Identifying the sick child- will the NHSE PEWS score help us in primary care?

How do we identify a sick child?  Wouldn’t it be great if there was a handy scoring system that could accurately predict those at risk of serious illness who may need admission? Or just as importantly reassure us that serious illness is very unlikely, and referral is not needed!

Could we pinch any of the tools used by our hospital colleagues?  A recent study looked to address this question!

Hang on a minute, I can hear you thinking, we already have a tool for assessing acute illness in children! 

The traffic light system is recommended by NICE (NG143) for assessing fever in under 5s and widely used. However, a recent retrospective cohort study (published in the BJGP) questioned how helpful this tool is for guiding referral decisions from primary care. They assessed the traffic light scores of 6703 acutely unwell children aged <5yrs presenting to general practice in England and Wales, and then looked at hospital admissions within 7 days of the consultation. Most children were categorised as red (31.6%) or amber (62.7%), with only a small proportion green (5.7%). The ‘red’ category alone had a sensitivity of 58.8% for identifying serious illness (specificity 68.5%), so use of this alone missed a substantial proportion. Combining red and amber increased the sensitivity to 100%, but also reduced the specificity to 5.7%, resulting in a large proportion falsely identified as at risk of serious illness. The authors concluded that the traffic light system did not accurately detect children admitted with a serious illness in this cohort, nor clearly identify the majority of those that could be managed in the community. 

So are there better options available? 

NHSE is rolling out the national PEWS (paediatric early warning system) for identification of deteriorating patients across inpatient clinical settings  (four different age appropriate charts used). It incorporates clinical signs (such as head bobbing/ recession etc) alongside objective measures (see here for the age 1-4 chart). Alternatively,  the Liverpool quick Sequential Organ Failure Assessment (LqSOFA) has been validated in emergency departments for predicting critical care admissions. It consists of a simple four variables: heart rate, respiratory rate, consciousness level and capillary refill time. 

Could either the national PEWS or LqSOFA be helpful in identifying patients at risk of serious illness/ admission from primary care?

A new retrospective cohort study published in the BJGP studied the LqSOFA and PEWS scores in same cohort as the study above. The PEWS was modified to exclude BP and oxygen requirement, as these were not routinely documented. The primary outcome was admission within 2 days of GP consultation (104/6703 children, 1.6%). 

Spoiler alert- neither score was great! The sensitivity of the LqSOFA using a threshold score of ³1 was 30.6% (specificity 84.7%), meaning it missed over 2/3 of admissions. The national PEWS with a threshold score of ≥1 had a much better sensitivity at 81.0% , but the specificity was only 32.5%, meaning use without additional consideration could result in a high volume of unnecessary referrals. 

The bottom line is that identifying the small proportion of seriously unwell children in primary care is inherently challenging, and none of the current tools do it well in isolation. That being said they can support our assessment, and a positive effect of the traffic light tool is that measurement of vital signs such as RR/HR has become routine; reinforcing good practice, helping to avoid nasty surprises (e.g.  the well looking child with an alarming tachycardia!), and providing an objective measure of deterioration when reviewed as part of the safety netting process. Any score needs to be balanced with assessment of the patient in front of you. Clinical judgment and careful safety netting remain key! 

In our upcoming new Urgent Care course on November 16^th we shall cover a broad range of urgent care topics including an update on recognising sepsis at the extremes of age, including the use of the PEWS score in children and recognition of sepsis in patients with frailty. We hope you can join us then!