• %JM is a SAS® macro to fit shared parameter models (SPMs) for the joint modelling of longitudinal and time-to-event responses
  • Four possible distributions are available for the longitudinal response:
                1. Normal
                2. Binary
                3. Binomial
                4. Poisson 
  • A range of options are available to fit subjects' evolutions of the longitudinal response over time:  
               1. A linear function (random intercepts/slopes model)
               2, Splines
               3. Natural cubic splines
               4. B-splines
  • For the time-to-event response, the %JM macro fits parametric, stratified or not, proportional hazard models. The baseline hazard function needs to be specified with one of the following options:
               1.  Exponential
               2.  Weibull
               3, Piecewise exponential
               4. Generalization of the Gompetz model based on splines
               5. Generalization of the Weibull model based on splines
               6. Royston-Parmar model
  • The macro offers multiple options to connect the time-to-event model and the longitudinal model:
                1. A (trajectory) current-value dependent shared parameter
                2. A (trajectory) slope dependent shared parameter
                3. Lagged effects
                4. Cumulative effects
                5. Random effects coefficients
                6. Interaction effects
                7. Any combinations of the above options
Imagen
Longitudinal and time-to-event data

Download the %JM Macro V02.02 Package (link below)

jmmacrov02.02.zip
File Size: 862 kb
File Type: zip
Download File
NEW LINK!
Macro Manuscript at the Journal of Statistical Software​
Developed by
     Alberto Garcia-Hernandez* and Dimitris Rizopoulos**

*   Astellas Pharma Global Development Europe, Leiden, The Netherlands
** Erasmus Medical Center, Erasmus University, Rotterdam, The Netherlands