Software

Three freely available R packages are available for fitting joint models, including advanced cases involving competing risks, multivariate data and meta-analyses.

joineR

Analysis of repeated measurements and time-to-event data via random effects joint models. Fits the joint models proposed by Henderson and colleagues (single event time) and by Williamson and colleagues (2008) (competing risks events time) to a single continuous repeated measure. The time-to-event data is modelled using a (cause-specific) Cox proportional hazards regression model with time-varying covariates. The longitudinal outcome is modelled using a linear mixed effects model. The association is captured by a latent Gaussian process. The model is estimated using am Expectation Maximization algorithm. Some plotting functions and the variogram are also included. This project is funded by the Medical Research Council (Grant numbers G0400615 and MR/M013227/1).

A reference manual and two vignettes tailored to the needs of medical and public health researchers are freely available.

For further details, please see the dedicated development page.

joineRML

Fits the joint model proposed by Henderson and colleagues (2000), but extended to the case of multiple continuous longitudinal measures. The time-to-event data is modelled using a Cox proportional hazards regression model with time-varying covariates. The multiple longitudinal outcomes are modelled using a multivariate version of the Laird and Ware linear mixed model. The association is captured by a multivariate latent Gaussian process. The model is estimated using a Monte Carlo Expectation Maximization algorithm. This project is funded by the Medical Research Council (Grant number MR/M013227/1).

A reference manual and two vignettes are freely available.

For further details, please see the dedicated development page.

joineRmeta

The joineRmeta package implements methods to analyse multi-study joint data consisting of a single continuous longitudinal outcome, and a single possible censored time-to-event outcome. The modelling framework for the longitudinal data is a linear mixed effects model (Laird and Ware, 1982). The modelling framework for the time-to-event outcome is a Cox proportional hazards model with an unspecified baseline hazard (Cox, 1972). The longitudinal and time-to-event sub-model are linked through an association structure. Currently only the random effects only proportional association is available (see Gould et al, 2015). The methodology used to fit the model is described in Henderson et al (2000) and Wulfsohn and Tsiatis (1997).

The joineRmeta package contains methods to perform the second stage of a two stage meta-analysis (MA) of study specific joint modelling fits, and one stage MA of multi-study joint data where between study heterogeneity can be accounted for using interaction terms with study membership variables, study level random effects, or baseline hazard stratified by study. The package also contains plotting and simulation functions.

A reference manual and vignette are freely available.

For further details, please see the dedicated development page.