Page 82 - 15Diarrhoeaandvomiting

Fluid management

Evidence summary

A well-conducted systematic review [EL = 1++] did not find any significant difference in the
incidences of hyponatraemia, hypernatraemia, the mean duration of diarrhoea, weight gain or total
fluid intake in children treated with ORT compared with IVT. Although ORT was associated with
a 4% higher risk of rehydration failure, when the analysis was conducted using a homogeneous
definition of rehydration failure, no statistically significant difference was seen. Dehydrated
children treated with ORT had a significantly shorter stay in hospital and those receiving IVT
had a higher risk of phlebitis but no statistically significant differences were found between the
ORT and IVT groups for the other complications – hypernatraemia, paralytic ileus, abdominal
distension, peri-orbital oedema or seizures. Methodologically, there was great variation between
the trials with regard to the study population characteristics, composition of ORS solution and the
modes of administration of ORS solution.

Cost-effectiveness evidence
The GDG identified two treatment alternatives for children with clinical dehydration as a
priority for economic analysis. The results are summarised below; further details are available in
Appendix A.

A decision-analytical model was developed which aimed to compare the cost-effectiveness of
ORT versus IVT. All children are ultimately rehydrated regardless of which treatment they have
and therefore the model assumed equal clinical effectiveness for both treatment methods. The
83
model probabilities were based on a Cochrane review where the primary outcome was failure
to rehydrate. For patients on ORT, failure to rehydrate implies a requirement for IVT. Theoretically,
IVT should be able to replace fluid lost and manage continuing losses and therefore, for the
purposes of this model, it was assumed that IVT treatment ‘failure’ is where IVT is required for a
longer period of time. Complications from treatment were included in the model but limited to
outcomes where a statistically significant difference was reported at the 5% level in the Cochrane
review. Costs were taken from standard NHS/UK sources and focused on resource use that
differed between the treatment alternatives.
A cost-minimisation approach was adopted for the base case analysis, as the cheapest option is
also the most cost-effective where effectiveness between alternatives is judged to be equivalent.
The base case analysis showed ORT to be the cheapest option. A ‘worst case’ analysis for ORT
relative to IVT was also undertaken. The rationale was to subject this cost-minimisation finding to
the most vigorous scrutiny by biasing model assumptions (within plausible limits) in favour of IVT.
Results of this ‘worst case’ analysis continued to favour the use of ORT as the most cost-effective
method of treating children with some dehydration. Further sensitivity analysis demonstrated
that the finding that ORT is cost-effective is not particularly sensitive to the baseline inputs of
the model. A threshold analysis was undertaken to assess the quality-adjusted life year (QALY)
gain that would be needed for IVT cost-effectiveness, given the differences in cost between the
alternatives. This showed that a larger QALY gain than could ever be expected from a small
improvement in time to cure (rehydrate) would be needed for the expensive treatment option
(IVT) to be considered cost-effective.

5.2.2 ORT versus IVT for children with severe dehydration

Evidence overview
Two randomised trials evaluated the effectiveness and safety of ORT versus IVT in severely
dehydrated children. Both of the trials were conducted in hospital settings – one in Iran and the
other in Indonesia. Owing to the nature of the treatment and control protocols, blinding and
allocation concealment was not done. One of these trials was included in the Cochrane review
84
described above. 83
In the RCT from Iran, the study population included 470 children (age range 1–18 months)
85
presenting with watery diarrhoea (>10 ml/kg per hour), vomiting (more than six times per 24 hours)
and two or more signs of severe dehydration (WHO criteria). They were recruited irrespective
of previous treatment and of their nutritional state, and included those presenting with shock.
Inclusion and exclusion criteria were not well defined and the method of randomisation was not

57