|
Abstract:
|
Joint models for longitudinal and time-to-event data are useful in situations where an association exists between a longitudinal measurement and a survival time. These models can become complicated because of shared random effects and multiple submodels. Several statistical software procedures exist to assist in the fitting of these models. We review the available functions for frequentist and Bayesian models in the statistical programming languages R, SAS, STATA, and WINBUGS as well as an option in the non-statistical program MATLAB. A description of each procedure is given including estimation techniques, input and data requirements to run the code, available options for customization, and some available extensions, such as competing risks models. The software options are discussed and compared citing differences, strengths and weaknesses. Finally the implementation is illustrated on a dataset from a clinical trial.
|
