Do you recall ever ringing / speaking to a more senior colleague, after finding a worrying blood test result, pointing at it and having the response “Hmm… But have you looked at the patient?”

(If you haven’t – you either need to work more or have reached clinical nirvana.)

It’s the sort of statement that echoes in a quite brilliant tweet as part of our last #ADC_JC

“@arivkan “Echo/ECG need to be strongly grounded by clinical acumen & ability to challenge reassuring test when suspicion is valid”

It was this that I had in mind when I looked at the interesting title of

Use of time from fever onset improves the diagnostic accuracy of C-reactive protein in identifying bacterial infections

This paper explicitly looked at the connection between the parent report of duration of fever, and an emergency department CRP value, to determine if this information when taken together was more valuable than the number on its own. Given the time it takes to produce & release CRP, it makes sense that a 24hr+ low value will be better at ruling-out significant infection than one taken 90mins into the illness.

And this paper quite elegantly shows this.

But – what it doesn’t do – and what almost no papers of inflammatory biomarkers in the neutropenic sepsis literature do either – is look at the patient first. They don’t give us what we really want – which is the added value of CRP to help us with ‘pretty damn well looking child with early septicaemia’. They all get lumped and the CRP is taken without context.

How can we help researchers to do studies that reflect what we all learn – to look at the patient first and add the results of tests to what we already know?

– Bob Phillips