Lastly (for now) we were able to externally validate these algorithms using data from the CPRD Gold database in Scotland, Wales and Northern Ireland. This helps generate the highest grade research evidence to support implementation into practice.

The study, published today in @natcomms.nature.com, can be read here: www.nature.com/articles/s41...

this fourth graph shows how well the prediction model discriminates between risk of different types of cancer. We were able to include 15 types of cancer in our analysis due to the excellent linkage with the cancer registry.
www.nature.com/articles/s41...

This third graph shows how cancer risk varies by age, body mass index and haemoglobin levels, using routinely collected NHS data. there is huge clinical value for patient care through research using anonymised data from GPs so our thanks to the many GPs who share data
www.nature.com/articles/s41...

and the humble white cell and platelet counts (again done many millions of time each year in UK general practice) also show important associations with risk of an existing cancer
www.nature.com/articles/s41...

The #research used the anonymised electronic health records of 7.4 million adults in England plus the results of 7 routine blood tests to create the new algorithms which could revolutionise how cancer is detected in primary care, and make it easier for patients to get treatment at earlier stages.

The NHS currently uses prediction #algorithms, such as the QCancer scores, to combine relevant information from patient data and identify individuals deemed at high risk of having a currently undiagnosed cancer, enabling GPs and specialists to call them in for further testing.

Many thanks Azeem, that would be super. I'll drop you a line once we have landed