Thread by @drbobphillips

It’s the end of a 7 year journey for our @Leeds_Childrens pharmacist Rachel Boys who has completed her DPharm research project and is telling us today!

She was funded by @CandlelightersT as well as the @LeedsHospitals to look at how we can routinely measure renal function without using complex testing

We estimate how well the kidney works looming at the levels of creatinine in the blood, balanced against the height and weight of the patient.

(In adults there are sex and “ethnicity” fiddling too .. lucky we don’t need to get into the dodgyness of that.)

There are around a dozen variations on the equation used to fiddle between the height and weight and creatinine values. All of them are a bit unreliable.

Nearly all of the equations have been developed in people with poor renal function. It’s not been tested well in the paediatric oncology population, and this might give different results.

(Spoiler - they are a bit pants in PaedOnc kids)

Throwing a new measurement into the mix - Cystatin C - was thought to be the best way forward and we would get better.

(Spoiler - it didn’t link.springer.com/article/10.100… )

But maybe adding them together would work? A couple of people had done work before and shown it looked ok ... but as I am sure you all remember the replication of these rules is vital as they sometimes don’t work so brill

So that’s what Rachel did ... collected new data over time looking at all these old rules and seeing if they were good a) on the population or b) when used on individuals....

47 different individuals agreed to be part of this research, with around 100 separate complex (isotope GFR) measurements to compare against these slightly simpler ways you can get from just a single routine blood sample.

The data suggest the average renal function over treatment falls a bit - not enough to affect life - but the old equations stay pretty poor at guessing the real function (eg a known iGFR of 100 might have an estimated value of 80 or 160 ... so ...)

When you get a simple test saying “100” the real correct value is between 90-110 in only FOUR PERCENT of the time

(96/100 people with an estimated value of 100 have a true value that’s lower than 90 or higher than 110)

When we throw the new measurement in and look at the super shiny new equations we end up with next to no diff- they are slightly better (only 75/100 patients have a value that’s more than 10% different than out guess) ... but that’s still a bit rubbish.

Examining individuals who had multiple measurements to see if you could “adjust” the value to be correct for them in particular... found the same. It was not consistent. Not reliable. A bit pants.

So. Rachel has shown fairly effectively that the new super measure and the fancy new equations don’t give us the answers we want. We need to keep doing the iGFR, 4-8hours, extra day in hospital tests.

Not all “innovative ideas” work. Just “implementing” won’t always help.

This thread is #WhyWeDoResearch

@threadreaderapp Unroll please