Page 133 - 15Diarrhoeaandvomiting

Diarrhoea and vomiting caused by gastroenteritis in children under 5 years

The RCT conducted in Venezuela 165 recruited children aged 6 months to 8 years with gastroenteritis
with emesis who had vomited twice within 1 hour (n = 36). Patients were hospitalised for a
minimum of 24 hours. Exclusion criteria were severe dehydration, seizures, rectal temperature
of 39 °C or more, parenteral anti-emetic medication in the 6 hours prior to the start of the study
or parasite-induced gastroenteritis. The trial had three arms and participants were randomised
to a 10 minute infusion with 0.3 mg/kg IV ondansetron (n = 12), 0.3 mg/kg IV metoclopramide
(n = 12) or a 15 ml bolus of normal saline (placebo) (n = 12).

The methods of randomisation were not reported although it was stated that pharmacy-controlled
medication preparation permitted the patients and outcome assessors to be blind to treatment
allocation. The proportion of patients randomised but lost to follow-up was less than 20%. No
sample size power calculation was provided.
The groups were different at baseline for age, height, weight and degree of hydration, with
comparability on gender and food intake only. There were no statistically significant differences
between the IV ondansetron and placebo groups in vomiting or diarrhoea episodes in the first
24 hours. Ten of the 12 children receiving IV metoclopramide had more than four episodes
of diarrhoea during the first 24 hours compared with 4/12 children in the placebo group. The
difference was statistically significant (RR 2.50; 95% CI 1.08 to 5.79). However, no statistically
significant difference was found for emetic episodes between the two groups in the first 24 hours.

Evidence summary
There was evidence from three RCTs [EL = 1+] that supported the effectiveness of oral ondansetron
in the treatment of gastroenteritis in children. The meta-analysis performed after extracting the
data from two RCTs showed that children with gastroenteritis and receiving oral ondansetron
along with rehydration solution were more likely to stop vomiting. Pooled data from three trials
demonstrated that the ondansetron group were less likely to require further IVT and less likely
to be hospitalised compared with children who had received only rehydration solutions and
placebo. No consistent results were found for diarrhoea outcomes. Two of the three trials reported
statistically significant results to show that children receiving ondansetron seemed to experience
more episodes of diarrhoea.
There was a lack of high-quality evidence for the effectiveness of IV ondansetron, IV metoclopramide
and IV dexamethasone in the treatment of children with gastroenteritis. A small RCT [EL = 1−]
showed no difference in the cessation of vomiting during the first 24 hours following treatment
in children receiving IV ondansetron or IV metoclopramide compared with children treated with
placebo. The risk of having more than four diarrhoeal episodes was higher in both the treatment
groups (IV ondansetron group and IV metoclopramide group) compared with the placebo group,
but the difference was statistically significant only for the IV metoclopramide group.
However, a second underpowered trial did show more children given ondansetron did not
require hospitalisation and tolerated ORT more quickly than those given placebo. No statistically
significant differences were found between the IV dexamethasone and placebo groups for
hospitalisation rates or ORT tolerance.

Cost-effectiveness evidence

A simple economic model was also developed which demonstrates potential economic
advantages of ondansetron if given to children with persistent vomiting in whom IV fluids are
being considered; further details are provided in Appendix B. Owing to limited evidence for the
efficacy of IV ondansetron, the economic analysis only considers the use of oral ondansetron.
The use of anti-emetics such as ondansetron may be effective in the cessation of vomiting and
may in turn help with the successful delivery of ORT, thereby reducing the need to treat with
IVT. This would have cost-saving implications for the NHS through fewer admissions for IVT. The
analysis uses a meta-analysis of three RCTs 160,163,164 to estimate the difference in effect between
placebo and ondansetron for three outcomes: cessation of vomiting, hospitalisation and the need
for IVT. Savings in downstream costs were calculated based on the difference in effect obtained
with ondansetron. The net cost was calculated by offsetting the cost of ondansetron against the
saving achieved with reduced downstream resource use. The analysis suggests that ondansetron,

108