Michael Kattan, Ph.D.
Assistant Attending Outcomes Research Scientist
Department of Epidemiology and
Biostatistics
Nomograms for Cancer Prediction
This presentation is motivated by personal experience as a patient at M. D. Anderson, and a resulting pursuit of cancer prediction models. Nomograms—graphical tools that implement regression models—will be presented as useful tools for calculating outcome probabilities for individual patients. Some of our work in computerizing these models and comparing them with clinician prediction will be described. Finally, I'll advocate the use of computerized prediction models over standard approaches, such as staging systems.