Prediction models for cardiovascular disease in diabetes mellitus

Throughout the history of medicine, physicians have diagnosed and treated patients relying on their complaints and symptoms. Today, to gauge a patient’s risk of future ill health, physicians rely in addition on patient characteristics expressed as numerical values from physical and laboratory measurements, and from family and past medical histories. Risk scores represent examples of mathematical equations that utilize this information to model reality. Although sometimes not recognized as such, models currently aid in the everyday care of patients with diabetes and include, for example, simple models for adiposity (e.g. body mass index), more complex models for glomerular function, and even more complex algorithms to calculate dosages for continuous subcutaneous insulin based on levels of blood glucose, insulin sensitivity, exercise, diet, and more.

Calculators such as the Framingham or United Kingdom Prospective Diabetes Study (UKPDS) risk equations are increasingly being used to predict the occurrence of cardiovascular disease (CVD) and death. Among the complications of diabetes, CVD, comprising cerebrovascular, coronary, and depending on the definition, peripheral arterial disease, occurs most frequently and generates the highest costs. Cardiovascular risk scores provide a numerical estimate of the risk of future CVD and death from CVD, conditional on the presence of a number of factors.