Translational risk models

With rapid progression of computing and other technological advances, the practice of modern medicine has moved from primarily anecdotal to largely quantitative. With due credit to the Internet and the new cyber-society, individuals have taken a more active role in the decision-making process concerning their health, from deciding whether or not to get screened for a disease to which treatment is best for their specific clinical profile. Treating physicians are more connected with latest medical breakthroughs through vast dissemination via the Internet. Statistical prediction models assembled on large well-designed cohorts, multiply validated and easily accessible through online calculators play a role in translating basic science results to implementation in the community for public health benefit. This chapter describes the risk model building process that forms the basis of modern medical decision-making, from statistical estimation to validation and implementation on the Internet. The early diagnosis of cancer is used as the context to illustrate principles, though the concepts immediately transcend to other disciplines as concluding examples in forestry and finance will show.