are associated with cocaine/benzoylecgonine concentrations in excess of 5mg/L, depending on the route of use. The IV route is the most dangerous. Cocaine may show significant post-mortem redistribution.

Further reading

McCord J, Jneid H, Hollander JE, et al.; American Heart Association Acute Cardiac Care Committee of the Council on Clinical Cardiology. Management of cocaine-associated chest pain and myocardial infarction: a scientific statement from the American Heart Association Acute Care Committee of the Council on Clinical Cardiology. Circulation 2008; 117: 1897–907.

Senthilkumaran S, David SS, Jena NN, Menezes RG, Thirumalaikolundusubramanian P. Acute myocardial infarction and cocaine toxicity: One step closer. Indian J Crit Care Med 2014; 18: 118.

Shields LB, Rolf CM, Hunsaker JC 3rd. Sudden death due to acute cocaine toxicity-excited delirium in a body packer. J Forensic Sci 2015; 60: 1647–51.

Cyanide

Cyanide poisoning can occur by deliberate inhalation of gas, ingestion of cyanide salts, or exposure in industrial fires.

Arterial blood gas estimation

Essential to determine the O₂ saturation and acid–base status of the patient.

Serum lactate

This is helpful in confirming suspected toxicity and can be used clinically as a surrogate for the cyanide assay. It is likely to exceed 7mmol/L in cases of significant exposure. Thus, a blood gas analyser in the emergency department can give a quick (indicative) answer to the question of whether exposure has taken place.

Electrocardiogram

All patients should have an ECG. It should be examined for evidence of ischaemic damage, e.g. ST depression, ST elevation, T wave inversion.

Cyanide assay

Blood cyanide concentrations are rarely of use in emergency management, because they cannot be measured quickly enough. However, a sample should be taken before antidote administration for assay at a later stage. Cyanide concentrations of <0.2mg/L are ‘normal’; 1.0–2.5mg/L causes obtundation and coma, and >2.5mg/L is potentially fatal.

►►Note: the antidote of choice for cyanide poisoning is hydroxocobalamin, together with O₂; these are antidotes that can safely be given without certainty of cyanide ingestion. An alternative is to give sodium thiosulfate and sodium nitrite, and dicobalt edetate. However, note that dicobalt edetate should only be given if cyanide poisoning is certain, i.e. a proper history is available; otherwise you may kill your patient with cardiotoxicity of the antidote. Excessive administration of sodium nitrite can cause significant methaemoglobinaemia.

Further reading

Borron SW, Baud FJ, Barriot P, Imbert M, Bismuth C. Prospective study of hydroxocobalamin for acute cyanide poisoning in smoke inhalation. Ann Emerg Med 2007; 49: 794–801.

Borron SW, Baud FJ, Megarbane B, Bismuth C. Hydroxocobalamin for severe acute cyanide poisoning by ingestion or inhalation. Am J Emerg Med 2007; 25: 551–8.

Hall AH, Saiers J, Baud F. Which cyanide antidote? Crit Rev Toxicol 2009; 39: 541–52.

Digoxin

Patients who are already taking digoxin and those with pre-existing CVD are more susceptible to digoxin toxicity after overdosage. Digoxin toxicity can also result from progressive renal impairment or due to interactions with other drugs, as well as overdoses. Acute digoxin toxicity is an indication for oral multi-dose activated charcoal and IV hydration.

Electrocardiogram

All patients with suspected digoxin poisoning should have a 12-lead ECG, and all symptomatic patients should be attached to a cardiac monitor. Digoxin poisoning can cause virtually any type of cardiac arrhythmia due to ↑ automaticity and ↓ AV conduction. The combination of heart block with tachyarrhythmia is very common.

Plasma digoxin concentration

Absorption of digoxin often peaks at 4–6h after ingestion. Its half-life is in excess of 30h. Digitoxin is a structurally related drug that has an even longer plasma half-life (6 days). A digoxin measurement is a useful, but not absolute, guide to toxicity, as plasma digoxin concentrations correlate poorly with the severity of poisoning, particularly early in the course of acute poisoning. However, it is desirable in acute poisoning (although not essential) if anti-digoxin antibody fragments (Fab) are to be used, as it is useful in calculating the dose of fragments required ( Digoxin, Indications for Fab fragments and doses of Fab fragments, pp. 718–719), as well as confirming exposure. Plasma digoxin concentrations cannot be interpreted after administration of digoxin Fab using normal assay procedures. Samples taken to investigate probable chronic digoxin intoxication should be taken at least 6h after dosing. They are not normally analysed urgently, unless life-threatening features of toxicity are present and use of Fab is being considered. The therapeutic range for digoxin is 0.8–2.0µg/L.

Urea and electrolytes, creatinine

It is important to ascertain if the patient has any renal impairment and plasma creatinine and urea are helpful, although of course do not exclude renal impairment completely. Hyperkalaemia is common in acute digoxin overdose and may be severe, e.g. >7mmol/L.

If possible, a magnesium level is helpful to exclude hypomagnesaemia, which contributes to risk of cardiotoxicity and is easily corrected.

Indications for digoxin antibody (Fab) fragments in acute toxicity and doses of Fab fragments

•Bradycardia or heart block associated with hypotension.

•Tachyarrhythmias associated with hypotension, especially ventricular arrhythmias

Fab fragment administration should be considered in less severe stages of poisoning in older patients and those with pre-existing CVD. At one time, the plasma K⁺ concentration was considered an indication for use of Fab fragments, but this is no longer used as an indication. Previously, use of Fab fragments was considered for chronic toxicity, but this tends not to be used now.

The dose of Fab fragments to give for an acute ingestion of digoxin can be calculated from either the dose of digoxin ingested or the plasma digoxin concentrations. If in doubt, ten ampoules of Digibind^® can be given, followed by an additional ten ampoules if clinically indicated.

	Number of 40mg vials of Fab = plasma digoxin concentration (ng/mL) × body weight × 0.0084
Or
	Ingested dose (mg) × 1.2
Or
	Best guess of 10–20 vials

OHCM 10e, p. 842. Please note that if other types of digoxin-binding antibodies are used, then the dosage may be different to those indicated above.

Further reading

BMJ Best Practice (last updated 2016). Digoxin overdose. http://bestpractice.bmj.com/best-practice/monograph/338.html.

Dasgupta A. Therapeutic drug monitoring of digoxin: impact of endogenous and exogenous digoxin-like immunoreactive substances. Toxicol Rev 2006; 25: 273–81.

Flanagan RJ, Jones AL. Fab antibody fragments: some applications in clinical toxicology. Drug Safety 2004; 27: 1115–33.

Ethylene glycol, ethanol, and methanol poisoning

A history of ingestion or the presence of a metabolic acidosis raises suspicion of poisoning with these substances. Calculation of the anion gap and osmolal gaps is helpful in the assessment of such patients.

Anion gap

Calculating the anion gap

The normal anion gap is 12 ± 2.

Many toxins cause a high anion gap acidosis and these include

•Ethanol.

•Methanol. (Note: the high anion gap is due to metabolites and may take several hours to develop.)

•Ethylene glycol. (Note: the high anion gap is due to the metabolites and may take 6–24h to develop.)

•Metformin.

•Cyanide.

•Isoniazid.

•Salicylates (aspirin).

This list can be further reduced by measuring the osmolal gap.

Osmolal gap

This is the difference between the laboratory estimation of osmolality (Om) and calculated osmolality (Oc).

Calculating the osmolal gap

The osmolal gap is measured osmolality (Om) minus calculated osmolality (Oc):

Oc = 2([Na⁺] + [K⁺]) + [urea] + [glucose]

The osmolal gap is normally <10.

Toxic causes of a raised osmolal gap include

The acronym ‘MEDIE’ can be a helpful mnemonic.

Ethylene glycol and methanol plasma concentrations

Often the diagnosis of ethylene glycol or methanol poisoning can be difficult, because assays for these substances are not widely available. If possible, their measurement can help manage severe intoxication. Other parameters may have to be used, i.e. anion gap, osmolal gap, and ABG analysis. A normal osmolal gap does not exclude poisoning with ethylene glycol or methanol, but if the osmolal gaps and anion gaps are both normal, and the patient is not symptomatic, then significant ingestion is unlikely to have occurred. In general, ethylene glycol or methanol measurements should not be carried out, unless metabolic acidosis is present and there is an anion gap.

Ethylene glycol and methanol concentrations in blood are useful to confirm ingestion and indicate when to stop antidotal treatment (with ethanol or 4-methylpyrazole) and/or when haemodialysis is needed (>500mg/L; Indications for haemodialysis in methanol or ethylene glycol poisoning are, p. 722). However, a low concentration may just mean that most of the parent compound has been metabolized. Formate (i.e. the methanol metabolite) levels can also be checked in patients who may have taken methanol.

Microscopy of urine for oxalate crystals

In suspected ethylene glycol poisoning, microscopy can be performed to look for oxalate crystals. However, they are only present in 50% of cases and often only many hours after ingestion. Treatment of a patient should not be delayed or dependent upon looking for crystals.

Brent J. Fomepizole for ethylene glycol and methanol poisoning. N Engl J Med 2009; 360: 2216–23.

McMartin K, Jacobsen D, Hovda KE. Antidotes for poisoning by alcohols that form toxic metabolites. Br J Clin Pharmacol 2016; 81: 505–15.

Megarbane B, Borron SW, Baud FJ. Current recommendations for treatment of severe toxic alcohol poisonings. Intens Care Med 2005; 31: 189–95.

Iron

Serum iron concentrations

In the UK, one 200mg tablet of ferrous sulfate contains 65mg of elemental iron. Iron is also found in varied amounts in many over-the-counter vitamin supplements. The elemental iron content of the preparation ingested should be checked carefully, as the important consideration is the amount of elemental iron ingested, not the weight of iron or vitamin tablets.

Serum iron concentrations should be measured urgently in all patients who may have ingested >30mg/kg of elemental iron acutely, those who have ingested an unknown quantity, or those with symptoms, e.g. GI. If a sustained-release preparation of iron has been taken, a later serum iron concentration should be taken. A blood sample taken late after ingestion may underestimate the iron, as it may have already started distributing to tissues, i.e. in a late-presenting patient, a low concentration cannot be interpreted, but a high one indicates toxicity.

If the antidote desferrioxamine is given before 4h have elapsed, it interferes with the colorimetric assay for iron, and so a serum sample for iron should be taken off before it is given. If atomic absorption spectrophotometry is available for measurement of serum iron, there is no interference from desferrioxamine.

It is essential to interpret the serum iron concentration result in the context of the clinical state of the patient. If <55µmol/L (<300mg/dL), mild toxicity is expected. If above 90µmol/L (500mg/dL), severe toxicity is expected and treatment with desferrioxamine is necessary. Do not wait for an iron concentration if altered conscious state, shock, or severe acidosis (pH <7.1) is present. Antidotal treatment is also indicated for patients with iron concentrations of >55µmol/L if there is additional clinical evidence of toxicity, e.g. GI symptoms, leucocytosis, or hyperglycaemia. Antidotal therapy with desferrioxamine is indicated without waiting for the serum iron concentration in patients with severe features (e.g. fitting, unconscious, or hypotensive). Desferrioxamine is usually continued until the urine has returned to a normal colour, symptoms have abated, and all radio-opacities of iron tablets on AXR have disappeared. Urine free iron estimation is the best test of when to stop chelation therapy with desferrioxamine but is not widely available.

Working out if the patient needs a serum iron level checked

If a patient has ingested <30mg/kg body weight of elemental iron (a 200mg ferrous sulfate tablet = 65mg of elemental iron), then no serum iron level is required. If in doubt, a plain AXR will usually indicate if lots of tablets are present. A serum concentration of <55µmol/L (<300mg/dL) also indicates low risk ( Iron, Serum iron concentrations, p. 724).

Abdominal X-ray

This is required in patients who have ingested in excess of 30mg of elemental iron/kg body weight. The AXR determines the need for gut decontamination either by gastric lavage or whole bowel irrigation with polyethylene glycol. Undissolved tablets appear radio-opaque, but they disappear once dissolved, so the absence of radio-opacities does not exclude the possibility of toxicity.

Full blood count

This is needed in all cases of iron poisoning. Leucocytosis (>15 × 10⁹/L) is common with significant toxicity. Cross-match is a wise precaution in potentially serious poisoning.

Urea and electrolytes and creatinine, baseline liver function tests, and clotting

This is needed in all cases.

Blood glucose

Hyperglycaemia is common in serious poisoning.

Arterial blood gases

These should be checked in symptomatic or severely poisoned patients. Metabolic acidosis can occur.

Total iron binding capacity

This has no role in the assessment of acute iron poisoning.

What to do if estimation of serum iron concentration is unavailable

If serum iron assay is not available, the presence of nausea, vomiting, leucocytosis (>15 × 10⁹/L), and hyperglycaemia (>8.3mmol/L) suggests significant ingestion and the need for treatment with desferrioxamine.

OHCM 10e, p. 842.

Further reading

Chang TP, Rangan C. Iron poisoning: a literature-based review of epidemiology, diagnosis, and management. Pediatr Emerg Care 2011; 10: 978–85.

Royal Children’s Hospital Melbourne.Iron poisoning. Clinical practice guidelines. http://www.rch.org.au/clinicalguide/guideline_index/Iron_poisoning/.

Lead exposure and poisoning

Blood lead concentrations

Blood lead concentrations are used to confirm exposure and decide on whether environmental exposure reduction measures or (rarely) iron chelation therapy is required. Samples are not ‘urgent’ (except in the case of suspected acute lead encephalopathy) and must be taken into an EDTA tube.

A blood lead level of 5µg/dL or more requires further testing and monitoring, and the source of lead to be found and removed. A lead level of >45µg/dL in a child usually indicates the need for chelation treatment. Occupational lead levels and appropriate responses for adults are enshrined in Worker/Occupational Health and Safety legislation.

There are two agents used for chelation therapy in lead poisoning—IV EDTA (disodium calcium edetate) and oral DMSA (2,3-dimercaptosuccinic acid). Before use, chelation therapy should be discussed with a toxicologist. In general, patients with a blood lead concentration of >45µg/dL should be treated with chelation therapy and removal from further exposure. Children with encephalopathy or a blood lead concentration of >75µg/dL require admission to hospital for urgent chelation therapy.

Abdominal X-ray

A plain AXR should be performed in all children, particularly if there is a history of pica, to exclude ingested paint or lead foreign bodies such as curtain pulls or fishing sinkers. Long bone X-rays in children may show lead lines.

Zinc protoporphyrin estimations

Zinc protoporphyrin (ZPP) estimations can be helpful in individuals with moderate (>20µg/dL) to high (>40µg/dL) blood lead concentrations in whom one is trying to determine the chronicity of exposure. There is a poor correlation between ZPP and blood lead at lower blood lead concentrations. There are other conditions (e.g. iron deficiency) that can ↑ ZPP, and there is significant inter-individual variation. ZPP has been proposed as a surrogate marker for blood lead, but blood lead is the best marker and should not be replaced by ZPP.

Other essential investigations

Patients should also have FBC and a blood film (for basophilic stippling), U&E, LFTs, and serum Ca²⁺ measured. All children should have their serum iron measured, as iron deficiency is an important diagnosis and, if corrected, can reduce ongoing lead absorption from the gut.

Further reading

Boreland F, Lesjak MD, Lyle DM. Managing environmental lead in Broken Hill: a public health success. NSW Public Health Bull 2008; 19: 174–9.

Gulson B, Korsch M, Matisons M, Douglas C, Gillam L, McLaughlin V. Windblown lead carbonate as the main source of lead in the blood of children from a seaside community: an example of local birds as “canaries in the mine”. Environ Health Perspect 2009; 117: 148–54.

Lithium

Blood lithium concentration

Lithium is available as sustained-release and non-sustained-release tablets and liquid. After ingestion of liquid preparations, plasma lithium concentrations peak at 30min. With sustained-release preparations, peak concentrations occur at 4–5h. The plasma half-life of lithium is often in excess of 24h. Interpretation of plasma lithium concentrations depends on the clinical circumstances of exposure ( Acute overdose in lithium-naive patient, p. 728, Chronic excess of lithium, p. 728, Acute-on-chronic lithium poisoning, p. 728). Do not take blood for lithium levels into a lithium heparin tube!

Acute overdose in lithium-naive patient

A single overdose in a lithium-naive patient is of low risk. However, onset of toxicity may be delayed for as much as 24h. Plasma samples for lithium assay should be taken at 6h post-ingestion and measured urgently. The patient should have IV fluids to facilitate lithium elimination. Consider haemodialysis if plasma lithium concentration is >7.5mmol/L.

Chronic excess of lithium

Lithium toxicity can occur if the patient has been taking too high a dose or is dehydrated, or if an interaction with thiazide diuretics, NSAIDs, ACE inhibitors, or tetracyclines has occurred. Risk of toxicity is further enhanced by the presence of hypertension, diabetes, cardiac failure, renal failure, or schizophrenia. Blood for plasma lithium assay should be taken at presentation. Often good IV hydration suffices to clear the lithium; rarely haemodialysis is needed. Consider haemodialysis if the plasma lithium exceeds 2.5mmol/L.

Acute-on-chronic lithium poisoning

A patient taking lithium chronically who then takes an acute overdose is at risk of serious toxicity, because tissue binding of lithium is already high. The plasma lithium levels should be measured urgently at 6h post-ingestion. Lithium measurements should be repeated 6- to 12-hourly in symptomatic patients until clinical improvement occurs. Consider haemodialysis if plasma concentrations exceed 4mmol/L.

Indications for haemodialysis or arteriovenous haemodiafiltration

Lithium is effectively removed by haemodialysis (preferred) or arteriovenous haemodiafiltration. It is indicated in all patients with severe lithium poisoning, i.e. coma, convulsions, respiratory failure, or ARF. Plasma lithium concentrations can also guide the need for haemodialysis/haemodiafiltration. Each hour of dialysis will reduce the plasma lithium by 1mmol/L, but plasma lithium often rebounds after haemodialysis has stopped, so the assay should be repeated at the end of dialysis and again 6–12h later.

Urea and electrolytes, serum creatinine

Hyponatraemia is common in lithium toxicity. It is also important to check the serum K⁺ concentration and urea, as lithium is renally excreted and renal failure delays its elimination.

ECG

Lithium poisoning may result in arrhythmias, and complete heart block has been reported. An ECG (and sometimes monitoring) is required. Chronic lithium toxicity frequently has non-specific and diffuse depressed ST-segments and T wave inversion, which are seldom of sinister consequence.

Further reading

Bellomo R, Kearly Y, Parkin G, et al. Treatment of life-threatening lithium toxicity with continuous arterio-venous hemodiafiltration. Crit Care Med 1991; 19: 836–7.

Eyer F, Pfab R, Felgenhauer N, et al. Lithium poisoning pharmacokinetics and clearance during different therapeutic measures. J Clin Psychopharmacol 2006; 3: 325–30.

Methaemoglobinaemia

Oxidizing agents convert Hb to methaemoglobin (MetHb), and this renders it incapable of carrying O₂. Common agents causing methaemoglobinaemia include: dapsone, sulfonamides, chlorates, nitrites, nitrates, and local anaesthetics including lidocaine. The onset and duration of symptoms will depend on the agent. Nitrites cause breathlessness and flushing within minutes of exposure, but dapsone may cause methaemoglobinaemia several hours after ingestion and the methaemoglobinaemia may then persist for days.

Essential investigations

Patients with suspected methaemoglobinaemia should have the following

•ABGs.

•FBC (especially if dapsone has been taken → haemolytic anaemia).

•Blood MetHb concentration.

MetHb can produce a normal PO₂ in the presence of reduced O₂ saturation. Pulse oximetry measures both MetHb and oxygenated Hb so can give false results.

Methaemoglobin estimation in blood

Measurement of blood MetHb is required to confirm the diagnosis and assess the severity of poisoning. The measurement must be done urgently when administration of the antidote methylthioninium chloride (methylene blue) is contemplated. Samples for MetHb estimation need to be analysed as soon as possible after collection, as if left to stand around, the MetHb will be falsely low owing to a reduction by endogenous MetHb reductase. The severity of symptoms correlates roughly with the measured MetHb concentrations. Anaemia and cardiac or pulmonary disease will lead to more severe symptoms at a lower MetHb level (see Table 11.2).

Table 11.2 Clinical effects

MetHb concentration (%)	Clinical effects
0–15	None
15–30	Mild: cyanosis, tiredness, headache, nausea
30–50	Moderate: marked cyanosis, tachycardia, dyspnoea
50–70	Severe: coma, fits, respiratory depression, metabolic acidosis, arrhythmias
>70%	Potentially fatal

If the patient has severe clinical features of toxicity or if the blood MetHb concentration is >30%, the patient should be given methylene blue. Methylene blue can be given at lower blood MetHb concentrations in those who are symptomatic.

Opioids

Classic features of opioid poisoning

•Depressed respiratory rate and volume.

•Pinpoint pupils.

•Coma

•Signs of parenteral drug use, e.g. needle marks.

Toxicity can be prolonged for 24–48h, particularly after ingestion of methadone, which has a long half-life. The lifesaving measure is prompt administration of adequate doses of naloxone, before waiting for results of any investigations. This often needs to be repeated or an infusion started, as the half-life of the antidote is much shorter than opioid drugs.

Adequacy of ventilation

O₂ saturation monitoring and/or ABG analysis demonstrates the adequacy of ventilation in those whose respiration is depressed, together with accurately measuring the respiratory rate.

Urine drug screening

Qualitative screening of the urine (group-specific immunoassay) confirms recent use. This may, however, not detect fentanyl derivatives, tramadol, and other synthetic opioids.

Measuring opioids in blood

Quantitative analysis is usually undertaken by LC/Q-TOF-MS or GC-MS or MS. This is often required for medicolegal purposes, particularly where a fatality or a childcare issue is involved.

‘Plasma morphine free and total morphine concentrations were an average of 88 and 277µg/L in 54 people treated for heroin overdose’. ⁸ Interestingly, the degree of poisoning correlated better with total morphine concentrations than free concentrations.

Post-mortem morphine levels in heroin overdose deaths vary, depending on prior narcotic history, but in general exceed 0.3mg/L.⁸

‘Blood tramadol concentrations in 105 people arrested for impaired driving performance averaged 850µg/L’.⁸ Post-mortem concentrations in five tramadol deaths averaged 6.1mg/L. Heroin metabolites and tramadol do not undergo significant post-mortem redistribution.

Paracetamol screening

Opioid tablets are frequently combined with paracetamol. All unconscious patients should therefore have plasma paracetamol level measured.

Further reading

Jones AL. Management of opioid poisoning. In: A Webb, D Angus, S Finfer, L Gattinoni, M Singer (eds.). Oxford Textbook of Critical Care, 2nd edn. 2016; pp. 1522–5.

Paracetamol (acetaminophen) poisoning

Overview

Paracetamol is the commonest drug taken in overdose in the UK. Paracetamol can be measured by a variety of assay methods, but HPLC or LC-MS is less susceptible to interference than some enzyme-based assays.

Measurement of plasma paracetamol concentration is essential for assessing the need for antidotal treatment (NAC) in most cases of paracetamol poisoning. The nomogram for paracetamol concentrations and protocols for when and how to give NAC differs between countries, and it is important to consult your own national guidelines. If NAC is given within 12h of the overdose, it provides complete protection against liver injury and renal failure. Beyond 12h after ingestion, the protection is less complete and assessment of liver damage is required. Paracetamol poisoning can be deceptive, as there is a latent phase of many hours where the patient remains well before liver damage develops. The co-ingestion of opioids may delay gastric emptying and peak plasma paracetamol concentrations.

INR/PT

The most sensitive marker of prognosis in paracetamol poisoning is the PT or INR. This often starts to ↑ within 24–36h of the overdose and peaks at 48–72h. Once the INR/PT starts to improve, this is a sign that hepatotoxicity is starting to improve and the patient will not go on to develop acute liver failure. Approximately half of patients with a PT of 36s at 36h post-ingestion will develop acute liver failure.

Plasma alanine and aspartate aminotransferases

These may begin to rise as early as 12h post-ingestion but usually peak at 72–96h. AST or ALT values in excess of 10,000IU/L are not unusual, and a plasma ALT of >5000IU/L is very suggestive of paracetamol poisoning (see Fig. 11.1). Serum bilirubin may peak after the aminotransferase, and this should not lead to concern for patients in whom the INR or PT have begun to fall.

Fig. 11.1 Time course of liver function tests in paracetamol poisoning.

Do not correct abnormalities in PT or INR with FFP or cryoprecipitate, unless life-threatening bleeding is taking place; otherwise the most sensitive marker of how the patient is progressing will be lost.

Other blood test abnormalities in paracetamol poisoning

Hypoglycaemia and metabolic acidosis are common. Early metabolic acidosis is often associated with very high plasma paracetamol concentrations, e.g. >400mg/L. Later, development of acidosis indicates incipient acute liver failure and the need to urgently check ABGs, LFTs, and INR/PTR.

Pancreatitis with ↑ serum amylase/lipase has been reported. Several cases of thrombocytopenia have been reported.

U&E and creatinine should be checked. Renal failure can occur in the context of hepatic failure, but also in its absence (in 1 in 100 patients). It is treated with NAC and supportive measures, e.g. haemodialysis, if needed. Full recovery with supportive care is common.

Investigating the patient who has taken a paracetamol overdose <4h ago

Ingestion of >75mg/kg of paracetamol or a paracetamol-containing product should be recognized as a potential hepatotoxic dose for most people. If ingestion of this amount or more has occurred within the last 1h, activated charcoal should be given orally (50g for an adult). A plasma paracetamol level should then be checked at 4h from the time of ingestion, to determine the need for NAC treatment from the nomogram (see Fig. 11.2).

Very rarely, e.g. after ingestion of 4 × 500mg tablets by an adult, a confirmatory plasma paracetamol level is not needed, but in general it is safer to be certain by checking a blood concentration.

Investigating the patient who has taken a paracetamol overdose between 4 and 8h ago

A plasma paracetamol level should be checked as soon as possible. If a single acute ingestion has taken place, then the result is plotted on the relevant national nomogram against the time since ingestion (e.g. see Fig. 11.2). This determines the need for NAC antidote treatment (i.e. if the level is above or very close to the line).

If the overdose is staggered or repeated, supratherapeutic, then specific toxicology, advice is needed. If in doubt, treat the patient with NAC. Some countries have modified this nomogram to treat patients at lower paracetamol levels.

Investigating the patient who has taken a paracetamol overdose between 8 and 24h ago

Start treatment with NAC straightaway. Take blood for a paracetamol level, INR/PT, creatinine, and plasma venous HCO₃⁻ (if plasma venous HCO₃⁻ is abnormal, check ABGs). Check results and refer to the graph to determine whether treatment with NAC needs to be continued (i.e. is the plasma level above the treatment line?) or can be stopped (below the line). ►► Beyond 16h after ingestion, the sensitivity of the assay for paracetamol may be too low to detect a treatable level—check and, if in doubt, treat the patient with NAC! On completion of NAC, check blood INR/PT, creatinine, and plasma venous HCO₃⁻ (if abnormal, check ABGs). If the patient is asymptomatic and the INR or creatinine is normal or falling, discontinue NAC. If the patient has symptoms (abdominal pain or vomiting) or the INR or creatinine is rising, continue maintenance NAC (50mg/kg in 500mL of glucose every 4h) until the INR improves. Contact a poisons centre/liver unit.

Fig. 11.2 An acetylcysteine treatment graph.

Note: The most up-to-date national nomogram should be accessed for your country.

Investigating the patient who has taken a paracetamol overdose >24h ago or the timing is not able to be established

Start treatment with the antidote NAC straightaway, unless a trivial amount has been taken.

Take blood for baseline INR/PT, creatinine, venous HCO₃⁻ (if abnormal, check ABGs), and paracetamol.

If the patient is asymptomatic and the laboratory tests normal (INR <1.3, paracetamol concentration <5 mg/L, and ALT <2 times the upper limit of normal), discharge the patient and advise to return if vomiting/abdominal pain develops. If the blood results are abnormal, continue NAC, and phone a liver unit/poisons centre for advice.

Investigating the patient who has taken a staggered overdose

The Commission on Human Medicines currently recommends that a Xmg/kg/24h dose ingested calculation is not used to guide therapy. Current advice is that all staggered overdoses should be treated with NAC and discussed with a toxicologist.

Blood is taken for paracetamol level, U&E, creatinine, HCO₃⁻, ALT, and INR. Hepatotoxicity is not likely if the patients is asymptomatic, the plasma paracetamol concentration is <5 mg/L, INR is <1.3, and ALT is <2 times the upper limit of normal. If all of the above are found, then NAC may be stopped and the patient discharged, with advice to return if they experience vomiting or abdominal pain.

Investigating the patient who has a repeated supratherapeutic ingestion

All patients with a supratherapeutic overdose should be considered for NAC treatment and discussed with a toxicologist. Currently, patients with symptoms or signs of hepatotoxicity or those who have definitely ingested 75mg or less of paracetamol require no treatment.

Further reading

BMJ Best Practice. Paracetamol overdose. http://bestpractice.bmj.com/best-practice/monograph/337.html.

TOXBASE. https://www.toxbase.org/.

UK National Poisons Information Service. http://www.npis.org/.

Salicylate poisoning

Features of severe poisoning

Ingestion of >150, 250, and 500mg/kg body weight of aspirin, respectively, produces mild, moderate, and severe poisoning, respectively. Aspirin poisoning is becoming increasingly rare in developed countries. Signs of serious salicylate* poisoning include metabolic acidosis, renal failure, and CNS effects such as agitation, confusion, coma, and convulsions. Death may occur as a result of CNS depression and cardiovascular collapse. The development of metabolic acidosis is a bad prognostic sign, as it also indicates ↑ CSF transfer of salicylate.

Note: (*) aspirin, oil of wintergreen.

Plasma salicylate concentration

Plasma salicylate should be measured urgently in all, but the most trivial overdose, i.e. all those thought to have ingested >150mg/kg body weight of aspirin or any amount of oil of wintergreen. It should be performed at 4h post-ingestion, because delayed absorption of the drug renders such levels uninterpretable before this time. As salicylates may form concretions in the stomach, which delay absorption, it is recommended that a salicylate level is rechecked 3–4h after the first sample, to catch the peak salicylate concentration. There is no evidence for indiscriminate requesting of salicylate concentrations in every unconscious patient (unlike paracetamol) or in conscious patients who deny taking aspirin and who have no features suggesting salicylate toxicity. The plasma salicylate concentration is not an absolute guide to toxicity, as paracetamol levels are in paracetamol poisoning, but should be interpreted together with clinical features and acid–base status of the patient.

Urinary alkalinization ( OHCM 10e, p. 844) is indicated for patients with salicylate concentrations of 600–800mg/L in adults and 450–700mg/L in children and the elderly. Metabolic alkalosis is not a contraindication to bicarbonate therapy, as patients may have a high base deficit in spite of an elevated serum pH.

Haemodialysis is very effective at salicylate removal and correction of acid–base and electrolyte abnormalities. It should be considered if the plasma salicylate levels are >700mg/L in children and >800mg/L in adults. Other indications for haemodialysis include resistant metabolic acidosis, severe CNS effects, such as coma, convulsions, pulmonary oedema, and ARF.

Arterial blood gases

Acid–base problems are common in salicylate poisoning. Respiratory centre stimulation causes respiratory alkalosis. Uncoupled oxidative phosphorylation and interruption of glucose and fatty acid metabolism by salicylates often cause concurrent metabolic acidosis. Serial ABGs are needed in severe salicylate poisoning.

Further reading

Ghosh D, Williams KM, Graham CG, et al. Multiple episodes of aspirin overdose in an individual patient: a case report. J Med Case Rep 2014; 19: 374.

Theophylline

Acute theophylline poisoning can carry a high mortality, and its management is best guided by the Shannon severity grading scheme, bearing in mind that delayed effects tend to occur after sustained-release formulations have been ingested. The adult therapeutic range is 10–20mg/L. Serious toxicity occurs at >100mg/L (770mmol/L).

Urea and electrolytes, creatinine, glucose

It is vital to check the plasma K⁺ concentration frequently, as hypokalaemia is a life-threatening complication and the serum K⁺ concentration is a useful guide to severity. If >2.5mmol/L, the patient is less severely poisoned (grade 1) than if it falls to <2.5mmol/L (grade 2). Check blood glucose since hyperglycaemia is a common complication.

Arterial blood gases

In potentially serious poisoning (e.g. ingestion of >20mg/kg body weight), abg analysis is helpful in optimizing the acid–base status of the patient. An initial phase of hyperventilation with respiratory alkalosis can be followed by a further stage of metabolic acidosis.

ECG

In potentially serious poisoning (e.g. ingestion of >20mg/kg body weight), an ECG is required. Cardiac monitoring is helpful for early identification of arrhythmias in such patients.

Plasma theophylline concentrations

Measuring plasma theophylline concentrations confirms theophylline ingestion if this is in doubt, and it is usually undertaken by HPLC or LC-MS/MS. However, for the vast majority of poisoned patients, obtaining a plasma theophylline concentration does not guide their management. Therapeutic levels rarely exceed 20mg/L (155mmol/L). Theophylline peak concentration in plasma may occur at 1–3h after ingestion of a standard-release formulation. However, overdose is often with sustained-release products, and delayed absorption can result in delayed peak plasma concentration and toxicity, often 12–24h later.

Plasma concentrations are also helpful in deciding when to employ charcoal haemoperfusion in seriously poisoned patients (particularly if plasma concentrations are >100mg/L (770mmol/L)). Charcoal haemoperfusion can be considered at lower concentrations, e.g. 80mg/L, in the elderly or those with pre-existing IHD, and those with persistent seizures or hypotension not responding to IV fluids. Charcoal haemoperfusion can also be decided on the basis of grade 3 or 4 severity grading alone. If charcoal haemoperfusion is not possible, then haemodialysis with multi-dose activated charcoal is probably an alternative.

Urine testing for myoglobinuria and measuring serum creatine kinase

Theophylline poisoning can be accompanied by rhabdomyolysis. Hence, the urine should be dipstick-tested, and if found +ve for blood, a serum CK should be obtained. This will then indicate that renal function should be closely monitored and whether the urine should undergo alkalinization.

Further reading

Shannon M. Predictors of major toxicity after theophylline overdose. Ann Intern Med 1993; 119: 1161–7.

Tricyclic antidepressants

The main risks of overdose with these drugs are the cardiovascular system (CVS) and CNS toxicity.

Electrocardiogram

An ECG should be performed in all but the most trivial cases of overdose.

ECG abnormalities are common in moderate to severe poisoning and include

•QRS prolongation: >110ms in adults predicts the risk of ventricular cardiac arrhythmias (and the need for IV sodium bicarbonate), and a QRS >160ms predicts the risk of fits. In children, a QRS >110ms is predictive of the risk of arrhythmias, but not fits.

•Note: ECG criteria are not the only factors assessing the risk of arrhythmias, fits, and acidosis—electrolyte disturbances contribute. Supraventricular, and potentially fatal ventricular, arrhythmias can occur.

Cardiac monitoring

This is essential if ingestion of >10mg/kg body weight has taken place. It is seldom necessary beyond 24h after ingestion.

Arterial blood gas analyses

These should be done on all patients with marked symptoms and signs, particularly those with a reduced GCS score. It should also be performed on those with widened QRS or seizures, not least because such patients are receiving IV sodium bicarbonate therapy and a pH of 7.5 should not be exceeded.

Plasma concentrations

This is of no value, as plasma concentrations of tricyclic antidepressants correlate poorly with clinical features of toxicity.

Warfarin and superwarfarins

INR/PTR

This is a key test in possible warfarin/superwarfarin overdose. In the case of warfarin, the INR often begins to rise from day 2 after the overdose and settles within a week, but with superwarfarins, the time course of clotting abnormality may stretch to weeks. Thus, repeated INR/PTR estimations may guide progress and the need for vitamin K1 antidote in warfarin toxicity. Individual factor assays have been done in some cases but are not routinely necessary.

Liver function tests

Assessment of liver function is helpful in warfarin overdose. A congested, dysfunctional liver would be expected to handle the consequences of a warfarin overdose less well.

Table of conversion factors between mass and molar units

(See Table 11.3.)

Table 11.3 Conversion factors between mass and molecular units

Drug	Mass units	Molar (SI) units	Conversion factor
Carbamazepine	mg/L	µmol/L	4.23
Digoxin	µg/L or ng/mL	nmol/L	1.28
Ethanol	g/L	mmol/L	1.28
Iron	mg/L	mmol/L	0.179
Lead	mg/L	mmol/L	0.0048
Paracetamol	mg/L	mmol/L	0.0066
Phenytoin	mg/L	mmol/L	3.96
Salicylate (aspirin)	mg/L	mmol/L	0.0072
Theophylline	mg/L	mmol/L	7.7

1 Wittebole X, Hantson P. Use of the molecular adsorbent recirculating system (MARS^TM) for the management of acute poisoning with or without liver failure. Clin Tox 2011; 49: 782–93.

2 Chue P, Singer P. A review of olanzapine-associated toxicity and fatality in overdose. J Psychiatr Neurosci 2003; 28: 253–61.

3 Baselt RC, Cravey RH.Disposition of Toxic Drugs and Chemicals in Man, 9th edn. San Francisco: Chemical Toxicology Institute, 2011.

4 Baselt RC, Cravey RH.Disposition of Toxic Drugs and Chemicals in Man, 9th edn, San Francisco, Chemical Toxicology Institute, 2011.

5 Lofton AL, Klein-Schwartz. Evaluation of lamotrigine toxicity reported to poison centers. Ann Pharmacother 2004; 38: 1811–15.

6 Sen S, Ratnaraj N, Davies NA, et al. Treatment of phenytoin toxicity by the Molecular Adsorbents Recirculating System (MARS). Epilepsia 2003; 44: 265–7.

7 Baselt RC, Cravey RH.Disposition of Toxic Drugs and Chemicals in Man, 9th edn. San Francisco, Chemical Toxicology Institute, 2011.

8 Baselt RC, Cravey RH.Disposition of Toxic Drugs and Chemicals in Man, 9th edn. San Francisco: Chemical Toxicology Institute, 2011.