Page 179 - 15Diarrhoeaandvomiting

Diarrhoea and vomiting caused by gastroenteritis in children under 5 years

the extent that ondansetron leads to lower hospitalisation and need for IVT, there will be savings
to at least partially offset some of the costs of treatment.
The analysis calculates the difference in effect between placebo and ondansetron for the
three outcomes from the pooled meta-analysis. Any savings in potential downstream costs is
then calculated and this is attributed to the difference in effect achieved by taking a dose of
ondansetron. This downstream saving is calculated by multiplying the difference in effect by the
costs of treatment. It is also important to note that this saving also needs to be offset against the
cost of treatment, i.e. the cost of ondansetron. Therefore:
Net savings = downstream savings − treatment cost of ondansetron

Results

The results of the analysis are listed in Table B.3. Note that the outcome of ‘cessation of vomiting’
was only reported in two of the trials. 163,164 Net savings are derived by multiplying the difference
in effect by the costs of treatment.
Once the treatment cost is taken into account, this gives a net saving of approximately £52 with
ondansetron compared with placebo. Given that ondansetron also leads to a benefit in terms of
increased cessation of vomiting, this implies that the treatment is dominant, providing additional
clinical benefit at reduced cost.

Sensitivity analysis

It is important in economic evaluation to take into account uncertainty. Therefore, a probabilistic
sensitivity analysis (PSA) was undertaken to reflect the fact that the meta-analysis only provides an
estimate of the ‘true’ effect size. To carry out the PSA, the model was run for 10 000 simulations
with the values of the model probabilities sampled from a probability distribution, as indicated
in Table B.1. The cost values were accepted as fixed as these are based on published cost data.
The results of the probabilistic simulations are shown in Figure B.1.

Table B.2 Costs used in the model

Item Value Source
Ondansetron £3.24 British National Formulary for Children (BNFC 2007) 209
IVT £93.01 a Cost-effectiveness of IVT versus ORT for children with
dehydration (Appendix A)
Hospitalisation £602.00 Cost-effectiveness of IVT versus ORT for children with
dehydration (Appendix A)
a This value of £93.01 is obtained by summing all costs (equipment, labour, tests and consumables) derived for IVT in
Appendix A.

Table B.3 Net savings with ondansetron using the pooled meta-analysis

Outcome Probabilities Net savings with ondansetron
Control Ondansetron Difference in effect
Cessation of vomiting 0.652 0.862 0.210 N/A a
Hospitalisation 0.099 0.039 0.060 £36.12
Need for IVT 0.137 0.339 0.202 £18.79
Total downstream £54.91
savings
Total net savings £51.68
a Vomiting has been assumed to have no associated costs and therefore no savings are made by the cessation of
vomiting.

154