Regarding how the individual risks have been assessed, this is clearly related to data mining - 'digging' through huge amounts of data, trying to correlate the 'circumstances' with clinical data.
Whether above statement has anything to do with your personal understanding of "data mining" I just do not know :)