Intended for healthcare professionals

Clinical Review

Lesson of the week: Deaths from low dose paracetamol poisoning

BMJ 1998; 316 doi: (Published 06 June 1998) Cite this as: BMJ 1998;316:1724
  1. S Bridger, clinical research fellowa,
  2. K Henderson, specialist registrarb,
  3. E Glucksman, consultantb,
  4. A J Ellis, specialist registrarc,
  5. J A Henry, professor (j.a.henry{at},
  6. Roger Williams, professorc
  1. aDepartment of Medicine, King's College School of Medicine, London SE5 9PJ,
  2. bAccident and Emergency Department, King's College Hospital, London SE5 9RS,
  3. cInstitute of Liver Studies, King's College Hospital,
  4. dAccident and Emergency Department, St Mary's Hospital, London W2 1NY
  1. Correspondence to: Professor Henry
  • Accepted 3 December 1997

Patients with serum paracetamol concentrations below the standard treatment line may develop acute liver failure Paracetamol is the most commonly used substance in self poisoning (about 70 000 cases annually in Britain1) and is the most frequent subject of inquiries to the National Poisons Information Services.2 Paracetamol overdose is the commonest cause of acute liver failure in the United Kingdom,3 accounting for at least half of all cases sent to tertiary referral units. To decrease the chance of liver damage in cases of paracetamol overdose, protocols and guidelines for treating patients with an antidote before referral to specialist care have been drawn up. The antidote acetylcysteine should be given to all patients with a serum paracetamol concentration >200 mg/l four hours after ingestion of the drug. A nomogram in which this value is joined to an end point of 25 mg/l at 16 hours allows identification over this period of the patients who should receive the antidote.4 If the antidote is not given, over 60% of patients with serum paracetamol concentrations above the treatment line may develop serious liver damage, and of these about 5% will die.5 Recent studies also suggest that acetylcysteine given after 16 hours, even at the stage of encephalopathy, can reduce the frequency of multiorgan failure and improve survival. 6 7 Factors that have been reported to enhance hepatotoxicity include chronic alcohol misuse,8 eating disorders,9 and enzyme inducing drugs,10 and in each of these contexts treatment is advisable below the treatment line.

No deaths have been reported in any of the major treatment trials of paracetamol overdose,11 12 13 14 however high the initial serum paracetamol concentration, provided acetylcysteine was given within 10 hours of the drug's ingestion. Furthermore, there was only a 2% incidence of serious liver damage (defined by an aspartate transaminase concentration >1000 IU/l) in patients with an initial serum paracetamol concentration above the 300 mg/l (2 mmol/l) line, when treated with intravenous acetylcysteine within 10 hours, compared with an expected incidence of 90%.15 These data suggest that serious hepatotoxicity should be uncommon and death extremely rare after a paracetamol overdose provided patients are treated within 10 hours; most cases present to an accident and emergency department in this time.16 Yet there have been reports of patients presenting within this time with serum concentrations below the treatment line who nevertheless develop fatal acute liver failure despite having no additional risk factors.17 Factors which may account for such cases include concealment of the real time of overdose and number of tablets taken or an enhanced susceptibility to liver damage.

Case reports

Case 1—A 16 year old girl presented to her accident and emergency department four hours after ingesting 20 paracetamol tablets (10 g). She had no additional risk factors for enhanced hepatotoxicity. Her serum paracetamol concentration was 156 mg/l, and she was discharged after receiving 50 g of a proprietary formulation of activated charcoal. Over the next two days her family noted that she was becoming progressively confused and drowsy. She represented 48 hours after the initial assessment. On examination she was encephalopathic (Glasgow coma score 9), icteric, dehydrated, and tachypnoeic, and had an international normalised ratio of 6.3 and an arterial pH of 7.16. Investigations showed serum concentrations as follows: alanine aminotransferase >7 500 IU/l, bilirubin 87 µmol/l, glucose 1.9 mmol/l, and creatinine 146 µmol/l. She was started on an acetylcysteine infusion and treatment for acute liver failure, and was electively ventilated before transfer. She continued to deteriorate with rising intracranial pressure, and underwent a total hepatectomy. A liver transplant was carried out 36 hours later but as there was no recovery of brain stem function by 10 days after the overdose ventilatory support was withdrawn.

Case 2 —A 24 year old man presented to his accident and emergency department within four hours of ingesting 64 paracetamol tablets (32 g). He had a history of alcohol misuse (about 100 units per week). Four hours after the overdose he had a serum paracetamol concentration of 178 mg/l. He was given 50 g of activated charcoal and discharged. Over the next two days he had worsening abdominal pain and was vomiting. He represented 48 hours after the overdose. He was retching profusely, was icteric, and had bilateral subconjunctival haemorrhages. Investigations showed a serum glucose concentration of 2.1 mmol/l, a serum creatinine concentration of 218 µmol/l, an international normalised ratio of 8, and an arterial pH of 7.3. He was started on acetylcysteine before transfer and, despite developing grade 3 encephalopathy, his liver recovered to the extent that his international normalised ratio fell to 2.2. His progress was complicated by oliguric renal failure, lobar pneumonia, and ventricular arrhythmias. He died from septicaemia and persistent multifocal seizures on the sixth day after the overdose.

Case 3 —A 38 year old man presented to his accident and emergency department after ingesting 50 paracetamol tablets (25 g). He had no apparent risk factors for enhanced hepatotoxicity. Four and a quarter hours after the overdose his serum paracetamol concentration was 178 mg/l. He was admitted to a psychiatric ward for further observation. Over the next two days he became increasingly confused and aggressive and was noted to be icteric and had periorbital bruising. Investigations showed a serum glucose concentration of <1 mmol/l, a serum creatinine concentration of 435 µmol/l, an international normalised ratio of 6, and an arterial pH of 7.45. He was electively ventilated and transferred. He continued to deteriorate as a result of pneumonia, fungal sepsis, and increasing intracranial pressure. He died on the 10th day after the overdose.

Case 4 —A 34 year old woman presented to her accident and emergency department after ingesting 30 paracetamol tablets (15 g). She had no known risk factors for enhanced hepatotoxicity. Six hours after the overdose her serum paracetamol concentration was 122 mg/l. After psychiatric review she was discharged but represented two days later with a 24 hour history of malaise, vomiting, and abdominal pain. At that time she was fully alert and orientated but clinically dehydrated. Her pulse was 92 sinus rhythm and she had a blood pressure of 90/60 mm Hg, diffuse abdominal tenderness, an international normalised ratio of 10, and an arterial pH of 7.4. Investigations showed serum concentrations as follows: glucose 2 mmol/l, sodium 130 mmol/l, potassium 4.5 mmol/l, urea 11.9 mmol/l, and creatinine 229 µmol/l. She was transferred but despite full supportive treatment died on the 15th day after the overdose while awaiting liver transplant.


The course and outcome of the four case histories raise questions about the initial management of these patients. Although the patients were medically reviewed within six hours of overdose, acetylcysteine was not given and three patients were discharged with no follow up.

The history of alcohol misuse in case 2 suggests that his serum paracetamol concentration of 178 mg/l four hours after the overdose should have been compared with the high risk line (100 mg/l at four hours) rather than the standard treatment nomogram. Cases 1, 3, and 4 had no known additional risk factors for enhanced hepatotoxicity, and the timing of the overdose in these cases was thought to be accurate, but despite this the patients developed acute liver failure within 48 hours of discharge. The most likely explanation is an error in the timing of the overdoses. If the overdoses in cases 1 and 4 had been taken two hours earlier an error in timing would have given non-toxic concentrations (156 mg/l at four hours and 122 mg/l at six hours) above the treatment line at presentation whereas a one hour error in case 3 would have had the same effect.

Alternatively the rapid onset of acute liver failure might have been the consequence of an overdose before the presenting overdose, or the patients may have been more susceptible to the hepatotoxic effects of paracetamol. Wide intersubject and ethnic differences in the metabolism of paracetamol have been reported: one report showed a 60-fold range in the metabolic activation of paracetamol between subjects, and a threefold variation in glucuronide and sulphate conjugation,18 suggesting a subgroup of patients profoundly more susceptible to the hepatotoxic effects of paracetamol. This has never, however, been proved in the context of an overdose.

Clinical course

The four cases developed signs of progressive liver failure over a 24 hour period before representing with severe and established liver disease. Cases 1 and 3 had increasing confusion, and cases 2 and 4 had abdominal pain and vomiting. If these patients and those accompanying them had been offered clear and written instructions on when to return, treatment of the incipient hepatic failure might have been more effective.

There is no evidence that activated charcoal is an effective treatment four hours after paracetamol overdose, 15 19 and apart from one case of a patient developing a toxic paracetamol concentration after an initial non-toxic concentration20 there is no evidence to recommend serial testing.

Of 42 treatment nomograms received from accident and emergency departments across the United Kingdom all used the 200 mg/l treatment line as recommended by the National Information Service. We are aware of one hospital in the south west which has adopted the 150 mg/l treatment line as a result of a fatality—this nomogram, widely used in the United States, joins 150 mg/l (1.0 mmol/l) at four hours with 30 mg/l (0.2 mmol/l) at 12 hours.12 We recommend that all patients presenting with a serum paracetamol concentration >150 mg/l should be treated with acetylcysteine, with a treatment threshold of 100 mg/l for those patients with known risk factors. The costs of this modification are small compared with the morbidity, the treatment costs of delayed recognition of a patient with acute liver failure, and prevention of a death.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
View Abstract