Andrew Tagg. An apple juice a day?, Don't Forget the Bubbles, 2016. Available at:
Pippi, aged 3, has been a little bit unwell of later. Most of her family have had viral gastroenteritis and she has now got it too. She’s been vomiting for the last 24 hours and is struggling to keep anything down. Her parents are concerned that she is becoming dehydrated so they bring her into the ED. She gets a sublingual ondansetron wafer and tries some oral rehydration solution. “Yeuch!” she says as she spits it out, “That tastes disgusting.” You wonder if there is anything else you can try.
Today we are going to take a look at the following paper:-
What population did they look at?
Children aged 6 to 60 months that presented to the study a centre who met these inclusion criteria
- 3 or more episodes of diarrhoea or vomiting in the preceding 24 hours
- less than 96 hours of symptoms
- weighed more than 8kg
- had minimal dehydration
The exclusion criteria essentially rule out children who have a number of pre-morbid conditions or who may have more serious underlying pathology.
What was the intervention they tried?
The intervention group had half-strength apple juice.
What did they compare this too?
This was compared to a standard apple-flavoured, sucralose sweetened paediatric electrolyte solution.
What were their outcome measures?
The primary outcome was treatment failure. This was a composite measure defined as any of the following occurring in the 7 days following enrolment.
- intravenous hydration
- subsequent unscheduled physician encounter for the same illness
- protracted illness
- physician request to crossover groups
- 3% or more weight loss or a worsening of Clinical Dehydration Score
It’s easy to read the abstract of a trial and just agree with the conclusion but we should be more sceptical of what we read. Using a validated tool such as that from the Best Evidence in Emergency Medicine group can help with critical appraisal.
Let’s go through the quality appraisal checklist for a randomized, control trial.
Quality Appraisal Checklist
Does the study population focus on the Emergency Department?
Yes. The children all presented to a tertiary care paediatric hospital in Ontario, Canada.
Were the patients adequately randomized?
Children were randomized using computer-generated block allocation.
Was the randomisation process concealed?
Block randomisation took place with allocation in identical, opaque, numbered envelopes that were kept in a locked cupboard.
Were the patients analyzed in the groups to which they were assigned?
Yes. An intention-to-treat analysis was performed.
Were the study patients recruited consecutively?
No. Patients were recruited 12 hours a day, 6 days a week, between the months of October and April of 2010 through to 2015. Interestingly, over the course of the study 3668 children were eligible but this was whittled down to just 647 after exclusions. 1297 patients were not enrolled as study personnel were not available.
Were the patients in both groups similar with respect to prognostic factors?
Yes, they were.
Were all participants unaware of group allocation?
No. Whilst initial enrollment was concealed the parents were aware of which group they child belonged to. Once they were discharged from hospital the group that received half-strength apple juice were encouraged to give their children their drink of choice (other than balanced electrolyte solutions) whereas the standard group were to continue with usual rehydration solutions.
It is also impossible to disguise the taste of the liquid they were given.
Were all groups treated equally except for the intervention?
Yes, they were.
Was follow-up complete?
Follow-up data was available for an impressive 644/647 patients. The majority of this was by telephone.
Were all patient-important outcomes considered?
Absolutely. Nobody wants their child to have to return to the ED for IV rehydration.
Was the treatment effect large enough and precise enough to be clinically significant?
This study was designed as a non-inferiority trial and powered appropriately.
So we can see that the trial appears to hold up to scrutiny with regard to its method and analysis. What we really want to know is whether diluted apple juice is as good as the usual rehydration solution. The bottom line, according the study authors, is this:-
Among children with mild gastroenteritis and minimal dehydration, initial oral hydration with dilute apple juice followed by their preferred fluids, compared with electrolyte maintenance solution, resulted in fewer treatment failures.
In order to reduce the need for intravenous rehydration the team focussed both on stopping the vomiting (with sublingual ondansetron) and replacing potential losses. Interestingly, 68% of the children in the study had no clinical evidence of dehydration equating to a Clinical Dehydration Score of zero but still received oral rehydration solution or diluted apple juice. Here lies the catch in this study. Many of the patients we see in Australian paeds EDs are minimally or mildly dehydrated and thus the results of this study can probably be extrapolated to them. ORS was designed for children with Cholera who had significant dehydration, ongoing fluid loss, and pathology that affected their ability to absorb enteral fluid (remember those glucose-sodium co-transporters from med school?). ORS is safe and effective in pretty much any degree of dehydration whereas this study only demonstrates diluted juice is useful in the mildest of cases. So by all means start handing out the dilute juice to many of the patients you see but don’t chuck out the Gastrolyte just yet!
Freedman SB, Willan AR, Boutis K, Schuh S. Effect of Dilute Apple Juice and Preferred Fluids vs Electrolyte Maintenance Solution on Treatment Failure Among Children With Mild Gastroenteritis: A Randomized Clinical Trial. JAMA. Published online April 30, 2016. doi:10.1001/jama.2016.5352 Full text here
Jauregui J, Nelson D, Choo E, Stearns B, Levine AC, Liebmann O, et al. (2014) External Validation and Comparison of Three Pediatric Clinical Dehydration Scales. PLoS ONE 9(5): e95739. doi:10.1371/journal.pone.0095739 Full text here