Pediatr Infect Dis J, 2002;21:41116

Copyright 2002 by Lippincott Williams & Wilkins, Inc.

Vol. 21, No. 5

Printed in U.S.A.

Effect of probiotic Lactobacillus strains in

young children hospitalized with acute
diarrhea
VIBEKE ROSENFELDT, MD, KIM FLEISCHER MICHAELSEN, MD, DR MED SCI, MOGENS JAKOBSEN, DR MED SCI,
CHARLOTTE NEXMANN LARSEN, PHD, PETER LANGE MLLER, PHD, PERNILLE PEDERSEN, MD,
MICHAEL TVEDE, MD, HEIKE WEYREHTER, MD, NIELS HENRIK VALERIUS, MD, DR MED SCI AND
ANDERS PRREGAARD, MD, DR MED SCI

duced by 48% (3.5 vs. 1.7 days, P 0.03). At the

end of the intervention, rotavirus antigen was
found in 12% of patients from the treatment
group vs. 46% from the control group (P 0.02).
Conclusions. The two probiotics, L. rhamnosus
19070-2 and L. reuteri DSM 12246, ameliorated
acute diarrhea in hospitalized children and reduced the period of rotavirus excretion. Oral
bacteriotherapy was associated with a reduced
length of hospital stay. The beneficial effects
were most prominent in children treated early in
the diarrheal phase.

Background. Oral bacteriotherapy promotes

recovery from acute childhood diarrhea, but few
strains have been shown to have therapeutic
potentials. We examined the effect of two newly
identified probiotic Lactobacillus strains in
acute childhood diarrhea.
Methods. Sixty-nine children were randomized
during hospitalization for acute diarrhea to receive a mixture of Lactobacillus rhamnosus
19070-2 and Lactobacillus reuteri DSM 12246,
1010 colony-forming units of each strain or placebo twice daily for 5 days. Before selection of
these stains their potential probiotic characteristics were demonstrated in vitro and in healthy
volunteers.
Results. In patients receiving probiotics, the
diarrheal phase was reduced by 20%. The duration of diarrhea was 82 h in the treatment group
vs. 101 h in the control group (not significant, P
0.07). However, 3 of 30 patients from the treatment group vs. 13 of 39 from the control group
still had loose stools at the end of the study
period (P 0.03). In patients with diarrhea for
<60 h before start of treatment (early intervention), a clear effect of the probiotics was demonstrated (80 h in the treatment group vs. 130 h in
the control group, P 0.003). After early intervention, the length of hospitalization was re-

INTRODUCTION

Probiotics are defined as live microorganisms which,

when ingested, may have a positive effect in the prevention or treatment of a specific pathologic condition.1
Probiotics are thought to work by altering the microbial ecology of the host. From this point of view a more
adequate definition, as proposed by Saavedra,2 would
be microorganisms capable of modifying the relationship with our immediate microbial environment in
ways that may benefit human health.
In the past two decades most scientific reports on
probiotics have focused on the lactic acid- producing
bacteria, particularly lactobacilli and bifidobacteria. In
1991 a Finnish group demonstrated beneficial effect of
Lactobacillus strain GG ATCC 53103 (L. GG) in young
children with viral gastroenteritis.3 Subsequently several placebo-controlled studies with L. GG, in particular, but also other probiotic strains, e.g. Lactobacillus
reuteri have been performed, mainly in hospitalized
well-nourished children with predominantly rotavirus
gastroenteritis.4 12 These studies concurrently demonstrate a reduction in the duration of watery diarrhea.
However, factors more specific to the severity of gastroenteritis, such as duration of vomiting and time to
normalization of acidosis, seem to be unaffected by
probiotic therapy. In one recent multicenter study11
the duration of hospital stay was significantly reduced
in patients receiving oral rehydration solution (ORS)

Accepted for publication Dec. 16, 2001.

From the Research Department of Human Nutrition (VR,
KFM) and the Department of Dairy and Food Science (MJ, CNL,
PLM), The Royal Veterinary and Agricultural University; the
Department of Clinical Microbiology, H:S Rigshospitalet (MT),
Copenhagen; the University Clinic of Pediatrics, H:S Hvidovre
Hospital, Hvidovre (VR, HW, NHV, AP); and Department of
Pediatrics, Copenhagen County Hospital, Glostrup (PP), Denmark.
Key words: Probiotics, Lactobacillus, childhood diarrhea.
Address for reprints: Vibeke Rosenfeldt, M.D., University
Clinic of Pediatrics, H:S Hvidovre Hospital, DK-2650 Hvidovre,
Denmark. E-mail VRosenfeldt@dadlnet.dk.
DOI: 10.1097/01.inf.0000014470.82881.25

411

412

THE PEDIATRIC INFECTIOUS DISEASE JOURNAL

containing L. GG compared with patients not receiving

probiotics.
The purpose of this study was to examine the efficacy
of a mixture of selected lactobacilli in children hospitalized with acute diarrhea. A logical management
approach to the evaluation and selection of probiotic
strains was intended. We have established collaboration between laboratory and clinical specialties to identify Lactobacillus strains with potential probiotic characteristics in vitro followed by in vivo evaluation in
randomized controlled trials.
MATERIALS AND METHODS

Patients. The study was performed at the Pediatric

Departments of H:S Hvidovre Hospital and Copenhagen County Hospital, Glostrup, Denmark, during a
6-month period, from December, 1998, through May,
1999. Children 6 to 36 mo old, hospitalized with acute
diarrhea, i.e. 2 or more consecutive loose stools during
24 h, with duration of no more than 7 days, were
eligible for inclusion. Exclusion criteria were underlying chronic disease and prescription of antibiotics during the study period. During the study period ingestion
of fermented milk products containing live bacteria
was not allowed.
Study preparation. The bacterial preparation consisted of lyophilized L. rhamnosus 19070-2 and L.
reuteri DSM 12246. Before selection of these stains
their potential probiotic characteristics, including adhesion and antimicrobial properties, were demonstrated in vitro and in vivo. As described previously,13
L. rhamnosus 19070-2 was chosen because of its extensive ability in vitro to adhere to CaCo-2 cells. L. reuteri
DSM 12246 was selected because of its substantial
antimicrobial activity against pathogenic strains, without suppression of the normal intestinal flora. None of
the strains was in commercial use. After in vitro
studies the safety and colonization abilities of the
strains were demonstrated in healthy volunteers13 and
in adult patients undergoing colonoscopy (V Rosenfeldt
and A Pregaard, unpublished data).
The study preparation was manufactured by Chr.
Hansen A/S, Hrsholm, Denmark, and each dose was
stored in airtight Alu-bags at 20C until the first day
of the intervention period. A dose of 1010 colonyforming units (CFU) of each strain was given twice
daily for 5 days. The identical looking placebo preparation consisted of skim milk powder and dextrose
anhydrate. When administered to the patient, the
study preparation (weight 1.0 g) was dissolved in 2.5 to
5 ml water, milk or diluted fruit syrup.
During intervention the unopened bags were kept at
4C. In June, 1999, the viability of both strains was
confirmed. The numbers of viable cells per dose were:
L. rhamnosus 19070-2, 1.7 1010 CFU; and L. reuteri
DSM 12246, 0.5 1010 CFU.

Vol. 21, No. 5, May, 2002

Study design. On admission the patients were

examined clinically, and the degree of dehydration was
assessed according to the method of Hochman et al.14
Before start of treatment one stool sample for the
detection of bacterial pathogens and rotavirus was
obtained. The children were randomly allocated in a
double-blind placebo controlled fashion to receive either the bacterial preparation or identical looking
placebo. For rehydration, the patients were offered
ORS with a composition as recommended by the European Society for Paediatric Gastroenterology, Hepatology and Nutrition15 (Semper rehydration solution,
Semper, Sweden). If necessary, because of excessive
vomiting and clinical signs of dehydration, parenteral
rehydration was established. The study preparation
was given as soon as possible after randomization and
did not await rehydration. To monitor viral excretion
one stool sample from each day in the intervention
period was requested. Furthermore a stool sample was
analyzed on Day 6 for recovery of the two administered
Lactobacillus strains.
Clinical observations. The duration of diarrhea,
i.e. the time from treatment start until the appearance
of the first normal stool, was chosen as the main
outcome criterion. After receiving instructions given by
the attending nurse, the parents each day completed a
registration scheme. For each child the number of
vomiting episodes, the number of bowel movements
and the rectal temperature were recorded. The parents
assessed stools as watery, loose or normal and were
instructed to record the appearance of the first stool of
normal consistency (time to recovery). In addition the
number of hospitalization days for each patient was
registered.
Microbiologic analyses. Identification of rotavirus
antigen was done by a latex agglutination test using
monoclonal antibody for rotavirus detection (Slidex
Rota-Kit 2).16
Stool specimens were cultured by routine microbiologic methods for bacterial pathogens, i.e. Salmonella
spp., Shigella spp., Yersinia enterocolitica, Campylobacter spp., Aeromonas spp., Escherichia coli (enteropathogenic and enterotoxigenic) and Clostridium difficile.
Lactobacilli were identified by culturing diluted stool
samples on MRS agar (Merck, Darmstadt, Germany).
The isolated strains were initially screened with PCR
by amplifying the internal transcribed spacer region
separating the 16S and 23S ribosomal genes. The
amplified DNA fragments were subjected to digestion
with TaqI restriction enzyme (New England Biolabs,
Inc., Beverly, MA) and then analyzed by agarose gel
electrophoresis. Strains with the same profile as the
two ingested strains were selected, and their identity
was confirmed by restriction enzyme analysis combined with pulsed field gel electrophoresis.

Vol. 21, No. 5, May, 2002

413

THE PEDIATRIC INFECTIOUS DISEASE JOURNAL

Statistics. Statistical analysis was performed with

SPSS for Windows Version 7.0. Students t test and and
the chi square test were used when appropriate. P
0.05 was chosen as the level of significance.
Ethics. The study was approved by The Ethical
Committee of the Municipality of Copenhagen and
Frederiksberg, The Ethical Committee of the County of
Copenhagen and the Medical Agency, Journal No.
5312-19. Informed consent from the parents of the
enrolled patients was obtained.
RESULTS

A total of 86 children were enrolled in the study.

Seventeen patients were excluded. Three patients from
the control group and 2 from the treatment group were
prescribed antibiotics during the intervention period.
One patient from the treatment group and 3 from the
control group who initially fulfilled the inclusion criteria recovered within the first day of intervention. These
4 patients did not pass any stools within the first 24 h
after enrollment and thus did not receive the test
preparation. Four patients from the treatment group
and 4 from the control group were noncompliant to the
protocol.
Sixty-nine children, mean age 17.6 months, completed the study. At the start of the study rotavirus was
identified as the only pathogen in 40 patients (58%). In
6 patients a bacterial pathogen with rotaviral coinfection was found. Rotavirus antigen was identified in
66% of the cases. In 3 patients Campylobacter jejuni,
and in 1 Salmonella typhimurium was cultured from
stool samples. In 20 patients no etiology was found.
Four patients, 2 from each study group, were partially
breast-fed. No significant differences regarding age,
gender or duration of diarrhea before intervention were
found between the study groups. The clinical characteristics and severity of gastroenteritis did not differ
between treatment group and control group (Table 1).
Twelve patients, 4 from the control group and 8 from
the treatment group, received parenteral rehydration
(not significant, P 0.09). Twenty-nine patients (mean
age, 18.4 months) were treated early, i.e. within 60 h
after diarrhea had started. In these patients the mean

TABLE 1. Clinical characteristics of patients who

completed the study
Variable
Age (mo)
M:F ratio
Duration of diarrhea before
intervention (h)
Rotavirus-positive
Vomiting
Fever on admission
(38.5C)
Watery diarrhea on
admission
Dehydration
None
Mild/moderate
Severe

Treatment
Group
(N 30)

Control
Group
(N 39)

18.5 14.4*
15:15
77.4 44.2

16.7 13.0
24:15
72.3 45.3

0.56
0.34
0.64

18 (60)
19 (63)
15 (48)

28 (71)
29 (76)
17 (45)

0.30
0.15
0.64

25 (83)

35 (92)

0.20

16 (52)
15 (48)
0

21 (55)
17 (45)
0

NS

* Mean SD.
Numbers in parentheses, percent.
NS, not significant.

duration of diarrhea before intervention SD was 35 h

16 h (median, 34 h; range, 7 to 60 h). Otherwise the
clinical characteristics of patients treated early were
comparable with characteristics registered for the total
study group. Ninety-three percent had watery diarrhea
on admission, and 82% suffered from vomiting. The
frequency of these symptoms did not differ significantly from the frequency found in the total study
group (see Table 1 for comparison). Eighteen patients
(64%) from this group were rotavirus-positive.
Effect of probiotics. For the total study group no
significant difference in duration of diarrhea after
intervention between the treatment group and the
control group was found (Table 2). In patients receiving
probiotics, the diarrheal phase was shorter, but the
difference did not reach significance (the duration of
diarrhea was 82 h in the treatment group vs. 101 h in
the control group; not significant; P 0.07). However,
only 3 of 30 patients from the treatment group vs. 13 of
39 from the control group still had loose stools 120 h (5
days) after enrollment (P 0.03).
For a closer evaluation of which patients might
benefit from bacteriotherapy, the duration of diarrhea
after treatment was plotted vs. duration before onset of

TABLE 2. Effect of probiotics

Whole Study Group
Variable

Duration of diarrhea after start of treatment (h)

Duration of watery diarrhea (h)
Duration of temperature 37.5C (days)
Loose stools on Day 5
Duration of vomiting (days)
Length of hospital stay (days)
* Mean SD.
Students t test.
Numbers in parentheses, percent.
Chi square test.

Early Intervention (within 60 h)

Treatment group
(N 30)

Control group
(N 39)

Treatment group
(N 10)

Control group
(N 18)

81.5 37.3*
50.7 30.7
1.9 1.7
3 (10)
1.2 1.2
1.6 1.0

101.1 47.6
52.5 32.5
2.1 2.0
13 (33)
1.6 1.4
2.7 2.0

0.07
0.26
0.78
0.02
0.29
0.02

79.6 44.0
42.3 33.8
2.0 2.0
1 (10)
1.3 1.4
1.7 0.7

129.7 23.4
64.8 34.6
2.1 2.0
11 (61)
2.3 1.5
3.5 2.2

0.003
0.12
0.70
0.009
0.13
0.03

414

THE PEDIATRIC INFECTIOUS DISEASE JOURNAL

treatment. The mean values of residual duration (i.e.

differences in response between treatment group and
control group the therapeutic gain) were estimated
(Fig. 1). A clear pattern of higher residual values was
found after early treatment. According to Figure 1 a
definition of early intervention as intervention within
the first 60 h of the diarrheal phase was made.
In patients treated within the first 60 h of illness
(early intervention), a more pronounced effect of bacteriotherapy was found. The duration of diarrhea after
intervention was 80 h in the treatment group vs. 130 h
in the control group (P 0.003). At the end of the
intervention period, 1 of 10 treated early with lactobacilli compared with 11 of 18 (61%) from the control
group still had loose stools (P 0.009).
The length of hospital stay was shorter in patients
treated with probiotics. The duration of hospitalization
was 1.6 days in the treatment group vs. 2.6 days in the
control group (P 0.02). After early intervention
(within the first 60 h of illness), the length of hospitalization was reduced by 48% (duration of hospital stay,
1.7 days in the treatment group vs. 3.5 days in the
control group; P 0.03).
Recovery of lactobacilli from stool samples. On
Day 6 stool samples to identify recovery of the test
strains were obtained from 57 patients (32 of 39 from
the placebo group and 25 of 30 from the treatment
group). Neither L. rhamnosus 19070-2 nor L. reuteri

FIG. 1. Scatter plot of duration of diarrhea after start of

treatment vs. duration of diarrhea before treatment. f, patients
allocated to active treatment; , patients allocated to placebo.
The estimated mean values of residual duration based on kernel
smoothing are shown as broken and solid lines. The distance
between the solid line (treatment group) and the broken line
(placebo group) indicates the therapeutic gain. Based on the
scatter plot early intervention was defined as the first 60 h
illness.

Vol. 21, No. 5, May, 2002

DSM 12246 was recovered from patients receiving the

placebo preparation. L. rhamnosus 19070-2 was identified in 17 patients from the treatment group, and L.
reuteri DSM 12246 was identified in 4 patients. In 2
patients treated with these probiotics, both strains
were recovered. Overall the test strains were identified
in 22 to 25 (88%) of examined and treated patients.
Viral excretion during intervention. Complete
data on viral shedding from Day 1 until Day 5 in the
intervention period was available in 41 patients (25 of
28 rotavirus-positive in the placebo group and 16 of 18
rotavirus-positive in the treatment group).
In patients treated with probiotics viral excretion
was reduced (Fig. 2). On Day 5 rotavirus antigen was
found in 12 of 26 patients from the control group vs. 2
of 16 from the treatment group (P 0.025). On Day 6 a
result of viral examination was available in 38 patients, among these 8 (7 from the control group) continued to excrete rotavirus antigen. After Day 6 no
analysis of rotavirus antigen was made. Thus an exact
measurement of the duration of viral excretion in the
two study groups was not known.
DISCUSSION

We have shown that a mixture of two Lactobacillus

strains (L. rhamnosus 19070-2 and L. reuteri DSM
12246) with distinct characteristics ameliorated acute
infantile diarrhea and reduced the length of hospital
stay. In the present study the beneficial effects of
probiotics were most pronounced in patients treated
early in the diarrheal phase. In previous studies the
association between the duration of diarrhea before
treatment and effect of probiotics has not been directly
assessed, although it has been suggested8 that probiotics might be more effective when given early. In a
European multicenter study11 the mean duration of
illness before intervention was as short as 1.5 days. In
the Finnish hospital-based trials, the mean duration of
diarrhea before intervention was 2.0 to 2.5 days in
most studies. In our study patients treated early may

FIG. 2. Percent distribution of patients excreting rotavirus

during the 5-day intervention period. Complete data from 41
patients were available. , statistically significant difference
between the study groups. Placebo group, n 25; treatment
group, n 16.

Vol. 21, No. 5, May, 2002

THE PEDIATRIC INFECTIOUS DISEASE JOURNAL

have been less ill when admitted, but clinical characteristics, including degree of dehydration, in patients
treated early or later were comparable. In a recently
published study17 on the effect of L. GG in acute
diarrhea, it was advocated that bacteriotherapy should
be administered as early as possible and concomitantly
with oral rehydration solution. Further studies designed to define the optimal time for initiating probiotic therapy are needed.
In industrialized countries infants with acute gastroenteritis are usually discharged after rehydration and
when sufficient fluid intake is obtained. At discharge
most infants still have loose stools. Although no effect
on the frequency of vomiting and the length of the
febrile period was found, probiotics may provide nonspecific benefits and promote general well-being. A
reduced duration of hospital stay after probiotic therapy has been demonstrated previously in the European
multicenter study.11
In a study on rotavirus diarrhea the mean duration
of viral excretion analyzed by enzyme-linked immunosorbent assay was 6 days.18 Applying molecular
biology methods the duration of viral shedding ranged
from 4 to 57 days after onset of diarrhea.19 In our study
data to measure the exact duration of viral excretion
were not available. However, in the treatment group
the number of patients excreting rotavirus 5 days after
initiation of therapy was reduced. Thus our data support previous findings. In children with mild diarrhea
oral bacteriotherapy with L. GG was associated with a
reduced viral excretion.12
Several studies on the effect of probiotics in childhood diarrhea have been performed. L. GG3 8, 11, 12 in
particular, but other strains of human origin as well,
i.e. L. reuteri,9, 10 have been shown in placebocontrolled studies to have a beneficial effect in children
with rotavirus diarrhea. In previous studies medical
staff did the assessment of stool consistency and other
symptoms of gastroenteritis, and registration was suspended after discharge from the hospital. We presumed
that after careful instructions, the parents would be
reliable observers and superior in distinguishing
changes from usual stool consistency in their own child.
The distinction between watery diarrhea and loose
stools may be difficult. Recognizing this potential bias
we chose as a main outcome criterion the first appearance of normal stool consistency, defining this as time
to recovery. This may turn out to be an important
issue in comparing results of our study with those of
many previous hospital-based studies. Furthermore we
have registered stool consistency not only during hospital stay but for at least 6 days or until diarrhea had
stopped. The risk for a protracted course of diarrhea
has been considered in only one larger study.11 Although in the current study we did not assess the risk
for persisting diarrhea, we found that the number of

415

patients with loose stools on Day 5 after enrollment

was significantly reduced after treatment with lactobacilli.
We have used a multidisciplinary approach to study
Lactobacillus strains with probiotic potentials. L.
rhamnosus 19070-2 and L. reuteri DSM 12246 were
selected from in vitro screening.13 After 10 days of
administration of L. rhamnosus 19070-2 and L. reuteri
DSM 12246, at least one of them was recovered in
colonic biopsies of 10 of 11 healthy adults ((V Rosenfeldt and A Pregaard, unpublished data). The importance of carefully selecting probiotic preparations was
recently emphasized. In a comparative study20 the
effect of a mixture of well-described probiotic bacteria
(Lactobacillus bulgaricus Streptococcus thermophilus Lactobacillus acidophilus Bifidobacterium
bifidum) was similar to the effect of L. GG in reducing
the duration of diarrhea in young children. In contrast
a preparation containing the yeast Saccharomyces boulardii and another containing Enterococcus SF68
showed no difference from a control group receiving
placebo and ORS only. We chose as test preparation a
mixture of 2 microorganisms, anticipating that the
combined effect of 2 microorganisms, each suggested to
work by separate mechanisms,21 would be most beneficial. However, a combination of two probiotic strains
did not seem to bring additional value to bacteriotherapy with a single Lactobacillus strain, e.g. L. GG.
For a probiotic effect in viral gastroenteritis, the capacity of Lactobacillus strains to adhere to the intestinal
epithelium and to transiently colonize the mucosa
(represented by L. rhamnosus 19070-2) might be more
important than antimicrobial activity (represented by
L. reuteri DSM 12246).
Our study confirms that selected probiotics can ameliorate acute diarrhea in hospitalized children and
reduce viral excretion. Oral bacteriotherapy was associated with a reduced length of hospital stay. The
beneficial effects seemed to be most prominent in
children treated early in diarrheal phase.
ACKNOWLEDGMENTS
We thank The Clinical Research Unit, Hvidovre Hospital for
statistical advice and for performing the computerized randomization lists; the attending nurse, Susanne Buus, for assisting the
clinical examinations; and Anette Gregersen and Birgit Jrgensen, Department of Clinical Microbiology, Rigshospitalet,
Copenhagen, for performing the analyses of viral excretion.

REFERENCES
1. Vanderhoof JA, Young R. Use of probiotics in childhood
gastrointestinal disorders. J Paediatr Gastroenterol Nutr
1998;27:32332.
2. Saavedra JM. Probiotics plus antibiotics: regulating our bacterial environment. J Pediatr 1999;135:5357.
3. Isolauri E, Juntunen M, Rautanen T, Silanaukee P, Koivula
T. A human Lactobacillus strain (Lactobacillus GG) promotes
recovery from acute diarrhea in children. Pediatrics 1991;88:
90 7.

416

THE PEDIATRIC INFECTIOUS DISEASE JOURNAL

4. Kaila M, Isolauri E, Virtananen E, Laine S, Arvilommi H.

Enhancement of the circulating antibody secreting cell response in human diarrhea by a human Lactobacillus strain.
Pediatric Res 1992;32:141 4.
5. Isolauri E, Kaila M, Mykkanen H, Ling W, Salminen S. Oral
bacteriotherapy for viral gastroenteritis. Dig Dis Sci 1994;39:
2595 600.
6. Majamaa H, Isolauri E, Saxelin M, Vesikari T. Lactic acid
bacteria in the treatment of acute rotavirus gastroenteritis.
J Pediatr Gastroenterol Nutr 1995;20:333 8.
7. Kaila M, Isolauri E, Arvilommi H, Vesikari T. Viable vs.
inactivated Lactobacillus strain GG in acute rotavirus diare .
Arch Dis Child 1995;72:513.
8. Shornikova A, Isolauri E, Burkanova L, Lukovnikova S,
Vesikari T. A trial in Karelian republic of oral rehydration
and Lactobacillus GG for treatment of acute diarrhea. Acta
Paediatr 1997;86:460 5.
9. Shorkinova AV, Casas I, Isolauri E, Mykkanen H, Vesikari T.
Lactobacillus reuteri as a therapeutic agent in acute diarrhea
in young children. J Pediatr Gastroenterol Nutr 1997;24:
399 404.
10. Shorkinova AV, Casas I, Mykkanen H, Salo E, Vesikari T.
Bacteriotherapy with Lactobacillus reuteri in rotavirus gastroenteritis. Pediatr Infect Disease J 1997;16:11037.
11. Guandalini S, Pensabene L, Zikri M, et al. Lactobacillus GG
administered in oral rehydration solution to children with
acute diarrhoea: multicenter study. J Pediatr Gastroenterol
Nutr 2000;30:54 60.
12. Guarino A, Canari R, Spagnuolo M, Albano F, Benedetto L.
Oral bacteriotherapy reduces the duration of symptoms and
of viral excretion in children with mild diarrhea. J Pediatr
Gastroenterol Nutr 1997;25:516 9.
13. Jacobsen CN, Rosenfeldt Nielsen V, Heyford A, et al. Screen-

14.

15.

16.

17.

18.

19.

20.

21.

Vol. 21, No. 5, May, 2002

ing of probiotic activities of 47 strains of Lactobacillus spp. by

in vitro techniques and evaluation of the colonisation ability
of five selected strains in humans. Appl Environ Microbiol
1999;65:4949 56.
Hochman H, Grodin M, Crone R. Dehydration, diabetic
ketoacidosis and shock in the paediatric patient. Pediatr Clin
North Am 1979;26:80326.
Brooth I, Cunha Ferreira R, Desjeux JF, et al. Recommendations for composition of oral rehydration solutions for children of Europe: report of an ESPGHAN Working Group.
J Pediatr Gastroenterol Nutr 1992;14:113 4.
Kohli E, Pothier P, Denis F, Freymuth F, Goudeau A.
Multicenter evaluation of a new commercial latex agglutination test using a monoclonal antibody for rotavirus detection.
Eur J Clin Microbiol Infect Dis 1989;8:2513.
Rautanen T, Isolauri E, Salo E, Vesikari T. Management of
acute diarrhea with low osmolarity oral rehydration solutions
and Lactobacillus strain GG. Arch Dis Child 1998;79:157 60.
Guarino A, Berni Canani R, Russo S. Oral immunoglobulin
for treatment of acute rotaviral gastroenteritis. Pediatrics
1994;93:1216.
Richardson S, Grimwood K, Gorrell R, et al. Extended excretion of rotavirus after severe diarrhea in young children.
Lancet 1998;351:1844 8.
Canani B, Albano F, Cesarano L, et al. Probiotics for acute
diarrhoea: a comparative study [Abstract.] Presented at the
34th Annual Meeting of the European Society for Paediatric
Gastroenterology, Hepatology and Nutrition, Geneva, May,
2001.
Lewis SJ & Friedman AR. Review article: the use of biotherapeutic agents in the prevention and treatment of gastrointestinal disease. Aliment Pharmacol Ther 1998;12:80722.