Wednesday, 21 October 2015

It's all about the wave - An expert's tip for recognising seizures in children

I was recently picking the brains of a paediatric neurology colleague and discussing how they distinguished epileptic seizures from other events.  I already knew that it wasn't through the black arts of EEG (electroencephalogram) interpretation.  I'm sorry to be the bearer of bad news but EEG has a very poor sensitivity and specificity for epilepsy (1).  By the time an EEG is requested, a diagnosis should already have been made on clinical grounds.  The EEG is then useful as part of deciding how to manage the epilepsy, not to screen for epilepsy.

Why do we front line clinicians need to be able to tell the difference?  There are so many things that children present with that could be seizures but probably are not.  One good example of this is absence seizures.

Absence seizures (no longer called petit mal) account for 10% of epilepsy in children.  The incidence of epilepsy in childhood is not well described but is probably about 1/2000 making absence seizures about 1/20,000.  Staring episodes are very common in children and if these occur frequently, it is understandable that parents or teachers want to know if this is a form of epilepsy.  Not all of these need to be referred.

If a paediatric neurologist makes the diagnosis from the history, then so can any clinician.  All we need is to know what their secret is.  It turns out that they are riding a wave.

The characteristic feature of absence seizure is the abrupt and brief impairment of consciousness with a complete inability to intrude on the episodes.  (2) What this means to the observer is that they will notice a sudden onset and then notice the event suddenly finish.

 Fig 1. If you were to plot the change over time, an inattentive staring episode would be like a sine wave.
Fig 2. The typical abrupt onset and cessation of an absence seizure gives a square wave.

It may just be that simple.

Of course, I am not suggesting that no examination is needed.  There are also other features of absence seizures that may be helpful:
  • The episodes occur during any activity and in any environment
  • There may be subtle muscle twitches or lip smacking (automations)
  • There may be a slight loss of tone leading to slumping of the head or trunk

There are two ways that you can get to see the episodes for yourself.  If they happen frequently then, through the magic of smartphones the parents will be able to capture an episode for you.  There is a better way though, as almost all absence seizures can be provoked through hyperventilation as demonstrated beautifully in this video:

So, I have to give thanks to Tony for telling me about the wave thing.  I find pictures are a great way to help me understand things that I know, when I don't necessarily know them well enough.

Edward Snelson
Wave junky

  1. EEG in the diagnosis, classification, and management of patients with epilepsy, S J M Smith, J Neurol Neurosurg Psychiatry 76:ii2-ii7 doi:10.1136/jnnp.2005.069245
  2. Typical absence seizures and their treatment, C P Panayiotopoulos, Arch Dis Child 81:351-355 doi:10.1136/adc.81.4.351

Tuesday, 13 October 2015

Reverse Engineering - Feeding Problems in the Under-One-Year-Old

What's in a name?  When it comes to the medical conditions affecting babies, there is so very much to be confused about.

Whatever your reaction to that, you are probably aware that the terminology used to describe infant feeding problems is confusing and inconsistent.  I have spent a little time recently looking at the relevant websites, guidelines and journal review articles available to professionals on this subject.  So far I have not found two that use exactly the same terminology to categorise the various clinical entities that exist.  I also know from professional experience that understanding varies and even conflicts between clinicians.

Let me give you an example: reflux.  "Reflux" in babies means to many doctors that a child has a problem with bringing up milk that needs treating.  However, bringing milk up is normal for babies, so many encourage the use of the words posseting or regurgitation to describe harmless reflux.  Many academic papers and guidelines use the medically correct term of gastro-oesophageal reflux (GOR) for the harmless type but distinguish this from gastro-oesophageal reflux disease (GORD) when there is a significant problem.  So there we have five different ways to describe just two clinical entities.  Two of those terms might be used interchangeably to describe either entity.

The conditions involving any sort of reaction to milk are even more confusing.  Some try to simplify things by insisting on having just two diagnoses- cow's milk protein allergy or lactose intolerance.  This goes a long way to help clinicians to avoid getting confused between the two, but then many children who are having a reaction to cows milk do not have lactose intolerance, and do not have any evidence of allergy to cow's milk protein either.  For this reason, the term Milk Intolerance is used by some including NHS choices.  Some feel that this helps to distinguish children who have a proven allergy, and so need to be managed as such, from those who do not.  The term milk intolerance does however confuse us again by it's vagueness.  Intolerance to what?  There is now a growing consensus that there is only milk protein allergy or lactose intolerance.

When you can't get started because nothing makes sense, the process that will get you there is called reverse engineering.  Start with the end point and work backwards.  So, lets have a go at this for the infant with feeding problems.  There is one outcome that we are aiming for: a well child.  There are only a few treatment options, so lets start with those and describe the children who should be having that treatment.  All scenarios rely on a thorough history and examination.  One of the most important parts of this is to take a feed history.

Lets start with my favourite outcome and reverse engineer that:

How to get there: reverse engineering option 1 - doing as much nothing as possible

Babies who regurgitate feeds without choking, gagging frequently, faltering growth or being significantly distressed during or after feeds do not require intervention.   It is important to safety-net so that parents know to return if problems such as feed refusal develop.

How to get there: reverse engineering option 2 - reducing the feeds

The baby that vomits and cries a lot may be over-feeding.  This is because, while most babies determine their own requirements, some get stuck in a loop called the feed-cry cycle and need to have their feeds restricted to something more sensible.  Take a feed history in any baby presenting with vomiting and distress and limit the feed to 150ml/kg/day.  If that doesn't work, look at the other options.
How to get there: reverse engineering option 3 - thickeners and alginates

This is where to start with the infant who fits these characteristics:

  • Onset of vomiting before the age of three months old
  • Gagging, choking or significant distress
  • Symptoms are not progressive over a few days and no signs of other illness
  • Adequate weight gain and still taking a sensible amount of feed

Thickened milks are the recommended first line of treatment in the NICE guideline for GORD in children.  If thickened feeds do not help or if the baby is breastfed then an infant alginate preparation is a thickener as well as an antacid.

How to get there: reverse engineering option 4 - what am I missing?

This is not really a feeding problem treatment option but an important reminder that we all need to re-think when a child does not improve.  Even if the initial assessment was reassuring, by the second or third visit we should always be asking "what am I missing?"  One thing not to miss is a urinary tract infection.  These low-grade UTIs may be sub-clinical apart from symptoms indistinguishable from GORD.  Other conditions such as pyloric stenosis should be considered if the vomiting escalates over the space of a few days.

How to get there: reverse engineering option 5 - acid suppression

H2 receptor agonists (H2RAs) and proton pump inhibitors (PPIs) are an option for the child who has not responded to option 3 and does not have anything to suggest a hidden diagnosis (option 4).  As long as there is a re-evaluation of the pathway, acid suppression is the next step assuming that the infant does not have sever symptoms or signs of an allergy.

How to get there: reverse engineering option 6 - try an extensively hydrolysed feed (EHF) (or revert to breastfeeding)

There are a few scenarios which might benefit from EHF.  These milks have the proteins broken down so that the feed is very unlikely to trigger an allergic reaction.  
If treatment option 5 has been tried then this is one possible way to go next when there is no response.  In this context the EHF is a trial to see if the child is reacting to the milk with a reaction that only manifests as symptomatic vomiting or distress around feeds.
If an infant who is initially treated with thickeners goes onto develop signs of atopy (such as severe eczema) or fails to improve and has a very strong family history of atopy, EHF can be tried earlier.  Most likely this will be after option 3 in an infant under three months old.
In an infant who presents later in the first year of life, the odds of the problem being a reaction to the milk increase.  In such children, ECF could be tried even without trying options 3 or 4.
The main differences from option 6 are that with option 5 you do not need to refer.  There are a variety of ways to go next.  NICE recommends re-challenging after 2-6 weeks (in this scenario) while local guidelines vary.

How to get there: reverse engineering option 7 – start an extensively hydrolysed feed (EHF) (or revert to breastfeeding) and refer to a specialist

In an infant who has a clear allergic reaction (e.g. develops urticaria after weaning from breast milk) then there is an urgent need to remove cow’s milk protein from the baby’s diet.  This can be done by using EHF or by returning the infant to being fully breastfed.  In these cases the next steps require the input of a specialist so they will need to be referred.  Breastfeeding mothers should exclude milk from their diet until a dietician can fully advise.

How to get there: reverse engineering option 8 – start a lactose free milk (or add lactase for breast feeding mothers)

In the UK, lactose free milk is usually needed when a child has partially recovered from a gastroenteritis,  The unwellness and vomiting settles but they have persistent and often explosive loose stools.  In these cases, it is likely to be the case that the child will dramatically improve with a trial of lactose free milk.  Once symptoms have been resolved for two weeks, the child can return to normal milk products.
For breast feeding mothers, there is no point excluding lactose from their diet as the breast naturally produces lactose anyway.  One option is to add lactase to the feeds in the form of drops.
Note: EHFs tend to be lactose free (because lactose also causes problems for babies reacting to the proteins in cow's milk) but this is a very expensive option if the problem is purely the lactose. Choose a simple lactose free milk in this scenario.

So, I have managed to describe the various options and the patients who should benefit from them, yet I only mentioned a diagnosis once (in the context of a NICE guideline).  When it comes to feeding problems in the under one year old, the diagnosis matters little and may never be known.  What really matters is to know what has the best chance of success for a given clinical entity and what to do next if that intervention does not work.

Having said all that, if you want to know how the names match up to the scenarios, here's a simplistic table that pretty much does that as well:

Edward Snelson
Paediatrician with sub-specialty training in having a short attention span

Disclaimer: I could have just given you the last picture, but what would be the fun in that?