Page 116 - 15Diarrhoeaandvomiting

Antibiotic therapy

salmonella continued for significantly longer in the ampicillin group compared with the placebo
group (days until last positive culture WMD 20.40 days; 95% CI 13.49 to 27.31 days) and the
amoxicillin group compared with placebo (days until last positive culture WMD 16.10 days;
95% CI 8.75 to 23.45 days), but differences across all three groups were not statistically significant
(Kruskal–Wallis nonparametric ANOVA P > 0.5). Eight patients in each antibiotic treatment
group developed bacteriological relapse (16/30), although there were no relapses in the placebo
group. This difference was statistically significant (P = 0.003). All but one of the relapses occurred
between days 4 and 20 after entering the study. Three patients with bacteriological relapse from
each antibiotic group also suffered a diarrhoeal relapse (6/30). There were no diarrhoeal relapses
in the placebo group (n = 14), but this difference was not statistically significant.
142
The second three-armed study, from Taiwan, recruited 42 children older than 6 months presenting
to hospital with suspected salmonella enteritis (defined as blood and/or mucoid diarrhoea with
or without fever). Confirmation and serotyping of salmonella was performed using stool culture.
Children were excluded if they had a negative salmonella stool culture, a toxic appearance, were
vomiting, had abdominal distension indicative of sepsis or ileus or if they had taken antibiotics in
the 72 hours prior to admission. Participants were assigned to treatment with oral azithromycin
10 mg/kg per day, in one dose daily for 5 days (n = 14), cefixime 10 mg/kg per day, in two doses
daily for 5 days (n = 14) or to no treatment (n = 14). A computer-generated random number
program was used to assign the children to treatment groups. No details of blinding of outcome
assessors or of a power calculation were given. Patients returned to clinic 7 days after completing
the course of treatment, and every week thereafter for stool sampling and culture, until two
consecutive normal stools were noted. The main outcomes assessed were duration of diarrhoea
and fever after initiation of therapy, and clinical or bacteriological relapse. [EL = 1−]
The three groups were similar at the baseline for sex and clinical parameters but children receiving
cefixime were younger those in the other two groups (P < 0.05). No statistically significant
differences were found between azithromycin or cefixime compared with each other or to no
treatment for the mean duration of diarrhoea post treatment or the proportion of patients with
positive cultures at week 3 post treatment.

A third three-armed treatment trial, from Canada, 143 included children aged between 10 months
and 15 years who were admitted to hospital with a history of diarrhoea, fever for 3 days or more
and/or mucus and blood from diarrhoeal stools. Only those with a subsequent positive culture
for salmonella remained in the study. Children who had received antibiotics in the previous
5 days or who had renal or hepatic disease, blood dyscrasia, or salmonella bacteraemia were
excluded. Participants were randomly assigned to treatment with 20 mg/kg per day trimethoprim
plus 100 mg/kg per day sulfamethoxazole oral suspension four times per day for 7 days (n = 14)
or ampicillin 100 mg/kg per day oral suspension or capsules four times per day for 7 days
(n = 10) or no antibiotic treatment (n = 12) for the management of salmonella enteritis. No
details regarding allocation concealment, sequence generation, blinding of outcome assessors or
a power calculation were provided. Follow-up was by daily physical examination and culture of
stool specimens, during treatment. After treatment had finished, two or three consecutive daily
stool samples were taken for culture at 1 week, 8 weeks and 6 months. Family contacts also had
stool cultures performed at admission and as for participants post therapy. [EL = 1−]
The treatment groups were comparable at baseline for age and clinical parameters. No statistically
significant differences were noted between the trimethoprim/sulfamethoxazole, ampicillin,
or no treatment groups for the mean duration of diarrhoea after start of therapy (2.8, 3.1 and
3 days, respectively), the mean duration of hospitalisation after start of therapy (5.3, 5 and
6 days, respectively) or the mean duration of fever after start of therapy (3.2, 1.6 and 2.6 days,
respectively).
The fourth trial, from Colombia, 144 examined the effect of ampicillin versus placebo on salmonella
infection. One hundred and ten of 282 malnourished infants and children younger than 2 years
admitted to hospital with diarrhoea as a major symptom had salmonella isolated from culture of
stool specimens. Children were recruited into the study once culture confirmation of shigella,
salmonella or E. coli was made from rectal swab and stool specimens taken 12–16 hours
previously. One patient without recognised pathogens was entered into the study for every two
patients with shigella, salmonella or E. coli. Treatments were given intramuscularly (IM ampicillin
versus IM sterile fructose) in the first year of the trial, and orally in the second (oral suspension of

91