Salicylate Toxicity Severity Interpreter
Salicylate Toxicity Severity Interpreter
Read a salicylate concentration against the arterial pH, the mental state and the kidney function the way EXTRIP and the AACT guidance do — and see why the Done nomogram, still reproduced everywhere, was validated at a predictive index of 0.42 and should not be used.
Salicylate severity
Level + pH + pictureSalicylate 62 mg/dL, arterial pH 7.40, alert and orientated, no new oxygen requirement, normal kidney function, acute single ingestion
What the inputs are doing, and in what order
It suggests it above 6.5 mmol/L (90 mg/dL), above 5.8 mmol/L (80 mg/dL) with impaired kidney function, or at a blood pH ≤ 7.20.
Stop when there is clinical improvement and the concentration is below 1.4 mmol/L (19 mg/dL).
- why mental status comes first
- the rules above are evaluated in EXTRIP’s order of grading rather than in order of concentration, so altered mental status fires before any threshold is tested. That is deliberate: it is a grade 1D recommendation that applies whatever the level, and it is the only rule that reliably catches chronic poisoning
- mmol/L against mg/dL
- EXTRIP publishes both. The molar figures use salicylate’s molar mass of 138.12 g/mol, so 7.2 mmol/L is 994 mg/L, which is 99.4 mg/dL and rounds to the published 100; and 1.4 mmol/L is 19.3 mg/dL, the published 19. The salicylate unit converter uses the same molar mass, so the two pages agree
- mg/L against mg/dL
- a factor of ten, and both are in routine use. 620 mg/L is 62 mg/dL. Reading one as the other is the commonest error with this assay and it errs in the dangerous direction as often as the safe one
- the acid-base triad
- AACT: respiratory alkalosis alone is mild; a near-normal pH of 7.40 ± 0.05 over a combined respiratory alkalosis and metabolic acidosis is moderate; acidaemia below 7.35 is severe, and the acidaemia itself worsens the course by increasing the un-ionised fraction that enters the brain
- urinary alkalinisation
- a urine pH of 7.5 to 8.0 increases urinary salicylate excretion more than tenfold. It needs adequate glomerular filtration and it fails in hypokalaemia, because the kidney will reabsorb potassium in exchange for hydrogen ions and acidify the urine regardless of the plasma pH
Worked example
Salicylate 62 mg/dL, arterial pH 7.40, alert and orientated, no new oxygen requirement, normal kidney function, acute single ingestion
Mental state alert → the two EXTRIP recommendations that fire on mental status alone do not apply
No new oxygen requirement → the pulmonary oedema recommendation does not apply
62 mg/dL is below 90 and below 100, and kidney function is normal, so no concentration threshold is crossed
pH 7.40 is above 7.20, so the acidaemia suggestion does not apply
Acute exposure, so the chronic rule does not apply — and pH 7.40 is below 7.45, so this is not a pure respiratory alkalosis
62 mg/dL with a pH of 7.40 → the moderate pattern: a normal-looking pH sitting on a respiratory alkalosis and a metabolic acidosis at once
This is the point of the page. A pH of exactly 7.40 in a patient with a salicylate of 62 mg/dL is not a normal blood gas. Check the bicarbonate and the pCO₂: both will be low, and the reader who stops at the pH will have missed both
Change one thing — set the mental state to any altered mental status — and the verdict becomes an EXTRIP grade 1D recommendation for dialysis at the same concentration. The number did not move. That is the relationship between the level and the patient that the Done nomogram cannot represent
The Done nomogram, and why it is not on this page
| What was found | Source | |
|---|---|---|
| How it performed when tested | 55 acute salicylate intoxications. Overall predictive index 0.42 — it classified fewer than half of the cases correctly. Best in the mild category, at 0.79 | Dugandzic et al, Ann Emerg Med 1989 |
| Which way it was wrong | “The nomogram tends to overpredict the severity of intoxication in the moderate and severe categories” — the two categories where the answer changes management | Dugandzic et al, 1989 |
| What the authors concluded | Management “should be based on clinical presentation and good judgment as well as the serum salicylate concentration in relation to the time of ingestion” — not on the nomogram | Dugandzic et al, 1989 |
| What the current guidance says | “The Done nomogram is not reliable for interpretation of concentrations in either acute or chronic toxicity” | AACT/ACMT Guidance Document, Management Priorities in Salicylate Toxicity |
| The structural problem | Its earliest plotted timepoint is six hours after ingestion, although earlier concentrations are advisable and clinically useful. Its axes require a single known ingestion time, which chronic exposure, staggered ingestion and unwitnessed overdose do not provide — and those are most of the patients who are actually in trouble | AACT/ACMT guidance; the nomogram’s own construction |
| What to use instead | The concentration read against the arterial pH, the mental state, the exposure pattern and the kidney function — which is what this page does — plus EXTRIP’s criteria for extracorporeal removal | AACT/ACMT guidance; EXTRIP workgroup |
EXTRIP’s criteria, as published
| Indication | Strength | mmol/L | mg/dL |
|---|---|---|---|
| Severe poisoning, including any patient with altered mental status | Recommended (1D) | any | any |
| New hypoxaemia requiring supplemental oxygen | Recommended (1D) | any | any |
| Failure of standard therapy — supportive measures, bicarbonate | Recommended (1D) | any | any |
| Concentration threshold | Recommended (1D) | > 7.2 | > 100 |
| Concentration threshold, impaired kidney function | Recommended (1D) | > 6.5 | > 90 |
| Concentration threshold | Suggested (2D) | > 6.5 | > 90 |
| Concentration threshold, impaired kidney function | Suggested (2D) | > 5.8 | > 80 |
| Blood pH | Suggested (2D) | pH ≤ 7.20 | pH ≤ 7.20 |
| Stop extracorporeal treatment | Recommended (1D) | < 1.4 and clinical improvement | < 19 and clinical improvement |
Four things that mislead, beyond the nomogram
| Trap | Why it happens | What to do |
|---|---|---|
| A normal pH | Moderate poisoning presents at 7.40 ± 0.05 because a respiratory alkalosis and a metabolic acidosis cancel. The reader who checks the pH and stops sees a normal gas in a patient who is decompensating | Read the bicarbonate and the pCO₂, not the pH. Both low together is the signature. The arterial blood gas interpreter and the Winter’s formula calculator separate the two disorders |
| A concentration that is still climbing | Enteric-coated aspirin and tablet bezoars absorb slowly and erratically. A first concentration taken early can rise for twelve hours or more afterwards | Serial concentrations every two to four hours until they are clearly falling. One reassuring level is not a discharge decision |
| Intubation | Sedation and paralysis abolish the hyperventilation the patient is using to hold their pH up. The AACT guidance reports rapid worsening and increased mortality unless a normal or slightly alkalaemic pH is maintained, and case reports describe arrest within minutes of induction | Avoid it where possible; if unavoidable, maintain hyperventilation and give bicarbonate through the procedure. This is a decision for the most experienced person available |
| A suppressed anion gap | Reported in ACEP Now’s salicylate pitfalls column, not in a peer-reviewed source, and included here as reported there rather than as established: salicylate can interfere with the chloride ion-selective electrode on some analysers, producing a pseudohyperchloraemia that narrows the calculated anion gap | Trust a low bicarbonate over a normal anion gap, and check the chloride before concluding there is no metabolic acidosis. The anion gap calculator shows what the chloride is doing to the result |
A discredited nomogram, a normal pH that is not normal, and EXTRIP
Start with the nomogram, because it is still everywhere and because removing it from a reader’s mental toolkit is the most useful thing this page can do. A. K. Done published it in 1960: plot the salicylate concentration against the hours since ingestion and read off mild, moderate or severe. It was tested properly once, by Dugandzic and colleagues in 1989, against 55 acute salicylate intoxications. Its overall predictive index was 0.42 — it classified fewer than half of the patients correctly — and the errors ran in a specific direction: it overpredicted severity in the moderate and severe categories, which are precisely the categories where a clinician is asking it for help. The AACT and ACMT guidance document states flatly that it “is not reliable for interpretation of concentrations in either acute or chronic toxicity”. On top of that it cannot be read before six hours, and its axes require a single known ingestion time — which the chronic, staggered and unwitnessed presentations that account for most serious salicylate poisoning simply do not have.
What replaced it is not another number but a way of reading the patient, and its centre is the acid-base picture. Salicylate stimulates the respiratory centre directly, so the first thing that happens is hyperventilation and a respiratory alkalosis — the AACT guidance calls that pattern, on its own, mild. Then salicylate uncouples oxidative phosphorylation, anaerobic metabolism raises lactate, and a metabolic acidosis develops underneath the alkalosis. The result is the pattern that gets people into trouble: a pH of 7.40 plus or minus 0.05, which looks like a normal blood gas and is in fact two serious derangements cancelling each other out. The bicarbonate and the pCO₂ will both be low, and the reader who checks only the pH will see nothing. Later, when the respiratory compensation fails, the pH falls below 7.35, and that acidaemia is not merely a marker of severity — it increases the un-ionised fraction of salicylic acid, which is the fraction that crosses into the brain. The same concentration is more dangerous at a lower pH, which is why the pH is an input on this page and not an afterthought.
For the decision that actually matters — whether to remove the drug extracorporeally — the current reference is the EXTRIP workgroup’s 2015 systematic review, not the older textbook thresholds. Its concentration criteria are a recommendation above 7.2 mmol/L, which it publishes as 100 mg/dL, and above 6.5 mmol/L (90 mg/dL) where kidney function is impaired, with weaker suggestions ten milligrams per decilitre below each. But three of its four strong recommendations need no concentration at all: any patient with altered mental status, new hypoxaemia requiring supplemental oxygen, and failure of standard therapy. That is why this interpreter evaluates mental status before it looks at the level, and it is the rule that rescues the patient the nomogram was worst at — the older person on regular aspirin who arrives confused and tachypnoeic with a concentration of 40 mg/dL and a working diagnosis of sepsis. The AACT guidance says the same thing in the same terms: significant central nervous system abnormalities at 40 mg/dL warrant consideration for haemodialysis.
Everything else is supportive and time-critical. Urinary alkalinisation to a urine pH of 7.5 to 8.0 increases urinary salicylate excretion more than tenfold, and it fails in hypokalaemia, because a potassium-depleted kidney will acidify the urine whatever the plasma pH is doing. Absorption is slow and erratic after overdose, especially from enteric-coated tablets, so a single concentration is a snapshot of a moving process and needs repeating every two to four hours until it is clearly falling. And intubation deserves its own warning: sedation and paralysis remove the hyperventilation the patient is relying on, and the AACT guidance associates it with rapid clinical deterioration and increased mortality unless a normal or slightly alkalaemic pH is maintained throughout. For the concentration itself in whichever of the four units your laboratory uses, the salicylate unit converter handles the tenfold mg/L against mg/dL trap; this page is about what to do with the number once you have it, which is never to read it alone.
Frequently asked questions
Why is the Done nomogram no longer used?
Because it was tested and it failed. Dugandzic and colleagues evaluated it against 55 acute salicylate intoxications in 1989 and found an overall predictive index of 0.42 — fewer than half the patients correctly classified — with the errors concentrated in the moderate and severe categories, where it overpredicted severity. The AACT and ACMT guidance document states that it is not reliable for interpretation of concentrations in either acute or chronic toxicity. There are also two structural problems: it cannot be read before six hours post-ingestion, and it needs a single known ingestion time, which chronic, staggered and unwitnessed ingestions do not supply. It is still widely reproduced, which is why it is worth knowing specifically that it should not be used rather than just not reaching for it.
What is the classic acid-base picture in salicylate poisoning?
A mixed respiratory alkalosis and metabolic acidosis. Salicylate stimulates the respiratory centre directly, producing hyperventilation and a respiratory alkalosis first; it then uncouples oxidative phosphorylation, raising lactate and generating a metabolic acidosis underneath. The AACT guidance grades it by what the combination does to the pH: respiratory alkalosis alone is mild, a near-normal pH of 7.40 plus or minus 0.05 with both disorders present is moderate, and acidaemia below 7.35 is severe. The moderate pattern is the dangerous one to read, because a pH inside the reference interval looks like a normal gas. Check the bicarbonate and the pCO₂ — both low together is the signature.
When is haemodialysis indicated in salicylate poisoning?
The EXTRIP workgroup recommends extracorporeal treatment for severe poisoning including any patient with altered mental status, for new hypoxaemia requiring supplemental oxygen, for failure of standard therapy, and for a concentration above 7.2 mmol/L (100 mg/dL) — or above 6.5 mmol/L (90 mg/dL) with impaired kidney function. It suggests it above 6.5 mmol/L (90 mg/dL), above 5.8 mmol/L (80 mg/dL) with impaired kidneys, and at a blood pH at or below 7.20. Intermittent haemodialysis is the preferred modality, and treatment stops when there is clinical improvement and the concentration is below 1.4 mmol/L (19 mg/dL). Three of the four strong recommendations do not involve a concentration at all.
Why can a concentration of 40 mg/dL be more serious than one of 80?
Because of the exposure pattern. In chronic or staggered ingestion the tissue compartment is already loaded and the ratio of tissue to plasma salicylate is high, so a modest plasma concentration can accompany severe central nervous system toxicity — while in an acute overdose a much higher plasma concentration may not yet have distributed. The AACT guidance gives the example directly: a patient with significant central nervous system abnormalities and a concentration of 40 mg/dL warrants consideration for haemodialysis. Chronic salicylate poisoning in an older person is routinely mistaken for sepsis, pneumonia, delirium or heart failure, and the normal-looking level is part of why.
Why is intubating a salicylate-poisoned patient dangerous?
Because the hyperventilation is load-bearing. The patient is holding their pH up by blowing off carbon dioxide, and sedation and paralysis abolish that, causing carbon dioxide retention and a respiratory acidosis on top of the metabolic one. The falling pH then increases the un-ionised fraction of salicylate crossing into the brain, so the poisoning worsens rapidly. The AACT guidance states that intubation and mechanical ventilation can be associated with rapid worsening and increased mortality unless a normal or slightly alkalaemic pH is maintained through hyperventilation and bicarbonate. Where intubation cannot be avoided it needs the most experienced operator available, a plan to hyperventilate immediately, and bicarbonate ready.
How do I convert between the units this assay is reported in?
Salicylate is reported in mg/L, mg/dL, µmol/L and mmol/L. The factor between mg/L and mg/dL is ten — 620 mg/L is 62 mg/dL — and that tenfold confusion is the commonest error with this assay. The molar conversions use salicylic acid’s molar mass of 138.12 g/mol, which is also the mass EXTRIP’s molar thresholds are built on: 7.2 mmol/L is 994 mg/L, or 99.4 mg/dL, which is the 100 mg/dL they publish alongside it. This page takes mg/dL; the salicylate unit converter handles all four.
Related calculators
References
- Juurlink DN, Gosselin S, Kielstein JT, et al; EXTRIP Workgroup. Extracorporeal treatment for salicylate poisoning: systematic review and recommendations from the EXTRIP workgroup. Ann Emerg Med. 2015;66(2):165–181. Extracorporeal treatment recommended above 7.2 mmol/L (100 mg/dL), above 6.5 mmol/L (90 mg/dL) with impaired kidney function, for severe poisoning including any patient with altered mental status, for new hypoxaemia requiring supplemental oxygen and for failure of standard therapy; suggested above 6.5 mmol/L (90 mg/dL), above 5.8 mmol/L (80 mg/dL) with impaired kidney function, and at pH ≤ 7.20; intermittent haemodialysis preferred; cessation when clinical improvement is apparent and the concentration is below 1.4 mmol/L (19 mg/dL).
- Dugandzic RM, Tierney MG, Dickinson GE, Dolan MC, McKnight DR. Evaluation of the validity of the Done nomogram in the management of acute salicylate intoxication. Ann Emerg Med. 1989;18(11):1186–1190. 55 acute salicylate intoxications; overall predictive index 0.42, best in the mild category at 0.79; the nomogram tends to overpredict the severity of intoxication in the moderate and severe categories; management should be based on clinical presentation and good judgment as well as the serum salicylate concentration in relation to the time of ingestion.
- American College of Medical Toxicology / American Academy of Clinical Toxicology. Guidance Document: Management Priorities in Salicylate Toxicity. J Med Toxicol. 2015. The Done nomogram is not reliable for interpretation of concentrations in either acute or chronic toxicity; respiratory alkalosis alone indicates mild toxicity, a pH of 7.40 ± 0.05 over a combined respiratory alkalosis and metabolic acidosis indicates moderate poisoning, and acidaemia below 7.35 is seen in severe poisoning and itself worsens the expected course; urine alkalinisation to pH 7.5–8.0 increases urinary excretion more than tenfold; intubation and mechanical ventilation can be associated with rapid worsening and increased mortality unless a normal or slightly alkalaemic pH is maintained.
- Salicylate Toxicity. StatPearls. NCBI Bookshelf. Mild 40–80 mg/dL, moderate 80–100 mg/dL, severe above 100 mg/dL; direct stimulation of the cerebral medulla causes hyperventilation and respiratory alkalosis, and uncoupling of oxidative phosphorylation with rising lactate produces the metabolic acidosis.
- Done AK. Salicylate intoxication: significance of measurements of salicylate in blood in cases of acute ingestion. Pediatrics. 1960;26(5):800–807. The original nomogram, cited here as the source of a tool this page recommends against using.
Medical Disclaimer: The tools and content provided here are for educational and reference purposes only. They are not intended to substitute for professional medical advice, diagnosis, or treatment. Clinical decisions should always be based on the comprehensive assessment of a qualified healthcare professional.