Periodically people (doctors/CCGs etc) ask QDiabetes about implementation. I will answer anything here as I go along…
Is there a sensible cut-off for screening risk?
I suspect 2016 will see a lot of interest in QDiabetes with the Diabetes Prevention Program coming to the fore.
Most of what we learnt is on the predm.co.uk website, but we found pragmatically that using >15% 10yr risk to give about the best “bang per buck” i.e. those with risks of 50% had a higher hit rate, but we were missing many more lower down the spectrum with lower risks but higher volume of people. Too low a cut off and you might as well screen everyone. I think in the QDiabetes website it refers to 14% risk which correlates well.
What is interesting is QDiabetes is not at all the same as QRisk scoring.
We found that writing to people and including via mailmerge what their %age risk score was, to be a very effective means of getting responses to an invitation to screening, though unfortunately if you tell someone they are low risk if you calculate it ad hoc, it can have a negative effect of falsely reassuring some folk!
Can QDiabetes be run natively on INPS and SystmOne systems
Currently no, we used a standalone data processor tool to run in our CCG using approximated values. I’ve heard talk of integration in INPS but nothing firm