Friday, 21 May 2021

Joining The Dots - How to recognise the seriously unwell child

In the previous post, I outlined how to tell the difference between abnormal signs that are part of a functional and fully compensated response and those that are part of an illness that is having a more significant clinical effect.

The child who is completely well and the child who is visibly seriously unwell and decompensating are both fairly straightforward scenarios.  The child who is very well causes no clinical concern.  The child who is decompensating causes unambiguous and immediate clinical concern.  If only decision making was always this easy.

The child who is febrile with tachycardia but a reassuring level of activity and interaction is also relatively straightforward.  The trouble is that we end up having to make decisions about those children who have a reduced activity level or whose interaction is not completely reassuring.  Often there is a disconnect between what guidelines tell us to worry about and how worried we actually are.  I believe that this often occurs when we intuitively include something in our decision making that is rarely featured in guidelines: the pattern of the illness.

Everything written tends to be geared towards the snapshot:

  • What is the heart rate?
  • What is the temperature?
  • How active and interactive is the child?

This is problematic in any acute specialty assessing febrile children.  Catch the child at the wrong time and they seem to trigger multiple red flags.  Base your assessment on the snapshot alone and you may be falsely reassured.

As covered in the previous post, physiology in young children (not so much babies and older children) responds to illness with what can be dramatic changes even in uncomplicated low-risk infections.  Unfortunately the same abnormalities can be seen in more clinically significant infections.  No one thing is particularly sensitive or specific when deciding whether to be worried.

So if a snapshot can be misleading and there is significant overlap between low-risk and high risk scenarios, how can we decide when a clinical presentation is high risk for sepsis or serious bacterial infection?

The pattern of symptom progression is probably the answer.  In the past, research has concentrated on the snapshot, over-emphasising the assessment of various parameters at a single point in time.  Human intelligence allows us to incorporate the more complex business of considering three very important factors that previous research has not often considered or emphasised:

  • The most recent worst state of the child
  • The most recent best state of the child
  • The pattern and progression of symptoms over time

The evaluation of the best and worst states and the pattern may be include retrospective (history) and prospective (a period of observation) information.  The likelihood is that you already put a great deal of weight on these factors in your decision making.  It is also likely that you have recognised that there are two main patterns of illness in children.  I have represented these in the following diagram, without labelling which is the low risk scenario and which is high risk.

Intuitively, I would assume that you recognise the dotted line as the high risk pattern and the solid line as the typical pattern of a low risk clinical picture of childhood febrile illness.

I suspect that the explanation for this is that the child represented by the solid line is exibiting a physiological reponse to an infection that does not cause organ dysfunction.  Therefore as their immunological response fluctuates, they swing from one extreme to the other.  The child represented by the dotted line is sufferng from an infection which is causing significant physiological dysfunction.  The result is an inability to return to normal.

If you buy that, have another look at the diagram and look at the lines in the first half from a time point of view.  In the left hand part of the diagram, severity of symptoms is often worse for the child who swings from being more unwell to a return to baseline.  I believe that this is the reason that formulaic risk-assessment for serious illness that is based on a snapshot is impractical and problematic.

The end results of this over-emphasis on clinical information taken at a single point in time are twofold:
  • Using a snapshot will over-diagnose serious illness, simply due to the pre-test probability of sepsis and serious bacterial infection (SBI) in the low-risk (which is most children) child.
  • Over time, the repeated realisation that children usually have an uncomplicated, self-limiting illness despite the severity of symptoms and abnormal physiological parameters risks de-sensitises the clinician to the possibility of sepsis/SBI.  Abnormality becomes normalised and we learn to ignore things that have poor specificity for a serious outcome.
I would strongly encourge stepping back from the snapshot to see the big picture over time.  Doing so, either through the history, observation or both might help you to recognise which children are best managed symptomatically and with good safety netting advice.  It migh also allow earlier recognition of the less common scenario of the child who persistently fails to return to (or close to) baseline and is more likely to have a more significant illness.

Telling the difference between self-liniting childhood illness and serious infection is complex and requires us to process an awful lot of information.  When it comes to features like best and worst states and the pattern of symptoms over time, your intuition is probably already joining these dots for you.  I think you should trust that intuition.

Edward Snelson
Dot-joiner but likes to colour outside the lines

Thursday, 1 April 2021

You are Entitled to Compensation - When does "abnormal" physiology mean that I should worry about an unwell child?

 Assessing the severity of a respiratory problem in children is not entirely straightforward.  Guidelines attempt to categorise according to specific parameters however it is not uncommon for the category given to contradict our gut feel.  Sometimes the answer suggested by the guideline seems to be contradicted by the appearance of the child in frontof us.  Why is that?  What weight can we put on our gut feel?  When should we be worried?

To understand any severity system and to explore the nuances, I would ask the question, “What is the correlation between the words (mild/ moderate/ severe) and the clinical state of the patient?”

The clinical state of the patient could be categorised as follows:

  • Normal physiology
  • Fully compensating
  • Partially compensating
  • Decompensating

Normal physiology

In a normal physiological state, circulation and respiration are unaffected.  The child’s activity and interaction are normal.  This normal behaviour is evidence of end-organ function – the brain is well perfused, hydrated and nourished.

Fully compensating

When a child becomes unwell, physiology alters to compensate for the illness.  For example, a baby with mild bronchiolitis will breathe faster and heart rate will increase to mitigate the effect of the viral illness on their lungs.  If this compensation is fully effective, activity and interaction will be unaffected.  In this context, the infant is often referred to as a “happy wheezer”.  Thus happy wheezer is not a diagnosis, but rather a clinical evaluation.

Partially compensating

As the effects of an illness become more significant, physiological compensation increases but there reaches a point where it is no longer fully effective.  The brain is a sensitive organ and the early effects of reduced perfusion, hydration and respiration are usually apparent in the gross neurology of a child.  Activity and interaction becomes reduced as an early sign that compensatory physiology is no longer fully effective.


If physiological compensation begins to fail, the child’s clinical state deteriorates rapidly.  If the problem is primarily respiratory, the deleterious effects on the central nervous system and the respiratory muscles create a vicious cycle: tired muscles and a dysfunctioning brain are unable to continue driving the compensatory mechanisms.  Respiration is further impeded which in turn reduces the oxygen delivery and CO2 clearance so badly needed for the child’s physiology to compensate.

If the problem is primarily circulatory the point of decompensation leads to reduced cerebral, cardiac, hepatic and renal perfusion.  This accelerates the effects of poor perfusion through a combination of worsening cardiac output and biochemical changes that impair the metabolic functions needed to cope.

In either case, the effects of decompensation are dramatic.  Conscious level is affected and the child is likely to be pale and look seriously unwell.  Do not mistake a falling heart rate or respiratory rate for clinical improvement.  What is happening is quite the opposite.  The child’s physiology is failing – impending cardiorespiratory arrest is inevitable without critical care support.

Using this model for assessing the unwell child, we can see that "abnormal" physiology which is fully effective (e.g. a child with vial wheeze who has tachypnoea and mild recession but is running around and happy) is very different to abnormal physiology with evidence that there is only partial compensation (e.g a child with viral wheeze, tachypnoea, recession and reduced activity level).

So when we assess a child and see that changes in physiology are fully effective, we can see this as a very different evaluation to that of the child with signs that may indicate compensation that is not fully effective.
So any score which gives excessive weight to more objective emasures will fail to distinguish the fully compensating child from the child who has a more clinically significant illness.  Normal end organ function, usually manifested in the behavior and activity of a child, is probably the more important part of the assessment.  Abnormal physiology tells us that the child is unwell, whereas the clinical effect is what should tell us when to be concerned. (1)

That, dear friends, is why you find yourself perplexed when a guideline says that you need to treat "abnormal numbers" as red flags when the child you see in front of you seems to be basically well apart from these physiological changes.  In other words, your gut feel that the child is fine is valid.  It is valid because the child who is fully mobile and normally interactive either has no problem or has fully compensated for the problem.  That's what they do.  When they are not fully compensating, then it's time to worry.

Edward Snelson

  1. Snelson E, Ramlakhan S, Which observed behaviours may reassure physicians that a child is not septic? An international Delphi study, Arch Dis Child, 2018 Sep;103(9):864-867. doi: 10.1136/archdischild-2017-314339. Epub 2018 Mar 15.

Wednesday, 10 March 2021

Newborn Presentations

People get worried about newborn babies.  When presented with a baby problem, there are a few basic rules to apply:

Most things that newborns present with are unlikely to have a significant cause.  Some presentations are more concerning that others and while there are no absolutes, it is good to know which things to be more suspicious of.

In many cases, it is possible to significantly change the index of suspicion by knowing the red flags to look out for.

Let's look at a few examples:

Imperfectly shaped head

Since the discovery that putting a baby to sleep on their back dramatically reduces the risk of sudden infant death, there has been a significant rise in the number of babies and infants with asymmetrically shaped heads.

Picture credit:

The vast majority of babies with this presentation will have plagiocephaly - the result of gravity on a compliant skull.  Plagiocephaly has no apparent risk of harm and requires no intervention.  However, this presentation nicely illustrates the general principles of assessing any presentation in a baby because there is a significant pathology that is rare but significant - craniosynostosis.

So while the head shape is very likely to be nothing to worry about, it is still necessary to look for evidence to support that decision.  The head should be palpated and measured (and plotted), the baby should be given a neurological examination and development should be assessed.

It also follows another rule of newborn presentations: where there is no credible pathology, we should do as much nothing as possible.  We should avoid unnecessary tests and treatments. Remember that the baby is your patient.  Don't do anything to them that does not stand to benefit them directly.

Sticky eyes

While the temptation is to presume that sticky eyes are due to infection, this is rarely the case in newborns.  In most cases the problem is a blocked tear duct.  True eye infections are uncommon and usually quite obvious.  This presentation nicely illustrates the general principles of:

  • Assess the problem - Is the conjunctiva red?  Is there periorbital redness and swelling?
  • Look at the baby - Do they appear well?
  • Decide if there is a significant problem (ophthalmia neonatorum)
  • If not, don't do an unnecessary test (swab) or give an unecessary treatment (antibiotic drops)

Excessive crying

Babies cry.  How much is excessive is hugely subjective and open to interpretation.  Although excessive crying in the absence of pathology has been given a name (colic) this scenario demonstrates the principle that we can't always offer a diagnosis.  Calling the problem colic implies that we know what causes the problem and validates an interventional approach.  There is no treatment for excessive crying in the absence of pathology that has a good evidence base.  Sometimes honesty about uncertainty and futility of intervention is the best policy.

Regurgitation of Feeds

Note the use of the word regurgitation.  People often use the word vomiting or even "projectile vomiting" when neither is what is happening.  Babies often bring back some milk after a feed.  This is generally a passive event, as opposed to vomiting, which is what happens when peristalsis works in reverse or against an obstruction such as pyloric stenosis.

By now, you probably know what I'm going to suggest.

  • Assess the problem - Is regurgitation the only symptom?  Are there red flag features (see above)?  How long has the problem been occurring?  Has it changed much and if so how rapidly?
  • Look at the baby - Do they appear well?  Do they appear well grown?  Plot their growth on a chart.
  • Decide if there is a significant problem (e.g. Pyloric Stenosis) or whether they could be in a feed-cry cycle.
  • If not, don't do an unecessary test or give an unnecessary treatment.  If you give alginates to every regurgitating baby, you'll double your workload as they come back the next week with constipation and without the original symptom having improved.

Not opening bowels for X number of days

Guess what?  Yes: in most cases, the baby who has not passed a stool for the past few days is usually going through something that is normal in the first few weeks of life.

As with all of the low risk newborn presentations, if the baby looks well, examines normally and is growing and developing normally, they are normal.  If something is significantly wrong, it should manifest itself in the history or examination.  In most cases, we should normalise this presentation rather that give something to treat it.

It can be really difficult to do nothing when faced with a baby and a concerned parent.  However, a careful clinical assessment and evaluation are the most important interventions you can offer.  If you have done that and not come up with something serious, explanation and safety netting are the premium service, not the economy class package.

Umbilical presentations

There are rare and significant pathologies that can affect the umbilicus.  Most of these will present at birth or in a way that alerts the clinician to the fact that something is clearly not right,  The more common presentations and the best approach to each are:

Skin presentations

Lots of things can happen to the skin of a baby in the first few days and weeks.  Most presentations are either normal phenomena (peeling skin), dysfunctional but harmless (erythema toxicum) or problematic but mostly uncomplicated (cradle cap).

Picture credit: Skin Deep - a DTFB project

In each case there are simple questions to be answered

  • Peeling skin - Is the baby well?  Is there dermis exposed by the peeling skin (if so then epidermolysis bullosa is a possibile diagnosis)?
  • Erythema toxicum - Is the baby well (erythema toxicum is completely harmless)?
  • Cradle cap (seborrheaic dermatitis capitis) - Is the baby well?  Does the skin have signs of infection?

Peeling skin and erythema toxicum are best left alone.  Cradle cap can be treated with olive oil in most cases.  Occasionally it can become infected.

Jittery Movements

Sleep myoclunus is a normal phenomenon at all ages.  In babies it can cause people to worry thet their child is having a seizure.  This is partly due to the protective reflexes that they are born with such as the moro reflex.

Here we go again with the standard procedure for a newborn presentation:

  • Assess the problem - Are the movemments occuring during or around sleep time?  Is the baby otherwise normal in between episodes?
  • Look at the baby - Do they appear well?  Are they developing normally. Measure and plot head circumference.
  • Decide if there is a significant problem (e.g. Infantile spasms/ West syndrome)
  • If not, don't do an unecessary test or give an unnecessary treatment.  Explain, reassure and give safetynetting advice.

Demedicalising infancy is a good thing to do.  The simple apprach of assess, look and decide will allow you to do that in the majority of cases.  In the rare cases of the discovery of a red flag or atypical presentation, there are always the options of advice or referral.

Edward Snelson

Thursday, 25 February 2021

Everything has changed - Non-blanching rash in children

If you told me that nothing has changed for you this past twelve months, I'd be quite surprised.  This year has been a rollercoaster both in and out of work for every healthcare professional I know.  As a finishing touch to the year that has changed everything, I have one more bit of news for you that will change your practice.  Strap in.  This one is huge.

So how did we get there from where we used to be? - "Fever with non-blanching rash is meningococcal sepsis until proved otherwise."

It has long been recognised that this outdated adage has become obsolete in a population with effective meningococcal vaccination.  When it was first coined, the pre-test probability of meningococcal disease (MD) in a child with fever and non-blanching rash was around 1 in 5.  In an unvaccinated population, the 20% chance of MD is more than enough reason to have a "treat in every case" approach in the absence of a rapid diagnostic test.

Since the introduction of a very successful meningococcal vaccination program, the prevalence of MD has dropped dramatically.  The absolute number of cases of MD (the numerator) became a fraction of the pre-vaccination years.  We continued to see large numbers of children with fever and non-blanching rash (the denominator) but no-one was recording how many.

Enter the PiC study, (1) possibly the most significant academic publication of recent years regarding the management of the unwell child.  This large UK based multicentre prospective study did a couple of very important things.  Firstly, it collected data about the prevalence of meningococcal disease in children with fever and petechial rash.   The number that it found was, as we had all hoped and expected, small.   That number was about 1%.

That was only part of the clinically important information that the study produced.  After all, people might say that a 1/100 risk of MD is enough to justify the continued blind treatment of all such children.

What if you could safely tell who to treat and not to treat though?  The PiC study had enough information about clinical features and outcomes to be able to test the validity of any guideline.  They simulated what would happen if all the children in the PiC study were managed according to a guideline's algorithm.  This allowed them to see how sensitive and specific each guideline is.

The NICE guideline (2) recommends treatment of all febrile children with non-blanching rash.  So no surprises that the PiC study found the NICE approach to have 100% sensitivity but only 1% specificity.  99% of children treated in this way have unnecessary tests, treatment and time in hospital. 

Most major paediatric emergency departments in the UK have been deviating from the "treat every time" approach for many years.  Most centres have guidelines which use a combination of clinical assessment and the use of inflammatory markers to select which children will not be treated.  The PiC study also evaluated the sensitivity and specificity of 6 of these local guidelines.  What this showed was that these guidelines retained 100% sensitivity but improved specificity.  The best guideline (Barts London) achieved a specificity of 36%.  That means that a lot of children are safely avoiding unnecessary treatment and time in hospital.

If you are thinking that this is all very nice but changes nothing for the pre-hospital clinician, the best bit is still to come.  A guideline that wasn't included in the PiC study was the Sheffield Children's Hospital Emergency Department (SCHED) Handbook.  (I believe that the reason that it was not included at the time was that the guideline was being changed.)  The direction of that change was away from using inflammatory markers as part of the decision making process.  The SCHED (3) guideline uses pattern recognition and experienced decision making.  Blood tests are not a recommended part of the process outside of specific circumstances (e.g. diagnosing haematological cause).

Although this guideline is not one of those in the PiC study, it has since been applied to the PiC study dataset of 1300 children with fever and non-blanching rash.  The exciting result of this is that the Sheffield guideline also retains 100% sensitivity (95% CI 82-100%) but achieves an even higher specificity at 69% (95% CI 66-72%).  (3)

The exciting thing about this approach is that it is a decision that can be made anywhere.  What the decision is made up of is the following

  • Continuing default treatment in a few cases (rare but important)
    • Fever and purpuric rash
    • Fever and petechial rash and clinically probable sepsis
  • Identifying other possible causes (such as mechanical cause from vomiting) using pattern recognition to identify those at low risk
  • For those children who do not have another diagnosis or deemed to need default treatment, allowing an experienced decision maker to choose whether to treat or discharge.
Here is an adapted flowchart from the Sheffield Children's Hospital Emergency Department Handbook:

The next question you might be asking yourself is, "Am I an experienced decision maker?" when it comes to the child with fever and petechial rash.  While there is no simple answer to this, the likey answer is yes if you have 5-10 years of postgraduate experience in a role that includes decision making about unwell children.

Everything has changed in the management of the well child with petechial rash and fever.  Thanks to vaccination and high quality research, we can take a very different approach and avoid overtreatment of what is now known to be a low risk clinical presentation.

Edward Snelson


  1. Waterfield T, Maney J-A, Fairley D, Lyttle MD, McKenna JP, Roland D, Corr M, McFetridge L, Mitchell H, Woolfall K, Lynn F, Patenall B, Shields MD, Validating clinical practice guidelines for the management of children with non-blanching rashes in the UK (PiC): a prospective, multicentre cohort study, The Lancet, November 2020
  2. NICE. Meningitis (bacterial) and meningococcal septicaemia in under 16s: recognition,diagnosis and management | Guidance and guidelines | NICE. 2015 [cited 2017 Oct 10]
  3. Snelson E, Waterfield T, Testing the limits of pragmatism in children with fever and non-blanching rash, Correspondence, The Lancet March 2021

Monday, 14 December 2020

Labels in child and adolescent mental health presentations - A Christmas stocking stuffer

 Here's another stocking stuffer.  In the same way as last time, it's just a mini-FOAMed post.

So next time you see a child or young person with a mental health presentation, I would suggest the following:
  • Don't think of behaviour as behavioral.  Assume it is a symptom of something more complex.
  • Don't feel pressured to give a label.  Many young people don't ever get a formal diagnosis.
  • See each contact as an opportunity to discover more about what is going on and why.  While this may not always be something massive, sometimes it takes a lot of feeling safe for a young person to disclose something.  The bigger the thing is to them, the more time and space it may take.

Mental health problems in young people are complex and that can be daunting for us as front line clinicians.  If you ever feel that you are not finding it easy, you are not alone.

Edward Snelson

Monday, 30 November 2020

Ipratropium for infant wheeze - a Christmas stocking stuffer

On the run-up to Christmas, this site will be delivering some rather minimalist FOAMed.  Instead of comprehensive explanations, there will be some short but hopefully useful posts for you to enjoy.  Think of them like a stocking stuffer rather than your main present.  Perhaps you'll like this format even better.  [I will never forget the year that my children played more with one of their stocking stuffers than with their main present.  That stocking filler present was a whoopee cushion.]

Here it is:

That little caveat at the end is about the use of ipratropium as an additional agent in the treatment of severe/ life-threatening brochospasm due to viral wheeze.  In that scenario, it's still very much all about the salbutamol.

That's all folks.  If you wanted something bigger, you'll have to wait until we're opening the main Christmas presents, or you could read this post that goes into more detail about infant wheeze diagnosis.

I hope you're looking forward to your next stocking stuffer.

Edward Snelson

Thursday, 12 November 2020

A Whole New World - Honesty in Paediatrics

Paediatrics is a specialty where lying about a diagnosis is normal practice.  It's not because we're bad people.  When you think about the challenges of diagnosis in children combined with the expectation of a diagnosis, it is completely unsurprising.  The adult accompanying the child would like a diagnosis (please and thank you) and the clinician would very much like to give one (you're welcome).

While that all seems very reasonable, in child health it often isn't entirely truthful.  It is one of the mantras of medicine that the diagnosis is going to come from history and examination in most cases.  Hurrah for clinical diagnoses.  In paediatrics, the history is often from a third party and will have an inevitable element of bias.   The examination will also contain more uncertainties more of the time.  You have to accept a significant lack of information when interpreting examination finding in children.

The result of this is that clinical diagnosis is more challenging in paediatrics.  Here's the paradox: clinical diagnosis is the default position in child health.  Why?  Because we don't want to do tests on children or give them treatments "in case" unless these investigations or therapies are very likely to benefit the child.

This week, something big happened and it didn't even hit the news.  The General Medical Council released some new and updated guidance: "Guidance on professional standards and ethics for doctors Decision making and consent."  While much of the content is old news, there is a new emphasis on honesty when there is diagnostic uncertainty that is hugely relevant to paediatric practice, thanks to the fact that uncertainty is where we work.

So, when are we lying to our patients or the adults that accompany them?  The truth is that there is a spectrum of how far what we tell people lies from the truth.  What we should probably do in the light of the new GMC guidance is to re-evaluate our approach to a variety of clinical presentations and ask, "Should I change what I say about this?"

You could argue that nothing is certain in medicine, so what are the thresholds of uncertainty that decide when we should be honest in this way?  That's a fair comment.  We need to apply some measure here - enter the certometer.

The certometer takes the things that we are already using in our diagnostic approach and gives us an idea of how truthful it is to give that diagnosis.  Last week, I asked the medical Twitter world for a few suggestions of diagnoses that we could feed into the Certometer and this seems like a good time to give this contraption a go.

First up is an intriguing suggestion:  Diagnosis - Viral illness.
In my experience this diagnosis is usually given to children with a fever and signs or symptoms of upper respiratory tract infection without signs or symptoms of a more specific diagnosis.

Let's imagine a common scenario then: a 2yr old previously healthy child with a fever for 2 days.  They have a runny nose but no cough.  They have no respiratory abnormality.  Pharynx and both tympanic membranes are inflamed.

The pre-test probability of this being a viral illness is high. It's a child with a fever so the probability that the illness is viral is around 90%.

Positive predictors of a viral cause do exist and include wheeze and urticarial rash in children. This child has none of these things.

Good negative predictors of a diagnosis of viral illness in this sort of case would be some signs of suppurative complications such as mastoiditis.  We haven't seen any signs to suggest this.

So having looked for something specific that truly discriminates and found none, what you are left with is your pre-test probability, dialled down slightly by virtue of the absence of signs of another diagnosis.  In other words, all we have truly achieved is to rule out complications.  Since complications are rare, we're essentially no more certain this is a virus than before we started.

Calling it a viral illness implies that we've added some certainty to the underlying cause that in reality, we haven't.  In fact, by calling it "a virus" we have admitted that there are no specific finding identifying a particular viral illness.

What we have done is far more important.  We have looked for signs of complications and more serious infection (sepsis, meningitis etc) and found none.  What we can say with honesty and certainty is that this is an uncomplicated upper respiratory tract infection.

To emphasise the point about how often the lack of specific signs and symptoms is the norm in paediatrics, I'll give a couple of examples of common, clinical diagnoses that are usually made with enough certainty to be considered completely honest.
  • Croup
  • Chickenpox
  • Febrile convulsion
  • Vasovagal syncope
Yep, that's pretty much it.  Most other common problems are really labels given with real uncertainty due to the lack of specific signs or symptoms with good positive or negative predictive value.

Here are a few other examples of diagnoses that are commonly given in what is in reality a great deal of uncertainty that this problem is causing the symptoms or signs.
  • Infant reflux disease
  • Cow's milk protein allergy (non-IgE)
  • Asthma
    • in the under 5 yr old child
    • where the diagnosis is based on chronic cough without wheeze
  • Mesenteric Adenitis
  • Hypermobility
Then there's a whole new level of diagnostic uncertainty.  At the beginning I used colic as an example.   Let's try feeding a classic colic presentation into the Certometer.  You see a three week old baby whose only symptom is "crying all the time".  The pregnancy and birth were uncomplicated.  The baby examines normally and is thriving.  They are feeding well and passing urine and stools normally.

What is the pre-test probability that this is colic?  Unfortunately there's no good answer to that because it's not an actual disease.  There is no pathology or treatment.   Colic is simply a label to be given to crying infants that have no pathology.  If you try to put this through the Certometer, you will break it because you can't have any certainty of something that doesn't exist.

It is often argued with colic that the label is therapeutic.  The new GMC guidance should give us an opportunity to re-evaluate that approach.  What would be wrong with telling the parent of the infant described above that their baby is normal and healthy?  That would be honest and potentially just as therapeutic.  We could then use the time that we might have spent (explaining a condition that doesn't exist) on being supportive and encouraging to the parent.  The crying excessively phase does settle and in the meantime, it's all about making sure that it doesn't break the parent.

Here are a couple of other examples of diagnostic labels in children that are without evidence for any disease process.  Neither of these has ever had any pathology associated or been shown to respond to any treatment:
  • Growing pains
  • Non-specific abdominal pain
Is it time to embrace the idea of greater honesty when we diagnose and explain symptoms in children?  I certainly find that an explanation without a diagnosis is entirely acceptable to families when it comes to a situation where in the past I might have given a non-diagnosis.  Changing that practice is relatively straightforward.  You simply stop saying the thing.

For the situations where we are dealing with an actual diagnosis but there is significant uncertainty, we've got a few options.  The infant with crying and regurgitation of feeds is a good example.  Perhaps we should be stricter about starting off with a label of "possible GORD"?  Perhaps we should go further and start with "Feeding symptoms under observation and follow-up."  Increasingly, I don't give a diagnosis.  Instead I tell the parents that (in the absence of red flags such as fatering growth) "crying and regurgitation can be normal, it can be early symptoms of reflux disease and it can be rarer problems such as allergy.  We don't want to give unnecessary treatments to babies but we also want to treat problems when it's going to help.  This is how we're going to try to get the right balance between those two things..."  

It's a whole new world being honest about our uncertainty but it does work and it works like this:
Edward Snelson

Sunday, 25 October 2020

The Decision Maker's Guide to Bronchiolitis Assessment

 This bronchiolitis season is going to be different.  While SARS-CoV2 virus does not seem to be a significant cause of wheeze in children (1), all the other usual viruses are still out there and will be causing wheeze soon in a child near you.  What might have changed is how we make decisions about that child.

For the purposes of exploring our decision making, it is important to define bronchiolitis as a condition that is a virally induced inflammation of the small airways of the lungs in a child, typically under the age of 1.  It is clinically distinctive from viral induced wheeze which is virally induced bronchospasm of the large airways, typically in a child over the age of 1.  For a separate article on differentiating these two conditions, click this link.

The necessary decisions regarding bronchiolitis haven't changed.  What might change during a global pandemic is the outcome of those decisions.  Hospitals have always been dangerous places, with a significant risk of hospital acquired infection.  That risk has escalated due to the prevalence of the highly infective SARS-CoV2 virus.  Though very unlikely to cause COVID-19 infection in children, there is that risk, the risk of PIMS-TS and the risk of COVID-19 to the accompanying adults.

The aim in bronchiolitis decision making has always been to keep as many children out of hospital as is safe to do so.  In order to do that expertly, we just need to make three decisions.
  1. Does this child have bronchiolitis?
  2. Should this child be managed at home or in hospital?
  3. What treatment should the child be given?
Question 1: Does this child have bronchiolitis?

Most children under the age of 1 year presenting with a tight cough, wheeze, respiratory signs and poor feeding have bronchiolitis.  There are other possible explanations for that presentation however and it is important to know about these other possibilities.

Viral induced wheeze, which involves bronchospasm is separate from bronchiolitis.  Clues that it may be viral induced wheeze include the age of the child (most commonly over 1 year) and previous episodes of viral induced wheeze.  The other clue is the onset of the respiratory changes.  Bronchiolitis is a slow accumulation of wetness in the airways and the history is typically of a gradual and progressive worsening of symptoms over days.  Viral induced wheeze, due to the bronchospasm involved, presents with a more sudden onset of wheeze and distress, often going from normal to significantly abnormal over a few hours.

Pneumonia is almost never associated with wheeze in children (2).  Focal crepitations are often heard in a viral lung infection of any kind.  The presence of wheeze strongly suggests that the signs and symptoms are virally induced in some way.  Infants with pneumonia will tend to be significantly unwell.  The simple rule of thumb is this:  If the infant has a wheeze and is well enough to be managed in a pre-hospital setting, they do not have bacterial pneumonia.

Congestive cardiac failure (CCF) due to haemodynamically significant yet undetected congenital cardiac abnormalities is a rare mimic of bronchiolitis but one that is important to be aware of.  The typical cause is a large ventricular septal defect (VSD) causing a significant left to right shunt.  This increased pressure through the lung circulation causes pulmonary odema which manifests as poor feeding, fine crepitations and wheeze.  Thankfully, most significant heart defects are detected before a baby is discharged from postnatal care, but occasionally one slips through and the signs and symptoms are easily mistaken for bronchiolitis.

There are usually clues however.  A murmur is the most obvious clue but this can be difficult to hear at >160bpm.  An excessive tachycardia is a possible sign of CCF.  A significant hepatomegaly (normal babies often have up to a centimetre of palpable liver) is highly suspicious of CCF.  Finally, the progression of symptoms does not fit for bronchiolitis as they continue to get worse after the 3-4 days in which bronchiolitis reaches its peak.

Putting these things together, it is usually possible to be confident in diagnosing bronchiolitis as long as the history and findings are consistent with bronchiolitis and not one of the other pretenders.
If the diagnosis is bronchiolitis, we can move onto our next question:

Question 2: Should this child be managed at home or in hospital?

Most children with bronchiolitis can be managed in the community.  Keeping people away from hospital where it safe to do so has never been more important.  In the UK, the NICE guidelines for bronchiolitis (3) give recommendations for when to refer and when to consider referral.

Referral is always recommended for red flags.  In the NICE guidelines, these are a combination of signs of potential respiratory failure.  Notably, apnoeas are included as a stand-alone red flag.  That means that a child without any chest signs of severe respiratory distress should still be referred if they have had episodes where they appear to stop breathing.  The reason for this is that in such cases, immature respiratory drive may be a factor.  Following an apnoea, a baby can temporarily seem much improved but may go on to have further events and deteriorate suddenly.

Feeding and hydration is probably the least well defined element of the decision making element.  The guidelines ask the clinician to consider a variety of factors, however being able to assess whether the amount of feeding is adequate is next to impossible apart from overt signs of dehydration.  We never know how much a breast fed baby is getting unless the answer is "nothing."  If the baby is bottle fed, applying a percentage to that as being adequate doesn't take into account the fact that many bottle fed babies take much larger volumes as a baseline.  As a result, the most objective measure of adequate feeding has to be signs of hydration or dehydration.  For that reason, I have included clinical dehydration in the list of red flags and beyond that, feeding difficulties remain a matter of clinical judgement when it comes to referral.

Possibly the most controversial element of the decision making is the presence of risk factors.  In the guidance, it is stated "When deciding whether to refer a child with bronchiolitis to secondary care, take account of any known risk factors for more severe bronchiolitis such as... (e.g.) premature birth, particularly under 32 weeks."  The guideline evidence statement lists the basis for each risk factor listed and with the exception of neuromuscular disease, the committee acknowledged that there is no credible published evidence for the other risk factors.  Apart from neuromuscular disease, they are all consensus opinion recommendations.

So what are you supposed to do when you see an 8 month old baby with mild bronchiolitis, no red flags and adequate feeding when you know that they were born at 31 weeks gestation?  Do you send them to secondary care in case because they have a risk factor for severe bronchiolitis or do you keep them well away from hospital because they don't have severe bronchiolitis and you don't want to add a hospital acquired infection to their list of problems?

Balancing risk vs benefit is what it is all about here.  There is a known risk of hospital acquired infection vs an unknown risk of severe bronchiolitis.  There is also no evidence that admitting high risk children with bronchiolitis is any safer than good safety-netting advice.

If the decision is made to manage a child with bronchiolitis at home, the third and final question is:

Question 3: What treatment should the child be given?

There has been a load of research done to try to find an effective treatment for bronchiolitis.  Supportive interventions (oxygen, CPAP etc) in the cases where respiratory support are needed have been shown to be effective.  Each and every other therapy have in turn shown to have no benefit for mild to moderate uncomplicated bronchiolitis.  Therapies proven to be ineffective include β-agonists, ipratopium, hypertonic saline, antibiotics and corticosteroids.  The bottom line is that for a child being managed in the community, no pharmacological treatment should be given.  This recommendation is consistent  across guidelines from the UK, USA and Australia (3,4,5).

That makes this flowchart nice and simple:

Finally, you might be asking yourself if you are an expert decision maker when it comes to a small person who has a cough and wheeze.  Hopefully this post helps you to feel that you are.  Decision making in such children is all about recognition, knowing the red flags and above all, learning that if in doubt, looking at the child will almost always give you your answer.

Edward Snelson


  1. Roland D, Teo KW, Bandi S, et al COVID-19 is not a driver of clinically significant viral wheeze and asthma Archives of Disease in Childhood Published Online First: 16 October 2020. doi: 10.1136/archdischild-2020-320776
  2. Hirsch A, Monuteaux M, Neuman M, Bachur R, Estimating Risk of Pneumonia in a Prospective Emergency Department Cohort, Paediatrics, Vol 204, p172-176.E1, Jan 01, 2019 doi:10.1016/j.jpeds.2018.08.077
  3. Bronchiolitis in children: diagnosis and management, NICE guideline [NG9] Published date: 01 June 2015
  4. American Academy of Pediatrics Clinical Practice Guideline: The Diagnosis, Management, and Prevention of Bronchiolitis, Pediatrics November 2014, 134 (5) e1474-e1502; doi: 10.1542/peds.2014-2742
  5. The Royal Children's Hospital Melbourne Clinical Practice Guidelines: Bronchiolitis

Wednesday, 19 August 2020

Periorbital cellulitis in children

Eye infections in children are common.  The majority consist of simple infections of the conjunctiva (the layer that covers the sclera and the inside of the eyelid).  While these infections can be viral or bacterial, the tendency is for both to self-resolve and so infection confined to the conjunctiva can be managed conservatively.  The likelihood of benefit from topical antibiotics is low and there is a significant risk of the ingredients of antibiotic eye drops creating a chemical conjunctivitis and making things worse.  As a result, NICE suggests a limited number of scenarios in which topical antibiotics may be worthwhile in conjunctivitis.

It is worth mentioning two things about conjunctivitis:
  • Atypical conjunctivitis infections are more problematic.  If herpetic, chlamydia or gonococcal infection is suspected specialist input is advisable.
  • Neonates are the exception to this conservative approach to conjuntivitis.  For a full explanation regarding why and how to manage the newborn baby with an eye infection, click this link.
When infection spreads to the periorbital tissues, it is a different matter.  Infection of the skin and subcutaneous infections around the eye is usually bacterial and is associated with more invasive infection.
Image from Wikimedia Commons, the free media repository
Complications of invasive infection include:
  • Loss of eyesight (optic nerve damage, retinal detachment, retinal artery thrombosis)
  • Meningitis
  • Intracranial abscess
  • Cavernous sinus thrombosis
Although these complications are serious, they are mainly associated with orbital, or post septal cellulitis - infection of the tissues in the orbit (eye socket).  Infection confined to the pre-septal tissues is usually uncomplicated and low risk.
[Medical illustration credit to Naomi Snelson]

Telling the difference between pre-septal cellulitis and orbital cellulitis is about looking for signs and symptoms that could indicate orbital cellulitis.  In the absence of red-flag symptoms, it is assumed that the infection is pre-septal.

The management of pre-septal cellulitis has evolved considerably over time.  Many centres used to treat even pre-septal cellulitis as an inpatient with antibiotics given intravenously to begin with.  It is now much more normal to treat pre-septal cellulitis with oral antibiotics.
While some who have made this move choose to follow up and review the patient (often at about two days into their oral antibiotics) there are strong arguments for safety-netting and no planned follow-up.  When a child is sent home on oral antibiotics for pre-septal cellulitis, things will go one of two ways.  If treatment is successful, there will be significant improvement within the first two days.  If that happens, follow-up adds nothing.

If treatment is unsuccessful, signs and symptoms will worsen.  If that happens, there is a risk that a planned follow-up will delay escalation of treatment.  If things are getting worse, the child needs to be seen and admitted for intravenous antibiotics immediately, rather than waiting for their review appointment.

Safety-netting advice for children discharged on oral antibiotics for pre-septal cellulitis
Return for immediate assessment if-
  • the child becomes febrile or unwell
  • the swelling becomes visibly worse
  • eye movements are affected
  • vision is affected
  • the child is unable to take their antibiotics
  • the child starts vomiting
This simple approach is another great way to safely keep children out of hospitals.  That has always been a good thing but there has never been a better time to avoid unnecessary admissions than now.

Edward Snelson

Wednesday, 22 July 2020

There, I've Said It: There is No Such Thing as Early Sepsis

Words are really important when it comes to communication.  Certainly, non-verbal communication counts for a lot, but words are very powerful ways of getting a message across.  We should be responsible about how we use them.

Diagnostic error in paediatrics is a very emotive issue.  If a clinician is wrong (always in retrospect) and a child is involved, it can be difficult to be objective.  No-one is right all the time though, and diagnoses  that are later apparent are sometimes easy to miss or even too elusive to reasonably detect in their early stages.

There is one important exception to this: sepsis, because there is no such thing as early sepsis.  This is really important because clinicians are both criticised and self-critical when a child has been diagnosed with sepsis and there was an earlier clinical contact.

You're going to want me to back this one up no doubt.  I haven't got any academic references, but I don't need them since it's a case of simple logic.  I'm going to Spock the heck ot of the idea of early sepsis.
Starting with the terminology that we are working with, there is already difficulty with the definition of sepsis.  Ever since the word ceased to mean infection in any form and came to be used for an effect of the infection, sepsis has remained impossible to decisively define.

The best definition that we have of sepsis "a life-threatening organ dysfunction caused by a dysregulated host response to infection," comes from the 2016 Third International Consensus Definitions for Sepsis and Septic Shock.  This definition is inherently subjective.  There is no formula and no test.  Note also the lack of attempt to define early sepsis at this event.
Why do I care if people use the term early sepsis?  Because words.

The term early sepsis implies that there is a clearly definable and therefore recognisable entity which has therefore earned its own name.  Implications are more dangerous than overt statements because they go unchallenged and become part of the profession's assumption that because a term is used, it must be valid.

The harm of the term being given validity is that (again with the logic) if it is a definable entity, whenever a child is diagnosed with sepsis and they had an earlier contact with a clinician, that person has failed to recognise early sepsis.  In reality, they will have seen a child with features that are attributable to sepsis.  These features are also commonly found in children who are febrile but not sepstic.

How damaging is that?  Apart from the medicolegal implications, the negative impact on a clinician's confidence and reputation is potentially huge.  Using the term early sepsis risks leaving a string of second victims in its wake.

Just to be clear, I am 100% behind the idea of recognising sepsis early.  Swapping those words around is all it takes to make them functional again.  I'm also all for anything that improves the early recognition and treatment of sepsis.  So far, no strategy has proved successful in achieving earlier diagnosis of sepsis in children.  Awareness, careful clinical assessment are key, as is treating every illness as dynamic.  That is why appropriate observation and good safety-netting are key interventions.  Whether the child is managed at home or in a health-care setting, no illness is 100% safe until the child is better.

The focus on sepsis over the past decade or so has improved the timely treatment when sepsis is diagnosed, but making that decision in the first place remains a complex business.  Here's a nice oversimplifiaction of that process:
The red and green patients are relatively easy in terms of decision making.  The amber patients represent a small uncertainty which needs to be managed expertly.  In the small proportion that later become red and therefore relatively easy to define as having sepsis, retrospectively calling the preceding illness "early sepsis" defies logic and undervalues the difficulties of managing a large volume of moderately unwell children.

Next time you hear someone talk about recognition of early sepsis, politely challenge them and explore whether they mean early recognition of sepsis.  Sepsis is a thing.  The point when an illness goes from not sepsis to sepsis is not sudden and therefore easily missed.  Implying that there is a clearly definable and recognisable thing called early sepsis risks the vilification of front line clinicians in both primary and secondary care.

Edward Snelson