Skip to content

Varicella and NSAIDs – are you too chicken to prescribe?


Jaxon is a 3-year-old boy brought to the ED by his mother with chickenpox because of decreased oral intake. On examination, he has a rash consistent with chickenpox but no red flags.

He is febrile and looks uncomfortable, and you note his heart rate is high. You go to prescribe paracetamol, but his mum has given him four doses in the past 24 hours already.

You wonder if you can prescribe ibuprofen instead, but your colleague warns against it.

Where did this controversy arise?

There is a long history of anecdotal evidence associating invasive group A Strep (GAS) complications or severe skin and soft tissue infections (SSTIs) with exposure to NSAIDs. The chickenpox story started with a series of case reports linking ibuprofen with invasive GAS, prompting an attempt to investigate the association more rigorously.

There are currently 5 papers, ranging from 1997 to 2008, which try to answer this question. I will not take these individual studies to pieces here, but feel free to delve deeper into each study using this fantastic resource by the UK Medicines information agency on this topic.

How did they answer the question?

Almost all the studies used a case-control method to try and answer this question.

What does that mean? They took a group of children who had varicella and the outcome of interest (invasive GAS infection, severe SSTI). They compared them to a group of children who had varicella and did not get these outcomes, seeing which groups were more likely to have had ibuprofen.

What did they find?

The studies are pretty heterogeneous, so unsurprisingly, the results varied. One study found no association. Most found an association between ibuprofen exposure and severe GAS/SSTI with an Odds Ratio of between 2 and 5 (one had an OR of 10, but the study was too small to trust).

So, what’s the controversy?

These studies all found an association, but they generally all suffer from the same big problem, which is confounding by indication.

What does that mean? Basically – if you have more severe varicella or are developing complications, you’re more likely to need ibuprofen. It might not be that ibuprofen causes complications, but rather bad varicella needs ibuprofen and is also more likely to get complications anyway. As the famous saying goes, “Correlation does not equal causation”.

There have also been suggestions that easing symptoms with ibuprofen may have delayed initial medical consultation for complications at an earlier stage, which could have prevented subsequent hospitalization.

So how can we definitively answer this question? We probably can’t without randomizing children to ibuprofen vs paracetamol in the setting of varicella, which is not likely to happen.

How do we manage this potential risk?

Most authorities have taken a risk-averse standpoint (including NICE) to avoid ibuprofen due to the presence of uncertainty regarding this association, given that a perfectly acceptable alternative exists in paracetamol. Seems reasonable.

However, professionally, we are often in a scenario when the child is in hospital, paracetamol hasn’t cut the mustard, and we want to use it. Are you running a risk? Are parents who are desperate for symptom relief putting their child in imminent danger if they give ibuprofen?

Allow me some back-of-the-envelope maths for the Worst Case Scenario (WCS).

The annual incidence of varicella in the UK is roughly 13,000/100,000 children <5 years per year.

The incidence of necrotising fasciitis from chickenpox is roughly 0.05/100,000 children per year.

If (WCS) we assume they ALL occur in the under 5 group, this gives a rate of roughly 0.04/10,000 cases of necrotising fasciitis in chickenpox per year.

Let’s say ibuprofen has an Odds Ratio of 5 for necrotising fasciitis in varicella (WCS). This means if you have chickenpox and you are given ibuprofen, your risk goes from 0.04 in 10,000 to 0.2 in 10,000.

An Absolute Risk increase of 0.0016% (WCS).

Bottom line

Ultimately, everyone admits there is a huge degree of uncertainty in the possibility of ibuprofen increasing your risk of complications in varicella. You must assess the risk for yourself and decide what you feel most comfortable advising people. A reasonable approach might be to advise using paracetamol in the first instance and to consider the use of ibuprofen with a low threshold for consulting a medical professional if the child remains distressed.


  • Alasdair Munro is a Paediatric registrar in the UK, currently working as a Clinical Research Fellow in Paediatric Infectious Diseases. His interests include evidence based medicine, diagnostics and antimicrobial resistance. @apsmunro | Ally's DFTB posts


No data was found

Leave a Reply

Your email address will not be published. Required fields are marked *

3 thoughts on “Varicella and NSAIDs – are you too chicken to prescribe?”

  1. Other common outcomes that are relevant are febrile convulsions. I bet this risk of convulsions is lower in the ibuprofen group.As a clinician you must way up a lot more than just the extremely small risk of necrotising fasciitis. This is so extremely rare in general practice; I have not seen a single case in 25 years.

  2. Thanks Tom for the response, and to Alasdair for the original posting. At the risk of sounding like a boring pedant I do think terminology is vital here; and I think is the cause of the original problem with this clinical conundrum.

    Absolutely everything revolves around the term ‘routine’. I am not sure Alasdair ever used that term but I can see the the title of the blog implies it. The core thing is, and I think we all agree on this, is that the routine treatment of distress for chicken pox is with paracetamol. This is because we know paracetamol effectively reduces fever (just google it) and there is clearly something about ibuprofen and chicken pox is relevant (however ill-defined) in studies. However because (and I should stop preaching this; I should do the study) we don’t have a recognised ‘distress’ scale in children; if you don’t appear to respond to paracetamol there is a dilemma. If you are a parent current advice suggests leaving your child distressed with the only option to seek medical help if you are concerned. If you are clinician you would like, because a ridiculously low proportion of children with chicken pox have serious bacterial illness, to be able to discharge children who don’t have current features of red flag sepsis. Normalising physiology (NOT FEVER!) is a methodology for this. My personal view is that considering the use of ibuprofen in a child with unrelenting distress secondary to chicken pox, after an appropriate dose of paracetamol has been given, is NOT routine. Understanding how black and white lawyers might be with the current advice however I would suggest defending yourself with the fact that the CKS is an evidence summary with expert opinion but not the open strength of consensus that a formal NICE guidance affords. Sadly NICE have chosen not to actively measure the economic impact of advice given in equivocal areas of practice (i.e does the cost of attendance at an ED because Ibuprofen has not been given outweigh the risk of invasive disease). This, along with the urgent need to develop a distress scale in children I think does need studying. Please let me know if you are interested in doing this work!

    Just giving ibuprofen without thought should not be recommended.

    But determining how much thought means you move away from being a ‘routine’ case far more difficult to define. But if it helps make your decision, the risk of severe skin infection is much lower than the reported risk of anaphylaxis from penicillin [1] and we don’t not prescribe that on that basis

    [1] The facts about penicillin allergy: a review

  3. Thomas Waterfield (PEM Belfast, Research Fellow QUB)

    This is an excellent summary. I agree with the findings in principle. I would however, warn against the routine use of Ibuprofen in Varicella. It should be used only in exceptional instances and the reasoning for this clearly documented including discussion of risks with the family. I would also encourage all junior doctors to avoid deviating from the guidance given by NICE in their CKS ( without first involving a senior clinician.

    Although the risk of severe skin reactions is low and the evidence poor NICE has chosen to advise against the use of Ibuprofen. Their take on the literature is pasted below.

    “CKS has not recommended the use of nonsteroidal anti-inflammatory drugs (NSAIDs) because of concerns that use of NSAIDs in children with varicella is associated with an increased risk of severe skin and soft tissue infections (usually caused by group A streptococcus and Staphylococcus aureus) [UKMi, 2016]. This is supported by expert opinion in review articles [Gould, 2014; BMJ Best Practice, 2016].A literature review by UK Medicines Information found a number of cohort and case-control studies and case series suggesting an association between use of NSAIDs in children with chickenpox and severe skin reactions. Although the confounding factor of the increased likelihood of NSAIDs being required for children with severe cases of chickenpox or complications was noted, on the basis of the available evidence the authors recommended avoiding NSAIDs in children with chickenpox. Paracetamol was recommended as the preferred alternative treatment [UKMi, 2016]. The risk of skin reactions in adults with chickenpox using NSAIDs is unclear as most of the available trials involved children [UKMi, 2016]. Until further evidence is available, CKS recommends avoiding NSAIDs such as ibuprofen in adults with chickenpox, in line with expert opinion in a review article [Gould, 2014].”

    In summary I agree with the DFTB summary and this makes for an interesting debate. I would however discourage the routine use of Ibuprofen and I would be mindful of the risk/benefits of using Ibuprofen in children with Varicella infection. Whilst an expert may be able to defend the decision to deviate from accepted practice a junior doctor/non-expert might find it a little harder to justify that decision.