Tuesday, 23 October 2018

The Practicalities of Croup Management in the Community

This post is in response to a very specific question from a local GP. The question wasn't about recognising croup or even about the best evidence based treatment.  Recognising croup is fairly straightforward. There is pretty much consensus on the best management of croup. The question was about the practicalities.

The evidence for the ideal management of croup has given us a fairly straightforward and reasonably robust answer: a single 0.15mg/kg dose of oral dexamethasone.  Sounds simple doesn't it?  The difficulty is that a single dose is actually quite problematic from a pharmacy point of view. As a result the decision isn't always about the best available evidence.  It might also be about the best available medication and formulation.  To determine the answer to this question, we need to go back a couple of steps.

Croup is a clinical presentation involving barking cough, with or without stridor and respiratory distress.  This usually occurs in a relatively well child, though they will have the symptoms of a viral upper respiratory tract infection.  Like so many presentations in childhood, the underlying cause is a viral illness but the problem is due to the effect or response to the virus.  In the case of croup, that effect is upper airway inflammation and swelling.

When should croup be treated?
Croup is usually classified into mild, moderate or severe.  This can be done with or without a croup score.  While it is a minor oversimplification of what happens next, the likelihood is that severe croup will be treated with steroids and often admitted to hospital while moderate croup will usually be treated with steroids and discharged home after a period of observation.

It is the management of mild croup which often generates the most discussion.  The first question is whether it should be treated at all.  There is evidence that treating mild croup with corticosteroids (1) reduces symptoms.  There is the suggestion that it is safer to treat mild croup in that there is a reduction in time spent in hospital and reduced readmission rate for those that are treated.  However there is no specific evidence that not treating mild croup leads to an increased risk of severe or life threatening croup.  This leads some clinicians to the conclusion that if a child has a barking cough but no stridor or respiratory distress, they prefer to provide safety-netting advice and reassess if the child develops new signs.

How should croup be treated?
There is also evidence regarding the most effective steroid treatment for croup in children.  Oral dexamethasone outperforms oral prednisolone.  Both oral treatments outperform nebulised budesonide.  The suspicion is that dexamethasone outperforms prednisolone because it is better tolerated.  It's difficult for a medication to be effective if it's just been puked onto the floor.

If that's all so well evidence based, what's the problem?  Lets's get on with giving them all dexamethasone 0.15mg/kg. The problem with this is that is that dexamethasone liquid has done itself out of a job.

Dexamethasone is given as a single dose in the vast majority of cases.  The evidence shows that this works well, quickly (2) and with an effect which is sustained over several days.  It is quite potent, so small doses are effective.  These factors, combined with an unpredictable demand and a relatively short shelf life make dexamethasone liquid something that doesn't make business sense for pharmacies to stock.

I recently asked the twitter community about what they had available and while many did have dexamethasone liquid, it certainly wasn't routinely available.  The question also sparked a smattering of stories from people who had been sent from place to place looking for one that had some dexamethasone available.

This then presents a dilemma for the clinician in the community.  Do you prescribe the best tolerated and most effective treatment and take the risk that it will be unavailable?  Do you prescribe an alternative (soluble prednisolone) that is known to be slightly less effective and less well tolerated on the grounds that a medication can only be effective if it's actually been given?

There is also an opportunity to be proactive about the issue.  You could get a member of your team to contact the local pharmacies and ask if any of them do stock liquid dexamethasone.  If not, perhaps one would in which case they would be where you sent your children with croup for their treatment.

On a larger scale, primary care groups (e.g. Clinical commissioning groups in the UK) could coordinate something so that each locality has a pharmacy that stocks liquid dexamethasone.

Another way of looking at it is that there is a vicious cycle to break.  Because dexamethasone is not always available, not everyone provides it.  Because it is not prescribed often enough, it is not always stocked by pharmacists.  More prescribing of dexamethasone should make it more likely that dexamethasone will be stocked.

It is possible that liquid dexamethasone will become a more commonly prescribed medication since it has recently been suggested that it is as effective as prednisolone for childhood wheeze. (3)

What about age banding and using soluble dexamethasone?

Dexamethasone has a large therapeutic window.  The current recommended dose of 0.15mg/kg is a quarter of the dose of 0.6mg/kg which was previously the most often used dose.

This is good because age banding doses is very difficult.  A four year old can be anything from 13-22kg based on the 9th-91st centiles of the WHO growth charts.  Knowing the age is therefore nowhere near as good as having an actual weight.  Obtaining a child's weight does not require any special equipment.  If a child will not stand on a set of scales, simply weigh an adult carrying the child and without holding the child.  The difference is the child's weight.

If using Using the 9th-91st weight centiles and aiming for a dose of 0.15-0.3mg/kg gives the following results:

The ideal is definitely to have a weight and to have a liquid suspension available that would allow the precise dose of 0.15mg/kg to be given.  However, when thinking about a plan B, it seems a shame to go to Prednisolone which is known to be less effective, has more side effects and can only be given in aliquots of 5mg.  Why not do the same with dexamethasone, even if it does mean that the dose may be over in some cases?  Again, the therapeutic window of dexamethasone allows this to be possible.

Although liquid dexamethasone is not always on the shelves of the local pharmacy, it probably should be and possibly would be if it was more often used and the pharmacist knew that the bottle would get used.

Edward Snelson
Pharmacoeconomist of the year 2020

Disclaimer - If treatments are better but do not make sense financially, children should have to pay for that themselves.  If necessary, there are some coal mines near me that could be reopened, giving the children an opportunity to earn the money to pay for all the wasted dexamethasone that they are responsible for.



  1. Wonderful thanks Ed.

    So we can get dex in. Lasts 3 months. No problems there.

    1 difficulty is, I'm ashamed to say, for it to prescribed at scale we need age related doses rather than per/kg doses as it's one extra step to make to doing something that is not culturally expected.

    However even if we had age related doses, if we carefully work out '2ml' this is put into a 30ml bottle how much will come out?

    I'm told we can write 'dose 2ml' supply 5ml. Another step likely to be perceived as messy.

    The best solution if poss is to know what age range could have the 2mg soluble tablet? Sorry if this feels lazy and bad medicine. We don't enjoy being rushed and 'new' management can take years to catch on.

    If this is not safe, age related doses for dex liquid and we can create some ccg templates that say how much to both give an supply.

    We mess around doing so many things that don't matter to miss something that with a pragmatic solution could become the norm

    1. You are unlikely to be giving such small volumes. The liquid comes in 2mg/5ml and most children with croup will have a dose of 2mg or more so I think that the error of such a volume is less of a concern. That said, I think that having more dispensed would be better to make sure they get a therapeutic dose since, as you suggest, if the exact dose goes in a bottle and not all of it comes out, the child hasn't had a proper dose.
      With regards to age banding, I'll add something to the article.