- Email: [email protected]
Predictors of outcome after acetaminophen poisoning in children and adolescents
Predictors of outcome after acetaminophen poisoning in children and adolescents Laura P. James, MD, Elayna Wells, Robert H. Beard, MD, and Henry C. Farrar, MD Objective: Shortened courses of N-acetylcysteine may be acceptable in patients with acetaminophen poisoning who are at low risk for toxicity. The goal of this study was to determine which clinical findings best identified patients at lowest risk for acetaminophen-related hepatotoxicity after an acute overdose. Study design: This was a retrospective analysis, throughout 10 years, of hospital admissions for acute acetaminophen poisoning, with inclusion criteria being an acetaminophen concentration above the possible toxicity line by nomogram, arrival within 24 hours, and an initial prothrombin time (PT), aspartate aminotransferase (AST), and alanine aminotransferase (ALT) measured within 24 hours of ingestion. Clinical parameters capable of identifying patients most and least likely to have hepatotoxicity were evaluated by using sensitivity and specificity testing. Results: Of 95 patient charts identified, 41 met all inclusion criteria, with 16 patients having hepatotoxicity. PT, AST, and ALT within the first 24 hours postingestion did not identify all patients who had hepatotoxicity. The best predictor of a low risk of toxicity was the presence of normal values for the PT, AST, or ALT within 48 hours of ingestion. Conclusions: These data suggest that all patients with an acute acetaminophen overdose should be observed and treated for at least 48 hours postingestion. (J Pediatr 2002;140:522-6)
N-acetylcysteine (NAC) is well-recognized as the treatment of choice for acetaminophen overdose, with early treatment after single, acute ingestions reducing the incidence of severe hepatotoxicity to ≤10%.1,2 Three treatment regimens are thought to be equivalent in efficacy: (1) a 20-hour continuous infusion of NAC, (2) a 48-hour regimen
of intermittent intravenous infusions, and (3) a 72-hour regimen of intermittent oral doses.1-3 In a retrospective study, the use of an abbreviated (ie, 2448 hours) oral treatment regimen in a limited number of patients appeared to be effective.4 However, none of these regimens has been compared in a clinical trial. Because oral NAC is the only
From the Division of Pediatric Clincial Pharmacology, University of Arkansas for Medical Sciences, and the Department of Pediatrics, Arkansas Children’s Hospital, Little Rock.
Submitted for publication May 25, 2001; revision received Oct 24, 2001; accepted Dec 20, 2001. Reprint requests: Henry Farrar, MD, Section of Pediatric Clinical Pharmacology, 800 Marshall St, Little Rock, AR 72202-3591. Copyright © 2002, Mosby, Inc. All rights reserved. 0022-3476/2002/$35.00 + 0 9/21/122936 doi:10.1067/mpd.2002.122936
522
approved formulation in the United States, one of the current questions in the management of acetaminophen poisoning is whether all patients receiving oral NAC require a full 72-hour course of treatment.5
See editorial, p 495. The plasma acetaminophen concentrationversus-time nomogram is well-recognized as being clinically useful in determining which patients should receive treatment with NAC after an acute acetaminophen overdose. However, the nomogram overestimates the development of acetaminophen-related hepatotoxicity because 40% of patients whose concentrations are above the line that predicts probable toxicity do not have hepatotoxicity.2,6 Therefore, many patients treated with NAC for acetaminophen poisoning will not be expected to have hepatotoxicity even if untreated. This has caused increased interest in defining which patients after an overdose would be at the least risk of having acetaminophen-related hepatotoxicity. Patients at low risk may be eligible for shortened courses of NAC rather than the full 72-hour oral course that is typically used in the United States.3,5 ALT AST NAC PT
Alanine aminotransferase Aspartate aminotransferase N-acetylcysteine Prothrombin time
The goal of this study was to determine which clinical findings would be most predictive of the absence of the development of hepatotoxicity in patients being treated with NAC after a significant acetaminophen overdose. Clinical
JAMES ET AL
THE JOURNAL OF PEDIATRICS
VOLUME 140, NUMBER 5 Table I. Patient charts included in sensitivity/specificity analysis
Acetaminophen-related hepatotoxicity None Patient charts (n) Age (y)* White race, n (%) Female sex, n (%) Suicide attempts, n (%) Reported acetaminophen dose (mg/kg)* Acetaminophen dose ≥250 mg/kg,† n (%) Time to first NAC treatment* (h) No. with delay in treatment ≥10 h,‡ n (%) No. above probable toxicity nomogram line,‡ n (%) No. with emesis at presentation, n (%) Peak AST (IU/L)* Peak ALT (IU/L)* Peak PT (sec)*
25 15 (1.5-17) 18 (72) 24 (96) 16 (64) 227 (96-543) 11 (44) 7 (2-19.5) 7 (28) 7 (28) 15 (60) 34 (15-69) 26 (17-58) 13.3 (12.0-20.9)
Mild 7 15 (8-16) 5 (71) 4 (57) 5 (71) 390 (146-500) 4 (57) 11 (9-24) 6 (86) 6 (86) 4 (57) 109 (72-507) 179 (28-941) 14.9 (13.9-16.9)
Severe 9 15 (14-17) 6 (67) 5 (55) 5 (55) 324 (190-819) 8 (89) 14 (6-22) 8 (89) 8 (89) 6 (67) 3193 (1194-25,650) 4420 (1257-17,590) 17.3 (15.2-72.6)
*Data reported as median (range). There is no significant difference between patients with and without toxicity unless otherwise noted. †Difference for patients with and without toxicity approached significance (P = .06, Fisher exact test). ‡Significant difference for patients with and without hepatotoxicity (P = .003, Fisher exact test).
parameters that were specifically evaluated were those that would be available early in the clinical course and would be readily available from the patient’s history, physical examination, and laboratory evaluation.
METHODS This study was a retrospective analysis of admissions for acetaminophen poisoning to Arkansas Children’s Hospital, Little Rock, during a 10-year period (1991-2001). This study was submitted to and approved by the local institutional review board and determined to be exempt from obtaining informed consent from patients. A computer search of discharge diagnoses identified medical records of pediatric patients admitted with acetaminophen poisoning. All charts were reviewed with regard to patient demographics, circumstances of the ingestion (eg, suicide attempt), initial acetaminophen concentration, the times of ingestion, presentation and first dose of NAC, emesis at the time
of presentation (not limited to treatment with NAC), and laboratory studies, including prothrombin time (PT), aspartate aminotransferase (AST), and alanine aminotransferase (ALT). The severity of the intoxication was defined by using the peak hepatic transaminase concentration. Severe toxicity was defined as either an AST or ALT level >1000 IU/L, whereas mild toxicity was defined as an AST or ALT level between 100 and 1000 IU/L. Final outcome of the patient was recorded as full recovery, transplantation, or death. Sensitivity, specificity, and predictive value were calculated by using standard equations. For these calculations, the outcome was the development of either mild or severe hepatotoxicity. Predictors of outcome selected for evaluation were the initiation of NAC therapy >10 hours postingestion, initial acetaminophen concentration above the probable toxicity line on the nomogram, emesis at the time of presentation, elevation of either the AST or ALT levels >100 IU/L in the first 24 and 48 hours, and a PT >14 seconds within the first 24 and 48 hours of ingestion. Because of
the period covered by the study, an international normalized ratio was not available in all patient charts. The reported reference range for PT at this institution was 11.5 to 13.8 seconds (mean, 12.6). Thus, a PT of >14 seconds was chosen as a predictor of outcome because it was consistently greater than the reference range for the test and it approximates a ratio of the patient-tocontrol PT of 1.2. Finally, the number of patients with severe hepatotoxicity correctly identified by each predictor or combination of predictors was determined. The charts of all patients identified through the search were reviewed and the data tabulated. Charts were excluded from sensitivity and specificity analysis for the following reasons: (1) the drug concentration on the acetaminophen nomogram was below the possible toxicity line (ie, the patient was admitted for other reasons), (2) the patient arrived >24 hours postingestion, (3) the time of ingestion was unknown, (4) the ingestion was chronic, or (5) no coagulation or transaminase studies were obtained within the first 24 hours 523
JAMES ET AL
THE JOURNAL OF PEDIATRICS MAY 2002
Figure. Changes in the median PT, AST, and ALT levels versus time after a single acute overdose of acetaminophen in pediatric patients who had severe hepatotoxicity.
of admission. Excluded charts in which the patient had severe hepatotoxicity (ie, AST or ALT levels >1000 IU/L) were used to assess the correct identification of patients who ultimately had severe hepatotoxicity to decrease the likelihood of missing a patient with hepatotoxicity.
RESULTS Patient charts (n = 95) were identified by the search and were abstracted. Of these, 41 charts met all inclusion criteria, had no exclusion criteria, and were therefore used in the sensitivity and specificity analysis. Of the 54 charts that were excluded from analysis, the reasons for exclusion were nontoxic ingestion by nomogram (n = 36), delayed presentation >24 hours (n = 4), unknown time of ingestion (n = 5), chronic ingestion (n = 6), and no PT, AST, or ALT obtained within the first 24 hours (n = 6). Of the charts that were excluded, 5 had severe and 4 had mild hepatotoxicity. Of the 5 patients with severe hepatotoxicity, 1 required transplantation and 4 recovered completely, whereas all 4 patients with mild hepatotoxicity had full recovery. Although these charts were not used in sensitivity 524
and specificity testing, the charts of patients with severe hepatotoxicity were reviewed to determine if the predictors correctly identified all patients with severe hepatotoxicity. Among the 41 charts included in the sensitivity and specificity analysis, 16 had hepatotoxicity (severe, 9; mild, 7) and 25 had no hepatotoxicity (Table I). The 3 patients who had hepatotoxicity and the highest values for PT (26.3, 43.8, and 72.6 seconds) received vitamin K ≥2 days after the ingestion. One patient with severe hepatotoxicity required transplantation, and the other 8 recovered completely. All 7 patients with mild hepatotoxicity recovered completely. Table II contains the results of sensitivity and specificity testing for various clinical predictors of outcome. Of particular importance was the inability of laboratory findings within the first 24 hours after ingestion to predict the eventual development of hepatotoxicity, including severe hepatotoxicity. The best predictor in this study of a low risk of hepatotoxicity was normal values for the PT, AST, or ALT within 48 hours of ingestion. An abnormal value for any of these tests was a highly sensitive indicator of pending hepatotoxicity. A variety of other combinations of predictors
were studied with no better results than those shown in Table II. The combination of an acetaminophen level below the probable toxicity line combined with the absence of elevations of the PT, AST, or ALT at 24 hours also predicted a low risk. The high degree of sensitivity but poor specificity for hepatotoxicity of this combination was consistent with the poor specificity of the nomogram. The PT began to rise and peaked before the AST and ALT peak (Figure). This resulted in a better sensitivity and negative predictive value for hepatotoxicity for the PT compared with the AST and ALT within the first 24 hours postingestion. The exclusion from the data of the 3 patients who received vitamin K resulted in a slightly lower median peak PT value (15.9 vs 16.9 seconds) and a slightly earlier peak (36 hours vs 48 hours) but did not otherwise change the PT-versus-time curve.
DISCUSSION There is growing interest in using shortened courses of NAC in the treatment of acetaminophen overdose, especially if it is possible to identify patients at lowest risk for hepatotoxicity who would be most eligible for shortened treatment courses. The acetaminophen concentration-versus-time nomogram provides a conservative estimate of patients who should receive NAC. However, it does not provide an estimation of the potential severity of hepatotoxicity, information that is vital to identifying the low-risk patient. For instance, a patient with “possible toxicity” by the nomogram, may develop “severe hepatotoxicity.” The goal of this study was to determine the most accurate method of identifying children and adolescents who do not have acetaminophen-related hepatotoxicity after an acute overdose. Patients with any hepatotoxicity (severe or mild) were included in the analysis for sensitivity, specificity, and predictive
JAMES ET AL
THE JOURNAL OF PEDIATRICS
VOLUME 140, NUMBER 5 Table II. Prediction of hepatotoxicity (AST or ALT levels >100 IU/L)
Clinical findings at presentation Probable toxicity by nomogram Time to treatment >10 h Emesis at presentation Laboratory findings PT >14 sec within 24 h AST or ALT levels >100 IU/L within 24 h PT >14 sec within 48 h AST or ALT levels >100 IU/L within 48 h Combination of predictors Elevated PT, AST, or ALT levels within 24 h Elevated PT, AST, or ALT levels within 48 h Probable toxicity by nomogram or AST or ALT levels >100 IU/L within 24 h
Sens* (%)
Spec* (%)
Neg PV* (%)
Severe toxicity missed† (No. patients)
88 88 67
72 72 38
90 90 76
3 4 6
75 50 88 81
88 100 84 100
85 76 91 89
5 4 3 None
81 100 100
88 88 72
88 100 100
3 None None
Sens, Sensitivity; Spec, specificity; Neg PV, negative predictive value. *Values were determined for patients with any hepatotoxicity (mild or severe) compared with patients without toxicity. †Number of patients who had severe hepatotoxicity that would have been incorrectly identified as low-risk for toxicity by using that variable. There were 14 charts of 95 patients with acetaminophen poisoning with severe hepatotoxicity.
value. Patients with mild hepatotoxicity might represent near-miss events and would eventually have severe hepatotoxicity if treated with a shortened course of NAC. Also, patients who had severe hepatotoxicity but did not meet all inclusion criteria for sensitivity and predictive value analysis were included when we evaluated the number of cases of severe toxicity that might have been missed. This was done to reduce the possibility of missing any patient with severe toxicity because of the small sample size and retrospective nature of this study. Clinical findings that were evaluated were those that we used in the management of these patients. At 48 hours postingestion, the presence of either an elevated PT, AST, or ALT had a high degree of sensitivity and correctly identified all patients who eventually had severe hepatotoxicity (Table II). Elevations of either the AST or ALT levels >100 IU/L at 24 and 48 hours had a high degree of specificity, because AST and ALT are clinical parameters used to measure liver injury. However, these indicators alone were not sensitive predictors of impending
hepatotoxicity, especially within the first 24 hours of ingestion. Our study suggests that all patients with an acute acetaminophen overdose should be observed for at least 48 hours postingestion to have the greatest potential of identifying and fully treating all patients who are at greatest risk of having severe hepatotoxicity. The actual duration of NAC therapy will vary among patients, depending on the time from ingestion to the initial dose of NAC. Other reports have also noted that some, but not all, patients with hepatotoxicity may have elevations in transaminase values within the first 24 hours.4,7 In the report of Singer et al,7 only 11 of 19 patients who developed hepatotoxicity had elevations of hepatic enzymes within 24 hours of ingestion. In a study by Woo et al,4 5 of the 6 patients who had severe hepatotoxicity had elevations of the AST or ALT levels within the first 24 hours. In our current report, one of 9 patients with severe hepatotoxicity had no elevation and 2 had only minor elevations in the AST or ALT levels at 24 hours (AST
or ALT <100 IU/L). Thus, whereas most patients who had hepatotoxicity demonstrated evidence of this early, there is a small percentage of patients who will not demonstrate any evidence of hepatotoxicity until >24 hours postingestion. A possible argument for longer courses of NAC therapy is that patients who develop hepatotoxicity may benefit from treatment. Keays et al8 reported that patients who had received NAC as a late therapy (mean delay >50 hours after overdose) had decreased rates of cerebral edema, less need for inotropic support, and better survival rates compared with patients who did not receive NAC. Also, the data of Smilkstein et al3 suggest that a 48-hour course of NAC therapy may be preferable to a 20-hour intravenous course in patients with delayed presentation of symptoms. The current study supports the use of shortened courses of NAC (ie, until 48 hours postingestion) in select patients with acetaminophen overdose, depending on the changes in the PT, AST, and ALT levels. However, other prospective 525
JAMES ET AL
THE JOURNAL OF PEDIATRICS MAY 2002
studies are needed in larger numbers of patients to further define the optimal duration of NAC therapy and patient observation. The results of this study could be useful in developing a treatment algorithm that can then be evaluated on a prospective basis. In conclusion, our data suggest that all patients at risk for the development of acetaminophen-related hepatotoxicity should be treated with NAC until at least 48 hours postingestion. Treatment beyond 48 hours postingestion should be individualized on the basis of the patient’s overall clinical status and the laboratory results at that point.
REFERENCES 1. Smilkstein MJ, Knapp GL, Kulig KW, Rumack BH. Efficacy of oral n-acetylcysteine in the treatment of acetaminophen overdose. N Engl J Med 1988;319:1557-62. 2. Anker AL, Smilkstein MJ. Acetaminophen concepts and controversies. Emerg Med Clin North Am 1994;12: 335-49. 3. Smilkstein MJ, Bronstein AC, Linden C, Augenstein WL, Kulig KW, Rumack BH. Acetaminophen overdose: a 48hour intravenous n-acetylcysteine treatment protocol. Ann Emerg Med 1991;20:1058-63. 4. Woo OF, Mueller PD, Olson KR, Anderson IB, Kim SY. Shorter duration of oral N-acetylcysteine therapy for acute
5.
6.
7.
8.
acetaminophen overdose. Ann Emerg Med 2000;35:363-8. Donovan JW. Medical fortune-telling: predicting acetaminophen toxicity. Acad Emerg Med 1999;6:1079-82. Prescott LF. Paracetamol overdosage: pharmacological considerations and clinical management. Drugs 1983;25: 290-314. Singer AJ, Carracio TR, Mofenson HC. The temporal profile of increased transaminase levels in patients with acetaminophen-induced liver dysfunction. Ann Emerg Med 1995;26:49-53. Keays R, Harrison PM, Wendon JA, Forbes A, Gove C, Alexander GJ, et al. Intravenous acetylcysteine in paracetamol-induced fulminant hepatic failure: a prospective controlled trial. BMJ 1991;303:1026-9.
Access to The Journal of Pediatrics Online is reserved for print subscribers! Full-text access to The Journal of Pediatrics Online is available for all print subscribers. To activate your individual online subscription, please visit The Journal of Pediatrics Online, point your browser to http://www.mosby.com/jpeds, follow the prompts to activate online access here, and follow the instructions. To activate your account, you will need your subscriber account number, which you can find on your mailing label (note: the number of digits in your subscriber account number varies from 6 to 10). See the example below in which the subscriber account number has been circled: Sample mailing label
This is your subscription account number
**************************3-DIGIT 001 SJ P1 FEB00 J009 C: 1 1234567-89 U 05/00 Q: 1 J. H. DOE, MD 531 MAIN ST CENTER CITY, NY 10001-001
Personal subscriptions to The Journal of Pediatrics Online are for individual use only and may not be transferred. Use of The Journal of Pediatrics Online is subject to agreement to the terms and conditions as indicated online.
526
N-acetylcysteine (NAC) is well-recognized as the treatment of choice for acetaminophen overdose, with early treatment after single, acute ingestions reducing the incidence of severe hepatotoxicity to ≤10%.1,2 Three treatment regimens are thought to be equivalent in efficacy: (1) a 20-hour continuous infusion of NAC, (2) a 48-hour regimen
of intermittent intravenous infusions, and (3) a 72-hour regimen of intermittent oral doses.1-3 In a retrospective study, the use of an abbreviated (ie, 2448 hours) oral treatment regimen in a limited number of patients appeared to be effective.4 However, none of these regimens has been compared in a clinical trial. Because oral NAC is the only
From the Division of Pediatric Clincial Pharmacology, University of Arkansas for Medical Sciences, and the Department of Pediatrics, Arkansas Children’s Hospital, Little Rock.
Submitted for publication May 25, 2001; revision received Oct 24, 2001; accepted Dec 20, 2001. Reprint requests: Henry Farrar, MD, Section of Pediatric Clinical Pharmacology, 800 Marshall St, Little Rock, AR 72202-3591. Copyright © 2002, Mosby, Inc. All rights reserved. 0022-3476/2002/$35.00 + 0 9/21/122936 doi:10.1067/mpd.2002.122936
522
approved formulation in the United States, one of the current questions in the management of acetaminophen poisoning is whether all patients receiving oral NAC require a full 72-hour course of treatment.5
See editorial, p 495. The plasma acetaminophen concentrationversus-time nomogram is well-recognized as being clinically useful in determining which patients should receive treatment with NAC after an acute acetaminophen overdose. However, the nomogram overestimates the development of acetaminophen-related hepatotoxicity because 40% of patients whose concentrations are above the line that predicts probable toxicity do not have hepatotoxicity.2,6 Therefore, many patients treated with NAC for acetaminophen poisoning will not be expected to have hepatotoxicity even if untreated. This has caused increased interest in defining which patients after an overdose would be at the least risk of having acetaminophen-related hepatotoxicity. Patients at low risk may be eligible for shortened courses of NAC rather than the full 72-hour oral course that is typically used in the United States.3,5 ALT AST NAC PT
Alanine aminotransferase Aspartate aminotransferase N-acetylcysteine Prothrombin time
The goal of this study was to determine which clinical findings would be most predictive of the absence of the development of hepatotoxicity in patients being treated with NAC after a significant acetaminophen overdose. Clinical
JAMES ET AL
THE JOURNAL OF PEDIATRICS
VOLUME 140, NUMBER 5 Table I. Patient charts included in sensitivity/specificity analysis
Acetaminophen-related hepatotoxicity None Patient charts (n) Age (y)* White race, n (%) Female sex, n (%) Suicide attempts, n (%) Reported acetaminophen dose (mg/kg)* Acetaminophen dose ≥250 mg/kg,† n (%) Time to first NAC treatment* (h) No. with delay in treatment ≥10 h,‡ n (%) No. above probable toxicity nomogram line,‡ n (%) No. with emesis at presentation, n (%) Peak AST (IU/L)* Peak ALT (IU/L)* Peak PT (sec)*
25 15 (1.5-17) 18 (72) 24 (96) 16 (64) 227 (96-543) 11 (44) 7 (2-19.5) 7 (28) 7 (28) 15 (60) 34 (15-69) 26 (17-58) 13.3 (12.0-20.9)
Mild 7 15 (8-16) 5 (71) 4 (57) 5 (71) 390 (146-500) 4 (57) 11 (9-24) 6 (86) 6 (86) 4 (57) 109 (72-507) 179 (28-941) 14.9 (13.9-16.9)
Severe 9 15 (14-17) 6 (67) 5 (55) 5 (55) 324 (190-819) 8 (89) 14 (6-22) 8 (89) 8 (89) 6 (67) 3193 (1194-25,650) 4420 (1257-17,590) 17.3 (15.2-72.6)
*Data reported as median (range). There is no significant difference between patients with and without toxicity unless otherwise noted. †Difference for patients with and without toxicity approached significance (P = .06, Fisher exact test). ‡Significant difference for patients with and without hepatotoxicity (P = .003, Fisher exact test).
parameters that were specifically evaluated were those that would be available early in the clinical course and would be readily available from the patient’s history, physical examination, and laboratory evaluation.
METHODS This study was a retrospective analysis of admissions for acetaminophen poisoning to Arkansas Children’s Hospital, Little Rock, during a 10-year period (1991-2001). This study was submitted to and approved by the local institutional review board and determined to be exempt from obtaining informed consent from patients. A computer search of discharge diagnoses identified medical records of pediatric patients admitted with acetaminophen poisoning. All charts were reviewed with regard to patient demographics, circumstances of the ingestion (eg, suicide attempt), initial acetaminophen concentration, the times of ingestion, presentation and first dose of NAC, emesis at the time
of presentation (not limited to treatment with NAC), and laboratory studies, including prothrombin time (PT), aspartate aminotransferase (AST), and alanine aminotransferase (ALT). The severity of the intoxication was defined by using the peak hepatic transaminase concentration. Severe toxicity was defined as either an AST or ALT level >1000 IU/L, whereas mild toxicity was defined as an AST or ALT level between 100 and 1000 IU/L. Final outcome of the patient was recorded as full recovery, transplantation, or death. Sensitivity, specificity, and predictive value were calculated by using standard equations. For these calculations, the outcome was the development of either mild or severe hepatotoxicity. Predictors of outcome selected for evaluation were the initiation of NAC therapy >10 hours postingestion, initial acetaminophen concentration above the probable toxicity line on the nomogram, emesis at the time of presentation, elevation of either the AST or ALT levels >100 IU/L in the first 24 and 48 hours, and a PT >14 seconds within the first 24 and 48 hours of ingestion. Because of
the period covered by the study, an international normalized ratio was not available in all patient charts. The reported reference range for PT at this institution was 11.5 to 13.8 seconds (mean, 12.6). Thus, a PT of >14 seconds was chosen as a predictor of outcome because it was consistently greater than the reference range for the test and it approximates a ratio of the patient-tocontrol PT of 1.2. Finally, the number of patients with severe hepatotoxicity correctly identified by each predictor or combination of predictors was determined. The charts of all patients identified through the search were reviewed and the data tabulated. Charts were excluded from sensitivity and specificity analysis for the following reasons: (1) the drug concentration on the acetaminophen nomogram was below the possible toxicity line (ie, the patient was admitted for other reasons), (2) the patient arrived >24 hours postingestion, (3) the time of ingestion was unknown, (4) the ingestion was chronic, or (5) no coagulation or transaminase studies were obtained within the first 24 hours 523
JAMES ET AL
THE JOURNAL OF PEDIATRICS MAY 2002
Figure. Changes in the median PT, AST, and ALT levels versus time after a single acute overdose of acetaminophen in pediatric patients who had severe hepatotoxicity.
of admission. Excluded charts in which the patient had severe hepatotoxicity (ie, AST or ALT levels >1000 IU/L) were used to assess the correct identification of patients who ultimately had severe hepatotoxicity to decrease the likelihood of missing a patient with hepatotoxicity.
RESULTS Patient charts (n = 95) were identified by the search and were abstracted. Of these, 41 charts met all inclusion criteria, had no exclusion criteria, and were therefore used in the sensitivity and specificity analysis. Of the 54 charts that were excluded from analysis, the reasons for exclusion were nontoxic ingestion by nomogram (n = 36), delayed presentation >24 hours (n = 4), unknown time of ingestion (n = 5), chronic ingestion (n = 6), and no PT, AST, or ALT obtained within the first 24 hours (n = 6). Of the charts that were excluded, 5 had severe and 4 had mild hepatotoxicity. Of the 5 patients with severe hepatotoxicity, 1 required transplantation and 4 recovered completely, whereas all 4 patients with mild hepatotoxicity had full recovery. Although these charts were not used in sensitivity 524
and specificity testing, the charts of patients with severe hepatotoxicity were reviewed to determine if the predictors correctly identified all patients with severe hepatotoxicity. Among the 41 charts included in the sensitivity and specificity analysis, 16 had hepatotoxicity (severe, 9; mild, 7) and 25 had no hepatotoxicity (Table I). The 3 patients who had hepatotoxicity and the highest values for PT (26.3, 43.8, and 72.6 seconds) received vitamin K ≥2 days after the ingestion. One patient with severe hepatotoxicity required transplantation, and the other 8 recovered completely. All 7 patients with mild hepatotoxicity recovered completely. Table II contains the results of sensitivity and specificity testing for various clinical predictors of outcome. Of particular importance was the inability of laboratory findings within the first 24 hours after ingestion to predict the eventual development of hepatotoxicity, including severe hepatotoxicity. The best predictor in this study of a low risk of hepatotoxicity was normal values for the PT, AST, or ALT within 48 hours of ingestion. An abnormal value for any of these tests was a highly sensitive indicator of pending hepatotoxicity. A variety of other combinations of predictors
were studied with no better results than those shown in Table II. The combination of an acetaminophen level below the probable toxicity line combined with the absence of elevations of the PT, AST, or ALT at 24 hours also predicted a low risk. The high degree of sensitivity but poor specificity for hepatotoxicity of this combination was consistent with the poor specificity of the nomogram. The PT began to rise and peaked before the AST and ALT peak (Figure). This resulted in a better sensitivity and negative predictive value for hepatotoxicity for the PT compared with the AST and ALT within the first 24 hours postingestion. The exclusion from the data of the 3 patients who received vitamin K resulted in a slightly lower median peak PT value (15.9 vs 16.9 seconds) and a slightly earlier peak (36 hours vs 48 hours) but did not otherwise change the PT-versus-time curve.
DISCUSSION There is growing interest in using shortened courses of NAC in the treatment of acetaminophen overdose, especially if it is possible to identify patients at lowest risk for hepatotoxicity who would be most eligible for shortened treatment courses. The acetaminophen concentration-versus-time nomogram provides a conservative estimate of patients who should receive NAC. However, it does not provide an estimation of the potential severity of hepatotoxicity, information that is vital to identifying the low-risk patient. For instance, a patient with “possible toxicity” by the nomogram, may develop “severe hepatotoxicity.” The goal of this study was to determine the most accurate method of identifying children and adolescents who do not have acetaminophen-related hepatotoxicity after an acute overdose. Patients with any hepatotoxicity (severe or mild) were included in the analysis for sensitivity, specificity, and predictive
JAMES ET AL
THE JOURNAL OF PEDIATRICS
VOLUME 140, NUMBER 5 Table II. Prediction of hepatotoxicity (AST or ALT levels >100 IU/L)
Clinical findings at presentation Probable toxicity by nomogram Time to treatment >10 h Emesis at presentation Laboratory findings PT >14 sec within 24 h AST or ALT levels >100 IU/L within 24 h PT >14 sec within 48 h AST or ALT levels >100 IU/L within 48 h Combination of predictors Elevated PT, AST, or ALT levels within 24 h Elevated PT, AST, or ALT levels within 48 h Probable toxicity by nomogram or AST or ALT levels >100 IU/L within 24 h
Sens* (%)
Spec* (%)
Neg PV* (%)
Severe toxicity missed† (No. patients)
88 88 67
72 72 38
90 90 76
3 4 6
75 50 88 81
88 100 84 100
85 76 91 89
5 4 3 None
81 100 100
88 88 72
88 100 100
3 None None
Sens, Sensitivity; Spec, specificity; Neg PV, negative predictive value. *Values were determined for patients with any hepatotoxicity (mild or severe) compared with patients without toxicity. †Number of patients who had severe hepatotoxicity that would have been incorrectly identified as low-risk for toxicity by using that variable. There were 14 charts of 95 patients with acetaminophen poisoning with severe hepatotoxicity.
value. Patients with mild hepatotoxicity might represent near-miss events and would eventually have severe hepatotoxicity if treated with a shortened course of NAC. Also, patients who had severe hepatotoxicity but did not meet all inclusion criteria for sensitivity and predictive value analysis were included when we evaluated the number of cases of severe toxicity that might have been missed. This was done to reduce the possibility of missing any patient with severe toxicity because of the small sample size and retrospective nature of this study. Clinical findings that were evaluated were those that we used in the management of these patients. At 48 hours postingestion, the presence of either an elevated PT, AST, or ALT had a high degree of sensitivity and correctly identified all patients who eventually had severe hepatotoxicity (Table II). Elevations of either the AST or ALT levels >100 IU/L at 24 and 48 hours had a high degree of specificity, because AST and ALT are clinical parameters used to measure liver injury. However, these indicators alone were not sensitive predictors of impending
hepatotoxicity, especially within the first 24 hours of ingestion. Our study suggests that all patients with an acute acetaminophen overdose should be observed for at least 48 hours postingestion to have the greatest potential of identifying and fully treating all patients who are at greatest risk of having severe hepatotoxicity. The actual duration of NAC therapy will vary among patients, depending on the time from ingestion to the initial dose of NAC. Other reports have also noted that some, but not all, patients with hepatotoxicity may have elevations in transaminase values within the first 24 hours.4,7 In the report of Singer et al,7 only 11 of 19 patients who developed hepatotoxicity had elevations of hepatic enzymes within 24 hours of ingestion. In a study by Woo et al,4 5 of the 6 patients who had severe hepatotoxicity had elevations of the AST or ALT levels within the first 24 hours. In our current report, one of 9 patients with severe hepatotoxicity had no elevation and 2 had only minor elevations in the AST or ALT levels at 24 hours (AST
or ALT <100 IU/L). Thus, whereas most patients who had hepatotoxicity demonstrated evidence of this early, there is a small percentage of patients who will not demonstrate any evidence of hepatotoxicity until >24 hours postingestion. A possible argument for longer courses of NAC therapy is that patients who develop hepatotoxicity may benefit from treatment. Keays et al8 reported that patients who had received NAC as a late therapy (mean delay >50 hours after overdose) had decreased rates of cerebral edema, less need for inotropic support, and better survival rates compared with patients who did not receive NAC. Also, the data of Smilkstein et al3 suggest that a 48-hour course of NAC therapy may be preferable to a 20-hour intravenous course in patients with delayed presentation of symptoms. The current study supports the use of shortened courses of NAC (ie, until 48 hours postingestion) in select patients with acetaminophen overdose, depending on the changes in the PT, AST, and ALT levels. However, other prospective 525
JAMES ET AL
THE JOURNAL OF PEDIATRICS MAY 2002
studies are needed in larger numbers of patients to further define the optimal duration of NAC therapy and patient observation. The results of this study could be useful in developing a treatment algorithm that can then be evaluated on a prospective basis. In conclusion, our data suggest that all patients at risk for the development of acetaminophen-related hepatotoxicity should be treated with NAC until at least 48 hours postingestion. Treatment beyond 48 hours postingestion should be individualized on the basis of the patient’s overall clinical status and the laboratory results at that point.
REFERENCES 1. Smilkstein MJ, Knapp GL, Kulig KW, Rumack BH. Efficacy of oral n-acetylcysteine in the treatment of acetaminophen overdose. N Engl J Med 1988;319:1557-62. 2. Anker AL, Smilkstein MJ. Acetaminophen concepts and controversies. Emerg Med Clin North Am 1994;12: 335-49. 3. Smilkstein MJ, Bronstein AC, Linden C, Augenstein WL, Kulig KW, Rumack BH. Acetaminophen overdose: a 48hour intravenous n-acetylcysteine treatment protocol. Ann Emerg Med 1991;20:1058-63. 4. Woo OF, Mueller PD, Olson KR, Anderson IB, Kim SY. Shorter duration of oral N-acetylcysteine therapy for acute
5.
6.
7.
8.
acetaminophen overdose. Ann Emerg Med 2000;35:363-8. Donovan JW. Medical fortune-telling: predicting acetaminophen toxicity. Acad Emerg Med 1999;6:1079-82. Prescott LF. Paracetamol overdosage: pharmacological considerations and clinical management. Drugs 1983;25: 290-314. Singer AJ, Carracio TR, Mofenson HC. The temporal profile of increased transaminase levels in patients with acetaminophen-induced liver dysfunction. Ann Emerg Med 1995;26:49-53. Keays R, Harrison PM, Wendon JA, Forbes A, Gove C, Alexander GJ, et al. Intravenous acetylcysteine in paracetamol-induced fulminant hepatic failure: a prospective controlled trial. BMJ 1991;303:1026-9.
Access to The Journal of Pediatrics Online is reserved for print subscribers! Full-text access to The Journal of Pediatrics Online is available for all print subscribers. To activate your individual online subscription, please visit The Journal of Pediatrics Online, point your browser to http://www.mosby.com/jpeds, follow the prompts to activate online access here, and follow the instructions. To activate your account, you will need your subscriber account number, which you can find on your mailing label (note: the number of digits in your subscriber account number varies from 6 to 10). See the example below in which the subscriber account number has been circled: Sample mailing label
This is your subscription account number
**************************3-DIGIT 001 SJ P1 FEB00 J009 C: 1 1234567-89 U 05/00 Q: 1 J. H. DOE, MD 531 MAIN ST CENTER CITY, NY 10001-001
Personal subscriptions to The Journal of Pediatrics Online are for individual use only and may not be transferred. Use of The Journal of Pediatrics Online is subject to agreement to the terms and conditions as indicated online.
526