Pediatric Acetaminophen Overdose: Title and subTitle BreakRisk Factors Associated With Hepatocellular Injury FREE

ACETAMINOPHEN IS widely available and is one of the most commonly used over-the-counter medications. Easy access and general familiarity with acetaminophen contributes to the high incidence of overdose in the pediatric population. In 1997, 75 969 poisoning cases involving acetaminophen were reported to the American Association of Poison Control Centers in patients younger than 19 years.¹ Despite its wide margin of safety in therapeutic doses, the ingestion of single acetaminophen doses exceeding 150 mg/kg in children has long been recognized as having the potential to produce serious hepatocellular injury.² Other studies have expressed concern that repeated "supratherapeutic" doses of acetaminophen may cause significant liver damage in some children.^{3 - 5} An adult study concluded that this accidental misuse with therapeutic intent may result in higher morbidity and mortality than in patients who ingest larger intentional overdoses.⁶ As a result of the various reported scenarios for acetaminophen-associated hepatocellular injury, practitioners and parents face confusion about the likelihood of injury associated with acetaminophen misuse in a therapeutic context in infants and children.

Previous studies of acetaminophen overdose in pediatric patients are limited in that they generally reflect the characteristics of the patients who develop signs or symptoms of toxic effects. To our knowledge, there has not been a comparison between pediatric patients who develop toxic effects from acetaminophen overdose and those who do not, nor have the relative frequencies of toxic effects from intentional overdosing, unintentional ingestion, and overdosing with therapeutic intent been established in children and adolescents. The purposes of this study were to define the demographic and clinical characteristics of the entire spectrum of acetaminophen overdose in a pediatric population and to determine the factors associated with hepatocellular injury.

The study group consisted of all patients who came with the chief complaint of acetaminophen overdose to the emergency department or inpatient units at Children's Mercy Hospital (Kansas City, Mo) and the Children's Hospital and Medical Center (Seattle, Wash) between January 1, 1988, and December 31, 1997. Both hospitals are urban pediatric referral centers, with approximately 10 800 annual admissions in Seattle and 7700 in Kansas City and approximately 20 500 annual emergency department visits in Seattle and 56 000 in Kansas City. Both hospitals began hepatic transplantation programs during the study period, but no patients were referred for this purpose and no liver transplants were performed in the study population.

Patients were identified by medical record review of International Classification of Diseases, Ninth Revision,⁷ diagnostic codes for acetaminophen ingestion (965.4), salicylate ingestion (965.1), and ibuprofen ingestion (965.61). The study protocol was approved by the investigational review boards of the participating institutions and was determined to be exempt from the requirement of informed consent. Patients with acetaminophen ingestion who were referred to the hospital only for phlebotomy and who were treated elsewhere were excluded from analysis. The study design did not exclude patients with other hepatic disease (ie, hepatitis A, hepatitis B); however, no patient (or guardian) reported these as a preexisting condition. To ensure complete case identification, a 20% sample of all medical records coded as ibuprofen or salicylate ingestion was reviewed to identify potential miscoded acetaminophen overdoses.

Data collected included age, sex, race (white, nonwhite), reported amount of acetaminophen ingested (milligrams per kilogram), time elapsed from reported ingestion to presentation at the hospital (hours), reason recorded in the chart for ingestion (unintentional, intentional, single or multiple dosing error with therapeutic intent), coingested substances, methods of gastrointestinal tract decontamination (ipecac, lavage, charcoal, cathartics), use of antidote therapy with N-acetylcysteine, and disposition of the patients (outpatient or inpatient medical or psychiatric care). Symptoms of vomiting or sleepiness noted at initial examination were also recorded. Laboratory data collected included serum acetaminophen concentrations (micrograms per milliliter), peak aspartate aminotransferase level (reference range, 11-31 U/L), peak alanine aminotransferase level (reference range, 4-25 U/L), peak bilirubin level (reference range, 2.9-25.1 µmol/L [0.17-1.47 mg/dL]), and peak prothrombin time (reference range, 10.7-14.2 seconds). "Peak" values for hepatic aminotransferase and bilirubin concentrations in serum were defined as the highest recorded value after acetaminophen ingestion in a given case.

Intentional overdose was defined as an ingestion that was determined to be a willful overdose, as in a suicide attempt or gesture. Unintentional overdose was defined as an accidental ingestion: the unsupervised toddler who ingested it without cognitive knowledge of potential for harm. Dose error with therapeutic intent referred to the ingestion that was taken with the belief that it would have therapeutic benefit although it exceeded the recommended dosage(s).

Hepatocellular injury was defined as an elevation of serum aminotransferase levels greater than 2 times the age-adjusted upper limit of the reference range. Severe hepatocellular injury was defined as aminotransferase levels exceeding 1000 U/L, as well as clinical evidence of altered mental status noted in the chart by the clinician and/or coagulopathy in addition to hepatocellular injury. Coagulopathy was defined as an elevation of prothrombin time to greater than 50% above the upper limits of normal.

The patients with and without hepatocellular injury were compared in relation to demographic factors, reason for ingestion, and time elapsed from ingestion to presentation to the hospital by means of odds ratios (ORs). Reference groups for calculation of OR were chosen to be the subgroup with lowest percentage of hepatocellular injury in the given category. The 95% confidence intervals (CIs) were calculated by the method of Greenland and Robins.⁸ Confidence intervals that include 1 are not considered to be statistically significant at P<.05.

During the 10-year study period, 322 patients (268 children from Kansas City and 54 from Seattle) were examined for acetaminophen overdose. Demographic and clinical characteristics of the patients are summarized in Table 1. The majority represented unintentional ingestions (n=172 [53.4%]), while 140 (43.5%) were intentional ingestions. Ten patients (3.1%) reported acetaminophen overdose resulting from a dosing error. As expected, intent groups differed by age and sex. Patients with intentional ingestion tended to be older (median age, 14 years) and female (78.6%), while those with unintentional ingestions were younger (median age, 2 years) and without significant sex predominance. The median age of the patient group that had acetaminophen overdose consequent to a dosing error was 3.5 years, with a female predominance (80.0%).

The range of dose ingested was very large for all patients (4-8333 mg/kg), but the median dose did not vary significantly between the intentional and unintentional groups (172 vs 150 mg · kg⁻¹ · 24 h⁻¹). Patients with dosing errors had a lower median dose (90 mg · kg⁻¹ · 24 h⁻¹). The majority (94.6%) came for treatment within 24 hours of ingestion; all 158 (100.0%) of the patients with unintentional overdose and 117 (89.3%) of the patients with intentional overdose presented on the day of ingestion. Five patients (71%) in the dosing error group also presented in the first 24 hours. Nearly half (41.3%) of the patients were admitted to the hospital. Of the intentional ingestion group, 111 (79.3%) were admitted: 41 for psychiatric assessment and 92 for medical management as well. Only 19 (11%) of the unintentional ingestion group and 3 (30%) of the dosing error group required hospitalization. One quarter of all patients received N-acetylcysteine: the intentional group predominated, with 65 patients vs 14 of the unintentional group and 3 of the dosing error group.

Hepatocellular injury developed in 27 patients (8.4%), and hepatic failure developed in 4 of these (0.1%). Their clinical and demographic characteristics are summarized in Table 2 and compared with characteristics of those in whom hepatocellular injury did not develop. Severe hepatocellular injury developed in 4 of these patients, and 1 died of multisystem organ failure. No patient underwent liver transplantation, although 2 were accepted for transplant but recovered before receiving an organ. The development of hepatocellular injury was associated with the following: presentation more than 24 hours after ingestion (OR, 335.0; 95% CI, 40.8-275.0), age 10 to 17 years (OR, 36.9; 95% CI, 4.9-275.4), intentional overdose (OR, 37.2; 95% CI, 5.0-278.2), dose greater than 150 mg/kg (OR, 17.9; 95% CI, 2.3-139.2), and white race (OR, 2.8; 95% CI, 1.1-7.2).

In our survey, 10 cases of dosing error (3%) were identified. Three cases represented adolescents self-treating pain, and 1 was an 11-year-old girl who repeatedly self-medicated for headache during a 30-hour period. She estimated ingesting a total acetaminophen dose of 120 mg/kg. A 4-year-old received four 100-mg/kg acetaminophen doses in a 6-hour period for pain from a caregiver. The remaining 5 cases were in patients younger than 3 years. The only patient in whom hepatocellular injury developed was the 11-year-old, who received N-acetylcysteine and had a peak aspartate aminotransferase level of 454 U/L. Severe hepatocellular injury did not develop in any of these 10 patients.

Ten patients (3.1%) were identified with acute ethanol coingestion, and 3 (0.9%) of these also admitted to long-term ethanol use. Ethanol was a component of the preparation of acetaminophen that 4 of the patients ingested. Hepatotoxic effects were seen in 2 of these patients (20%) who coingested ethanol: a 16-year-old who reported long-term ethanol use took 40 g of acetaminophen over 2 hours and presented more than 72 hours after last ingestion. She was treated with N-acetylcysteine and had a peak aspartate aminotransferase level of 5030 U/L. An acutely ethanol-intoxicated adolescent also ingested a single 30-g dose of acetaminophen and presented 12 hours after ingestion. She was treated with N-acetylcysteine therapy and her peak aspartate aminotransferase level was 8192 U/L.

Acetaminophen was first introduced as a nonprescription analgesic and antipyretic in 1960 and has since had a therapeutic profile that reflects widespread safety and efficacy. For more than 2 decades, prompt use of N-acetylcysteine after acetaminophen overdose has averted the potentially fatal hepatic damage for which acute intoxication with this drug is known.⁹ Nonetheless, serious acetaminophen-associated hepatocellular injury still occurs. Literature suggests that in addition to the acute suicidal or accidental intoxication, long-term "supratherapeutic" doses of acetaminophen in patients compromised by illness or ethanol use may also cause serious injury.^{10 - 12}

As recently reviewed by Kearns et al,¹³ therapeutic doses of acetaminophen are primarily metabolized by sulfation and glucuronidation to nontoxic intermediates that are renally excreted. In overdose, cytochromes P450 (CYP3A4, CYP1A2, CYP2E1) produce N-acetyl-p-benzoquinoneimine (NAPQI), which, after glutathione depletion, covalently binds to hepatocytes to produce cellular injury and death. Conditions that reduce glutathione stores or produce induction of the bioactivating P450 cytochromes can potentiate the hepatotoxic effect of acetaminophen. Malnutrition, concomitant illness, long-term ethanol use, genetic predisposition, and coingestion with substances metabolized by the liver have all been proposed as potentiating factors for acetaminophen-associated hepatocellular injury.^{14 - 15} Thus, it is possible that certain patients who receive only moderate repeated overdoses of acetaminophen may be at increased risk for hepatocellular injury.

For this reason, we undertook a broad, detailed examination of pediatric acetaminophen intoxication from a general, urban referral base. In this retrospective study we examined a large and representative series of all patients seeking medical care for acetaminophen overdose at 2 pediatric hospitals. Overall, our study showed a relatively low incidence of hepatocellular injury of 27 cases (8.4%). Of the ingestions reported, 128 cases (39.8%) did not require gastrointestinal tract decontamination or medical therapy because the acetaminophen dose ingested was considered nontoxic (<150 mg/kg). Of the remaining reported cases (n=194), 82 (42.3%) were treated with N-acetylcysteine. All but 1 of the 27 patients with hepatocellular injury were in the older age group (10-17 years); hepatocellular injury developed in a 3-year-old patient who accidentally ingested acetaminophen, 163 mg/kg, and presented 8 hours after ingestion. He was treated with N-acetylcysteine and had a peak aspartate aminotransferase level of 966 U/L. Hepatocellular injury did not develop in any younger patient (0-9 years) with acetaminophen dosing error. The low incidence of toxic effects in the young children supports previous findings by Rumack,¹⁶ who documented that only 3 (1%) of 417 children younger than 5 years developed hepatocellular injury after accidental acetaminophen overdose. A recent series of acetaminophen-associated hepatocellular injury in children younger than 12 years after ingestion of multiple doses given with therapeutic intent supports the assertion that certain patients may have a predisposition for the development of hepatocellular injury. In that series, Heubi et al⁵ reported 21 (45%) of 47 cases in children aged 2 years or younger; most occurred in children who were febrile and acutely malnourished. The case definition of dosing error used in our review was quite broad, including children aged 0 to 17 years, and single acetaminophen overdose as well as multiple if there was therapeutic intent. Our data suggest that acetaminophen dosing error with therapeutic intent resulting in hepatocellular injury is unusual. In contrast to the assertion that repeated dosing errors of acetaminophen occur in febrile, acutely malnourished children,^{5 ,11 - 12} our findings show that they are equally likely to occur in older children who are self-medicating for pain relief.

Late presentation was an important factor in the development of hepatocellular injury in our study. Schiodt et al⁶ reported on 71 hospitalized adults with acetaminophen-induced hepatocellular injury and demonstrated that those with late presentation (>24 hours after ingestion) had more severe liver involvement. Just as intentional ingestions accounted for 89% of the pediatric hepatotoxic cases, suicide attempts or gestures accounted for the majority of the adult patients. However, unlike our pediatric population, a significant proportion of the adult patients (30%) claimed to have taken repetitive smaller overdoses in an attempt to relieve pain, and these cases resulted in higher morbidity and mortality. The median doses ingested in both adult groups were much higher than in our pediatric population (adult accidental overdose cases, 12 g [approximately 170 mg/kg] vs suicidal adult cases, 20 g [approximately 285 mg/kg] vs all pediatric cases, 151 mg/kg). Unlike the pediatric cases, the adult cases were often associated with short-term and long-term ethanol ingestion. Larger doses, long-term ethanol use, and/or occult underlying hepatic disease may partially explain the differences seen in the pattern of injury between adults and children with acetaminophen-associated hepatocellular injury.

Our study has limitations inherent in a retrospective medical record–based survey. First, it relied on information recorded by multiple caregivers and patient self-report and therefore may have contained inaccuracies, such as presence of preexisting disease that could affect hepatic function. Second, common clinical practice is to estimate the dose ingested on the basis of worst-case scenario, which may represent overestimations. This is especially evident in the younger age group when the dose is considered on a per-kilogram basis. Third, given that our analysis reflects cases representative of those with suspected acetaminophen intoxication seen at a specialized tertiary care institution, our data may be skewed in the direction of greater severity compared with a community hospital. Finally, the small number of cases with significant hepatocellular injury during the 10-year review period prohibited our performing a multivariate analysis necessary to critically examine associations between risk factors and control for confounding. Without this type of analysis, it is not possible to identify the independent risk factors. Larger studies will be necessary to answer this important question. However, despite these inherent limitations, our data provide the practitioner with an accurate perspective about the likelihood of hepatocellular injury in the various scenarios of pediatric acetaminophen overdose.

Even when very large amounts of acetaminophen are consumed in pediatric overdose, remarkably few toxic effects are seen, especially in the young child. Pediatric patients do not appear to experience the same pattern of injury seen in adults with acetaminophen intoxication with or without coingestion of ethanol. Therapeutic misadventures producing acetaminophen toxic effects in pediatric patients appear to be unusual. Nonetheless, the potential for harmful misuse of acetaminophen should be included in the practitioner's anticipatory guidance, particularly with the older child who may self-medicate without adult supervision. Finally, the practitioner should be alert for serious sequelae in the suicidal adolescent patient who comes for treatment more than 24 hours after ingestion.

Accepted for publication September 30, 1999.

This study was supported in part by grant 1 U10 HD 31313-06 (Dr Kearns) from the National Institute of Child Health and Human Development, Bethesda, Md.

We thank Steven Simon, PhD, for his statistical assistance and Gary Wasserman, DO, for his toxicology and editing expertise.

Reprints: Sarah W. Alander, MD, Division of Emergency Medicine, Children's Mercy Hospital, 2401 Gillham Rd, Kansas City, MO 64108 (e-mail: salander@cmh.edu).

Editor's Note: According to this study, the major headache involves adolescents who intentionally overdose and/or wait more than a day to seek medical care. Why am I not surprised?—Catherine D. DeAngelis, MD