Page 123 - 15Diarrhoeaandvomiting

Diarrhoea and vomiting caused by gastroenteritis in children under 5 years

106.5 hours; P = 0.025). However, there were no statistically significant differences between the
antibiotic and placebo groups for the mean number of unformed stools in a 5 day period and
in week 1 and week 2 post treatment for all patients and the subgroups (those with fever and
patients with >3/HPF).

Evidence summary
Four clinical trials performed in South Africa and Mexico examined the use of antibiotics in
children presenting with diarrhoea in whom the results of stool microbiological investigations
were not yet available to inform management. Only one trial was of good quality and its results
suggested that erythromycin treatment reduced the mean duration of abnormal stool consistency.
The other three trials had methodological limitations, evaluated different antibiotics and reported
contrasting results. While one small trial reported that the use of trimethoprim/sulphonamide
had no effect on duration of clinical symptoms (diarrhoea, vomiting or pyrexia) or on the length
of hospital stay, another trial found evidence that this drug reduced the duration but not the
severity of diarrhoea in the first 5 days of treatment. A fourth trial reported that furazolidone and
trimethoprim/sulfamethoxazole achieved a ‘clinical cure’ in all patients within 3 days of starting
treatment. This effect was not seen for patients who were subsequently found to have negative
stool cultures. A protective effect of antibiotic administration was only seen for patients with
positive stool cultures when the data for both antibiotics were combined.

7.8 Traveller’s diarrhoea

Evidence overview
No trials including children were identified but a Cochrane systematic review of antibiotic
treatment for traveller’s diarrhoea in adults was found. [EL = 1+] The authors included all trials
155
in any language in which travellers older than 5 years were randomly allocated to antibiotic
treatment for acute non-bloody diarrhoea and where the causative organism was not known at the
time of treatment allocation. Patients with acute bloody diarrhoea for longer than 14 days were
excluded. Twelve trials were included in the Cochrane review in total, but only nine trials that
compared antibiotic therapy with placebo were relevant to this guideline (n = 1174). Participants
were students, soldiers, tourists, hotel guests or volunteers who had travelled to Mexico (five
trials), Morocco (one trial), the Gambia (one trial), Belize (one trial) and unspecified developing
countries (one trial). The antibiotics used in the trials were ofloxacin, bicozamycin, ciprofloxacin
(two trials), trimethoprim and trimethoprim-sulfamethoxazole, norfloxacin (two trials), fleroxacin
and aztreonam. Although all nine trials reported the mean duration of diarrhoea (assessed by
time to last unformed stool), only three reported the mean and standard deviation and one trial
reported the mean and P value from which a pooled standard deviation was derived.
Four comparisons of antibiotic (n = 199) with placebo (n = 264) were made in three trials and
statistically significant reductions in the mean duration of diarrhoea were found in those receiving
antibiotics (WMD −25.86 hours; 95% CI −32.58 to −19.14 hours). One study reported a mean
duration of diarrhoea of 26 hours in the antibiotic group (n = 8) compared with 60 hours in the
placebo group (n = 9) (pooled SD = 28 hours)
Six trials reported the number of patients cured at 72 hours. There were statistically significantly
more in the antibiotic groups who were cured at 72 hours (330/391) compared with the placebo
groups (154/306) (OR 5.90; 95% CI 4.06 to 8.57).
Change of severity of diarrhoea (number of unformed stools per 24 hour period) over 72 hours
was reported by two trials. There was a small but statistically significant reduction for those
receiving antibiotics (n = 117) compared with those receiving placebo (n = 106) sustained over
72 hours (0–24 hours WMD −1.59; 95% CI −2.66 to −0.52, 25–48 hours WMD −2.10; 95% CI
−2.78 to −1.42, and 49–72 hours WMD −1.38; 95% CI −1.94 to −0.82).

Five studies reported side effects from treatment: 110/523 participants receiving antibiotics
experienced a side effect compared with only 38/339 in the placebo groups (OR 2.37; 95% CI
1.50 to 3.75) although these were said not to have been clinically serious or to have resolved on
withdrawal of the treatment.

98