I think the field of 'personalized health' in the form of genetic and environmental specific diets, disease prevention strategies and treatment programmes is going to be a huge field in the future. The problem arises when theoretical improvements are inefficient to implement. A good example is the above breast cancer screening. To (fairly) impiment our new knowledge we'd have to test the genotype of all 15 regions in all the women in our population (already an expensive feat) and then (not accounting for false negatives and false positives) calculate the risk of that women developing breast cancer at a specific age. After this we can have the discussion with the women (imagine if at this point she simply turns around and says 'not interested' or worse yet, doesn't show up to the appointment) and hope to god she listens and heeds our medical advice...

I think there's a huge future for personalised healthcare, but I also think we need to be smarted about how we approach it because it's far too easy to spend a lot of time and effort but make no impact at the end of the day.