Author: Nikki Abela / Codes: SLO2, SLO5, SLO6 / Published: 12/09/2018

Co-Authored and reviewed by: Damian Roland and Edward Snelson

EM physicians need to be decision makers, quick thinkers and risk balancers. This is what we are good at. In fact, many will tell you that the purpose of the PEM rotation is to learn to spot the sick child. And that is very true. But if that is all we take away from this rotation, we are letting ourselves down.

Lots of unwell children are obviously sick, many of them obviously aren’t (yes, mum with the toddler eating a packet of crisps in the waiting room who was triaged as “decreased oral intake”, I’m looking at you), but then there are the ones that sit on that grey area in the middle. You have all heard of them – the ones which “hold their own”, “cope well”, until they suddenly don’t any more. Children are renowned for doing this. This is why ‘spotting the sick child’ in some ways is a misnomer. It’s a convenient phrase as, “recognising using history and examination the potentially unwell child at a stage of their illness before decompensation occurs,” doesn’t quite have the same ring to it.

So what do we do? Keep them all in and treat them all or send them all home and accept a “margin of error”? Of course, none of these options are acceptable for a multitude of reasons.

Yet, daily, PEM physicians manage to strike a balance between the two through decision making processes which are important skills to learn. However they cannot be taught through a classroom or formula or guide. One size will not fit all and ultimately all are underpinned by experiences, often both positive and negative.

In a recent article the two excellent co-authors of this blog analysed what goes into these decisions:

  1. The nature of paediatric physiology
  2. The variability of communication
  3. The heuristics used by clinicians and how they are affected by cognitive biases
  4. The impact of external factors: parents/carers and/or the clinical environment

Complex huh? Fantastic! Having a growth mindset myself, I love a challenge, and I encourage you to look at it the same way.

Let’s start with paediatric physiology. How many times have you heard the term, “You shouldn’t send home a child with a tachycardia”? Many departments will (rightly) observe to see a trend, or, if accompanied by a fever, give anti-pyretics to see if it settles. We can argue right and wrong till the cows come home here as the evidence is scanty, but one thing is certain, when my baby wants a feed, fever or not, if I deny her that milk for five minutes, her heart rate will be triggering a senior review on any paediatric observation scoring system (e.g. PEWs/POPS). So what am I saying? You need to take the whole picture into consideration.

The principle here is that it is not a binary tachycardia or not, it is the context of that tachycardia. “You shouldn’t send a child home with a tachycardia,” implies you can send a child home without one. This corollary is nonsense of course. So not all children should be sent home if they have a tachycardia, nor should you observe or treat them all. The decision to do so is aided by documenting your reasoning and reading it back to yourself, i.e  if the child is giggling away and covered in chocolate but has a mild fever, mild tachycardia, cough, snotty red ears and a red throat,  then the giggling and chocolate bit need to go down in the notes. This is not defensive medicine but gives a good description of the patient you examined in case they did later, say, fall off their precipice and re-present in a sicker state. As Edward says, “a picture is worth a thousand words”.

Guidelines tend to be simplistic and will under-represent the intangible elements of our decision making. Numbers based decision making implies that there is a line to cross from normal to abnormal. The reality is that all of the elements are valid – the numbers, the symptoms, the signs and the global gut feel.

In the same way that normality of things like heart rate is not binary, nor is the clinical assessment. Our evaluation of the likelihood of a significant diagnosis is a spectrum. Significant head injury is a good example. If a child has a minor bump and no worrying signs or symptoms, it is obvious to us that they do not have a brain injury. If they have a concerning mechanism and a reduced conscious level, it is reasonable to assume that they do have a brain injury. In many ways, it is those patients in the middle that are the most challenging. What happens when there is a plausibly significant mechanism, an alert child and a bit of headache or vomiting? In such cases, the decision making is both more complex and likely to be a longer process than in the first two cases.

And this is where communication comes in (which is the second point in the Roland-Snelson article). One of my favourite lines to tell parents is, “I can only comment/decide on the child I am seeing in front of me now (happily eating that chocolate ice-cream). If they change and he or she develop these symptoms {list core symptoms relevant to the presentation}, please bring them back”. Learning to safety-net appropriately is essential in any part of our practise – however, in paeds, communication needs to take us that bit further. We need to break the barriers down for a child and their parents to trust us to examine them, and then figure out how to adapt what we know (e.g. a cranial nerve examination) to the patient in front of us (e.g. a four day old infant).

As Damian and Edward explain in their piece: Innovate, Extrapolate and Economise.

Moving on to heuristic thinking, i.e. mental shortcuts used to make decisions and judgments quickly without having to spend a lot of time researching and analysing information. There’s lots of literature on this and how many physicians use it to aid decision making. We’ve skimmed over this before but you may not have noticed – remember when we discussed that patient in the waiting room stuffing his face with the crisps while his mum was worried that his oral intake has decreased? I (and probably you) had already decided his discharge destination using this technique.

But while this type of thinking is efficient (something EM physicians love), it is also full of traps. Because this type of thinking opens us up to cognitive biases. St.Emlyn’s excellent Natalie May has a whole blog on this, which I encourage you to read here. Remember to always use clinical decision aids and tools to help you structure the way you think and decide – most paeds EDs will be rife with guidelines, you just need to know where to look.

Lastly, we move on to external factors – parents, carers and/or the clinical environment. External factors could be changing the decision you make, and you need to be aware of that so that it does not limit you. As I said before, every EM clinician knows about risk, but the level of risk you are happy to take may not be the same as the parents and you may decide to lower the level of risk after discussion with them. Although every clinical case should be the same, none ever is, and flexibility to find what is right for the child in front of you should be your top priority. Again, communication here is paramount.

Moreover the clinical environment may be effecting you and the way you think, try to find ways not to let it. For example you cognitive load is increased when you have a parent looking at you when you go to waiting room/are at desk – that parent is worried about their child, find ways to go about it. Think about when you need to HALT.

I repeat, decision making in paediatrics is complex. But most of the time, people are right, even when they don’t invest so much grey matter.

As Damian says,

The low incidence of serious disease in children protects the unwary.

Just because you’re right, doesn’t mean your methods or reasons are. Be smart and decide well.

Damian and Edward are excellent PEM physicians with blogs of their own which are really worth following. Edward runs the site (which is not just for GPs) and Damian hosts a series of blogs on his website Both of them have an eminent social media presence and should definitely be followed on twitter to keep up to date with all things PEM.


  1. Roland D, Snelson E. ‘So why didn’t you think this baby was ill?’ Decision-making in acute paediatrics. 
  2. St. Emlyn’s – When is a Door not a Door? Bias, Heuristics and Metacognition