I found this website today: http://ubble.co.uk/
It was highlighted on cnn.com under the provocative headline: ‘quiz claims
to predict your death’
I probably should have instead found it by reviewing the scholarly
literature from which it was published:
http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(15)60175-1/abstract
Essentially, these reseachers have released a website of the ‘all-by-all’,
of all baseline characteristics relative to a large array of mortality
causes, providing age-adjusted predictions associated with each variable.
They’ve provided a fairly easy-to-use interface to explore the data and
build up a predictive model.
I’m impressed that this team has put all the pieces together and released
in an effective communications campaign. I think it serves as a powerful
example for what’s possible for making data useful both for academic
community and for the general public.
I’ll been keen to hear everyone’s thoughts on this development. Is this
the kind of direction you’d like to see the OHDSI community pursue? Are
there methodological issues with the approach they’ve taken that give you
pause? Wouldn’t it be amazing to have results from around the world,
across our 50+ databases and 600m+ patients all working together toward a
common goal of informing patient care?
Cheers,
Patrick
)