The tool
The NICE guideline on feverish illness in children was first published in 2007. One of the major remits for the guideline was to establish a system to aid differentiation of serious bacterial infection (SBI) from minor self-limiting infective illnesses – this is a common problem as there may be few if any indicators of serious disease in the early stages of a febrile illness. It was clear from previous attempts such as the Yale Observation Score (YOS) that there was no single parameter which would provide definite identification of serious illness, and review of available evidence confirmed this. A number of features were, however, identified which might indicate the presence or absence of SBI. There is a wide variation in both sensitivity and specificity for each of these parameters and it was felt that a traffic light table rather than a formal scoring system would be more helpful to doctors and nurses in assessing the likelihood of SBI.
The guideline was updated and a revised version was published in 2013. The traffic light system was retained but with some modifications – most significantly it has now been possible to define significant tachycardia allowing for age and level of fever.
It was noted during the update development that the traffic light system had been widely adopted in the UK and elsewhere and had generally been regarded as a helpful tool.
When to use it
The scope of the guideline was to look at feverish illness in children up to the age of five years, where no focus of infection has been identified. This is an everyday consultation in general practice, but increasingly initial assessments are made in other healthcare settings such as telephone help lines, out of hours services and A&E departments (both general and paediatric). The traffic light system is intended for use in any of these settings with the hope that wherever the consultation takes place there will be equal opportunity to identify children at higher risk of SBI.
The system is designed to identify three levels of risk of SBI based on a standard history and examination, with additional information from basic investigations if these are available within the setting. It covers four areas which should be part of the examination:
- the colour or general appearance of the child
- the activity and interaction displayed by the child
- the respiratory and cardiovascular status
- hydration and feeding
- plus an “other” category to include specific identified risks that do not fit into these main categories
The ‘scoring’ for each of these categories is shown below but in general normal findings result in allocation to the green column, mild to moderately abnormal findings result in allocation to the amber column and severely abnormal findings result in allocation to the red column.
A child who has only green scores is very unlikely to have a SBI. It is reassuring to know that evidence supports the intuitive impression that a child who has a fever but is otherwise well is at low risk of SBI.
Any feature which places a child in the red category puts them at significantly higher risk of SBI and means that they warrant referral to a paediatric unit for specialist assessment. Again, in this category available evidence supports the clinical impression that a feverish child who is generally unwell requires further investigation and assessment.
Referral for further assessment should also be considered if two or more amber scores are registered. The amber features do require clinical judgement and interpretation and should be regarded as the equivalent at a traffic signal of ‘proceed with caution’. Each of the amber parameters covers the range from mildly abnormal (not quite green) to moderately severe (almost red). Clinical interpretation is necessary as a child with a single, but moderately severe amber feature may warrant referral to a paediatric unit, while another child with two or more amber features all at the mild end of spectrum may warrant observation rather than immediate referral.
‘Safety net’ advice should always be given to parents or carers of children not immediately referred – this should include advice about what changes in symptoms and appearance would alter their traffic light category and warrant an urgent further consultation.
The evidence
The process of developing NICE guidelines is evidence based. Once the precise questions to be asked are determined, an extensive search of available published data is made. Evidence is strictly graded with a strong emphasis on data from randomised controlled trials and meta-analyses. The best available evidence of clinical and cost-effectiveness is then used as a basis for the recommendations made in the guidelines and there should be a clear and logical flow from the evidence to the recommendations. Each of the parameters used in the traffic light system was subjected to this degree of scrutiny.
Details of the evidence used can be found by accessing the appendix of the NICE guideline.
Dr John Crimmins is a GP in the Vale of Glamorgan and a member of the guideline development group for the NICE Feverish Illness in Children guideline.
