etvtrssmsccscsdraoirsrbaspt

RA

The Journal of Emergency Medicine, Vol. xx, No. x, pp. xxx, 2009Copyright © 2009 Published by Elsevier Inc.

Printed in the USA0736-4679/09 $–see front matter

ARTICLE IN PRESS

doi:10.1016/j.jemermed.2008.06.015

ClinicalReviews

THE MANAGEMENT OF CHILDREN WITH GASTROENTERITIS ANDDEHYDRATION IN THE EMERGENCY DEPARTMENT

James E. Colletti, MD,* Kathleen M. Brown, MD,† Ghazala Q. Sharieff, MD,‡ Isabel A. Barata, MD,§ andPaul Ishimine, MD;� for the ACEP Pediatric Emergency Medicine Committee

*Department of Emergency Medicine, Mayo Clinic College of Medicine, Rochester, Minnesota, †Department of Emergency Medicine,Children’s National Medical Center, Washington, DC, ‡Division of Pediatric Emergency Medicine, Rady Children’s Hospital and HealthCenter/University of California San Diego, San Diego, California, §Department of Pediatrics, New York School of Medicine, New York,New York, and �Division of Pediatrics and Medicine, University of California, San Diego/Rady Children’s Hospital and Health Center,

San Diego, CaliforniaReprint Address: James E. Colletti, MD, Department of Emergency Medicine, Mayo Clinic College of Medicine, 200 First Street SW,

Rochester, MN 55905

itsihvi

eh

AbnGd(pdidp

Abstract—Background: Acute gastroenteritis is charac-erized by diarrhea, which may be accompanied by nausea,omiting, fever, and abdominal pain. Objective: To reviewhe evidence on the assessment of dehydration, methods ofehydration, and the utility of antiemetics in the child pre-enting with acute gastroenteritis. Discussion: The evidenceuggests that the three most useful predictors of 5% orore dehydration are abnormal capillary refill, abnormal

kin turgor, and abnormal respiratory pattern. Studies areonflicting on whether blood urea nitrogen (BUN) or BUN/reatinine ratio correlates with dehydration, but severaltudies found that low serum bicarbonate combined withertain clinical parameters predicts dehydration. In mosttudies, oral or nasogastric rehydration with an oral rehy-ration solution was equally efficacious as intravenous (i.v.)ehydration. Many experts discourage the routine use ofntiemetics in young children. However, children receivingndensetron are less likely to vomit, have greater oralntake, and are less likely to be treated by intravenousehydration. Mean length of Emergency Department (ED)tay is also less, and very few serious side effects have beeneported. Conclusions: In the ED, dehydration is evaluatedy synthesizing the historical and physical examination,nd obtaining laboratory data points in select patients. Noingle laboratory value has been found to be accurate inredicting the degree of dehydration and this is not rou-inely recommended. The evidence suggests that the major-

ECEIVED: 13 January 2008; FINAL SUBMISSION RECEIVED:

CCEPTED: 4 June 2008

1

ty of children with mild to moderate dehydration can bereated successfully with oral rehydration therapy. Ondan-etron (orally or intravenously) may be effective in decreas-ng the rate of vomiting, improving the success rate of oralydration, preventing the need for i.v. hydration, and pre-enting the need for hospital admission in those receiving.v. hydration. © 2009 Published by Elsevier Inc.

Keywords—pediatric; dehydration; gastroenteritis; re-ydration; antiemetics

INTRODUCTION

cute gastroenteritis (AGE) is an illness characterizedy acute diarrhea, which may or may not be accompa-ied by nausea, vomiting, fever, and abdominal pain.astroenteritis is a relatively common diagnosis for pe-iatric patients presenting to the Emergency DepartmentED). Worldwide, diarrhea causes 1.4–2.5 million deathser year (1,2). In children under the age of 5 years,iarrhea is responsible for as many as 150,000 hospital-zations (approximately 10% of hospitalizations in chil-ren between 1 and 5 years of age) and 3.7 millionhysician visits annually in the United States (3,4). Es-

y 2008;

12 Ma

ta

adftsctgHsmniltddmdaamc

AtutadLtl“iecitdyasea

ic

ldEgtstins(2ea

AS

CqwahTwqbhoffhhm

iifcvhptdsott

b

2 J. E. Colletti et al.

ARTICLE IN PRESS

imated costs of caring for these patients in the hospitalnd outpatient setting exceed 2 billion dollars annually.

A common reason for hospitalization in children withcute gastroenteritis is greater degrees of severity ofehydration or mild dehydration accompanied by socialactors. It is important to differentiate the child with gas-roenteritis and dehydration from the child with a moreinister cause of vomiting such as an intra-abdominalatastrophe, an inborn error of metabolism, pyelonephri-is, or diabetic ketoacidosis. The presence of diarrhea isenerally a reassuring historical finding in these patients.owever, diarrhea does not exclude other conditions

uch as appendicitis. Patients with acute gastroenteritisay have abdominal pain and diffuse abdominal tender-

ess as well as increased bowel sounds. Other exam-nation findings such as distention, peritoneal signs, orocalized tenderness are rare in children with gastroen-eritis and should lead the clinician to consider otheriagnoses. After the diagnosis of gastroenteritis has beenetermined, the emergency physician must make severalanagement decisions such as whether to obtain laboratory

ata, how best to rehydrate the child, and if an antiemeticgent should be administered. This article will discuss thevailable evidence regarding the assessment of dehydration,ethods of rehydration, and the utility of antiemetics in the

hild presenting with acute gastroenteritis.

METHODS

literature review was conducted for all published ar-icles relevant to the assessment of dehydration, thetility of laboratory studies in the evaluation of dehydra-ion, oral rehydration therapy (ORT), and the use ofntiemetics in pediatric medicine. The search was con-ucted over a 40-year period from 2006 back to 1966.imits used in the PubMed search included all children 0

o 18 years of age and articles written in the Englishanguage. Search terms that were used were “pediatric,”dehydration,” “oral rehydration therapy,” “gastroenter-tis,” and “antiemetics.” Further, the reference list forach article was examined and additional relevant arti-les were reviewed. Articles and abstracts were includedf they concerned a human study that met one or more ofhe following criteria: assessment of dehydration, rehy-ration therapy, the use of antiemetics in patients � 16ears of age diagnosed with acute gastroenteritis. Therticles and abstracts were reviewed for patient age, inclu-ion and exclusion criteria, method of rehydration, anti-metics used, adverse reaction to antiemetics, and hospitaldmission rates in children administered antiemetics.

When the PubMed search for the three sub-topicsncluded in this manuscript was performed using the

ombinations of the search terms listed above, the fol- a

owing numbers of articles were found: assessing dehy-ration (n � 299), oral vs. intravenous rehydration in theD (n � 352), and the use of antiemetics in children withastroenteritis (n � 118). Articles and abstracts werehen considered based on applicability to one or moreub-topics. Articles deemed by the authors to be relevanto the use of antiemetics in children were appraised fornclusion based on relevance to the topic. The followingumbers of investigations were ultimately included: as-essing dehydration and the role of laboratory studiesn � 30), oral vs. intravenous rehydration in the ED (n �4), and the use of antiemetics in children with gastro-nteritis (n � 15). A hierarchy based on quality was notssigned.

DISCUSSION

ssessing Dehydration and the Role of Laboratorytudies

linical assessment of the degree of dehydration doneuickly and accurately in infants and young childrenith gastroenteritis often determines patient treatment

nd disposition. Certain clinical signs and symptoms canelp quantify the degree of dehydration (Table 1) (5,6).he clinical history in a child who presents to the EDith vomiting or diarrhea should assess the onset, fre-uency, quantity, and character (i.e., the presence of bile,lood, or mucous) of the vomiting and diarrhea. Theistory should also include weight before illness, recentral intake (including breast milk and other fluids andood), urine output, and associated symptoms, includingever or changes in mental status. The past medicalistory should identify underlying medical problems,istory of other recent infections, medications, and hu-an immunodeficiency virus status.Porter et al. determined that parental report of histor-

cal details and physical signs (emesis, diarrhea, fluidntake, urine output, ill appearance, weak cry, sunkenontanelle, sunken eyes, decreased tears, dry mouth, andool extremities) related to dehydration have a predictivealue for abnormal physiologic states and the outcome ofospital admission (8). The authors found parental re-orted data to have a greater sensitivity (range 73–100%)han specificity (range 0–49%) for the prediction ofehydration of 5% or greater. Further, the likelihood ofignificant dehydration was decreased by a historyf normal fluid intake, normal urine output, and normalearing state, whereas a sunken fontanelle and decreasedear production were associated with hospital admission.

The physical examination should include an accurateody weight, temperature, heart rate, respiratory rate,

nd blood pressure. The gold standard for diagnosis of

dpamteaTtaPp(ssmg“To(ea

tsrttcrp

tf

fcttdcaehwiogl

ofrdbfiesdsussf

T

MTH

QETMBSCEU

AaAGA C. Mam 3 Nov

Pediatric Gastroenteritis and Dehydration 3

ARTICLE IN PRESS

ehydration is measurement of acute weight loss. Aatient’s true pre-illness weight is rarely known in thecute care setting; as such, an estimate of fluid deficit isade based on clinical assessment. Thus, historically,

he measurement of dehydration has been based on sev-ral clinical variables scaled into three categories. Therere two classification scales for estimating fluid deficit.he first is derived from the 1992 recommendations of

he Centers for Disease Control and Prevention (CDC)nd the 1996 guidelines of the American Academy ofediatrics (AAP) (5,9). In this classification system, aatient’s degree of dehydration is subdivided into mild3–5% fluid deficit), moderate (6–9% fluid deficit), andevere dehydration (� 10% fluid deficit) (5,10). Theecond classification system was defined by the 1995anual by the World Health Organization and the 2001

uidelines of the European Society of Paediatrics titled,Gastroenterology, Hepatology and Nutrition” (6,11).his classification system divides patients into no signsf dehydration (� 5 %), some signs of dehydration5–10%), and severe dehydration (� 10%) (6,11). Thisstimate is employed to determine the need for therapynd the type of therapy to be administered (6,7,11).

During the physical examination, the general condi-ion of the patient should be assessed. Specifically, ithould be noted if the child is listless, apathetic, or lesseactive. Other important examination findings includehe appearance of the eyes, including the degree to whichhey are sunken, the presence or absence of tears, and theondition of the lips and mouth. The rate and quality ofespirations should be noted as they can be a clue to the

able 1. Symptoms Associated with Dehydration

SymptomMinimal or No Dehydration

(� 3% Loss of Body Weight)

ental status Well; alert Norhirst Drinks normally; might refuse liquids Thieart rate Normal Nor

uality of pulses Normal Noryes Normal Sligears Present Decouth and tongue Moist Dryreathing Normal Norkin fold Instant recoil Recapillary refill Normal Proxtremities Warm Coorine output Normal to decreased Dec

dapted from Ref (5): Duggan C, Santosham M, Glass RI. The mand nutritional therapy. MMWR Recomm Rep 1992;41(No. RR-1dapted from Ref (6): World Health Organization. The treatmenteneva, Switzerland: World Health Organization, 1995. Availabldapted from Ref (7): King CK, Glass R, Bresee JS, Dugganaintenance, and nutritional therapy. MMWR Recomm Rep 200

resence of metabolic acidosis. Examination of the ex- f

remities should include an estimation of systemic per-usion and capillary refill time (12).

Dehydration assessment conventionally has been per-ormed by evaluating clinical signs thought to be asso-iated with dehydration, such as general appearance,achycardia, abnormal radial pulse, sunken eyes, absentears, dry mucous membranes, abnormal respirations,ecreased skin elasticity, capillary refill � 2 s, and de-reased urine output. It is important to note that thessessment of dehydration, by using such clinical param-ters, is quite inaccurate and usually overestimates de-ydration when tested against the standard of acuteeight gain with rehydration (13). When studied, these

ndividual findings have not been shown to be sensitiver specific predictors of the degree of dehydration. Ineneral, individual findings in the studies generally had aow sensitivity and high specificity (14).

Gorelick et al. conducted a trial in which clinical signsf dehydration were assessed for their predictive valueor dehydration in children by using weight gain afterehydration as a standard (14). They concluded thatiagnosis of clinically important dehydration should beased on the presence of at least three of 10 clinicalndings: decreased skin elasticity, capillary refill, gen-ral appearance, tears, respirations, mucous membranes,unken eyes, radial pulse, tachycardia, urine output. In-ividual findings generally had a low sensitivity and highpecificity (an exception was parental report of decreasedrine output, which was found to be sensitive but notpecific). The presence of any three or more of theseigns had a sensitivity of 87% and a specificity of 82%or detecting a deficit of 5% or more. A subset of four

Moderate DehydrationLoss of Body Weight)

Severe Dehydration (� 9% Lossof Body Weight)

tigued or restless, irritable Apathetic, lethargic, unconsciousger to drink Drinks poorly; unable to drinkincreased Tachycardia, with bradycardia in

most severe casesdecreased Weak, thready, or impalpablenken Deeply sunken

AbsentParched

st Deep2 s Recoil in � 2 s

Prolonged; minimalCold; mottled; cyanoticMinimal

ent of acute diarrhea in children: oral rehydration, maintenance,0.rhoea: a manual for physicians and other senior health workers.tp://whqlibdoc.who.int/publications/2005/9241593180.pdf.naging acute gastroenteritis among children: oral rehydration,21;52(RR-16):1-16.

Mild to(3–9%

mal, farsty; eamal to

mal tohtly sureased

mal; faoil in �longedlreased

nagem6):1–2of diare at ht

actors (capillary refill � 2 s, absent tears, dry mucous

mdala

dwWspghh

pstNbpri92bfiHotst(

stdd9dcdhdOopmnfo

mi

(g

dnraadsrvuflrpfeTwdhbcdpad

ldbecwacbicnAsc(ldVbstse

4 J. E. Colletti et al.

ARTICLE IN PRESS

embranes, and ill general appearance) predicted dehy-ration as well as the entire set, whereas the presence ofny two or more of these signs indicated a deficit of ateast 5%. Interobserver reliability was good to excellent forll but one of the findings studied (quality of respirations).

Friedman and colleagues prospectively evaluated chil-ren � 3 years of age with gastroenteritis (12). Childrenere assessed for dehydration based on 12 clinical signs.eight changes from pre- to post-rehydration were mea-

ured and used as the gold standard to measure theercentage of dehydration. The authors concluded thateneral appearance, eyes, mucous membranes, and tearsad the most significant measurement properties for de-ydration (12).

Steiner et al. systematically reviewed 13 studies (1246atients) to determine the precision and accuracy ofigns, symptoms, and laboratory testing for the evalua-ion of dehydration of children � 5 years of age (15).one of the investigation-fulfilled criteria were found toe of high methodologic value. The three most usefulredictors of 5% dehydration were abnormal capillaryefill time (likelihood ratio [LR] 4.1; 95% confidencenterval [CI] 1.7–9.8), abnormal skin turgor (LR 2.5;5% CI 1.5–4.2), and abnormal respiratory pattern (LR.0; 95% CI 1.5–2.7). Diagnostic accuracy was improvedy the use of clinical score systems or the combination ofndings. The authors advocate for use of the 1995 Worldealth Organization (WHO) and 2001 European Societyf Paediatrics Gastroenterology, Hepatology and Nutri-ion guidelines that divide patients into categories of noigns of dehydration (� 3–5%), some signs of dehydra-ion (5–10%), and severe dehydration (� 10%) (Table 1)6,11).

Trials that have evaluated the correlation of clinicaligns of dehydration with post-rehydration weight reporthat the first signs of dehydration are evident at 3–4%ehydration, with increasing clinical signs noted at 5%ehydration and signs of severe dehydration noticed at–10% dehydration (7,13,15). It is especially difficult toistinguish between mild and moderate dehydration be-ause the first signs of dehydration appear at 3–4%ehydration, increase at 5% (the threshold for mild de-ydration), and there is not another significant clinicalifference until 9–10% (the threshold for severe) (7,16).verall, the ability of clinical signs to predict the degreef dehydration (mild, moderate, or severe) is, at best,roblematic. Employing the WHO classification groupsild to moderate dehydrated patients together; it elimi-

ates the need to distinguish between them and allowsor the fact that signs of dehydration might be apparentver a wide range of fluid losses.

Several laboratory studies have been proposed as aeans to help predict the degree of fluid deficit. These

nclude blood urea nitrogen (BUN), serum bicarbonate a

CO2) base excess, electrolytes, glucose, urine specificravity, and end-tidal carbon dioxide.

Evidence for the predictive value of BUN in dehy-rated children is mixed. Teach et al. studied a conve-ience sample of 40 children requiring intravenous fluidesuscitation (17). Laboratory variables of BUN to cre-tinine (Cr) ratio (BUN/Cr), CO2, serum uric acid, serumnion gap, urine anion gap, venous pH, venous baseeficit, urine specific gravity, and fractional excretion ofodium were individually assessed in a simple linearegression model with fluid deficit as the dependentariable. The authors found serum BUN/Cr and serumric acid to be significantly associated with increasinguid deficit (r � 52, p � 0.0005 and r � 0.35, p � 0.03,espectively). However, the sensitivity, specificity, andositive predictive value of these two laboratory studiesor the detection of � 5 % fluid deficit were poor. Shaoult al. retrospectively reviewed 300 pediatric cases (18).hey found a BUN � 14.3 mmol/L in 5% of childrenho were not dehydrated, in 26% of children with mildehydration, and in 38% of children with moderate de-ydration. In this study, BUN concentration was found toe 95% specific for dehydration status. Creatinine con-entration and mean pH were similar whether or notehydration was present. In contrast, Bonadio et al. re-orted that the magnitude of BUN concentration was notn accurate method of assessing hydration status in chil-ren with dehydration due to gastroenteritis (19).

Other investigators have looked at serum bicarbonateevel and base excess concentrations as predictors ofehydration in children. Narchi compared serum bicar-onate levels to clinical assessment of dehydration bymergency physicians (20). A serum bicarbonate con-entration � 22 mmol/L was more common in childrenith severe dehydration. Although decreased bicarbon-

te concentrations occurred more frequently with in-reasing degrees of dehydration, the magnitude of bicar-onate reduction was not significantly different withncreasing degrees of dehydration. The authors con-luded that the decrease in bicarbonate concentration isot reflective of the severity of fluid deficit. Vega andvner, and Ylimaz et al. evaluated the sensitivity of

erum bicarbonate in predicting degree of dehydration byomparing pre- and post-rehydration weights in children21,22). Vega and Avner found that a serum bicarbonateevel � 17 mEq/L was 77% sensitive for moderateehydration and 94% sensitive for severe dehydration. Inega’s investigation, when clinical impression was com-ined with a bicarbonate concentration of � 17 mEq/L,ensitivity for prediction of severe dehydration increasedo 100%. The findings of Yilmaz et al. suggest that aerum bicarbonate level of � 15 mEq/L coupled with anlevated serum urea concentration may be a valuable

djunct to clinical evaluation in predicting the degree of

dcbfcodvca

sdWwshio1

pdmtmoa

icphwfahgcf2gwictpBdafiwha

dhomfodc

tgdot(wSSmoawtgtesdigti(pohg

dTooattaip

isec

Pediatric Gastroenteritis and Dehydration 5

ARTICLE IN PRESS

ehydration. However, in each of these studies, manyhildren who were judged to have only mild dehydrationy clinical evaluation or rehydration weight gain wereound to have bicarbonate levels below the thresholdhosen for calculation of the sensitivity. The specificityf a low bicarbonate levels as a predictor of degree ofehydration at the thresholds defined by these authors isery poor. Both investigations suggest that a bicarbonateoncentration may be a useful adjunct to clinical evalu-tion in the assessment of dehydration.

Serum electrolytes alone may not be sufficiently sen-itive or specific enough to predict the degree of dehy-ration, but they may yield other useful information.athen et al. prospectively investigated 182 patientsith dehydration and found that 88 had an abnormal

erum electrolyte value such as hypoglycemia 9.9%,ypokalemia 6%, and hypernatremia 3 % (23). However,t should be noted that Wathen and colleagues found thatbtaining serum electrolytes changed management in0.4% of cases (23).

Overall, in the vast majority of patients with uncom-licated AGE, serum electrolytes are not helpful in pre-icting degree of dehydration or determining appropriateanagement of the patient. The AAP recommends ob-

aining electrolytes in cases of AGE associated with alteredental status, moderate to severe dehydration, clinical signs

f hypokalemia or hypernatermia, infants � 6 months ofge, and suspicious presentations (9,24).

Hypoglycemia may accompany dehydration. Severalnvestigations have been performed in underdevelopedountries where, unlike developed countries, a bacterialathogen is usually responsible for diarrhea causing de-ydration, and the dehydration often occurs in childrenho are chronically malnourished. Hirschhorn et al.

ound 2% of Pakistani children between 1 and 6 years ofge with dehydration secondary to gastroenteritis to beypoglycemic (25). The majority had a bacterial patho-en as the etiology of their gastroenteritis. None of thehildren in the study was malnourished. Glyn-Jonesound 7.9% of South African pediatric patients between

and 35 months of age with dehydration secondary toastroenteritis to be hypoglycemic (26). Hypoglycemiaas more common in the hypothermic and malnour-

shed. Daral and colleagues determined 14% of Indianhildren (� 3 months of age) presenting with dehydra-ion from diarrhea to be hypoglycemic (27). Fifty-fiveercent of children had a bacterial pathogen isolated.ennish et al., in the largest prospective investigation toate, found 4.5% of Bengali patients � 15 years olddmitted with diarrhea to be hypoglycemic (28). Sixty-ve percent had a bacterial pathogen identified and 39%ere malnourished. The children who were found to beypoglycemic had longer fasting times, seizures (35%),

nd altered mental status. Huq et al. determined 11% of y

ehydrated children at their center in Bangladesh to beypoglycemic (29). Seven percent were bacteremic andne-half to three-quarters of hypoglycemic patients werealnourished. Seizures and a higher mortality rate were

ound in the hypoglycemic children. The above data aref interest but are difficult to apply to dehydration in aeveloped nation where malnutrition and bacterialauses of diarrhea are less common.

Reid and Losek undertook a retrospective investiga-ion with a goal of estimating the prevalence of hypo-lycemia among pediatric patients with dehydration in aeveloped nation (30). The authors reviewed the recordsf 196 children, older than 1 month of age and youngerhan 5 years of age. They reported that 18 children9.2%) were hypoglycemic. The duration of vomitingas longer for the children with hypoglycemia (2.6 days,D � 1.5) than for those without hypoglycemia (1.6,D � 1.8; 95% CI 0.13–1.88). None of the hypoglyce-ic children was found to have an altered mental status

r to be hypotensive. A second investigation by Reid etl. enrolled 184 children to identify variables associatedith hypoglycemia (31). The authors found an associa-

ion with female gender, neurologic symptoms of hypo-lycemia, and a greater amount of vomiting vs. diarrheao be more closely associated with hypoglycemia. How-ver, these clinical variables did not have an adequateensitivity or specificity to accurately predict which chil-ren with AGE were hypoglycemic. To date, a definitivenvestigation describing indications to obtain a rapidlucose in the dehydrated child does not exist. Accordingo Reid and Losek, hypoglycemia is relatively commonn children � 5 years of age with dehydration from AGE30). A reasonable approach would be to adopt a liberalolicy on obtaining a rapid glucose, as the risks ofbtaining one are low and it seems that identifyingypoglycemia in the dehydrated child based on clinicalrounds alone is challenging.

The role of urine specific gravity to judge the dehy-ration status in the acute care setting is questionable.he reason for this is twofold. First, the dehydrated childften does not urinate until rehydration has begun. Sec-nd, as determined by investigations by Oppliger et al.nd Popowski et al., the value of urine specific gravityends to lag behind actual hydration status (32,33). Fur-hermore, the use of urine specific gravity in the neonatend young infant is unreliable, as the concentrating abil-ty of the kidney does not reach adult values until ap-roximately 1 year of age (34).

A study of 130 children with gastroenteritis compar-ng the use of end-tidal carbon dioxide monitoring toerum bicarbonate as a measure of acidosis showed thatnd-tidal carbon dioxide levels and serum bicarbonateoncentrations were correlated linearly in bivariate anal-

sis (10). The mean end-tidal carbon dioxide level for

pluct

tdrrdlascusSpcsaisn

OE

Cco(dPa(ti

aoopicptdcsi

rsblmsrtsitaiiotadtmftgdmi

T

WPRICGA

SPRe

6 J. E. Colletti et al.

ARTICLE IN PRESS

atients who had a return visit (33.0 � 4.0 mm Hg) wasower than the level for those who did not seek reeval-ation (36.6 � 3.6 mm Hg). The authors concluded thatapnography offers an objective non-invasive measure ofhe severity of acidosis among patients with gastroenteritis.

In summary, no one laboratory value has been foundo have great accuracy in predicting the degree of dehy-ration in children with gastroenteritis. Most expertsecommend that routine laboratory studies, including se-um electrolytes, to assess dehydrated patients with acuteiarrhea are unnecessary (7,9). However, serum electro-yte levels may provide useful information about otherbnormalities such as hypoglycemia, hypokalemia, andodium abnormalities. Other laboratory studies such as aomplete blood count, blood cultures, urine analysis, andrine culture should be obtained only if other pathology,uch as sepsis or urinary tract infection, is suspected.tool cultures are not routinely indicated in immunocom-etent patients with diarrhea. The majority of domesti-ally acquired cases of diarrhea are viral and, therefore,tool cultures are of low yield and will not likely changecute management. The presence of gross or occult bloodn the stool may raise suspicion of a bacterial etiology,uch as hemorrhagic Escherichia coli, Shigella, Salmo-ella, Yersinia, or Campylobacter species (35).

ral vs. Intravenous Rehydration in themergency Department

ontroversy exists as to the best route for rehydration ofhildren who present to the ED with dehydration sec-ndary to gastroenteritis. Although intravenous therapyIVT) remains the therapy of choice for severely dehy-rated patients, the AAP, CDC, European Society foraediatric Gastroenterology and Nutrition, and the WHOll strongly support the use of oral rehydration therapyORT) as the first-line therapy for the treatment of mildo moderate dehydration (6,7,36,37). Despite this, ORTs still underutilized and used incorrectly in many EDs

able 2. Composition of Commercial Oral Rehydration Solu

Solution Sodium (mEq/L) Carbohydrat

orld Health Organization 90 2edialyte 45 2.5ehydralyte 75 2.5

nfalyte 50 3eraLyte 70 4atorade 21 5.9pple juice 0.4 11.9

ources: Ref (46): Colletti JE. Diarrhea. In: Hendey GW, Hendry Practice of Emergency Medicine, 4th edition. Philadelphia: Lippi

ef (47): Stone B. Fluids and electrolytes. In: Robertson J, Shilkofski N.dn. St. Louis, MO: Mosby; 2005.

cross the United States (38–42). Barriers to utilizationf ORT include concern regarding the effectiveness ofral therapy, unfamiliarity with published guidelines, theerception that oral rehydration takes more time thanntravenous hydration, the misperception that ORT isontraindicated with vomiting, and parental or referringhysician preferences (38,39,43,44). There are few con-raindications to ORT; these include patients with hemo-ynamic instability, suspected ileus (suggested eitherlinically by abdominal distension or absent bowelounds, or if seen on radiography), or patients withmpaired protective airway reflexes (45).

ORT is a form of enteral rehydration therapy. Oralehydration solutions typically contain sodium, potas-ium, chloride, carbohydrates (typically glucose) and aase. The WHO recommends a reduced osmolarity so-ution for ORT. Table 2 lists the composition of com-ercial oral rehydration solutions and commonly con-

umed beverages (Gatorade and apple juice are notecommended as solutions for ORT but are included ashey are commonly consumed). ORT has been the main-tay of treatment of dehydration caused by gastroenteritisn underdeveloped countries out of sheer necessity, givenhe overwhelming burden of diarrheal illness and lack ofccess to more advanced medical interventions, such asntravenous therapies. Oral rehydration therapy is a strik-ng example of reverse technology transfer (i.e., technol-gy developed and embraced by lesser developed coun-ries that is now returning to more technologicallydvanced countries) (48). ORT is informally used toescribe rehydration both by mouth and by nasogastricube. Although more invasive than rehydration byouth, rehydration via nasogastric tube is safe and ef-

ective and is an alternative route in children who cannotolerate rehydration by mouth (49–51). Although notenerally recommended for patients with severe dehy-ration, nasogastric hydration can serve as a temporizingethod of hydration for children in whom intravenous or

ntraosseous access is difficult.

nd Commonly Consumed Beverages

) Potassium (mE/L) Chloride (mEq/L) mOsm/kg H2O

20 80 31020 35 25020 65 25025 45 20020 60 2202.5 1.7 377

26 — 700

en CH, Rosen CL, Schaider J, editors. Hardwood-Nuss’ Clinicalilliams and Wilkins; 2005:1221.

th

tions a

e (g/dL

L, Lindncott W

The Johns Hopkins Hospital: The Harriet Lane Handbook, 17

thttTtraote

oavr1(trrofrtgris0efr

tlasO4CcEIrocSdt

nt

rb(ssedrarrnl

Udttcfrrtr

wsvstawpdwwe

nctfinog(rA6ecC

Pediatric Gastroenteritis and Dehydration 7

ARTICLE IN PRESS

Studies comparing enteral and intravenous rehydra-ion therapy in mildly to moderately dehydrated childrenave reported numerous outcome measures. Some ofhese studies were performed in less-developed coun-ries, and most compared these therapies in inpatients.hree studies of rehydration methods enrolled ED pa-

ients (51–53). One additional study compared patientseceiving ORT in an ED holding area with ED patientsnd inpatients receiving IVT (54). The most relevantutcome measures for emergency physicians includeherapy efficacy, time required for therapy, and adverseffects.

A recent systematic review compared the failure ratef enteral rehydration to rehydrate or maintain hydrationfter initial rehydration of dehydrated children to intra-enous rehydration. This analysis found that the failureisk for enteral rehydration therapy was 4.9%, and was.3% for intravenous rehydration. For every 25 children95% CI 14–100) receiving oral or nasogastric rehydra-ion therapy, one would fail and require intravenousehydration. However, for patients treated with the cur-ently recommended WHO reduced osmolarity solutions,ne patient out of 100 children would need i.v. fluid afterailing enteral rehydration. The authors caution that theseesults may be applicable only to children with dehydra-ion from diarrhea and not necessarily to children withastroenteritis (i.e., vomiting and diarrhea), although theisk difference for trials between enteral rehydration andntravenous therapy that included children who had per-istent vomiting was 4% (95% CI �5–13) compared to% (95% CI �3–3) for children enrolled in trials thatxcluded persistently vomiting patients (45). This enteralailure rate is similar to previously published failure rateseported in meta-analyses (55,56).

Most studies comparing ORT with IVT studied inpa-ients; however, a few studies have investigated EDength of stay based on rehydration methods. An equiv-lent percentage of children receiving ORT or IVThowed improvement in dehydration scores at 2 h (79%RT vs. 80% IVT) and were successfully rehydrated ath (55% ORT vs. 57% IVT; difference �1%; 95%

I �24–22%) (53). In one pediatric ED-based study,hildren receiving oral rehydration therapy had a shorterD length of stay when compared to patients receiving

VT (225 min [SD 78 min] vs. 358 min [SD 160 min],espectively) (52). Nasogastric rehydration via continu-us infusion was as efficacious as intravenous therapy inorrecting moderately dehydrated patients at 3 h (51).uccessful i.v. placement is frequently difficult in dehy-rated children and may require multiple attempts, fur-her delaying rehydration via the intravenous route (53).

In the systematic review comparing oral vs. intrave-ous rehydration, 33 children (95% CI 20–100) needed

o be treated with i.v. fluid rather than oral or nasogastric i

ehydration to prevent one case of paralytic ileus. Phle-itis developed in 2.5% of patients who received IVT45). Most ORT vs. IVT trials excluded patients withevere dehydration and shock. Not surprisingly, deaths intudy patients were exceedingly rare and were reportedxclusively in low-middle-income countries, and theseeaths could not be directly attributed to the mode ofehydration therapy (45). In one trial, 1 patient developed

seizure 17 h after the onset of ORT, thought to beelated to a rapid decrease in total osmolality (57). Otherelatively minor complications (i.e., complications withasogastric tube placement and multiple attempts at i.v.ine placement) were reported as well (50,51,53).

There are limited cost-effectiveness data from thenited States (54). One study demonstrated that rehy-ration through a nasogastric tube was less expensivehan IVT (51). The cost of WHO oral rehydration solu-ion in 2005 was $0.07 per liter (6). Patients who re-eived ORT and IVT and were treated and dischargedrom the ED were equally likely to make an unscheduledeturn visit to the ED (52,53). Caregivers of patients whoeceived ORT were at least as likely to be satisfied withheir ED treatment when compared to those patientseceiving IVT (52,53).

All of these studies discussed above enrolled patientsho were dehydrated and had diarrhea, but there was

ome heterogeneity in the approach to children who hadomiting as well. Some studies included patients withignificant vomiting, whereas others excluded these pa-ients. Most studies excluded children � 3 months ofge. Although ORT is usually successful even in childrenho are vomiting, intractable or persistent vomiting mayreclude successful ORT. Some children with carbohy-rate malabsorption may develop worsening diarrheaith ORT, and ORT should be discontinued in childrenho have significant worsening of their diarrhea. How-

ver, there is no clear definition of ORT failure.Treatment of dehydration consists of three compo-

ents: rehydration, replacement of ongoing losses, andontinuation of normal feeding. The method of rehydra-ion therapy depends on the degree of dehydration; there-ore, the degree of dehydration should be assessed beforenitiating therapy (see above). Patients with minimal oro dehydration need therapy directed at replacement ofngoing losses. For each watery stool, patients should beiven 10 cc/kg body weight of oral rehydration solutionORS) and for each episode of emesis, children shouldeceive 2 cc/kg body weight of replacement fluid (7).lternatively, children weighing � 10 kg should receive0–120 cc of ORS for each diarrheal stool or episode ofmesis, whereas children weighing � 10 kg should re-eive 120–250 cc of ORS for each event (7) (Table 3).hildren who are breastfed should continue breastfeed-

ng ad lib, whereas patients eating solids should continue

tnbmcrie(os

bosiwcrft

adflrfastHa

Tw

Taalyat

T

N

M

S

S g acuta R-16):

T

0611235811

C

8 J. E. Colletti et al.

ARTICLE IN PRESS

o receive their usual diet, with the exception of elimi-ating foods high in simple sugars. Children should note taking any diluted or special formula. Patients withild to moderate dehydration should receive 50–100

c/kg of ORS over 3–4 h, supplemented by fluids toeplace ongoing losses (7) (Table 3). Patients shouldnitially receive small amounts (e.g., 5–15 cc) of ORSvery 5 min, increasing the volumes of ORT if toleratedTable 4) (7). Patients should also continue to breastfeedr eat solids (with the exception of limiting foods high inimple sugars) normally.

Patients assessed to have severe dehydration shoulde resuscitated initially with Lactated Ringer’s solutionr normal saline intravenously to restore hemodynamictability (58). The role of dextrose-containing solution in.v. rehydration is unclear. Levy and Bachur studiedhether the amount of i.v. dextrose administered to

hildren with gastroenteritis and dehydration affectedeturn visits warranting admission (59). The authors per-ormed a case control investigation of children 6 monthso 6 years of age who presented with acute gastroenteritis

able 3. Therapy Based on Degree of Dehydration

Degree of Dehydration Rehydration The

o dehydration to minimal None

ild to moderate dehydration ORS, 50–100 cc/kg body weORS therapy is initiated with

every 1 to 2 minIf not improving with ORS, co

(i.v.) fluid therapy.evere dehydration 0.9 normal saline or lactated

i.v. until mental status andfollowed by 5% dextrose ½at twice maintenance rates

ource: Ref (7): King CK, Glass R, Bresee JS, Duggan C. Managinnd nutritional therapy. MMWR Recomm Rep 2003 Nov 21;52(R

able 4. Oral Replacement Therapy Dosing Based on Weigh

Age Weight Initial Dosing

–6 mo 8 kg 5 cc every 5 min 6–12 mo 10 kg 10 cc every 5 min 122–18 mo 12 kg 10 cc every 5 min 128–24 mo 13 kg 10 cc every 5 min 12–3 years 15 kg 10 cc every 5 min 12–5 years 20 kg 15 cc every 5 min 18–8 years 25 kg 15 cc every 5 min 18–10 years 35 kg 15 cc every 2 min 450–12 years 40 kg 15 cc every 2 min 452–15 years 50 kg 15 cc every 2 min 45

ourtesy of Mark Hostetler, MD, University of Chicago, Illinois.

nd dehydration. The investigation concluded that chil-ren who received more i.v. dextrose, independent of theuid amount, were less likely to present for a return visitequiring admission. In patients receiving i.v. hydrationor severe dehydration, oral rehydration should be initi-ted as soon as it is tolerated. Table 5 summarizesuggestions for treatment of acute gastroenteritis fromhe European Society for Paediatric Gastroenterology,epatology and Nutrition (60). Indications for hospital

dmission are contained in Table 6.

he Use of Antiemetics in Childrenith Gastroenteritis

he third controversy in the treatment of children withcute gastroenteritis in the ED is whether or not to usentiemetics. In 1996, the AAP published practice guide-ines discouraging the use of antiemetics in children � 5ears of age (9). Administration of antiemetics has beenssociated with adverse reactions, most commonly ex-rapyramidal side effects such as acute dystonic reactions

Replacement of Losses

� 10 kg: 60–120 mL oral rehydration solution(ORS) per vomiting or diarrheal episode

� 10 kg: 120–240 mL ORS for per vomitingor diarrheal episode

er 3 to 4 hteaspoon)

intravenous

As above

20 cc/kgon improve,al saline i.v.

As aboveIf unable to tolerate oral fluids, administer

through nasogastric tube or administer 5%dextrose ½ normal saline with 20 mEq/Lpotassium chloride i.v.

e gastroenteritis among children: oral rehydration, maintenance,1–16.

Age

me/h First Advance Next Advance

10 cc/kg) 15 cc every 15 min 30 cc every ½ h10 cc/kg) 30 cc every 15 min 60 cc every ½ h10 cc/kg) 30 cc every 15 min 60 cc every ½ h10 cc/kg) 30 cc every 15 min 60 cc every ½ h10 cc/kg) 30 cc every 15 min 60 cc every ½ h10 cc/kg) 45 cc every 15 min 90 cc every ½ h10 cc/kg) 60 cc every 15 min 90 cc every ½ h10 cc/kg) 90 cc every 15 min 120 cc every ½ h10 cc/kg) 90 cc every 15 min 120 cc every ½ h10 cc/kg) 90 cc every 15 min 120 cc every ½ h

rapy

ight ov5 cc (1

nsider

ringersperfusi

norm.

t and

Volu

0 cc (0 cc (0 cc (0 cc (0 cc (0 cc (0 cc (0 cc (0 cc (0 cc (

appndd(tchoaFwicrftcrtptpstc

aapaag

tir(rcereaEawvsepzcwearpw(Tposo

stcz(bia

T

DS

ISP

T

III

I

VVVV

AMoGGe

Pediatric Gastroenteritis and Dehydration 9

ARTICLE IN PRESS

nd apnea. In one prospective investigation, the use ofrochlorperazine was associated with akathisia in 16% ofatients and dystonia in 4% of patients (61). Centralervous system depression manifesting as sedation androwsiness may interfere with oral hydration and make itifficult to assess the child for changes in mental statuse.g., lethargy). These dystonic-like reactions and seda-ive effects are reported to occur more frequently inhildren and young adults than in older patients (62). Itas been suggested by Kahn and colleagues that the usef antiemetics in infants with vomiting may be associ-ted with Sudden Infant Death Syndrome (63,64). Theood and Drug Administration has required a black boxarning on promethazine hydrochloride (65). This warn-

ng cautions against the use of promethazine inhildren � 2 years of age due to the potential for fatalespiratory depression. Furthermore, the warning callsor exercising caution when administering promethazineo children aged 2 years or older and to avoid the con-omitant administration of other medications with respi-atory depression effects. Concern also has been raisedhat the use of antiemetics in children with emesis couldossibly mask a more serious illness (66). It is importanto note that the above-mentioned investigations wereerformed mostly with phenothiazine antiemetics beforeerotonin type 3 (5-hydroxytryptamine3, 5-HT3) recep-or antagonists were widely employed in the vomitinghild.

The concerns regarding antiemetic use describedbove, combined with the knowledge that vomiting incute gastroenteritis is self-limited in nature and im-roves with correction of dehydration, led to the beliefmong many pediatric emergency medicine experts thatntiemetics should not be used in children with acute

able 5. The Essential Pillars of Good Treatment of AcuteGastroenteritis

. Use of oral rehydrating solution (ORS) for dehydrationI. Hypotonic solution (Na 60 mmol/L, glucose 74–111 mmol/L)II. Fast oral rehydration over 3–4 h

50–100 cc/kg maintenance plus:10 cc/kg for every watery stool2 cc/kg for every emesis

V. Rapid realimentation with normal feeding (including solids)thereafter

. Use of special formula is unjustifiedI. Use of diluted formula is unjustifiedII. Continuation of breast-feeding at all timesIII. Supplementation with ORS for ongoing losses

dapted from Ref (60): Szajewska H, Hoekstra JH, Sandhu B.anagement of acute gastroenteritis in Europe and the impact

f the new recommendations: a multicenter study. The Workingroup on acute Diarrhoea of the European Society for Paediatricastroenterology, Hepatology, and Nutrition. J Pediatr Gastro-nterol Nutr 2000;30:522–7.

astroenteritis (9,35). However, more recently, two au-

hors have published results revealing that antiemetic uses common in children with gastroenteritis and that theate of side effects may be less than previously reported67,68). Kwon et al. performed a cross-sectional surveyandomly sent to physicians board certified in the spe-ialties of emergency medicine, pediatrics, and pediatricmergency medicine (67). The majority (361, 60.9%) ofesponders reported using antiemetics for acute gastro-nteritis in the last year. A � 50% utilization rate ofntiemetics was reported for all three specialty groups.mergency physicians had the highest administration ofntiemetics (79%). The most commonly used antiemeticas promethazine, followed by trimethobenzamide. Ad-erse reactions were noted by 73 (20.2%) survey re-ponders. The most common adverse reactions werextrapyramidal. The medications most commonly re-orted as causing adverse reactions were prochlorpera-ine (42%) and promethazine (40%). The authors con-luded that use of antiemetics was common in childrenith acute gastroenteritis. They cautioned that an anti-

metic should be chosen with an understanding of itsdverse reactions. Li et al. performed a retrospectiveeview of 20,222 children, ages 1 month to 18 years,resenting with acute gastroenteritis (68). Nine percentere given a prescription for an antiemetic. The majority

90%) of these children were prescribed promethazine.he adverse reaction rate was 0.39%. Subgroup analysiserformed in children � 5 years of age failed to dem-nstrate a statistically significant difference in the pre-cription rate or the number of adverse reactions thatccurred.

Serotonin type-3 receptor antagonists, such as ondan-etron, are effective antiemetics. They have been showno have superior antiemetic effects in children receivingancer chemotherapy when compared with phenothia-ine-type antiemetics, droperidol, or metoclopramide69). Serotonin type-3 receptor antagonists also haveeen shown to be effective antiemetics in postsurgical-nduced emesis in children (70–72). All of these studieslso noted a lack of major side effects, including extra-

able 6. Potential Indications for Admission Criteria

ehydration � 5% and unable to tolerate oral fluidserious underlying diagnosis such as:SepsisUrinary tract infectionInborn error of metabolismIntra-abdominal catastropheDiabetic ketoacidosis

ntractable or bilious emesisignificant electrolyte disturbancesoor social situation such as:Parental noncompliance or inability to follow prescribed

therapy

Suspected neglect

peg

yciswn5rro(dt0dmseo

pd(matdrrisTstEotao

pn(otdacts

ttcmngsrCfgp

cydmbk81wdlirlovpsgr

sppndpt(oodfrciBad

10 J. E. Colletti et al.

ARTICLE IN PRESS

yramidal reactions. Several authors have looked at thefficacy and safety of ondansetron in patients with acuteastroenteritis.

Cubeddu et al. evaluated children aged 6 months to 8ears who had vomited twice within 1 h (73). Thirty-sixhildren were enrolled: 12 receiving placebo, 12 receiv-ng metoclopramide 0.3 mg/kg, and 12 receiving ondan-etron 0.3 mg/kg. All patients were hospitalized for 24 hith oral hydration attempted. The mean and the medianumber of emetic episodes over a 24-h time course wereand 3, respectively, for the placebo group; 5 and 6,

espectively, in the metoclopramide group; and 2 and 0,espectively, among the ondansetron group. Emesis wasbserved in 10 (83%) children receiving placebo, 867%) in the metoclopramide group, and 5 (42%) chil-ren receiving ondansetron. The authors found a statis-ical difference between ondansetron and placebo (p �.039), in ondansetron’s favor. Conversely, they did notetermine a significant difference between metoclopra-ide and placebo. Patients in the investigation received

imilar amounts of hydration fluid, but there were morepisodes of diarrhea in both the metoclopramide and thendansetron groups.

Ramsook et al. performed a randomized double-blindlacebo-controlled trial of oral ondansetron in 145 chil-ren aged 6 months to 12 years with acute gastroenteritis74). The patients also received oral hydration at 5 cc/in. For patients able to be discharged from the ED, five

dditional doses were administered at home. Ondanse-ron was effective in decreasing the episodes of emesisuring oral hydration, as well as the hospital admissionate. The frequency of emesis during ED observationanged in the placebo group from 0 to 7 and from 0 to 2n the ondansetron group. The rank sum of emesis epi-odes was lower in the ondansetron group (p � 0.001).he proportion of patients who were emesis-free aftertudy enrollment was greater for the ondansetron grouphen it was for the placebo group (p � 0.004) during theD stay. However, those who were discharged in thendansetron group were more likely to revisit the EDhan those in the discharged saline group. There was alson increase in the rate of diarrhea in those who receivedndansetron.

Reeves et al. performed a randomized, double-blind,lacebo-controlled trial of 107 children (54 to intrave-ous ondansetron, 53 to placebo) with gastroenteritis75). This investigation compared 0.15 mg/kg of i.v.ndansetron to placebo (saline) in children from 1 montho 22 years of age. Ondansetron was associated with aecrease in the number of episodes of emesis. After drugdministration, 38 (70%) of the ondansetron group had aomplete cessation of emesis, compared to 27 (51%) ofhe placebo group. The authors were unable to demon-

trate a decrease in the rate of overall hospitalization in d

he entire ondansetron-treated group vs. the placebo-reated group. However, ondansetron was shown to de-rease hospitalization in first-time treated children with aeasured serum carbon dioxide level � 15 mEq/L. Of

ote, the ondansetron group, compared to the placeboroup, had a greater proportion of children with a mea-ured serum carbon dioxide (CO2) � 15 (20% and 4%,espectively). At the author’s institution, patients with aO2 � 15 were routinely admitted and the authors there-

ore expected an admission bias against the ondansetronroup. Ondansetron was found to be cost-effective com-ared to saline hydration alone in their setting.

Freedman et al. performed a double-blind trial in 215hildren with gastroenteritis aged 6 months through 10ears (76). Enrolled patients were randomized to an oralose of ondansetron or placebo. Ondansetron was ad-inistered as an oral disintegrating tablet in a weight-

ased fashion (children 8–15 kg received 2 mg, 15–30g received 4 mg, and over 30 kg were administeredmg). Fifteen minutes after medication administration, a-h period of intense oral rehydration was initiated thatas continued until the point of disposition. The authorsetermined that children who received ondansetron wereess likely to vomit, vomited less often, had greater oralntake, and were less likely to be treated by intravenousehydration than those who received placebo. Meanength of stay in the ED was reduced by 12% in thendansetron group. Rates of hospitalization and returnisits to the ED did not significantly differ between thelacebo and ondansetron group. They concluded that aingle dose of oral ondansetron in children with acuteastroenteritis decreases vomiting and facilitates oralehydration.

Stork et al. compared ondansetron with dexametha-one and placebo in a randomized double-blind trial inatients aged 6 months to 12 years presenting to aediatric ED with acute gastritis and receiving intrave-ous hydration (77). Ondansetron was associated with aecrease in the rate of hospitalization (4% vs. 20 % forlacebo). The number needed to treat with ondansetrono prevent one hospitalization compared with placeboi.v. hydration alone) was 6. There was also a higher ratef patients who tolerated oral hydration at 2 h in thendansetron group. The authors concluded that in chil-ren presenting with dehydration secondary to vomitingrom acute viral gastritis, ondansetron with intravenousehydration improved tolerance of oral fluids and de-reased the hospitalization rate when compared withntravenous rehydration with or without dexamethasone.ased on this investigation, administration of dexameth-sone for dehydration secondary to acute gastroenteritisoes not seem to improve tolerance of oral fluids or

ecrease hospital admission rates. As such, dexametha-

si

vvtm

odtattpterTt

IvdwdhgrgdlpeoadH

id

mmoaitTsssiEts

ioisHsepAwrmnipnd

T

RRSR

i

Pediatric Gastroenteritis and Dehydration 11

ARTICLE IN PRESS

one cannot be recommended at this time for routine usen the vomiting child with acute gastroenteritis.

Overall, the above-mentioned investigations are fa-orable in regards to administration of ondansetron foromiting secondary to AGE, but it should be noted thathey are of limited power to detect important but uncom-on side effects.An argument that has been made against the use of

ndansetron in gastroenteritis, which is a self-limitedisease, is its relatively high cost. However, as noted inhe above studies, the use of ondansetron can be associ-ted with a decrease in need for hospitalization andherefore may be cost-effective (35,67–69). Cost-effec-iveness studies comparing ondansetron with metoclo-ramide in oncology patients have shown that ondanse-ron administered three times a day is at least as cost-ffective as metoclopramide, whereas twice-a-dayegimens of ondansetron are more cost-effective (78).he doses and route of administration of ondansetron in

he above-mentioned investigations can be seen in Table 7.

CONCLUSIONS

n the ED, dehydration is evaluated by synthesizing thearious historical, physical examination, and laboratoryata points in select patients. These signs and symptoms,hen combined into a score, may be helpful in predictingegree of dehydration. However, no one single findingas adequate specificity or sensitivity to predict the de-ree of dehydration. A subset of four factors (capillaryefill � 2 s, absent tears, dry mucous membranes, and illeneral appearance) have been demonstrated to predictehydration as well as the entire set. Similarly, no singleaboratory study has been found to be clinically useful inredicting degree of dehydration. Measurement of serumlectrolytes may provide useful information regardingther abnormalities such as hypoglycemia, hypokalemia,nd sodium abnormalities in children who are dehy-rated and are undergoing intravenous rehydration.owever, the routine measurement of serum electrolytes

able 7. Doses and Route of Administration ofOndansetron

Investigation Dose Route

amsook et al. (74) 0.3 mg/kg i.v.eeves et al. (75) 0.15 mg/kg (maximum 8 mg) i.v.tork et al. (77) 0.15 mg/kg i.v.amsook et al. (74) 6 months to 1 year: 1.6 mg p.o.

1 to 3 years: 3.2 mg4 to 12 years: 4 mg

.v. � intravenous; p.o. � per oral.

n the majority of children presenting to the ED with dehy-ration due to acute gastroenteritis is not recommended.

The available evidence supports the assertion that theajority of children with gastroenteritis and mild tooderate dehydration can be treated successfully with

ral rehydration therapy. ORT is highly effective, safe,nd inexpensive. ORT can be initiated more quickly thanntravenous therapy, and the rates of successful rehydra-ion at 2 h and 4 h after initiation of therapy are similar.his mode of therapy leads to high degrees of parentalatisfaction as well. With the exception of children withevere dehydration requiring resuscitation and childrenuspected of having paralytic ileus, ORT should be thenitial therapy for dehydrated children presenting to theD. It should also be initiated as early as possible in the

reatment of those receiving intravenous rehydration forevere dehydration due to gastroenteritis.

Traditionally, use of antiemetics has been discouragedn the pediatric population with acute gastroenteritis sec-ndary to adverse reactions and the possibility of mask-ng a more sinister illness. The exact rate of side effectsuch as sedation and extrapyramidal effects is unclear.owever, there is ample evidence that the use of ondan-

etron is associated with a much lower rate of these sideffects than the more traditional antiemetics such asromethazine and prochloperazine. Furthermore, theAP recommendation not to administer an antiemeticas made over a decade ago (1996), before 5-HT3

eceptor antagonists were widely utilized. There is accu-ulating evidence that ondansetron (oral or intrave-

ously) may be effective in decreasing the rate of vom-ting, improving the success rate of oral hydration,reventing the need for i.v. hydration, and preventing theeed for hospital admission in those receiving i.v. hy-ration.

REFERENCES

1. Kosek M, Bern C, Guerrant RL. The global burden of diarrhoealdisease, as estimated from studies published between 1992 and2000. Bull World Health Organ 2003;81:197–204.

2. Parashar UD, Bresee JS, Glass RI. The global burden of diarrhoealdisease in children. Bull World Health Organ 2003;81:236.

3. Malek MA, Curns AT, Holman RC, et al. Diarrhea- and rotavirus-associated hospitalizations among children less than 5 years of age:United States, 1997 and 2000. Pediatrics 2006;117:1887–92.

4. McConnochie KM, Conners GP, Lu E, Wilson C. How commonlyare children hospitalized for dehydration eligible for care in alter-native settings? Arch Pediatr Adolesc Med 1999;153:1233–41.

5. Duggan C, Santosham M, Glass RI. The management of acutediarrhea in children: oral rehydration, maintenance, and nutritionaltherapy. Centers for Disease Control and Prevention. MMWRRecomm Rep 1992;41:1–20.

6. World Health Organization. The treatment of diarrhoea. a man-ual for physicians and other senior health care workers, 4th revedn. Geneva, Switzerland: World Health Organization; 2005.

Available at: http://whqlibdoc.who.int/publications/2005/9241593180.pdf. Accessed April 2008.

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

3

3

3

3

3

3

3

3

3

3

4

4

4

4

4

4

4

4

4

4

5

5

5

5

5

12 J. E. Colletti et al.

ARTICLE IN PRESS

7. King CK, Glass R, Bresee JS, Duggan C. Managing acute gastro-enteritis among children: oral rehydration, maintenance, and nu-tritional therapy. MMWR Recomm Rep 2003;52:1–16.

8. Porter SC, Fleisher GR, Kohane IS, Mandl KD. The value ofparental report for diagnosis and management of dehydration in theemergency department. Ann Emerg Med 2003;41:196–205.

9. Practice parameter. the management of acute gastroenteritis inyoung children. American Academy of Pediatrics, ProvisionalCommittee on Quality Improvement, Subcommittee on Acute Gas-troenteritis. Pediatrics 1996;97:424–35.

0. Nagler J, Wright R, Krauss B. End-tidal carbon dioxide as ameasure of acidosis among children with gastroenteritis. Pediatrics2006;118:259–67.

1. Sandu BK; European Society of Pediatric Gastroenterology, Hepa-tology and Nutrition Working Group on Acute Diarrhoea. Practicalguidelines for the management of gastroenteritis in children. J Pe-diatr Gastroenterol Nutr 2001;33(Suppl 2).S36–9.

2. Friedman JN, Goldman RD, Srivastava R, Parkin PC. Develop-ment of a clinical dehydration scale for use in children between 1and 36 months of age. J Pediatr 2004;145:201–7.

3. Mackenzie A, Barnes G, Shann F. Clinical signs of dehydration inchildren. Lancet 1989;334:605–7.

4. Gorelick MH, Shaw KN, Murphy KO. Validity and reliability ofclinical signs in the diagnosis of dehydration in children. Pediatrics1997;99:E6.

5. Steiner MJ, DeWalt DA, Byerley JS. Is this child dehydrated?JAMA 2004;291:2746–54.

6. Duggan C, Refat M, Hashem M, Wolff M, Fayad I, Santosham M.How valid are clinical signs of dehydration in infants? J PediatrGastroentrol Nutr 1996;22:56–61.

7. Teach SJ, Yates EW, Feld LG. Laboratory predictors of fluiddeficit in acutely dehydrated children. Clin Pediatr (Phila) 1997;36:395–400.

8. Shaoul R, Okev N, Tamir A, Lanir A, Jaffe M. Value of laboratorystudies in assessment of dehydration in children. Ann Clin Bio-chem 2004;4:192–6.

9. Bonadio WA, Hennes HH, Machi J, Madagame E. Efficacy ofmeasuring BUN in assessing children with dehydration due togastroenteritis. Ann Emerg Med 1989;18:755–7.

0. Narchi H. Serum bicarbonate and dehydration severity in gastro-enteritis. Arch Dis Child 1998;78:70–1.

1. Vega RM, Avner JR. A prospective study of the usefulness ofclinical and laboratory parameters for predicting percentage ofdehydration in children. Pediatr Emerg Care 1997;13:179–82.

2. Yilmaz K, Karabocuoglu M, Citak A, Uzel N. Evaluation oflaboratory tests in dehydrated children with acute gastroenteritis. JPaediatr Child Health 2002;38:226–8.

3. Wathen JE, MacKenzie T, Bothner JP. Usefulness of the serumelectrolyte panel in the management of pediatric dehydrationtreated with intravenously administered fluids. Pediatrics 2004;114:1227–34.

4. McCollough M, Sharieff GQ. Abdominal pain in children. PediatrClin North Am 2006;53:107–37.

5. Hirschhorn N, Lindenbaum J, Greenough WB, Alam SM. Hypo-glycaemia in children with acute diarrhoea. Lancet 1966;2:128–33.

6. Glyn-Jones R. Blood sugar in infantile gastro-enteritis. S Afr MedJ 1975:1474–7.

7. Daral TS, Singh HP, Sachdev HP, Mohan M, Mathur M, BhargavaSK. Acute dehydrating diarrhea: clinical profile in neonates andyoung infants. Indian Pediatr 1985;22:333–8.

8. Bennish ML, Azad AK, Rahman O, Phillips RE. Hypoglycemiaduring diarrhea in childhood: prevalence, pathophysiology andoutcome. N Engl J Med 1990;332:1357–63.

9. Huq S, Hossain MI, Malek MA, Faruque ASG, Salam MA. Hy-poglycaemia in under-five children with Diarrhoea. J Trop Pediatr2007;53:197–201.

0. Reid SR, Losek JD. Hypoglycemia complicating dehydration inchildren with acute gastroenteritis. J Emerg Med 2005.;29:141–5.

1. Reid S, McQuillan S, Losek J. Hypoglycemia complicating dehy-

dration due to acute gastroenteritis. Clin Pediatr (Phila) 2003;42:641–6.

5

2. Oppliger RA, Magnes SA, Popowski LA, Gisolfi CV. Accuracy ofurine specific gravity and osmolality as indicators of hydrationstatus. Int J Sport Nutr Exerc Metab 2005;15:236–51.

3. Popowski LA, Oppliger RA, Patrick Lambert G, Johnson RF, KinJohnson A, Gisolf CV. Blood and urinary measures of hydrationstatus during progressive acute dehydration. Med Sci Sports Exerc2001;33:747–53.

4. Arant BS Jr. Postnatal development of renal function during thefirst year of life. Pediatr Nephrol 1987;1:308–13.

5. Burkhart DM. Management of acute gastroenteritis in children.Am Fam Physician 1999;60:2555–63.

6. Walker-Smith JA, Sandhu BK, Isolauri E, et al. Guidelines pre-pared by the ESPGAN working group on acute diarrhoea. Recom-mendations for feeding in childhood gastroenteritis. European So-ciety of Pediatric Gastroenterology and Nutrition. J PediatrGastroenterol Nutr 1997;24:619–20.

7. American Academy of Pediatrics. Managing acute gastroenteritisamong children. oral rehydration, maintenance, and nutritionaltherapy. Pediatrics 2004;114:507.

8. Ozuah PO, Avner JR, Stein RE. Oral rehydration, emergencyphysicians, and practice parameters: a national survey. Pediatrics2002;109:259–61.

9. Conners GP, Barker WH, Mushlin AI, Goepp JG. Oral versusintravenous: rehydration preferences of pediatric emergency med-icine fellowship directors. Pediatr Emerg Care 2000;16:335–8.

0. Kleinman RE. We have the solution: now what’s the problem?Pediatrics 1992;90:113–5.

1. Snyder JD. Use and misuse of oral therapy for diarrhea: compar-ison of US practices with American Academy of Pediatrics rec-ommendations. Pediatrics 1991;87:28–33.

2. Bezerra JA, Stathos TH, Duncan B, Gaines JA, Udall JN Jr.Treatment of infants with acute diarrhea: what’s recommended andwhat’s practiced. Pediatrics 1992;90:1–4.

3. Reid SR, Bonadio WA. Outpatient rapid intravenous rehydrationto correct dehydration and resolve vomiting in children with acutegastroenteritis. Ann Emerg Med 1996;28:318–23.

4. Reis EC, Goepp JG, Katz S, Santosham M. Barriers to use of oralrehydration therapy. Pediatrics 1994;93:708–11.

5. Hartling L, Bellemare S, Wiebe N, Russell K, Klassen TP, CraigW. Oral versus intravenous rehydration for treating dehydrationdue to gastroenteritis in children. Cochrane Database Syst Rev2006;3:CD004390.

6. Colletti JE. Diarrhea. In: Hendey GW, Hendry PL, Linden CH,Rosen CL, Schaider J, eds. Hardwood-Nuss’ clinical practice ofemergency medicine, 4th edn. Philadelphia, PA: Lippincott Wil-liams and Wilkins; 2005:1221.

7. Stone B. Fluids and electrolytes. In. Robertson J, Shilkofski N. TheJohns Hopkins Hospital: the Harriet Lane handbook, 17th edn. St.Louis, MO: Mosby; 2005:581.

8. Santosham M. Oral rehydration therapy: reverse transfer of tech-nology. Arch Pediatr Adolesc Med 2002;156:1177–9.

9. Sharifi J, Ghavami F, Nowrouzi Z, et al. Oral versus intravenousrehydration therapy in severe gastroenteritis. Arch Dis Child 1985;60:856–60.

0. Gremse DA. Effectiveness of nasogastric rehydration in hospital-ized children with acute diarrhea. J Pediatr Gastroenterol Nutr1995;21:145–8.

1. Nager AL, Wang VJ. Comparison of nasogastric and intravenousmethods of rehydration in pediatric patients with acute dehydra-tion. Pediatrics 2002;109:566–72.

2. Atherly-John YC, Cunningham SJ, Crain EF. A randomized trialof oral vs intravenous rehydration in a pediatric emergency depart-ment. Arch Pediatr Adolesc Med 2002;156:1240–3.

3. Spandorfer PR, Alessandrini EA, Joffe MD, Localio R, Shaw KN.Oral versus intravenous rehydration of moderately dehydratedchildren: a randomized, controlled trial. Pediatrics 2005;115:295–301.

4. Listernick R, Zieserl E, Davis AT. Outpatient oral rehydration inthe United States. Am J Dis Child 1986;140:211–5.

5. Gavin N, Merrick N, Davidson B. Efficacy of glucose-based oralrehydration therapy. Pediatrics 1996;98:45–51.

5

5

5

5

6

6

6

6

6

66

6

6

6

7

7

7

7

7

7

7

7

7

Pediatric Gastroenteritis and Dehydration 13

ARTICLE IN PRESS

6. Fonseca BK, Holdgate A, Craig JC. Enteral vs intravenous rehy-dration therapy for children with gastroenteritis: a meta-analysis ofrandomized controlled trials. Arch Pediatr Adolesc Med 2004;158:483–90.

7. Tamer AM, Friedman LB, Maxwell SR, Cynamon HA, Perez HN,Cleveland WW. Oral rehydration of infants in a large urban U.S.medical center. J Pediatr 1985;107:14–9.

8. Holliday MA, Ray PE, Friedman AL. Fluid therapy for children:facts, fashions and questions. Arch Dis Child 2007;92:546–50.

9. Levy JA, Bachur RG. Intravenous dextrose during outpatientrehydration in pediatric gastroenteritis. Acad Emerg Med 2007;14:324–30.

0. Szajewska H, Hoekstra JH, Sandhu B. Management of acutegastroenteritis in Europe and the impact of the new recommenda-tions: a multicenter study. The Working Group on Acute Diar-rhoea of the European Society for Paediatric Gastroenterology,Hepatology, and Nutrition. J Pediatr Gastroenterol Nutr 2000;30:522–7.

1. Olsen JC, Keng JA, Clark JA. Frequency of adverse reactions toprochlorperazine in the ED. Am J Emerg Med 2000;18:609–11.

2. Murray L. Physician’s desk reference. In: Wyeth. Montvale, NJ:Thompson PDR; 2005:3362.

3. Kahn A, Blum D. Phenothiazines and sudden infant death syn-drome. Pediatrics 1982;70:75–8.

4. Kahn A, Hasaerts D, Blum D. Phenothiazine-induced sleep apneasin normal infants. Pediatrics 1985;75:844–7.

5. Hampton T. Promethazine warning. JAMA 2005;293: 921.6. Anquist KW, Panchanathan S, Rowe PC, et al. Diagnostic delay

after dimehydrinate use in vomiting children. Can Med Assoc J1991;145:965–8.

7. Kwon KT, Rudkin SE, Langdorf MI. Antiemetic use in pediatricgastroenteritis: a national survey of emergency physicians, pedia-tricians, and pediatric emergency physicians. Clin Pediatr 2002;

41:641–52.

8. Li ST, DiGiuseppe DL, Christakis DA. Antiemetic use for acutegastroenteritis in children. Arch Pediatr Adolesc Med 2003;157:475–9.

9. Hahlen K, Quintana E, Pinkerton CR, Cedar E. A randomizedcomparison of intravenously administered granisetron versuschlorpromazine plus dexamethasone in the prevention of ifosf-amide-induced emesis in children. J Pediatr 1995;126:309 –13.

0. Fujii Y, Toyooka H, Tanaka H. Antiemetic efficacy of granisetronand metoclopramide in children undergoing ophthalmic or ENTsurgery. Can J Anaesth 1996;43:1095–9.

1. Patel RI, Davis PJ, Orr RJ, et al. Single-dose ondansetron preventspostoperative vomiting in pediatric outpatients. Anesth Analg 1997;85:538–45.

2. Shende D, Mandal NG. Efficacy of ondansetron and metoclopra-mide for preventing postoperative emesis following strabismussurgery in children. Anesthesiology 1997;52:496–500.

3. Cubeddu LX, Trujillo LM, Talmaciu I, et al. Antiemetic activity ofondansetron in acute gastroenteritis. Aliment Pharmacol Ther1997;11:185–91.

4. Ramsook C, Sahaun-Carreon I, Kozinetz CA, Moro-Sutherland D.A randomized clinical trial comparing oral ondansetron with pla-cebo in children with vomiting from acute gastroenteritis. AnnEmerg Med 2002;39:397–403.

5. Reeves JJ, Shannon M, Fleisher GR. Ondansetron decreases vom-iting associated with acute gastroenteritis: a randomized, con-trolled trial. Pediatrics 2002;109:e62.

6. Freedman SB, Adler M, Seshadri R, Powell EC. Oral ondansetronfor gastroenteritis in a pediatric emergency department. N EnglJ Med 2006;354:1689–705.

7. Stork CM, Brown KM, Reily TH, Secreti L, Brown LH. Emer-gency department treatment of viral gastritis using IV ondan-setron or dexamethasone in children. Acad Emerg Med 2006;13:1027–33.

8. Cox F, Hirsch J. Ondansetron: a cost-effective advance in anti-

emetic therapy. Oncology 1993;50:186–90.