Generalized additive models for medical research

Abstract

This article reviews flexible statistical methods that are useful for characterizing the effect of potential prognostic factors on disease endpoints. Applications to survival models and binary outcome models are illustrated.