Survival analysis (or time to an event analysis), and the Cox regression model--methods for longitudinal psychiatric research

Abstract

A multivariate statistical instrument is presented, which combines the life table with regression analysis: the Cox proportional hazards regression model. It is suitable for analyzing longitudinal clinical data, and to find predictors of different outcomes in psychiatric research such as death, relapse or cure. All patient information can be used, since the time to the outcome event is included in the model. Treatments can be compared, and prognostic indices created.