Dimenhydrinate in Children With Infectious
Gastroenteritis: A Prospective, RCT
WHATS KNOWN ON THIS SUBJECT: Dimenhydrinate has
traditionally been used for children with gastroenteritis and
vomiting in Canada and Germany, but there has been no study to
evaluate its efcacy.
WHAT THIS STUDY ADDS: We performed a randomized,
controlled trial to determine the efcacy of dimenhydrinate.
Dimenhydrinate signicantly reduced vomiting but did not
improve oral rehydration and did not decrease the rate of
hospitalization.
abstract
OBJECTIVE: Vomiting is a common symptom in children with infectious
gastroenteritis. It contributes to uid loss and is a limiting factor for
oral rehydration therapy. Dimenhydrinate has traditionally been used
for children with gastroenteritis in countries such as Canada and Ger-
many. We investigated the efcacy and safety of dimenhydrinate in
children with acute gastroenteritis.
METHODS: We performed a prospective, randomized, placebo-
controlled, multicenter trial. We randomly assigned 243 children with
presumed gastroenteritis and vomiting to rectal dimenhydrinate or
placebo. Children with no or mild dehydration were included. All chil-
dren received oral rehydration therapy. Primary outcome was dened
as weight gain within 18 to 24 hours after randomization. Secondary
outcomes were number of vomiting episodes, uid intake, parents
assessment of well-being, number of diarrheal episodes, and admis-
sion rate to hospital. We recorded potential adverse effects.
RESULTS: Change of weight did not differ between children who re-
ceived dimenhydrinate or placebo. The mean number of vomiting epi-
sodes between randomization and follow-up visit was 0.64 in the di-
menhydrinate group and 1.36 in the placebo group. In total, 69.6% of
the children in the dimenhydrinate group versus 47.4% in the placebo
group were free of vomiting between randomization and the follow-up
visit. Hospital admission rate, uid intake, general well-being of the
children, and potential adverse effects, including the number of diar-
rhea episodes, were similar in both groups.
CONCLUSIONS: Dimenhydrinate reduces the frequency of vomiting in
children with mild dehydration; however, the overall benet is low,
because it does not improve oral rehydration and clinical outcome.
Pediatrics 2009;124:e622e632
e622 UHLIG et al
AUTHORS: Ulrike Uhlig, MD,a,b Nicole Pfeil, MD(sci),a,c
Gotz Gelbrich, PhD,c Christian Spranger, MD,a,c Steffen
Syrbe, MD,a Boris Huegle, MD,a Barbara Teichmann, MD,d
Thomas Kapellen, MD,a,c Peggy Houben, PhD,c Wieland
Kiess, MD,a and Hans Holm Uhlig, MD, DPhila
aSection of Pediatric Gastroenterology, University Hospital for
Children and Adolescents, University of Leipzig, Leipzig,
Germany; bSection of Pediatric Gastroenterology, University
Hospital for Children and Adolescents, Martin Luther University
Halle, Halle, Germany; cCoordination Centre for Clinical Studies
Leipzig, Leipzig, Germany; and dPrivate Pediatric Practice,
Leipzig, Germany
KEY WORDS
gastroenteritis, vomiting, dimenhydrinate, children, randomized,
controlled trial
ABBREVIATIONS
CI condence interval
NNTnumber needed to treat
Drs U. Uhlig and H. Uhlig initiated the study, developed the initial
study design, contributed to day-to-day work, were involved in
the data analysis, and wrote the rst draft of the manuscript;
Drs Gelbrich and Kapellen contributed to the study design; Dr
Gelbrich developed the statistical aspects of the study design,
performed the data analysis, and wrote parts of the manuscript;
Drs Spranger, Kapellen, Houben, and Kiess contributed to the
coordination of the study; and Drs U. Uhlig, Spranger, Syrbe,
Huegle, Kapellen, and H. Uhlig and Ms Pfeil recruited patients. All
authors discussed and contributed to the manuscript.
This trial has been registered at European Union Drug
Regulating Authorities Clinical Trials (identier 2005-003943-30;
international standard randomized, controlled trial No.
53730137).
www.pediatrics.org/cgi/doi/10.1542/peds.2008-1650
doi:10.1542/peds.2008-1650
Accepted for publication Jun 1, 2009
Address correspondence to Hans Holm Uhlig, MD, DPhil,
University Hospital for Children and Adolescents, University of
Leipzig, Section of Pediatric Gastroenterology, Liebigstrasse 20a,
D-04103 Leipzig, Germany. E-mail: holm.uhlig@medizin.uni-
leipzig.de
PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).
Copyright 2009 by the American Academy of Pediatrics
FINANCIAL DISCLOSURE: The authors have indicated they have
no nancial relationships relevant to this article to disclose.

Infectious gastroenteritis is 1 of the
most frequent infectious diseases in
childhood.1 The majority of children
develop vomiting in the initial phase
of the disease. Vomiting causes dis-
comfort and contributes to uid loss.
Vomiting is a limiting factor for oral
rehydration therapy in children with
infectious gastroenteritis. Oral rehy-
dration is the recommended therapy
in children with mild to moderate de-
hydration that is caused by infec-
tious gastroenteritis.13 Ondansetron
has been shown to reduce vomit-
ing in children with infectious
gastroenteritis.412
Rectal dimenhydrinate has been used
for children with gastroenteritis and
vomiting in Canada and Germany,13 but
no studies of its effect have been pub-
lished. Rectal administration of anti-
emetics in children with diarrhea may
not be efcient.14 The sedative effect of
dimenhydrinate might counteract the
antiemetic effect by reducing overall
uid intake. We performed a random-
ized, controlled trial to determine
whether dimenhydrinate is effective in
improving oral rehydration, reducing
vomiting, and decreasing the rate of
hospitalization.
METHODS
The Vomiting-Enteritis-Dimenhydrinate
study (VomED) was a prospective,
double-blind, placebo-controlled, mul-
ticenter phase IV trial that investigated
the efcacy of dimenhydrinate on oral
rehydration therapy during infectious
gastroenteritis and vomiting in chil-
dren. Patients were recruited in 5 chil-
drens hospitals (University Childrens
Hospital Leipzig, University Childrens
Hospital Mainz, St Georg Hospital
Leipzig, St Elisabeth and Barbara Hos-
pital Halle/Saale, and Childrens Hospi-
tal Wurzen) and 6 pediatric practices
(Leipzig and Rotha) in Germany. Study
recruitment was performed during
2 consecutive autumn-to-spring sea-
PEDIATRICS Volume 124, Number 4, October 2009
sons (December 1, 2005, to May 31,
2006, and October 1, 2006, to May 31,
2007). Written informed consent was
obtained from at least 1 parent before
enrollment in the study. The study fol-
lowed the principles of good clinical
practice (European good clinical prac-
tice guidelines www.emea.europa.eu/
pdfs/human/ich/013595en.pdf and as
specied by German law as of Septem-
ber 8, 2004).
The study was approved by the Univer-
sity of Leipzig Ethics Committee (196-
05), by local ethics committees, and by
the German Federal Institute for Drugs
and Medical Devices. Data validation
and quality control procedures in-
cluded external monitoring of study
sites.
Patients
Inclusion criteria were (1) age be-
tween 6 months and 6 years, (2) sus-
pected infectious gastroenteritis, (3)
acute (24 hours) onset of vomiting
with at least 2 episodes within the last
12 hours, (4) outpatient attendance,
and (5) body weight of 7 kg. We ex-
cluded children with moderate to se-
vere gastroenteritis, on the basis of 1
of the following criteria: (1) acute
weight loss of 5% of body weight; (2)
bloody stools; (3) necessity of intrave-
nous rehydration therapy; or (4) meta-
bolic acidosis (pH 7.25) and/or elec-
trolyte disturbances (blood tests were
conducted at the discretion of the
treating physician, only for a minority
of children). We also excluded children
with preexisting diseases for which di-
menhydrinate is contraindicated (eg,
epilepsy, glaucoma, acute asthma,
porphyria, pheochromocytoma), cur-
rent treatment with drugs that are
suspected to interact with dimenhydri-
nate, or treatment with antiemetics or
secretion inhibitor racecadotril within
this episode of gastroenteritis. Simul-
taneous participation in other medical
trials was not allowed. Children for
ARTICLES
whom intravenous rehydration ther-
apy was indicated were hospitalized.
Randomization and Stratication
Patients were randomly assigned to
receive dimenhydrinate or placebo
suppositories (Table 1). Randomiza-
tion was stratied by category of body
weight (15 and 15 kg). An indepen-
dent biostatistician from Sandoz Phar-
maceuticals, Germany (GmbH, Holz-
kirchen, Germany), generated the
randomization list according to our
study protocol. For each stratum, a
blocked random allocation sequence
was generated (block randomization
in blocks of 4). Medication kits were
consecutively numbered and assigned
to the patients according to their stra-
tum and the order of recruitment in
the respective study site.
Study Intervention
The rst suppository (40 mg of dimen-
hydrinate or placebo) was given in the
outpatient department or pediatric
practice. Additional suppositories of
the same type (40 mg of dimenhydri-
nate or placebo) were provided, de-
pending on body weight (15 kg: 1
suppository; 1525 kg: 2 supposito-
ries; 25 kg: 3 suppositories). Caregiv-
ers were instructed to give additional
suppositories only in case of persis-
tent vomiting or in case of visible ex-
cretion of the suppository immediately
after administration. Both dimenhydri-
nate and placebo suppositories were
manufactured and supplied by Sandoz
Pharmaceuticals, Germany (formerly
Hexal AG, Holzkirchen, Germany).
Concomitant Treatment
Each patient received 10 sachets of
powdered oral rehydration solution (1
sachet corresponds to 200 mL of re-
constituted solution containing 100
mmol/L glucose, 60 mmol/L sodium,
and 20 mmol/L potassium; osmolarity
240 mosm/L [Stada Arzneimittel, Bad
Vilbel, Germany]). All parents were ex-
e623
TABLE 1 Description of the Study Groups at the Time of Randomization scale), adverse events, and use of co-
Parameter Dimenhydrinate Placebo medication, each recorded for the in-
(N 122) (N 115) terval between randomization and
Male gender, n (%) 68 (55.7) 68 (59.1) follow-up visit. Long-term outcome re-
Age, mean SD, y 2.5 1.5 2.3 1.5
Weight, mean SD, kg 13.3 4.4 13.0 4.3 corded in the structured telephone in-
15, n (%) 86 (70.5) 81 (70.4) terview included severe adverse
15, n (%) 36 (29.5) 34 (29.6) events that required hospitalization,
Vomiting episodes within 24 h before enrollment, n (%)
2 33 (27.0) 25 (21.7) time until recovery, and parent satis-
3 28 (23.0) 31 (27.0) faction with treatment (numeric scale
45 34 (27.9) 30 (26.1) from 1 [best] to 6 [worst]).
5 27 (22.1) 29 (25.2)
Episodes of diarrhea within 24 h before enrollment, n (%)
0 53 (43.4) 49 (42.6) Adverse Events
13 42 (34.4) 42 (36.5)
Sedation; paradoxic excitement; drowsi-
3 25 (20.5) 20 (17.4)
Unknown 2 (1.6) 4 (3.5) ness; skin reaction; and hospitalization
Dehydration, n (%) as a result of gastroenteritis, were
None 65 (53.3) 65 (56.5) predened as expected adverse
Mild 52 (42.6) 47 (40.9)
Moderate 5 (4.1) 2 (1.7) events. Fever, vomiting, and diarrhea
Severe 0 (0.0) 0 (0.0) are symptoms of gastroenteritis and
Unknown 0 (0.0) 1 (0.9) hence were not considered to be ad-
Co-medication, n (%)
Antipyretics 37 (30.3) 33 (28.7) verse events.
Antibiotics 9 (7.4) 6 (5.2)
Probiotics 13 (10.7) 11 (9.6) Data Collection and Statistical
Other 30 (24.6) 34 (29.6)
Analysis
All patient data were entered in a study
plicitly instructed on oral rehydration tion to follow-up visit after 18 to 24 database at the coordination center
therapy and early feeding and received hours: (weightfollow-up weightrando)/ for clinical studies of the University
an information leaet. weightrando. This means that the spec- of Leipzig. The database was not ac-
trum relative weight gain is between cessible to non-clinical study investi-
Follow-up Visit and Telephone gators. At no point did any of the com-
1 and 1. Weight loss would lead to
Interview panies involved have access to pa-
negative and weight gain to positive
A short-term follow-up visit in the tient data, case report forms, or the
values. When a child was hospitalized
study center was scheduled 18 to 24 study database. After recruitment of
as a result of gastroenteritis within 24
hours after randomization. Patients the last patient, after recording of all
hours after randomization, the end
underwent physical examination and patient data, and after response to all
point was formally set at 1 for pri-
weight measurement, and history queries that arose during the monitor-
mary analysis. This means that hospi- ing process, the study database was
was recorded by using a structured
talization was counted as the worst closed. The rst analysis was per-
interview.
possible outcome. By including cases formed blinded to the study medica-
For long-term follow-up, a research as- with hospitalization (ie, presumably
sistant called the families for a tele- tion as an intention-to-treat analysis,
severe dehydration and dramatic knowing the allocation to 2 groups but
phone interview 7 to 14 days after en- weight loss) into analysis, a potential
rollment. Again, a history was obtained not the allocation to dimenhydrinate
selection bias was prevented. or placebo.
by a structured questionnaire. Case
report forms of the randomization Secondary early outcome measures Primary analysis was conducted by the
visit, the follow-up-visit, and the tele- were the number of episodes of vomit- Mann-Whitney U test. Because we ex-
phone interview are provided in Ap- ing, the number of diarrheal episodes, pected that relative changes of weight
pendices 1 to 3. the volume of uid intake, hospitaliza- might be larger in small children, the
tion as a result of gastroenteritis, well- stratied version of this test was used.
Outcome Measures being of the child as assessed by the Stratication was done by baseline
The primary clinical end point was the parents on a 6-point smiley scale weight, dividing children into 4 groups
relative weight gain from randomiza- (equivalent to the Wong-Baker faces according to the quartile of weight. The

e624 UHLIG et al
 ARTICLES

test statistic was then computed using

the ranks within the strata (instead of
global ranks); thereby, only children of
similar initial weights were compared
with each other.
For secondary analysis, t tests were
used to evaluate quantitative out-
comes, and 2 statistics were used for
frequencies. Posthoc analysis was per-
formed to describe the outcome mea-
sure as a function of the severity of
diarrhea (Appendix 5).

Sample Size
A mean difference of 1.5% of relative
change of weight was considered a
clinically relevant effect. SD was antic-
ipated to be 3% on the basis of clinical
experience. With these assumptions,
210 patients were needed to achieve a
power of 0.95 at type I error level of
0.05. We planned to terminate recruit-
ment when 210 patients completed the
study or 270 patients were randomly
assigned.

RESULTS
Participants
A total of 243 eligible children were
randomly assigned to 2 study groups:
124 to dimenhydrinate and 119 to pla-
cebo (Fig 1). Six patients were ex-
cluded because they did not match the
eligibility criteria. After the follow-up
visit, 208 patients were available for
analysis of primary and 199 for analy-
sis of secondary end points. A total of
224 patients could be reached for tele-
phone follow-up. Reasons for missing
the follow-up visit were full recovery of
the child (5 in the dimenhydrinate
group versus 2 in the placebo group),
illness of family members (2 vs 1), and
personal logistic problems (3 vs 5).
Outcome at the Follow-up Visit
The change of body weight between
randomization and follow-up visit was
similar in both groups. Mean relative
gain of body weight was 0.14% (SD:
FIGURE 1
Enrollment of patients, randomization, stratication, follow-up visit, telephone interview, withdrawal,
and completion of the study.
1.96%) in the dimenhydrinate group
and 0.06% (SD: 1.74%) in the placebo
group (P .452; Table 2), with mar-
ginal differences across weight quar-
tile subgroups. Four children in the di-
menhydrinate group and 5 in the
placebo group were hospitalized for
gastroenteritis within 18 to 24 hours.
The mean number of episodes of vom-
iting between randomization and
follow-up visit was 0.64 in the dimenhy-
drinate group and 1.36 in the placebo
group (difference: 0.72 [95% con-
dence interval (CI): 1.16 to 0.29]).
At the follow-up visit, 69.6% in the di-
menhydrinate versus 47.4% in the pla-
cebo group were free of vomiting (P
.001). Numbers needed to treat
(NNTs) were 2 (95% CI: 1 to 4) to avoid
1 episode of vomiting and 5 (95% CI: 3
to 12) for complete cessation of vom-
iting. Additional use of the study
medication was reported in 30.4% of
children in the dimenhydrinate
group and in 54.6% of the placebo
group (P .001).
Mean frequencies of diarrheal epi-
sodes were 1.75 and 1.74, respectively
(P .720). The amount of uid intake
and the improvement of well-being of
the child according to parents assess-
ment were similar in both groups.
Sedation occurred in 22 (21.6%) chil-
dren who received dimenhydrinate
and in 18 (18.6%) children who re-
ceived placebo. One (1%) child in each
group had rash, and drowsiness was
reported for 1 (1%) child in the dimen-
hydrinate group. Three severe adverse

PEDIATRICS Volume 124, Number 4, October 2009 e625

TABLE 2 Main and Secondary Outcomes and Adverse Events
Parameter Dimenhydrinate Placebo
Primary outcome, n 106 102
Absolute weight change, mean SD, kg 0.02 0.23 0.01 0.25
Relative weight change, mean SD, % 0.14 1.96 0.06 1.74
Hospitalizations, n (%) 4 (3.8) 5 (4.9)
Primary test: rank by baseline weight, mean, kg
7.2 to 9.6 25.2 27.7
9.6 to 12.5 27.0 25.9
12.5 to 16.0 25.7 27.3
16.0 to 30.0 27.2 25.7
Secondary outcomes, n 102 97
Frequency of vomiting, mean SD 0.64 1.27 1.36 1.79
Vomiting episodes, n (%)
0 71 (69.6) 46 (47.4)
1 16 (15.7) 19 (19.6)
2 7 (6.9) 8 (8.2)
3 8 (7.8) 24 (24.7)
Frequency of diarrhea, mean SD 1.75 1.93 1.74 1.81
Oral uid intake, mean SD, mL 701.5 361.5 679.0 282.4
Repetitive administration of suppositories, n (%) 31 (30.4) 53 (54.6)
Parents reported outcome measure: improvement 1.35 1.37 1.14 1.50
of general well-being, mean SDa
Adverse events, n (%)
Sedation 22 (21.6) 18 (18.6)
Drowsiness 1 (1.0) 0 (0.0)
Exanthema 1 (1.0) 1 (1.0)
Others 0 (0.0) 1 (1.0)
Severe adverse events, n
RSV bronchiolitis 0 1
Upper airway infection 0 1
Unspecied circulatory system disorder 0 1
Parameters refer to the period between visits 1 and 2. RSV indicates respiratory syncytial virus.
a Scale: 1 indicates very well; 6, bad.
events were reported in the placebo to their child care facility 2.91 vs 3.21
group. Two children were hospitalized days (difference: 0.29 [95% CI 1.35
for respiratory tract infections, and to 0.77]; P .588) after randomization.
the third child was hospitalized for Mean parental satisfaction rating was
mild cardiovascular disorder and so- 2.39 vs 2.31 (difference: 0.08 [95% CI:
cial problems. All children fully recov- 0.28 to 0.45]; P .651).
ered. In no case was hospitalization at-
tributed to the study medication (Table Intention-to-Treat and Subgroup
2). Analysis
We performed an intention-to-treat
Outcome at Telephone Follow-up analysis. Similar to the main analysis
Overall, 10 dimenhydrinate versus 13 in the dimenhydrinate group, we found
placebo patients were hospitalized; a reduced number of vomiting epi-
among those, 9 vs 11 were hospitalized sodes (mean SD: 0.68 1.37 vs
for gastroenteritis. Mean duration of 1.41 1.81; P .001), a decreased
vomiting was 0.60 days in the dimenhy- risk for repeated vomiting (33 [30.0%]
drinate group versus 0.94 days in the vs 56 [53.8%]; P .001), and a de-
placebo group (difference: 0.34 [95% creased risk for repeated administra-
CI: 0.66 to 0.02]; P .036) and 1.40 tion of suppositories (34 [30.9%] vs 58
vs 1.40 days among those who vomited [55.8%]; P .001). No other signicant
after the rst dose of study medication differences were observed. For de-
(P .99). Children were able to return tailed analysis, see Appendix 4.
e626 UHLIG et al
Because not all children who were en-
P rolled as a result of vomiting devel-
oped diarrhea during the observation
.471 period, we performed a posthoc suba-
.452
.744 nalysis for children with diarrhea epi-
.863 sodes (24 hours before randomization
until follow-up visit). A total of 78% of
children who were available for the
intention-to-treat analysis developed
diarrhea. Similar to the intention-to-
.001
.001
treat analysis and the main analysis,
we found a reduced number of vomit-
ing episodes (mean SD: 0.64 1.26
vs 1.38 1.85; P .004), a decreased
.720 risk for repeated vomiting (24 [31.2%]
.627 vs 38 [51.3%]; P .002), and a de-
.001 creased risk for repeated administra-
.669
tion of suppositories (23 [29.9%] vs 43
[58.1%]; P .001). No other signicant
.596 differences were observed. In children
1.000
1.000 with more pronounced diarrhea, simi-
.980 lar results were obtained. For detailed
analysis, see Appendix 5.
DISCUSSION
Dimenhydrinate reduced the number of
vomiting episodes in children who had
vomiting as a result of gastroenteritis
and had no or mild dehydration. The
number of children with complete cessa-
tion of vomiting was 1.5 times higher in
the dimenhydrinate group (70% dimen-
hydrinate versus 47% placebo; NNT 5).
Dimenhydrinate was well tolerated, and
the number of adverse effects was not
different in the 2 groups.
The primary outcome measure was
the difference of body weight between
initial presentation and follow-up visit.
We wanted to investigate whether uid
loss as a result of vomiting is reduced
and rehydration is made easier by di-
menhydrinate. The degree of dehydra-
tion was measured by body weight de-
velopment and by a standardized
assessment scale.2 We could not ob-
serve any effect on oral rehydration.
Antiemetic treatment in children with
gastroenteritis-associated vomiting is
a widespread phenomenon.13 Dimen-

hydrinate is a classic H1-antihistaminic
drug that also shows antiallergic activ-
ity and benecial effects on motion
sickness and drowsiness. It has been
shown to reduce postoperative vomit-
ing in children.15 In Germany and Can-
ada, dimenhydrinate is the preferred
antiemetic treatment in children with
gastroenteritis13; however, concerns
have been raised that dimenhydrinate
might delay other diagnoses in chil-
dren with vomiting,16 and dimenhydri-
nate intoxication has been reported.17
We did not encounter a delay of other
diagnoses as a result of medication
with dimenhydrinate. Our data suggest
that dimenhydrinate is safe in children
with mild gastroenteritis.
Dopamine receptor antagonists such
as domperidone and metoclopramide
are frequently used in Italy, Spain, and
France, whereas, in the United States,
promethazine is the antiemetic treat-
ment of choice in children with gastro-
enteritis.13,18,19 None of those drugs
showed a consistent antiemetic bene-
t in children with gastroenteritis in
randomized, controlled trials.8 The
5-HT3 receptor antagonist ondanse-
tron plays only a marginal role in clin-
ical practice13; however, according to a
recently published meta-analysis of 6
randomized, controlled trials, ondan-
setron decreased the risk for hospital
admission (NNT 14 [95% CI: 9 44]),
for intravenous rehydration (NNT 5
[95% CI: 4 8]), and for persistent eme-
sis (NNT 5 [95% CI: 4 7]).8 In our
study, the benecial antiemetic effect
of dimenhydrinate (NNT 5 [95% CI:
312]) was comparable to that re-
ported for ondansetron. In contrast to
studies on ondansetron, dimenhydri-
nate did not decrease the rate of hos-
pitalization. In contrast to 3 of 6 studies
on ondansetron,8 the rate of diarrhea
was not increased by dimenhydrinate.
According to guidelines for treatment of
infectious gastroenteritis, we did not
perform routine stool smears to identify
PEDIATRICS Volume 124, Number 4, October 2009
the infectious agent, and laboratory
tests such as blood gas analysis were
not performed routinely. The study was
conducted during 2 winter seasons,
when the annual peak incidence of viral
gastroenteritis occurs and bacterial in-
fections are less common.
Because we investigated children with
no or mild dehydration in our study, it is
possible that there is a spectrum effect.
Recently, Goldman et al20 validated a clin-
ical dehydration scale for children with
gastroenteritis. A total of 57% of children
who presented in the emergency depart-
ment had no dehydration, 41% of chil-
dren had some dehydration, and 2% had
moderate or severe dehydration. Ac-
cording to Goldman et al, a mild degree
dehydration is a typical nding in North
America, because children present early
during the course of illness. Our study
included a spectrum of children that is
typical for the majority of children who
have gastroenteritis and present to pedi-
atricians, whereas the studies on ondan-
setron were performed on moderately ill
patients. The authors of a recent meta-
analysis consequently stated that there
is a need to investigate patients with
mild disease.8
We did not perform a cost-effectiveness
or cost benet analysis for using di-
menhydrinate in children with acute gas-
troenteritis. In the United States, costs of
ondansetron have been estimated to be
$15 to $25 (in-house pharmacies up to
$50) per pill,8 whereas costs of dimenhy-
drinate are approximately 0.4 to 0.5 per
suppository of 40 mg. This price differ-
ence is most likely one of the main rea-
sons that dimenhydrinate is frequently
prescribed to children with acute gas-
troenteritis in Germany. Although the
medication costs of ondansetron are
much higher, the overall health care
costs might still be reduced because of
reduced need for hospitalization. In a pi-
lot study that compared ondansetron
with dimenhydrinate in children with
postoperative vomiting, despite higher
ARTICLES
medication costs of ondansetron, the
overall costs were reduced.21 Prospec-
tive, randomized, controlled trials are
needed to compare the antiemetic ef-
fects and therapeutic benets and to es-
timate the overall cost/benet relation of
ondansetron and dimenhydrinate in
children with gastroenteritis-associated
vomiting.
CONCLUSIONS
Dimenhydrinate signicantly reduces
vomiting in children with gastroenter-
itis and is well tolerated; however, the
antiemetic effect of dimenhydrinate is
mild and oral rehydration therapy and
clinical outcome are not improved. It is
not clear whether dimenhydrinate
would have more pronounced effects
in children with a higher degree of
dehydration.
ACKNOWLEDGMENTS
The investigator-initiated study was
funded by the nonprot Hexal-Initiative
on Childrens Medication after a com-
petitive grant application procedure
and independent expert decision. The
grant was provided unconditionally.
Sandoz Pharmaceuticals GmBH Ger-
many (formerly Hexal AG, Holzkirchen)
supplied dimenhydrinate and placebo
suppositories, and Stada Arzneimittel
AG (Bad Vilbel, Germany) supplied the
oral rehydration solution. The compa-
nies had no role in the conception, de-
sign, or conduct of the study or in the
analysis or interpretation of the data.
We thank all the participating parents,
children, and pediatricians. The follow-
ing pediatricians contributed to the
study: S. Beblo, M. Behrens, M. Bern-
hard, A. Bigl, H. Eichner, C. Falkenberg,
L. Fischer, A. Galler, K. Frenzel, D. Har-
tig, S. Herbertz, C. Jorck, A. Keller, A.
Klossek, A. Korner, M. Landgraf, B.
Lesener, S. Liebermann, C. Henn, U.
Mutze, F. Schlensog, A. Siegler, V.
Schuster, P. Suchowerskyj, and F.
Wild (University Childrens Hospital
e627
Leipzig); U. Dietz (St Georg Hospital
Leipzig); T. Kofer, A. Fensterer, and K.
Steul (University Hospital Mainz); and
A. Fiegert, F. Hornemann, A. Rentzsch,
K. Mock, S. Springer, C. Kurzke, G.
Schleicher, J. Streidl, and M. Thole
(Hospital Wurzen). We are grateful to O.
Brosteanu, M. Vissiennon, N. Bauern-
feind, A. Franke, K. Schosnig, and M.
Dorschmann for discussions, monitor-
ing, and data management support.
We also thank P. Schoettler and B. Mu-
hlbauer for continuous support and S.
Koletzko and N. Blanchette for critical
comments on the manuscript.

REFERENCES
1. Elliott EJ. Acute gastroenteritis in children. BMJ. 2007;334(7583):35 40
2. World Health Organization. The Treatment of Diarrhoea: A Manual for Physicians and Other Senior
Health Workers. 4th revision. Geneva, Switzerland: World Health Organization; 2005
3. Guarino A, Albano F, Ashkenazi S, et al. European Society for Paediatric Gastroenterology, Hepa-
tology, and Nutrition/European Society for Paediatric Infectious Diseases evidence-based guide-
lines for the management of acute gastroenteritis in children in Europe. J Pediatr Gastroenterol
Nutr. 2008;46(suppl 2):S81S122
4. Cubeddu LX, Trujillo LM, Talmaciu I, et al. Antiemetic activity of ondansetron in acute gastroenter-
itis. Aliment Pharmacol Ther. 1997;11(1):185191
5. Freedman SB, Adler M, Seshadri R, Powell EC. Oral ondansetron for gastroenteritis in a pediatric
emergency department. N Engl J Med. 2006;354(16):1698 1705
6. Ramsook C, Sahagun-Carreon I, Kozinetz CA, Moro-Sutherland D. A randomized clinical trial com-
paring oral ondansetron with placebo in children with vomiting from acute gastroenteritis. Ann
Emerg Med. 2002;39(4):397 403
7. Reeves JJ, Shannon MW, Fleisher GR. Ondansetron decreases vomiting associated with acute
gastroenteritis: a randomized, controlled trial. Pediatrics. 2002;109(4). Available at:
www.pediatrics.org/cgi/content/full/109/4/e62
8. DeCamp LR, Byerley JS, Doshi N, Steiner MJ. Use of antiemetic agents in acute gastroenteritis: a
systematic review and meta-analysis. Arch Pediatr Adolesc Med. 2008;162(9):858 865
9. Roslund G, Hepps TS, McQuillen KK. The role of oral ondansetron in children with vomiting as a
result of acute gastritis/gastroenteritis who have failed oral rehydration therapy: a randomized
controlled trial [published correction appears in Ann Emerg Med. 2008;52(4):406]. Ann Emerg
Med. 2008;52(1):22.e6 29.e6
10. Vreeman RC, Finnell SM, Cernkovich ER, Carroll AE. The effects of antiemetics for children with
vomiting due to acute, moderate gastroenteritis. Arch Pediatr Adolesc Med. 2008;162(9):866 869
11. Szajewska H, Gieruszczak-Bialek D, Dylag M. Meta-analysis: ondansetron for vomiting in acute
gastroenteritis in children. Aliment Pharmacol Ther. 2007;25(4):393 400
12. Alhashimi D, Alhashimi H, Fedorowicz Z. Antiemetics for reducing vomiting related to acute gas-
troenteritis in children and adolescents. Cochrane Database Syst Rev. 2006;(4):CD005506
13. Pfeil N, Uhlig U, Kostev K, et al. Antiemetic medications in children with presumed infectious
gastroenteritis: pharmacoepidemiology in Europe and Northern America. J Pediatr. 2008;153(5):
659 662, 662.e1 662.e3
14. van Hoogdalem E, de Boer AG, Breimer DD. Pharmacokinetics of rectal drug administration: part
I general considerations and clinical applications of centrally acting drugs. Clin Pharmacoki-
net. 1991;21(1):1126
15. Kovac AL. Management of postoperative nausea and vomiting in children. Paediatr Drugs. 2007;
9(1):47 69
16. Anquist KW, Panchanathan S, Rowe PC, Peterson RG, Sirnick A. Diagnostic delay after dimenhydri-
nate use in vomiting children. CMAJ. 1991;145(8):965968
17. Scharman EJ, Erdman AR, Wax PM, et al. Diphenhydramine and dimenhydrinate poisoning: an
evidence-based consensus guideline for out-of-hospital management. Clin Toxicol (Phila). 2006;
44(3):205223
18. Albano F, Bruzzese E, Spagnuolo MI, De Marco G. Antiemetics for children with gastroenteritis:
off-label but still on in clinical practice. J Pediatr Gastroenterol Nutr. 2006;43(3):402 404
19. Kwon KT, Rudkin SE, Langdorf MI. Antiemetic use in pediatric gastroenteritis: a national survey of
emergency physicians, pediatricians, and pediatric emergency physicians. Clin Pediatr (Phila).
2002;41(9):641 652
20. Goldman RD, Friedman JN, Parkin PC. Validation of the clinical dehydration scale for children with
acute gastroenteritis. Pediatrics. 2008;122(3):545549
21. Piwko C, Lasry A, Alanezi K, Coyte PC, Ungar WJ. Economic evaluation of ondansetron vs dimenhy-
drinate for prevention of postoperative vomiting in children undergoing strabismus surgery.
Paediatr Anaesth. 2005;15(9):755761

e628 UHLIG et al
 ARTICLES

APPENDIX 1
Case Report Form for Time of Randomization (English Translation)

PEDIATRICS Volume 124, Number 4, October 2009 e629

APPENDIX 2
Case Report Form for the Follow-Up Visit (English Translation)

e630 UHLIG et al
 ARTICLES

APPENDIX 3
Case Report Form for the Telephone Interview (English Translation)

PEDIATRICS Volume 124, Number 4, October 2009 e631

APPENDIX 4 Main and Secondary Outcomes According to Intention-to-Treat Analysis
Parameter Dimenhydrinate Placebo P
Available for primary end point, n 114 109
Hospitalized, n 4 5
Absolute weight change, mean SD 0.02 0.23 0.01 0.24 .346
% weight change, mean SD 0.15 1.99 0.08 1.71 .368
Available for secondary end points, n 110 104
Vomiting episodes, mean SD 0.68 1.37 1.41 1.81 .001
Episodes of vomiting 0/1/2/3, n 77/16/7/10 48/20/9/27 .001
Diarrheal episodes, mean SD 1.83 1.99 1.76 1.79 .788
Oral uid intake, mean SD, mL 707 358 672 283 .426
Repeated administration of suppositories, n (%) 34 (30.9) 58 (55.8) .001
Parents reported outcome measure: 1.30 1.35 1.15 1.49 .436
improvement of well-being, mean SD
Parameters refer to the period between visits 1 and 2.

APPENDIX 5 Main and Secondary Outcomes for Various Degrees of Gastroenteritis

Parameter Dimenhydrinate Placebo P
Patients with at least 1 episode of diarrhea in total
Remaining in study ow, n 90 84
Available for primary end point, n 79 76
Hospitalized, n 2 2
Absolute weight change, mean SD 0.03 0.24 0.00 0.24 .425
% weight change, mean SD 0.25 2.07 0.06 1.79 .326
Available for secondary end points, n 77 74
Vomiting episodes, mean SD 0.64 1.26 1.38 1.85 .004
Vomiting episodes 0/1/2/3, n 53/13/5/6 36/14/5/19 .002
Episodes of diarrhea, mean SD 2.32 1.90 2.26 1.76 .820
Oral uid intake, mean SD, mL 713 356 704 299 .884
Repeated administration of suppositories, n (%) 23 (29.9) 43 (58.1) .001
Parents reported outcome measure: 1.17 1.39 0.97 1.50 .402
improvement of well-being, mean SD
Patients with at least 2 episodes of diarrhea in total
Remaining in study ow, n 75 75
Available for primary end point, n 68 67
Hospitalized, n 2 2
Absolute weight change, mean SD 0.01 0.24 0.00 0.23 .850
% weight change, mean SD 0.18 2.14 0.02 1.81 .573
Available for secondary end points, n 66 65
Vomiting episodes, mean SD 0.74 1.33 1.52 1.90 .007
Vomiting episodes 0/1/2/3, n 42/13/5/6 28/14/5/18 .004
Episodes of diarrhea, mean SD 2.64 1.87 2.48 1.76 .616
Oral uid intake, mean SD, mL 714 370 723 304 .879
Repeated administration of suppositories, n (%) 20 (30.3) 40 (61.5) .001
Parents reported outcome measure: 1.14 1.46 0.98 1.58 .564
improvement of well-being, mean SD
Patients with at least 3 episodes of diarrhea in total
Remaining in study ow, n 59 65
Available for primary end point, n 53 59
Hospitalized, n 2 2
Absolute weight change, mean SD 0.02 0.25 0.02 0.23 .380
% weight change, mean SD 0.30 2.21 0.02 1.81 .231
Available for secondary end points, n 51 57
Vomiting episodes, mean SD 0.84 1.46 1.39 1.76 .086
Vomiting episodes 0/1/2/3, n 32/9/4/6 25/13/5/14 .043
Episodes of diarrhea, mean SD 3.04 1.89 2.68 1.76 .315
Oral uid intake, mean SD, mL 746 399 731 292 .838
Repeated administration of suppositories, n (%) 16 (31.4) 35 (61.4) .002
Parents reported outcome measure: 1.18 1.49 1.05 1.48 .659
improvement of well-being, mean SD
Parameters refer to the period between visits 1 and 2.

e632 UHLIG et al