 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
,
,||,¶,||,¶,||,,||
* Section of Medical Toxicology
Division of Emergency Medicine, The Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania
Poison Control Center, Philadelphia, Pennsylvania
|| Department of Pediatrics, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania
¶ Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONREFERENCES |
|---|
18 years of age and to examine the relative contributions of several risk factors to the occurrence of vomiting. Methods. Data were collected on a prospective cohort of 275 consecutive children who were treated with activated charcoal for acute poisoning exposure. The study was set in the emergency department of an urban, tertiary-care children’s hospital. Sorbitol content of the charcoal was alternately assigned. Potential risk factors for vomiting were recorded prospectively, and the occurrence of vomiting within 2 hours of charcoal administration was measured.
Results. A total of 56 (20.4%) of 275 patients vomited. Median time to vomiting was 10 minutes. Previous vomiting (relative risk: 3.41; 95% CI: 1.48–7.85) and nasogastric tube administration (relative risk: 2.40; 95% CI: 1.13–5.09) were found to be the most significant independent risk factors for vomiting. The increased risk among children >12 years of age, compared with younger children, approached significance. Sorbitol content, large charcoal volumes, or fast administration rates did not increase vomiting risk significantly.
Conclusions. One of every 5 children who are given activated charcoal within our pediatric emergency department vomited. Children with previous vomiting or nasogastric tube administration were at highest risk, and these factors should be accounted for in future investigation of antiemetic strategies. Sorbitol content of charcoal was not a significant risk factor for emesis.
Key Words: charcoal • decontamination • vomiting • sorbitol • poisoning
Abbreviations: ED, emergency department • CI, confidence interval • RR, relative risk
Enteral administration of activated charcoal is the most common method of gastrointestinal decontamination prescribed for poisoned patients within the United States. Charcoal was administered to patients in at least 141 068 of the poison exposure cases reported to poison control centers in 2001.1 Clinical experience suggests that a significant proportion of acutely poisoned patients will vomit after the administration of activated charcoal. Some investigators2–4 and clinicians have begun using antiemetic strategies in an attempt to reduce the incidence of vomiting.
Currently, the incidence of emesis after charcoal administration in the pediatric population is unknown. Minimal knowledge pertaining to the factors that may be associated with the risk of vomiting exists. Identification of a subset of patients with an increased risk of vomiting may permit direction of future clinical use or study of antiemetic therapies. Additional understanding of this process will also aid the clinician in trying to weigh the risk versus benefit of charcoal therapy in the clinical setting.
The primary aim of this study was to determine the incidence of vomiting after administration of activated charcoal to acutely poisoned patients in an urban pediatric emergency department (ED). We also sought to examine the influence of several potential patient-specific, poison-specific, and procedure-specific risk factors that may be associated with the occurrence of vomiting of therapeutic activated charcoal.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Data Analysis
Risk factors for the occurrence of vomiting after administration of activated charcoal were analyzed in a prospective cohort manner. Each cohort was defined by the presence or absence of 1 or more host-, poison-, or procedure-specific risk factors. The outcome of vomiting was defined as the forceful regurgitation of stomach contents, as judged by the patient’s registered nurse, within 2 hours of the initiation of charcoal administration. Two hours was chosen as an upper limit for monitoring for vomiting as previous experience in adults suggested that the majority of patients who vomit after charcoal administration do so within 1 hour.6 In addition, it was believed that vomiting that occurs beyond 2 hours is unlikely to be of clinical importance with regard to prevention of drug absorption.
Data were stored, with patient identifiers removed, within a Microsoft Excel 97 spreadsheet. Data analysis was performed using Stata 6.0 (Stata Corporation, College Station, TX). The incidence of emesis and of several potential risk factors for vomiting within the study population were calculated with construction of 95% confidence intervals (CIs). Descriptive elements of the study were defined through use of mean, median, and range for continuous variables or through use of frequencies for discrete variables. Univariate estimation of factor-related differences in vomiting rates was calculated via the t test for normally distributed continuous variables, the Wilcoxon rank sum test for nonnormally distributed continuous variables, and 
2 or exact tests for discrete variables. Ordinal variables (level of consciousness), categorical variables (degree of agitation and route of charcoal administration), and continuous variables (age, volume of charcoal administered, and rate of charcoal administered) were dichotomized, for reported relative risk (RR) estimates, in a manner indicated by data distribution or perceived clinical utility. Factors that were identified as potentially significant with regard to risk of vomiting were analyzed, to determine their independent contribution to risk, by means of logistic regression.
P .05, or 95% CIs around RR calculations that excluded 1.0 were considered statistically significant for the purposes of this study. RRs detectable by this study with 80% power, assuming that 20% of unexposed patients vomit after charcoal administration, ranged from 1.8 to 2.9 depending on the prevalence of the risk factor in the study sample.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Fifty-six of 275 patients vomited after the initiation of administration of activated charcoal. Therefore, the overall incidence of vomiting in pediatric patients who were treated for an acute poisoning exposure, after activated charcoal, within our ED was 20.4% (95% CI: 15.8%–25.6%). The median time between initiation of charcoal administration and vomiting was 10 minutes (range: <1–120 minutes), with 95% of vomiting episodes occurring within 95 minutes. A total of 55.4% of vomiting patients did so before the entire dose of charcoal had been administered. Age, as represented in our study population, was bimodal, with groups primarily characterized by toddlerhood and adolescence. Age and gender distributions within the study population, with corresponding rates of vomiting, are displayed in Table 2.
|
1.5; P .20). Agitation or crying, depressed consciousness, large charcoal-slurry volumes, or quick charcoal administration times did not increase the risk of vomiting. Administration of activated charcoal with sorbitol did not pose a significantly increased risk of vomiting relative to charcoal without sorbitol. Risk factors such as the previous administration of syrup of ipecac, presence of concurrent illness or preexisting gastrointestinal disease, and co-administration of antiemetic or sedative drugs were too infrequent within the study sample to allow for confident estimation of their association with the risk of emesis.
|
2 analysis), and both were significantly more common among children 6 years of age. Within this age group, tube administration of charcoal demonstrated a RR of vomiting of 2.3 (1.1–4.6) and physical restraint a RR of 2.3 (1.1–4.7). For evaluating their independent contribution to the RR for vomiting of therapeutically administered activated charcoal, the 5 most significant unadjusted risk factors were modeled with logistic regression (Table 4). After multivariable modeling, the occurrence of previous vomiting and naso- or orogastric charcoal administration emerged as the most significant independent risk factors for vomiting, and signs and/or symptoms of toxicity were no longer significant risk factors. The presence of nausea and age >12 years neared statistical significance (P < .08). When entered into the logistic regression model, the use of physical restraint was not found to be an independent risk factor. As a single predictive variable, a history of previous vomiting yielded a positive predictive value of 48.6% (31.9–65.6) but a sensitivity of only 32.1% (20.3–45.9).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Anecdotally, emesis rarely occurs in healthy adult volunteers after enteral administration of activated charcoal. Reports from adult EDs, poorly controlled for potential confounding variables, suggest that emesis occurs after 12.5% to 34.4% of instances of charcoal administration.4,6,12–15 In what is perhaps the only such study previously performed within a pediatric ED, 6 (16%) of 38 children younger than 6 years vomited after being administered activated charcoal.16 During an assessment of the feasibility of home administration of a dose of 12 g of activated charcoal to children, only 8 of 116 children vomited at home17; however, children were able to self-limit administration and the ability to ascertain outcome correctly was questionable.18 We provide the most precise estimation of vomiting among poison-exposed children who were treated in a pediatric ED published to date. In our ED, 56 (20.4%) of 275 children 18 years of age vomited within 2 hours of administration of activated charcoal, with half of the patients vomiting within 10 minutes of the initiation of charcoal administration. Among the subgroup of children 6 years of age, we found, similar to the results of Kornberg and Dolgin,16 that 15.9% vomited (28 of 176).
The American Academy of Clinical Toxicology Position Statement19 on single-dose activated charcoal states that the influences of rate and volume of charcoal administration, ingested toxic substances, and premorbid conditions on the incidence of vomiting are unknown. When surveyed, nurses have offered their perceptions that a previous history of vomiting, a large volume of charcoal, a rapid speed of charcoal administration, a nasogastric tube administration, a rapid removal of a nasogastric tube, or crying and agitation all might predispose to vomiting.5 Previously, one study of 250 adults who were treated with 50 g of activated charcoal for potential poisoning found an incidence of vomiting of 17%, with no differences detected between route or between bolus-dosing versus slow-intermittent dosing.6
This study examined several potential risk factors for their association with vomiting of activated charcoal administered to children for the treatment of potential poisoning within a pediatric ED. The independent RRs for vomiting of 3.4 for previous vomiting and 2.4 for naso- or orogastric tube administration were statistically significant. A previous study found that 6 (29%) of 21 children who were administered charcoal orally vomited, whereas 18 (37%) of 49 who were administered charcoal by gastric tube vomited, but that study may not have had sufficient power to show statistical significance.16 The propensity of young children to vomit charcoal administered by nasogastric tube is an important therapeutic consideration, as pediatric reports of serious morbidity and mortality after pulmonary aspiration of charcoal almost invariably involve tube administration. The presence of nausea and age >12 years approached statistical and clinical significance as risk factors, but this could not be confirmed with this sample. The presence of signs or symptoms of poisoning and the "emetogenic" nature of the toxic substance did not add independent predictive value to the risk of emesis. In a discussion of the potential merits of so-called "superactivated" charcoal with increased adsorptive capability, it has been suggested that smaller volumes of oral activated charcoal might minimize the chance of vomiting.20 This concept is not supported by the study data. Likewise, timing of nasogastric tube removal, drugs that slow gastrointestinal motility, gender, patient position, level of patient agitation, and level of patient consciousness did not seem to influence significantly the rate of vomiting in our study.
In 1993, an uncontrolled comparison of vomiting rates between different studies of charcoal with or without sorbitol suggested that sorbitol seemed to increase the emetogenicity of activated charcoal.21 This comparison and conclusion were reiterated in the American Academy of Clinical Toxicology Position Statement on single-dose activated charcoal.19 However, Minocha et al12 found the incidence of emesis to be similar in these groups: 15.5% after administration of activated charcoal plus sorbitol and 12.5% after plain activated charcoal. In our sample, 23.2% of the patients who received sorbitol vomited, compared with 17.5% of those who were given plain charcoal. The unadjusted, univariate RR for vomiting with sorbitol was 1.32 (95% CI: 0.82–2.13), suggesting that the association is not strong. Similarly, the practice of adding chocolate milk, soda, or fruit juice, as flavoring agents, to charcoal does not seem to increase or decrease the incidence of vomiting.
Previous data in children found that 56% of children who first were administered syrup of ipecac vomited after receiving activated charcoal, in comparison with only 15.9% of children who were not treated with ipecac.16 Among adults, studies have shown that 43% to 55% of patients who received ipecac also vomited after subsequent charcoal administration.13,15 Our study included only 4 patients who had received syrup of ipecac before arrival at the ED, and it is interesting that none of them vomited their charcoal. The clinical importance of syrup of ipecac is questionable in 2003, as the percentage of poisoning exposures treated with syrup of ipecac has fallen from 13.4% in 1983 to only 0.7% in 2001.1 Too few patients in our study had concomitant infections or chronic gastrointestinal illness or antiemetic, paralytic, or sedative administration to allow us to determine their association with emesis after activated charcoal administration. However, their infrequency within our patient population may indicate that they are trivial in regard to the observed clinical phenomenon of vomiting of charcoal within a pediatric ED.
Anecdotal reports support the use of ondansetron to prevent emesis after N-acetylcysteine administration for acetaminophen intoxication,22 and high-dose metoclopramide has been studied retrospectively in this regard.23 Ranitidine plus droperidol was reported to be successful in allowing charcoal administration in 2 patients with vomiting after theophylline overdose.2 Currently, it is believed that many physicians administer antiemetic drugs before activated charcoal in an effort to reduce nausea and vomiting.4 Pilot data suggest that intravenous granisetron may lead to fewer episodes of emesis after administration of activated charcoal with sorbitol,3 yet criteria for antiemetic use in this situation have never been defined. More recently, vomiting was found to occur in 15 (14.2%) of 106 patients who were treated with acupressure before administration of activated charcoal and in 21 (25.9%) of 81 historical control subjects.4 However, between-group comparison of potential confounding factors was impossible.
The time from exposure to outcome is very brief (2 hours) in this cohort study, so that loss to follow-up was not problematic. However, other limitations merit discussion. Because we did not analyze all patients who were treated for acute poisoning within the ED, there may have been selection bias pertaining to those who were chosen to receive activated charcoal. Any such bias should be minimal as this study was performed over 2.5 years, and attending physicians were not involved in patient enrollment. As the same nurses who assign risk factors were responsible for assessing the outcome of vomiting, a potential for information bias exists. This bias should have been minimized by the required recording of risk factors before charcoal administration and by the relatively easily defined outcome of interest. Although we collected data on many demographics and potentially associated factors, the interactions between patients, poisons, and therapeutic regimens are complex. It was impossible to identify and record all of the subtleties of these interactions. Last, some sacrifice of generalizability of the results may have occurred as a price for the internal validity of the study. All patients were recruited from a single tertiary care children’s hospital, and only 1 brand of charcoal was used.
On the basis of our data, future studies that evaluate strategies to reduce the rate of vomiting after therapeutic charcoal should account for history of precedent vomiting and administration by naso- or orogastric tube, which were found to be significantly associated with vomiting. It may also be prudent to attend to the presence of nausea and age >12 years within study populations, as both showed strong inclination toward positive association with vomiting. Sorbitol does not seem to add much to the risk of vomiting of charcoal when compared with charcoal without sorbitol. Likewise, contrary to published speculation, large charcoal volumes and short charcoal administration times were not associated with increased risk.
In comparison with assessments regarding presence of nausea or presence of signs of toxicity, the previous occurrence of vomiting is a tangible assessment that is easily ascertained clinically. Previous vomiting is the strongest single risk factor for vomiting in this study, with an independent RR of 3.4 (1.5–7.9). Using previous vomiting alone as a predictor of vomiting yielded a positive predictive value of 48.6% (31.9%–65.6%); however, a sensitivity of only 32.1% (20.3%–45.9%) suggests that many patients who will vomit with charcoal administration will be missed with this strategy. Results of this study may provide direction for future development of predictive models to guide antiemetic therapy. Knowledge of the factors associated with increased risk of vomiting should enable practitioners to make more informed decisions regarding the benefit/risk ratio of administration of activated charcoal to potentially poisoned children or the adjuvant use of antiemetic pharmacotherapy.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Address correspondence to Kevin C. Osterhoudt, MD, MSCE, Division of Emergency Medicine, Children’s Hospital of Philadelphia, 34th St & Civic Center Blvd, Philadelphia, PA 19104. E-mail: osterhoudtk{at}email.chop.edu
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  Risk of Vomiting After Use of Activated Charcoal Journal Watch Emergency Medicine, June 30, 2004; 2004(630): 6 - 6. [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2004 American Academy of Pediatrics. All rights reserved. |
||
|
||
