Wednesday, 3 February 2016

Who should write your guidelines? Sepsis and the happy GP

In January 2016 NICE announced the consultation period for the draft guideline about sepsis and soon afterwards a sad news story hit the headlines.  The story was about a boy who had tragically died of sepsis.  There is a temptation to allow such emotive events to be drivers when writing guidelines but this needs to be balanced by considering the effect of any referral or treatment threshold on the whole population of children presenting to GPs and EDs.

As a user of guidelines, I would like to explore the complexities of any protocolised approach to recognising sepsis.  There are elements of the process that are simple, and elements that are less so.

First of all, let me be clear.  I welcome a guideline that aims to raise the awareness of clinicians to the possibility that a child may have sepsis.  I welcome the push to speed the diagnosis and shorten the time to receiving intravenous antibiotics.  However, greater sensitivity usually comes at the cost of poorer specificity.  In this case that probably means referring more well children and treating more well children in hospital.  That does not just have implications for the workload of emergency departments and inpatient wards.  It also affects every child who did not have sepsis but was flagged up on the flowchart as being high risk.  Referral to hospital and admission as an inpatient both have an associated morbidity.  We need to be careful not to repeat the bronchiolitis admission effect. (2)

In a recent BMJ, there was an article about why Dutch GPs are happier.  One of the reasons given was that Primary Care write their own guidelines in Holland.  In the UK, guidelines are more often developed in a more secondary-care-centric fashion.   I wonder whether having Primary Care take the lead on writing their own guidelines would help avoid the over-diagnosis effect.

What would a guideline for sepsis look like if it was being written by GP’s for GP’s?  I think it might look a bit like this…

  • Always consider sepsis if a person presents with signs or symptoms that indicate possible infection.  In primary care the presumption is that the febrile patient is unlikely to have sepsis.  However a small proportion do and detecting this requires vigilance.
  • Finding a simple explanation for signs and symptoms does not mean that a person does not have sepsis.  The possibility of sepsis is less about the specifics and more about the severity of signs (such as tachycardia) and symptoms (such as feeling very unwell or being significantly affected by the illness).
  • Be especially thorough if a patient or parent is especially concerned or attends repeatedly for the same illness.  This does not necessarily indicate sepsis but should prompt a thorough re-evaluation of the patient’s wellness.
  • Be aware that communication difficulties put patients at higher risk.
  • Be aware of high risk groups and have a lower threshold for referral.
  • The seriously unwell patient is usually easy to recognise.  The difficulty lies in recognising the patient who is in the early stages of sepsis.  It is also these patients who are the most likely to benefit from early recognition.
…or something similar.

I think that most of that would be agreeable to most clinicians.  After that, I find myself struggling to agree with myself over the nitty-gritty of the what to do when in each circumstance.  The reason for that is that there are so many common scenarios that make a sepsis pathway virtually impossible to write.  Some of these scenarios are the post immunisation baby with a fever, the six month old with bronchiolitis and tachycardia and the 2 year old with viral wheeze.  Each of these cases bring an interesting sensation to the mix.  I inherently feel that sespis is much less likely in these children and yet each will be spat out of my decision tool with a tequila coloured label of some kind.

What is the solution?  I feel that sepsis guidelines need to promote the rule-out value of wellness.  Wellness seems vague but it is really a combination of gestalt (gut feel) and what we see a child do.  Gestalt is valid as long as it is built on plenty of valid clinical experience.  What we see a child do is valid because when a child smiles and plays, it tells us that the frivolous centres of the brain are perfused and not toxic.

Will wellness be in your sepsis guideline?  I don't know, but there is one way to be sure.  Write it yourselves.  Meanwhile, the NICE guideline is still in consultation and if you are in the UK there is time to register as a stakeholder and send comments.

There is also one last chance to take part in the survey that is seeking to gain consensus on the factors that clinicians everywhere use as part of their rule-out process in ill children.  Please do take part.  It takes less than five minutes.

Edward Snelson

Disclaimer: Simple is easy when you're not actually writing the guidelines.


  1. Sepsis - NICE guideline in development (closing date for comments 22nd Feb 2016)
  2. Green et al., Admission to hospital for bronchiolitis in England: trends over five decades, geographical variation and association with perinatal characteristics and subsequent asthma, Arch Dis Child doi:10.1136/archdischild-2015-308723

No comments:

Post a Comment