Kullback-Leibler-Based Discrete Relative Risk Models for Integration of Published Prediction Models with New Dataset

Existing literature for prediction of time-to-event data has primarily focused on risk factors from an individual dataset. However, these analyses may suffer from small sample sizes, high dimensionality and low signal-to-noise ratios. To improve prediction stability and better understand risk factors associated with outcomes of interest, we propose a KullbackLeibler-based discrete relative risk modeling procedure. Simulations and real data analysis are conducted to show the advantage of the proposed methods compared with those solely based on local dataset or prior models.