Joint modeling of time-to-event data such as survival or time to disease progression incorporating with the longitudinal data has been an active research topic. Software programs are also available in the public domain to perform this kind of data analysis. However, most of the programs are limited to one longitudinal data series and to extend this to multiple longitudinal series is remaining as a challenge.
In this research, we estimate the treatment effects on disease progression using joint modeling with multiple data series of laboratory tests. We also estimate the variations of estimates via bootstrap. We compare the results from the existing software programs with respect to their consistency. Data from a recent clinical trial is used to illustrate the proposed approach.
|