“This American Life” report on acetaminophen sensationalistic and inaccurate

September 26, 2013, 1:18 pm

The entire hour-long current episode of public radio’s This American Life  is taken up by an investigation into risk of inadvertent accidental acetaminophen toxicity, carried out in association with the investigative public interest group ProPublica. The episode — titled “Use Only As Directed” — makes five major points, as summarized on the ProPublica site:

  1. About 150 Americans die a year by accidentally taking too much acetaminophen, the active ingredient in Tylenol, federal data from the CDC shows.
  2. Acetaminophen has a narrow safety margin — the dose that helps is close to the dose that can cause serious harm, according to the FDA.
  3. The FDA has long been aware of studies showing the risks of acetaminophen. So has the maker of Tylenol, McNeil Consumer HealthCare, a division of Johnson & Johnson.
  4. Over more than 30 years, the FDA has delayed or failed to adopt measures designed to reduce deaths and injuries from acetaminophen. The agency began a comprehensive review to set safety rules fro acetaminophen in the 1970s, but still has not finished.
  5. McNeil, the maker of Tylenol, has taken steps to protect consumers. But over more than three decades the company has repeatedly opposed safety warnings, dosage restrictions and other measures meant to safeguard users of the drug.

The report argues that acetaminophen kills more Americans than another other over-the-counter drug, while admitting that existing data does not exist to allow an accurate estimate of the risks of alternative OTC analgesics such as ibuprofen. The reporter also states that it is “very easy to overdose on acetaminophen”, a claim that seems to me sensationalistic, alarmist,  and inaccurate.

ProPublica has been accused in the past of reporting a “one-sided series of facts arrayed to support a point of view“, and I think this tendency is apparent here. Although “Use Only As Direct” contains interesting and important information, I found a number of major weaknesses in its argument:

  • There is very little effort to put the actual risk of accidental acetaminophen overdose in perspective, given that 50 million Americans take the drug every week.
  • There was almost no attempt to evaluate the risks of nonsteroidal drugs.
  • There is no adequate reference to support the figure of 150 inadvertent acetaminophen deaths per year. I could not find where the FDA claims this figure.
  • Even if the FDA has claimed the 150 deaths per year, this number is not reflected in recent data from the National Poison Data System.
  • The report views the FDA in contradictory ways as it finds convenient: either as a huge inefficient bureaucracy that takes decades to finalize a simple warning label, or as an agency that can thoroughly and accurately evaluate hundreds of cases a year to determine if documented deaths are directly related to minimal acetaminophen overdose.

To be sure, inadvertent therapeutic acetaminophen overdose can be fatal — I am aware of one case in Chicago of a young patient who of acute liver failure died after taking about twice the recommended dose for several days after surgery. Clinicians should be aware of this possibility. But such cases are very rare.

ADDENDUM: (9/26/2013  9:47 pm CST): @ferrisjabr tweeted a link to a sidebar post in which @ProPublica explains how they used various databases to estimate the number of people seriously injured or killed each year by acetaminophen. The entire post is a exercise in reading comprehension, and after going through it several times I still have no idea about their methods. Apparently they combined the information in 3 flawed data sets: the FDA Adverse Events Reporting System, CDC’s Multiple Cause of Death database,  and the American Association of Poison Control Centers National Poison Data System. In the end they conclude: “All we can say is just a fraction of people who take acetaminophen appear to suffer injuries or fatalities as a result”. It agree with that statement, but it is a far cry from the attention-grabbing claim that 150 Americans yearly die of inadvertent acetaminophen overdose.


  1. milkshaken Says:

    one group of problematic patients prone to acetaminophen toxicity is homeless people – my friend, an experienced physician from Europe, did his US residency in ER in a hospital in Phoenix – serving a slum urban area, and they had plenty cases of neglected poor patients, many of them habitual alcohol and drug users. He was furious about fellow residents giving acetaminophen freely to these patients without bothering to check their liver function…

  2. Kevin H Says:

    This is a bizarre perspective. The program was scrupulous in pointing out that the fatal cases were very rare. The issue was lack of guidance to the public that exceeding the recommended dose by a relatively small amount can be fatal.

    Your citation of a case of a young person with a fatal outcome taking twice the recommended dose for several days PROVES as the reporter stated that it is “very easy to overdose on acetaminophen”, a claim that is NOT sensationalistic, alarmist, or inaccurate.

    The FDA advisory panel years ago recommended stronger labeling which was not implemented. Do you disagree?

  3. Leon Says:


    Thank you for your comment. Malnutrition can increase the risk for acetaminophen toxicity, probably by depleting glutathione. However, I’m not sure that checking liver function tests in all these patients is justified as a way to determine whether or not to dispense a short course of the drug.

  4. Leon Says:


    I disagree that TAL was meticulous in communicating how rare these events are. To my ears, the overwhelming impression left by the show was that acetaminophen is an extraordinarily dangerous drug that causes much more morbidity and mortality than nonsteroidals such as ibuprofen. I don’t believe that there is good data to back that up, or that it is even true.

    I do agree that stronger labeling (as is now being initiated) was scandalously slow in coming, and should have been mandated years ago.

  5. Mitch Says:

    Long-time reader, first-time poster.

    I can quickly clear up some of the issues:

    First, the number of acetaminophen overdose deaths each year in the U.S. is closer to 500 (Nourjah et al. Pharmacoepidemiol Drug Saf. 2006). Personally, I’ve seen about 20 patients pass away in the last 2.5 years. It is, hands down, the single greatest cause of acute liver failure in the U.S. (Larson et al. Hepatology. 2005). Even though most patients survive without a liver transplant (~90%), the number of ODs still makes it the most common cause of acute liver failure related deaths in this country. So it’s nothing to sneeze at. However, the fact remains that most patients do survive. Moreover, the antidote N-acetylcysteine is so cheap, so available, and so effective.

    Second, the minimum toxic dose of acetaminophen (about 150 mg/kg for an average person) is about 2.5 to 3x the recommended therapeutic dose. It is true that this is a much narrower therapeutic index than most drugs currently out there. However, it is enough wiggle room for the vast majority of users of the drug to be just fine. The real note of concern for accidental abuse is for people with chronic pain who may be taking opioid co-formulations along with OTC Tylenol. Otherwise, you have no need to be concerned about using acetaminophen. It’s safe enough. And frankly, the standards used by government agencies, as well as by pharmaceutical companies who have billions of dollars riding on the safety and efficacy of their products, to decide what is a safe therapeutic index are quite strict. It’s actually possible that many effective drugs (and pesticides, industrial chemicals, etc.) have been unjustifiably shelved before hitting the market due to unfortunate government regulations. It would be a shame to lose a good analgesic like acetaminophen because a few people overdo it.

    Third, most people who OD on acetaminophen do it intentionally (~50% in the U.S. and ~70% in the U.K.). Warning labels won’t prevent this from happening (they might even encourage it). There is very little evidence that pack size restrictions in the U.K. or the use of blister packs have made any difference. And making acetaminophen illegal won’t prevent people from trying to commit suicide (and with a 90% survival rate, acetaminophen OD isn’t the most effective option anyway).

    So although acetaminophen overdose is a major clinical concern, we shouldn’t try to restrict or outlaw it as some would like. It would probably be a waste of tax dollars and would deprive the millions of people who use it safely and responsibly of an effective and inexpensive way to control some of their pain.

  6. Mitch Says:

    Let me clarify a couple of things real quickly:

    First, when I say the toxic dose is 2.5-3x a therapeutic dose, I’m referring to the 24 h maximum therapeutic dose (4 g in 70 kg person is about 50-60 mg/kg). However, a single therapeutic dose is lower still (about 15 mg/kg), so the toxic dose could actually be said to be as much as 10x higher than a therapeutic dose. Could have been some confusion there.

    Second, note that 150 is the toxic dose, not the lethal dose. A lethal dose is harder to determine in humans, but in mice it is around 500 mg/kg. So the normal therapeutic dose is even further away from the lethal dose.

    Just thought I should be a little more clear on those two points.

  7. Leon Says:


    Thank you for your comments. According to the abstract of the Noujah article, there are about 100 deaths each year from non-intentional acetaminophen overdose ( at least as of 2006). This is an estimate based on several national databases. I tried to retrieve the article to see how the authors arrived at this figure, but unfortunately the Wiley online library was down. The figure still seems high to me, but I’m beginning to think I may be wrong about that.

    I think that there can be extraordinary situations where the lethal dose can be significantly lower than usual. I am aware of a case many years ago in Chicago in which a young woman took an acetaminophen overdose of approximately 220 mg/kg. She was also on daily isoniazid. (There was no evidence of an isoniazid overdose.) She developed severe and ultimately fatal acute liver failure. Isoniazid induces the P450 system, resulting in enhanced production of the toxic metabolite NAPQI.

    As you point out, I think there is considerable confusion in much of the discussion of this topic. Many still claim that acetaminophen is the leading cause of liver failure in this country. It isn’t, That dubious distinction probably belongs to hepatitis B and C. Acetaminophen is, as you note, the leading cause of acute liver failure.

    Some also confuse liver toxicity with lethality. Much of the literature defines liver toxicity as a hepatic transaminase level > 1000 IU/L. Since of acetaminophen toxicity that recover do so completely, that definition to my mind is not worth much. The important patient-oriented outcomes are recovery, death, or transplantation.

    I agree with you that acetaminophen is still a valuable and generally safe drug. Certainly, more can be done to highlight the potential risks in taking more than the recommended dose, or multiple drugs containing acetaminophen. Although I had some problems with the tone of the radio episode, both TAL and @ProPublica are to be commended for raising the issue.

  8. Mitch Says:

    Hi Leon,

    In my first read through your post, I hadn’t noticed the word “accidentally” regarding the number of deaths. My mistake. Sorry about that.

    I agree that there probably are cases in which patients are more susceptible to acetaminophen tox. People always cite alcoholics, though I think this is controversial. Another example is Gilbert’s syndrome patients.

    I also agree with your point that ultimate outcomes need to be emphasized more in the literature. That is something to which I’ve given a lot of thought. The difficulty in the case of acetaminophen though (which is actually a good thing) is that so few patients don’t recover. It’s hard to get large enough cohorts of both survivors and nonsurvivors to draw solid conclusions. So we’re often forced to use “outcomes” like max ALT, bilirubin, etc.

  9. D5F Says:

    It really amazed me that they never once mentioned the number of people taking the drug compared with the number who die. Isn’t this important? If it is ‘number one killing drug’, yet it is the most taken drug by a huge margin, wouldn’t it still be the safest?

  10. Leon Says:


    I’d have to go back and check, but I think they did mention that 50 million Americans take acetaminophen every week. They just didn’t emphasize the point, and the overwhelming impression left by the show, despite caveats, was that acetaminophen is an extraordinarily dangerous drug that should be avoided.

  11. keith Says:

    I’m a fan of TAL, but when I heard this show my conclusion was completely in agreement with Leon’s characterization of sensationalistic, alarmist, and inaccurate. I might have substituted unbalanced for inaccurate, but otherwise spot on to me. I couldn’t really decide if the point of the story was to bash the FDA or raise alarm/educate among listeners. If it was to educate listeners then a major limitation of the piece in my opinion was not pointing out the concept of an analgesic ceiling effect, and stressing that medication many times does not allow for a “if one is good then two is better” thought process.

  12. Leon Says:


    Thanks for the comment. Actually, I agree with you that “unbalanced” might have been a better description than “inaccurate”. The show hit all the bases, and did mention that acetaminophen is still a great drug in many situations, as well as the fact that millions of people use it every day. But I still feel that the main impression left in most listeners’ minds was that acetaminophen is an exceedingly dangerous drug and that it’s easy to overdose on it accidentally.

  13. Kapil Says:

    In response to Mitch’s comment that “there is very little evidence that pack size restrictions in the U.K. or the use of blister packs have made any difference” to suicide attempts or severity. There is actually good evidence suggesting they have made a difference: http://www.bmj.com/content/322/7296/1203 .

  14. Leon Says:


    Absolutely. Thank you for referencing that important BMJ article.

  15. Linus Says:

    Acetominopen(Tylenol) is consumed in combination with opiates: Tylenol+oxycodone=Percocet, Tylenol+hydrocodone=Vicodin, Tylenol+codeine=T3. “Abuse of prescription drugs exceeded the # of heroin, cocaine, inhalants, and hallucinogens combined” (Substance abuse and Mental Health Services Administration, 2010)2009. A lot of medication induced liver damage is due to abuse of Tylenol/opiate combination prescription medications. Why is the Elephant in the room not mentioned? It’s a major part of this problem!!!! These patients’ deaths are not necessarily “intentional suicides”. They are eventual “overdose” death due to a 50 tablet a day Vicodin habit.

  16. Leon Says:


    I think that most of the clinically significant hepatotoxicity from inadvertent acetaminophen exposure is due, not to abusing combination products, but from taking higher than recommended doses to treat pain, or from taking multiple products containing the drug.

    Although overuse and abuse of opiates is a huge problem, the most prominent cause of morbidity and mortality in those cases is opiate overdose and respiratory depression.