Page 181 - 15Diarrhoeaandvomiting

Diarrhoea and vomiting caused by gastroenteritis in children under 5 years

US studies may not be the most relevant to UK practice, particularly with respect to subsequent
resource use. It cannot be assumed that the savings imputed from US models of care would be
transferable to an NHS setting.

The inclusion criteria for the studies included in the meta-analysis also varied. One study 163
included children with at least one reported episode of vomiting within the 4 hours preceding
triage whereas another 164 enrolled patients with five episodes of vomiting in the preceding
24 hours. The authors felt that the inclusion criteria for those being given ondansetron should be
restrictive in order to best identify the group of patients who would benefit from ondansetron.
This is in line with the view of the GDG who also agreed that not every child should be given
ondansetron but only those who are most likely to benefit, i.e. those patients who would otherwise
fail ORT and go on to IVT. It is in this group of patients that savings could be made.
None of the three above-mentioned studies reported any significant adverse events or complicating
side effects from the use of ondansetron and the economic analysis has not taken into these
into account. Nevertheless, the BNFC 209 reports several possible side effects from ondansetron,
ranging from headaches to chest pain and seizures. The chance of these side effects occurring,
although small, could lower the health-related quality of life improvement of ondansetron. It is
therefore important to remember the importance of any potential harms that may be of clinical
importance and may differ systematically between those who are treated with ondansetron and
those who are not.
Changes in diarrhoea in response to treatment was not included in this model because it was
unclear whether ondansetron worsened this outcome and more importantly because the GDG
queried the clinical significance of diarrhoeal outcomes reported in the studies. Two trials in
the meta-analysis 163,164 reported a statistically significant increase in the frequency of diarrhoea
as an adverse event of ondansetron. A third study 160 showed that the number of children with
fewer episodes of diarrhoea was less with ondansetron but the difference was not statistically
significant. To more accurately determine the cost-effectiveness of ondansetron, it would be
important to know the clinical significance of any increased diarrhoea and whether it led to a
concomitant increase in the use of healthcare resources.
In the economic model, it was assumed patients are given a single oral dose of ondansetron in order
to reduce vomiting. This is consistent with two of the studies included in the meta-analysis. 163,211
However, the third study 164 gave a single oral dose of ondansetron in hospital but also provided
discharged patients with an additional five doses of ondansetron to be used every 8 hours for
a total of 2 days. Although this approach would increase the cost of ondansetron, repeated
home doses of ondansetron may also help in delivering persistent benefit and consistently reduce
hospital admission. This would clearly have implications for the economic analysis.

Finally, treatment costs were restricted to the cost of ondansetron. To the extent that other costs,
such as staffing, in administering oral ondansetron have been omitted, there will be a bias, albeit
small, in favour of ondansetron in this analysis.
Although ORT has been proven to be a clinically effective and cost-effective treatment for
children suffering from dehydration, it remains underused, especially when the child is vomiting.
Clinicians are more likely to choose IVT in scenarios where vomiting is a major symptom,
and therefore a safe and effective method of controlling vomiting, such as ondansetron, may
increase the use of ORT. The simple model in this appendix is suggestive of potential clinical
and economic benefits of ondansetron; however, more evidence, particularly with regard to
diarrhoeal outcomes, was felt to be necessary by the GDG to justify its use in routine practice.

156