The traffic light system is 'not able to accurately detect or exclude serious illness in unwell children presenting to general practice', according to researchers writing in the British Journal of General Practice (BJGP).
NICE's traffic light tool - introduced to help GPs differentiate between serious and non-serious cases in large numbers of children presenting to general practice with acute illness - categorises children into green, amber or red categories, indicating low, intermediate or high risk.
One in four GP appointments and more than 2m emergency department attendances a year in the UK are accounted for by children with acute illnesses such as a fever or cough.
Traffic light system
The study found that prevalence of serious illness in 6,703 acutely unwell children presenting to UK general practice was 0.3% - although 31.6% of cases were were categorised as red by the traffic light system, with 62.7% categorised as amber and 5.7% as green.
The traffic light system's red category had a sensitivity of 58.8% and specificity of 68.5% for identification of children admitted to hospital with serious illness, the researchers showed. Combining the red and amber categories led gave a sensitivity of 100% but specificity of 5.7%.
The researchers warned: 'The conclusion reached is that the NICE traffic light system is not able to accurately detect or exclude serious illness in acutely unwell children presenting to general practice when the red category is used as a positive threshold.
'If this cohort’s traffic light classifications had been followed by GPs, a third of children (categorised into the red category) would have been urgently referred to hospital. Additionally, using the red category as a threshold for hospital referral would have missed 41.2% of children with a serious illness who were in the amber category, although NICE does recommend that clinicians should refer children in the amber category if indicated.
'Combining red and amber categories improved the sensitivity of the traffic light system, such that all seriously ill children were identified. This threshold would allow GPs to be confident sending children in the green category home, but at the cost of referring a substantial number of children to hospital.'
The researchers said the traffic light system was 'unable to accurately achieve' the goal for which it was created - helping GPs to confidently assess unwell children and aid decision making about who to refer and who to send home.
It called for more research to develop a tool that could 'correctly identify the most unwell children, while preventing unnecessary hospital referrals for children who are more likely to have a self-limiting illness', based on data from UK general practice.
RCGP chair Professor Martin Marshall said: 'Ensuring that children get appropriate and timely care, particularly if they are seriously unwell, is a top priority for GPs. We rely on evidence-based clinical tools, as well as our expert clinical knowledge and understanding of our patients in order to do this, so it’s vital that the tools we use are reliable.
'This is important research, the first of its kind conducted in a primary care setting. More such research needs to be conducted in this area, and it should be taken into account as clinical guidelines are updated and developed.'
A NICE spokesperson said: 'The early symptoms of acute illness in children under five, such as a fever, can be similar to those of a non-serious illness. This can make early diagnosis of a serious illness challenging for clinicians.
'The traffic light table included in our guideline aims to improve clinical assessment of children under five with fever to help clinicians diagnose serious illness. Our clinical guidelines are evidence-based and include recommendations developed by independent committees of experts.
'The population for the main study is ‘acutely unwell children’ (25% of whom did not have a fever). This is different from the population covered by our clinical guideline, which is ‘children under five with a fever with no obvious cause’. The conclusions drawn from the main study in this paper therefore do not affect the recommendations in our guideline.
'We recommend clinicians continue to refer to our guideline on the assessment and initial management of fever in under fives when carrying out clinical assessments.'