Are Acetaminophen Levels Necessary in All Overdose Patients?

Are Acetaminophen Levels Necessary in All Overdose Patients?

pills SS (1)ExpertPeerReviewStamp2x200Intentional overdose patients are notorious for giving inaccurate histories. “I took 14 tablets of this and 8 capsules of that. No, wait. It was 3 tablets of this and a handful of capsules of that… This happened about 2 hours ago. Actually, I think it was last night.” Round and round the merry-go-round we go.

  • How should we risk-assess whether acetaminophen is involved? 
  • If the patient provides no history of acetaminophen ingestion, do we need to order a level?

Acetaminophen poisoning remains the leading cause of liver failure in the United States. Early recognition and treatment of patients are essential to minimize morbidity and mortality. Goldfrank’s Toxicologic Emergencies textbook tells us that:

“The dose history should be used in the assessment of risk only if there is reliable corroboration or direct evidence of validity. Although the amount ingested by history roughly correlates with risk of toxicity and an acetaminophen concentration over the treatment line, historical information is NOT sufficiently reliable in all patients to exclude significant ingestion, particularly in patients with self-harm or drug abuse.” 1,2

Wait! You mean my suicidal patient may not provide an accurate history of what they took, how much, and when? Unfortunately, yes.

Utility of Universal Acetaminophen Screening

  • Sporer et al, retrospectively reviewed 1,820 ED patients who presented with a history of suicidal ingestion or an altered mental status with a strong suspicion of ingestion. They found 3% of patients had measurable acetaminophen levels, and 0.3% had potentially toxic levels not suggested by history. Universal screening in this cohort identified 1 potentially toxic acetaminophen level for every 333 patients screened. 3
  • Lucanie et al, retrospectively reviewed 320 poison center records for suicidal ingestions without a history of acetaminophen exposure. They found 7.2% of patients had measurable acetaminophen levels, and 2.2% had potentially toxic levels not suggested by history. Universal screening in this cohort identified 1 potentially toxic acetaminophen level for every 45 patients screened. 4
  • Ashbourne et al, prospectively studied 365 ED patients with intentional drug ingestion and no history of acetaminophen. They found 1.9% of patients had measurable acetaminophen levels, and 0.2% had potentially toxic levels not suggested by history. Universal screening in this cohort identified 1 potentially toxic acetaminophen level for every 365 patients screened. 5

Patients with suicidal/intentional ingestions who do not confirm an ingestion of acetaminophen MAY have a toxic level in 0.2-2.2% cases. While none of the studies had any patients develop liver failure and few patients were treated with acetylcysteine, that was not the primary endpoint for any of them.

On the Flip Side…

  • Dargan et al, retrospectively reviewed 411 ED patients in the UK. All patients with acetaminophen levels were included. Of 136 patients able to provide a history who denied acetaminophen ingestion, not one had a measurable level. Of note, the authors did find that 3.4% of patients presenting after collapse had elevated acetaminophen levels. They concluded that acetaminophen levels are not recommended universally in patients able to provide a history, but should be obtained in those presenting after collapse. Of note, this cohort included non-suicidal patients who may be more likely to provide an accurate history. 6
  • An Israeli study by Bentur et al, evaluated 154 intentional drug overdose patients admitted to an ED. They found that denial of both acetaminophen and multidrug ingestions after a thorough history taking can be considered reliable for acetaminophen ingestion. They concluded that in facilities with limited resources, these patients may not require routine acetaminophen screening. 7

What Does it All Mean?

  • When the history suggests possible risk or the patient is altered/unresponsive, further assessment using determination of acetaminophen levels should be performed.
  • Some data does not seem to support universal screening for every overdose patient if they can provide a history. However, from a healthcare cost and liability standpoint, preventing one case of liver failure seems worth it.
  • My practice, and my institution’s, is to send acetaminophen levels on all intentional overdose patients.

Image 1

Waring W, Robinson O, Stephen A, Dow M, Pettie J. Does the patient history predict hepatotoxicity after acute paracetamol overdose? QJM. 2008;101(2):121-125. [PubMed]
Nelson L, Lewin N, Howland M Ann, Hoffman R, Goldfrank L, Flomenbaum N. Acetaminophen. In: Goldfrank’s Toxicologic Emergencies, Ninth Edition. 9th ed. Mcgraw-Hill; 2010:483-499.
Sporer K, Khayam-Bashi H. Acetaminophen and salicylate serum levels in patients with suicidal ingestion or altered mental status. Am J Emerg Med. 1996;14(5):443-446. [PubMed]
Lucanie R, Chiang W, Reilly R. Utility of acetaminophen screening in unsuspected suicidal ingestions. Vet Hum Toxicol. 2002;44(3):171-173. [PubMed]
Ashbourne J, Olson K, Khayam-Bashi H. Value of rapid screening for acetaminophen in all patients with intentional drug overdose. Ann Emerg Med. 1989;18(10):1035-1038. [PubMed]
Dargan P, Ladhani S, Jones A. Measuring plasma paracetamol concentrations in all patients with drug overdose or altered consciousness: does it change outcome? Emerg Med J. 2001;18(3):178-182. [PubMed]
Bentur Y, Lurie Y, Tamir A, Keyes D, Basis F. Reliability of history of acetaminophen ingestion in intentional drug overdose patients. Hum Exp Toxicol. 2011;30(1):44-50. [PubMed]

Expert Peer Review

Whether to order a stat acetaminophen (APAP) level in every drug overdose patient has been debated for decades. Unfortunately, there is no definitive answer in the literature, as succinctly summarized by Dr. Bryan Hayes. Granted, it is rare to miss a serious APAP overdose if you can get a good and reliable history --- but that is not always possible.

Consider that:

  • Acetaminophen (APAP) ingestion is common: in our 1989 prospective study it was either present or suspected in nearly 25% of  all overdose patients.
  • APAP overdose can go unrecognized because symptoms early after overdose are absent or nonspecific (nausea, vomiting).
  • There is an effective antidote (N-acetylcysteine) but it needs to be given early to be effective.
  • An inexpensive APAP assay is available in virtually every ED.

With these facts in mind, I think it is prudent to order a stat APAP in any acute intentional drug overdose.

Kent Olson, MD
Medical Director, San Francisco Division California Poison Control System

Bryan D. Hayes, PharmD, FAACT, FASHP

Bryan D. Hayes, PharmD, FAACT, FASHP

Chief Science Officer, ALiEM
Creator and Lead Editor, Capsules series, ALiEMU
Attending Pharmacist, EM and Toxicology, MGH
Assistant Professor of EM, Harvard Medical School
  • Seth Trueger

    Even though the yield is low, it’s a simple, inexpensive test, for a dangerous overdose that has BL syndrome, and we have an effective antidote.

    • Michelle

      Yup, that’s what I was always taught by Karl Sporer (author of one of the papers and previous faculty at our site). Agreed.

  • Derrick Lung

    I think the literature supports checking in patients:
    1) altered mental status or evidence of active toxicity
    2) any analgesic overdose — patients are well-documented to have poor understanding of the difference (and names) of non-opioid analgesics and poor understanding of the acetaminophen content in opioid-combination analgesics

    I would personally include patients:
    3) with true lethal, suicidal intent

    So, thus, I occasionally OMIT checking acetaminophen levels in the non-suicidal, non-analgesic, recreational drug overdose. This is actually more than occasionally since my patient population includes a large number of recreational drug users, so I actually violate my rule #1 quite often.

    • Michelle

      Ditto I follow this same practice — even down to the occasional violation of rule #1. Glad we’re on the same page.

  • Thomas Dalton

    Another factor to consider is getting suicidal patient dispositioned, especially in a timely manner. At the two different institutions I work at even if my suspicion for an acetaminophen ingestion is low often times psychiatry likes to see a negative or non-life threatening level prior to admission to the psychiatry service or transfer of the patient to a psychiatric facility.

    Might not be the case everywhere but should be something to consider.

    • Bryan D. Hayes

      Thank you Thomas for this great point. Our hands are often tied as well to facilitate disposition.

  • Bruce Anderson

    Nice summary of the literature, Dr. Hayes! My impressions:

    There are not many studies that specifically look at this particular clinical problem. Of the available literature, the single largest study involves assessment of < 2000 patients. The rest of the studies involve a couple hundred patients at most. Those are VERY small studies (and each study use different assessment criteria) from which to attempt to come up with reasonable conclusions for universal screening. How do these studies compare to other preventative clinical assessments??? Example: Mamography for prevention of breast cancer: some studies have documented screening of 414 women for 7 years to prevent one cancer death (

    There are multiple studies looking to answer the question of how many mamography screens are needed in a specific population to prevent one death … and those studies involve thousands of patients that are followed for many years. We simply don't have that type of data available for acetaminophen overdose.

    So, what is one to do in situations where there is NOT a great deal of data? A reasonable question to ask is, what's the cost of doing an APAP level vs. the impact of NOT obtaining an acetaminophen blood level in suicidal patients or those with intentional exposures? How would you feel if you missed one case of APAP poisoning that you know could have been treated?

    Obtaining an APAP blood level involves very low risk to the patient and costs very little. It seems the risk-benefit ratio favors universal APAP screening.

    • Bryan D. Hayes

      Bruce, this is a great perspective AND the reason why I do recommend an acetaminophen level for all intentional overdose patients. It all depends on your definition of low yield. To me, preventing one case of liver failure seems pretty cost-effective even if screening thousands of patients to find one toxic level (but that hasn’t been officially studied).

  • So now we have proven (sort of) that we can’t trust patients on what they took. Then do you trust them on when they took it? Where do you plot them on the nomogram, or do you also treat them?

    Slightly tongue in cheek, certainly, but at what point in the believability scale do you stop?

    • Bryan D. Hayes

      Great questions. Acetaminophen risk assessment is very challenging. I wrote a UMEM pearl on this a while back. It may be helpful here as well:

      If the level is zero and liver function tests are normal, you’re done. If the level is detectable, you really need to try and pin down time of ingestion as much as you can to plot it on the nomogram. If there is any question or doubt, it’s best to start treatment. Newer data suggests a tailored approach to acetylcysteine treatment, so you can always stop it early if laboratory results remain normal.

  • Justin

    Doesn’t the UK (not sure about Israel) have restrictive policies regarding acetaminophen purchase? If so, I would suspect that the Dargan and Bentur studies lack generalizability. That being said, I agree with Dr. Lin. Also, does anyone know what the approximate cost of an APAP level is?

    • Bryan D. Hayes

      Justin, thanks for your comments. The Dargan study was published before the changes to labeling in the UK. Acetaminophen level costs are institution-specific, and we must different cost from the actual charge to the patient. Most ‘cost’ less than $40 to send.

      • Justin

        Got. Thanks for the great article.