Correspondence
Acetaminophen Toxicity in an Urban County Hospital
N Engl J Med 1998; 338:543-545February 19, 1998
- Article
To the Editor:
Untenable case definitions entrapped Schiødt and associates (Oct. 16 issue)1 in their study of acetaminophen toxicity in an urban county hospital. They recognized their predicament but nonetheless trumpeted a conclusion that they knew was questionable. How many million elderly people fear acetaminophen unnecessarily now, and how many have increased their use of gastrotoxic nonsteroidal antiinflammatory drugs (NSAIDs) to compensate for the loss of acetaminophen in their analgesic regimens?
Describing 21 patients hospitalized for excessive acetaminophen intake without suicidal intent, the authors noted that “only five patients were admitted to the hospital because of a specific history of excess acetaminophen ingestion; the majority had abnormal aminotransferase levels or prothrombin times that mandated their admission.” They contrasted these patients with accidental overdoses with 50 patients whose acetaminophen overdoses were considered to represent suicide attempts and who were admitted on behavioral grounds (43 had “histories consistent with such an attempt,” and 7 had “taken large amounts of acetaminophen as an angry gesture”).
Is it any wonder that “patients hospitalized with acetaminophen toxicity related to accidental misuse had higher rates of morbidity and mortality than those who attempted suicide, even though the latter had taken more acetaminophen”? Of course they did: the patients with accidental overdoses became eligible for analysis (were hospitalized) because of morbidity at presentation. With respect to patients with suicidal overdoses, by contrast, the authors took all comers. The comparison of these groups says nothing about the risk or determinants of acetaminophen toxicity. Schiødt and colleagues concede that letting the outcome define the cohort “may explain the high frequency of severe toxic effects in the patients who were admitted.” If they truly harbored these doubts, which are obviously well founded, their conclusions should not have been allowed to stand.
1 ReferencesAlexander M. Walker, M.D., Dr.P.H.
Harvard School of Public Health, Boston, MA 021151
Schiodt FV ,Rochling FA ,Casey DL ,Lee WM . Acetaminophen toxicity in an urban county hospital. N Engl J Med 1997;337:1112-1117
Full Text | Web of Science | Medline
To the Editor:
Schiødt et al. reach the startling conclusion that accidental consumption of even modest amounts of acetaminophen may be more likely to cause severe liver damage and death than an overt suicide attempt. Yet the methods used in the study violate several basic principles of pharmacoepidemiologic research.
Observational studies of drug effects can be even more conceptually demanding than randomized, controlled trials. The authors make the mistake of including patients in the study on the basis of both outcome (hepatotoxicity) and exposure (use of acetaminophen). This is a strange approach for either a case–control or follow-up study and is likely to bias findings toward “discovering” an association between the exposures and outcomes being investigated.
Because the study was conducted long after most events occurred, the authors were not able to collect their own data concurrently or uniformly. Fewer than 10 percent of patients in the accidental-overdose group met the three study criteria for acetaminophen hepatotoxicity, and over half these patients had acetaminophen levels of 10 mg per liter or less. The authors attribute this finding to long-term use of small amounts of acetaminophen. However, it is more likely that in a county-hospital population in which 63 percent had chronic ethanol ingestion, 44 percent had ingested ethanol acutely, 29 percent had taken other medications in addition to acetaminophen, and 19 percent were drug abusers, hepatotoxicity may well have resulted from other causes. It would be unfortunate if these findings caused patients or physicians to avoid appropriate use of acetaminophen, replacing it with NSAIDs, which have more substantial and better-documented toxicity, or forgoing analgesia altogether.
Acetaminophen has been one of the most commonly used drugs in the country for decades. Why does confusion continue to surround such life-and-death questions about its side effects? The Food and Drug Administration, the National Institutes of Health, and the Agency for Health Care Policy and Research have not paid adequate attention to the study of risk–benefit relations of drugs in common use, and it is naive to expect the pharmaceutical industry actively to seek new opportunities to document the toxicity of successful products. This study, one of the most provocative recent papers on drug safety in the United States, was supported by the Danish Medical Research Council and two small private philanthropies. Surely our trillion-dollar-a-year health care system could mount a systematic approach of its own to such pivotal questions.
Jerry Avorn, M.D.
Harvard Medical School, Boston, MA 02115To the Editor:
Acetaminophen hepatotoxicity is generally defined on the basis of an aspartate aminotransferase level above 1000 IU per liter.1,2 Schiødt et al., however, used an aspartate aminotransferase value of more than 3500 IU per liter, which may have significantly biased their sample. In addition, acetylcysteine was not used in 24 percent of patients with unintentional overdoses and 20 percent of patients with intentional overdoses. Acetylcysteine reduces toxic metabolites of acetaminophen, increases oxygen delivery and use, decreases inflammation, and enhances tissue perfusion.2,3 Well-established data demonstrate that these effects decrease morbidity and mortality, even when acetylcysteine therapy is initiated a median of 53 hours after the ingestion of acetaminophen.3
The authors' conclusions should be interpreted with caution. The higher frequency of chronic ethanol abuse among the patients with unintentional overdoses may reflect the lack of defined criteria for chronic ethanol abuse or the inherent selection bias of including patients with alcoholic hepatitis.
3 ReferencesRama B. Rao, M.D.
Robert S. Hoffman, M.D.
New York City Poison Control Center, New York, NY 100161
Rumack BH ,Matthew H . Acetaminophen poisoning and toxicity. Pediatrics 1975;55:871-876
Web of Science | Medline2
Smilkstein MJ ,Knapp GL ,Kulig KW ,Rumack BH . Efficacy of oral N-acetylcysteine in the treatment of acetaminophen overdose: analysis of the national multicenter study (1976 to 1985). N Engl J Med 1988;319:1557-1562
Full Text | Web of Science | Medline3
Keays R ,Harrison PM ,Wendon JA , et al. Intravenous acetylcysteine in paracetamol induced fulminant hepatic failure: a prospective controlled trial. BMJ 1991;303:1026-1029
CrossRef | Web of Science | Medline
To the Editor:
The large doses of acetaminophen taken in the study by Schiødt et al. raise serious doubts as to whether they represented accidental overdoses, but the authors do not address this issue. In a recent series of 38 patients with accidental overdoses, 30 percent were depressed at the time of the overdose, of whom all but 1 had made previous, and often multiple, suicide attempts (unpublished data). It is also well documented that overdose behavior is closely associated with excessive alcohol consumption. Despite these findings an overdose — not necessarily of acetaminophen — by a person with alcoholism is more likely to be assessed as accidental than an overdose by someone who does not drink.1 Although there have been numerous case reports showing that chronic alcohol consumption potentiates the toxic effect of acetaminophen, Schiødt et al., in keeping with the findings of other published series,2,3 were not able to establish a specific synergy.
The finding of an elevation of the serum aminotransferase level is of little clinical relevance as an indicator of the severity or clinical course of acute liver failure after acetaminophen overdose. The prothrombin time, in conjunction with the serum creatinine concentration, the grade of encephalopathy, and the pH, is the single most important means to establish the degree of hepatic failure.4
4 ReferencesAlastair J. Makin, M.D.
Roger Williams, M.D.
University College London, London WC1E 6HX, United KingdomWilliam Bernal, M.R.C.P.
King's College Hospital, London SE5 9RS, United Kingdom1
Suokas J ,Lonnqvist J . Selection of patients who attempted suicide for psychiatric consultation. Acta Psychiatr Scand 1991;83:179-182
CrossRef | Web of Science | Medline2
Makin AJ ,Wendon J ,Williams R . A 7-year experience of severe acetaminophen-induced hepatotoxicity (1987-1993). Gastroenterology 1995;109:1907-1916
CrossRef | Web of Science | Medline3
Whitcomb DC ,Block GD . Association of acetaminophen hepatotoxicity with fasting and ethanol use. JAMA 1994;272:1845-1850
CrossRef | Web of Science | Medline4
O'Grady JG ,Alexander GJ ,Hayllar KM ,Williams R . Early indicators of prognosis in fulminant hepatic failure. Gastroenterology 1989;97:439-445
Web of Science | Medline
Author/Editor Response
The authors reply:
To the Editor: Previous clinical studies of acetaminophen toxicity have been limited to case reports. The aim of our study was to characterize the overall scope of the problem (patients with actual or potential hepatotoxicity). In 60 percent of the patients with suicidal overdoses, no toxic effects occurred, but this factor did not preclude their being hospitalized and incurring substantial costs and risk, given the large doses that many had taken.
If elderly people shy away from taking acetaminophen as a result of our paper, as Walker suggests, they missed the points: all but three patients with accidental overdoses had taken more than the maximum of 4 g of acetaminophen every 24 hours that is suggested on the package label, and in the case of these three patients, substantial alcohol consumption was most likely a cofactor. Limiting daily doses of acetaminophen to 2 g (four extra-strength tablets) is a solution for those with heavy alcohol intake, chronic liver disease, or an extra concern about safety. We do not recommend that patients switch to NSAIDs, although these drugs may be safer for acute pain. All our patients had acute pain, not chronic pain. The Food and Drug Administration has recently proposed a change in the recommendations on the package label concerning the concomitant use of alcohol, making it apply to all over-the-counter analgesics, including acetaminophen and NSAIDs, although the risks of the use of NSAIDs in combination with alcohol are less clearly understood.1
In the accidental-overdose group, the aminotransferase levels were a strong determinant of the diagnosis of hepatotoxicity (52 percent of patients had values above 3500 IU per liter), since such high levels are well recognized as virtually unique to patients taking acetaminophen, though these values do not imply the presence of liver failure. We did, however, require other features, as described, to be certain that our patients had bona fide acetaminophen toxicity. Alcoholic liver disease is never associated with high aminotransferase levels (>500 IU per liter), so there should be no overlap. Viral hepatitis was ruled out by specific tests. The absence of very elevated acetaminophen levels in many patients in the accidental-overdose group is expected; only the ingestion of a large amount at a single time causes high acetaminophen levels. Lower levels are merely confirmatory of acetaminophen ingestion.
We agree with Avorn that prospective case-controlled studies of analgesic use and abuse would be helpful.
Makin and colleagues imply a blurring of the distinction between suicidal overdoses of acetaminophen and accidental overdoses for pain relief; we had no overlap between groups or patients whose motives for ingestion were unclear. Each of the patients with accidental overdoses had a specific cause of pain and denied having suicidal ideation. The ingestions, though of excessive size, did not occur at single time points and thus were more likely to have been related to attempted pain relief. None of our patients were considered to have acute liver failure unless criteria for coagulopathy and altered mental status were met.
1 ReferencesFrank V. Schiødt, M.D.
William M. Lee, M.D.
University of Texas Southwestern Medical Center, Dallas, TX 75235-9151- Citing Articles (3)
Citing Articles
1
Darren G N Craig, Caroline M Bates, Janice S Davidson, Kirsty G Martin, Peter C Hayes, Kenneth J Simpson. (2011) Staggered overdose pattern and delay to hospital presentation are associated with adverse outcomes following paracetamol-induced hepatotoxicity. British Journal of Clinical Pharmacologyno-no
CrossRef2
Darren G. N. Craig, Caroline M. Bates, Janice S. Davidson, Kirsty G. Martin, Peter C. Hayes, Kenneth J. Simpson. (2011) Overdose pattern and outcome in paracetamol-induced acute severe hepatotoxicity. British Journal of Clinical Pharmacology 71:2, 273-282
CrossRef3
G. Randall Bond, Ladonna K. Hite. (1999) Population-based Incidence and Outcome of Acetaminophen Poisoning by Type of Ingestion. Academic Emergency Medicine 6:11, 1115-1120
CrossRef
