[Nomogram as predictive model in clinical practice].

A nomogram which is developed based on logistic regression analysis with multiple factors provides accurate prediction in various situations. The ability of the nomograms to predict diagnosis, staging and prognosis in prostate cancer and other disease has been confirmed to be better than other predictive models such as risk stratification and artificial neural network. Making a nomogram requires a fixed number of patients and multiple steps such as validations and calibrations. And when nomograms are developed at other institutions, validations are essential for physicians before use at the actual clinical level. We review the clinical significance of nomograms and introduce the process of making a nomogram.