I was recently asked, “How do you recognise sepsis?” Answering that question would be so much easier if only we knew what sepsis was. A recent convention of experts recently met in an attempt to define the term. (1) What they came up with was: “life-threatening organ dysfunction caused by a dysregulated host response to infection.” All we need now is a definition of organ dysfunction and we’ve got this thing sorted. (Sigh)
- We are poor at recognising sepsis in children
- Recognising sepsis early saves lives
- Sepsis is recognisable
We are poor at recognising sepsis in children
Sepsis is diagnosed at the point in the illness when it is clear that the infection has had a significant dysfunctional and systemic effect. Since it is always preceded by an infection that is having some effect, that moment is difficult to define. As a result some of the following things may occasionally happen:
- Someone will diagnose sepsis and than say that the last clinician to see the patient 'missed the diagnosis'.
- People make assumptions without speaking to the clinician who made the initial assessment.
- Something definite will happen such as a growth on a blood culture. Bacteraemia, interestingly, does not equal septicaemia. This sometimes causes confusion.
Sepsis, to be clear, is a response to infection. It is a subjective global assessment of the effect of an infection on a child. Unfortunately this does not wash when it comes to academia. As a result people resort to things that are definable or binary. SIRS is a perfect example of a clumsy attempt to define an intangible entity. Many publications use positive blood cultures as evidence of sepsis. The two things may sometimes go together but they are not at all the same.
All of that said, it is true that a large proportion of children later deemed to be septic have seen a primary care clinician in the 24 hrs before sepsis was recognised. In many cases there is retrospective evidence that sepsis was present. In many cases the child was probably not septic yet. However, it is very difficult to remain constantly vigilant for a syndrome which is initially only subtlety different from all the non-sepsis. Do we miss sepsis? Of course we do, which is why we look for ways to improve the sensitivity of our assessment.
Logic dictates that sepsis left untreated is bad for you. What is unknown is the potential harm caused by over-referral, over-investigation and over-treatment. If we lower our threshold for treating presumed sepsis, how many children will come to harm for every child saved? No-one has meaningfully looked at that. Meanwhile, the only direction we seem to go in is towards caution, without a great deal of consideration for the possible dangers.
Sepsis is recognisable
This is where it gets tricky. Those who are trying to improve recognition of sepsis through the writing of guidelines have to give the reader something solid. There is little point in a guideline telling someone that they should make a gestalt assessment. There are also learning tools such as spotting the sick child. Reading the guidelines and these websites will raise as many questions as give answers. That is because recognising sepsis is just not that easy.
I now return to the original question, “How do you recognise sepsis?” Mainly, I do three things.
Although guidelines may emphasise the importance of abnormal physiology, I think that experienced clinicians quite rightly give weight to the child’s activity and behaviour. That doesn’t mean that the heart rate is unimportant, just not the only or most important thing.
What about blood tests? Well, this is also in the journals quite often at the moment. In children, white cells go up quickly in any infection, making that unreliable. CRP lags behind the infection so that by the time this is raised the child is often already clinically unwell. Inflammatory markers can not be relied upon to rule in or rule out sepsis. With one or two rather orthopaedic exceptions, I simply do not use blood tests to help me recognise serious infection. I make a clinical decision and take blood tests as baseline markers when intravenous antibiotics are given for presumed sepsis.
As clinicians, what we are good at is pattern recognition. So, I am going to tell you what you already know. Children with serious infections have a different pattern to their illness. It looks a bit like this:
If a child is returning to baseline and doing things that reassure you, you can say that they are not septic. That doesn’t mean that they cannot become septic of course. That can happen to any child. That is where good safety-netting comes in.
Acknowledgement - this post was originally requested by and published on the network locum blogsite. Thank you for that.
- The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3), JAMA, February 23, 2016, Vol 315, No. 8
- Spotting the sick child