Cox	Shared	Nested	Additive	Joint standard (Bivariate: 1 RE + 1 TE)	Joint cluster (Bivariate: 1 RE + 1 TE)	Joint general (Bivariate: 1 RE + 1 TE)	Joint nested (Bivariate: 1 RE + 1 TE)	Joint longitudinal (Bivariate: 1 LO + 1 TE)	Joint trivariate (Trivariate: 1 LO + 1 RE + 1 TE)	Joint non linear trivariate (Trivariate: 1 LO + 1 RE + 1 TE)	Joint Multivariate (Trivariate: 2 RE + 1 TE)	Joint surrogate (Bivariate: 2 TE)	Conditional two-part joint (Trivariate: 2 LO 1 TE)	Marginal two-part joint (Trivariate: 2 LO 1 TE)
Available options
Gamma distribution
Log-Normal distribution													<img src=“blue_x.png” title=“#– Surrogacy evaluation based on ganerated data with a combination of Monte Carlo and classical Gaussian Hermite integration :
jointSurroPenal(data=data.sim,int.method=2,nb.mc=300,nb.gh=20)”/>
Left-truncation
Interval Censoring
Two strata
More strata (max=6)
Time-dependant covariates
Calendar timescale
Weibull
Piecewise
Available output
Predicted frailties			<img src=“blue_x.png” title=“fit$frailty.pred.group
fit$frailty.pred.subgroup”/>	<img src=“blue_x.png” title=“fit$frailty.pred
fit$frailty.pred2”/>				<img src=“blue_x.png” title=“fit$frailty.pred
fit$frailty.fam.pred”/>
Variances of the frailties
Martingale residuals					<img src=“blue_x.png” title=“fit$martingale.res
fit$martingaledeath.res”/>			<img src=“blue_x.png” title=“fit$martingale.res
fit$martingaledeath.res”/>		<img src=“blue_x.png” title=“fit$martingale.res
fit$martingaledeath.res”/>
Prediction methods
Marginal prediction of a terminal event	<img src=“blue_x.png” title=“#– prediction of death for two patients between 100 and 100+w, with w in (50,100,…,1900) :
prediction(cox.fit,datapred,t=100,window=seq(50,1900,50))”/>				<img src=“blue_x.png” title=“#– prediction of death between 100 and 100+w given relapses (with confidence intervals) :
prediction(joint.fit,datapred,t=100,window=seq(50,1500,50),event="Terminal”,MC.sample=100)“/>			<img src="blue_x.png” title=“#– Compute prediction over the individuals 274 and 4 :
prediction(JointNested.fit,data=DataPred,t=500,window=seq(100,1500,200),conditional=FALSE,individual=c(274,4))”/>	<img src=“blue_x.png” title=“#– prediction of death between 1 year and 1+2 given history of the biomarker :
prediction(JointLongi.fit,datapredj,datapredj_longi,t=1,window=2)”/>	<img src=“blue_x.png” title=“#– prediction of death between 1 year and 1+w :
prediction(trivPenal.fit,datapredj,datapredj_longi,t=1,window=seq(0.5,2.5,0.2),MC.sample=100)”/>	<img src=“blue_x.png” title=“#– prediction of death between t and t+0.5 :
prediction(trivPenalNL.fit,datapredj,datapredj_longi,t=seq(2,3,0.2),window=0.5,MC.sample=100)”/>
Conditional prediction of a terminal event
Marginal prediction of a new recurrent event					<img src=“blue_x.png” title=“#– prediction of death between 100 and 100+w given relapses (with confidence intervals) :
prediction(joint.fit,datapred,t=100,window=seq(50,1500,50),event="Recurrent”,MC.sample=100)“/>
Conditional prediction of a new recurrent event
Model evaluation
Cmeasures
Epoce
Model structure
STE
Prediction of treatment effects
Statistical model
Mechanistic model (ODE)

Table 1: Package characteristics. Blue cross is for option available for a given type of model in the package on CRAN, orange cross is for option included in the package but not on CRAN yet. Empty cells mean that an option is not available for a given type of model (either not coded yet or simply not applicable). RE = Recurrent Event; TE = Terminal Event; LO = Longitudinal Outcome; STE = Surrogate threshold effect; ODE = Ordinary Differential Equation.

Reference

Agnieszka Krol, Audrey Mauguen, Yassin Mazroui, Alexandre Laurent, Stefan Michiels, Virginie Rondeau (2017), Tutorial in Joint Modeling and Prediction: A Statistical Software for Correlated Longitudinal Outcomes, Recurrent Events and a Terminal Event. Journal of Statistical Software, 81.

Casimir Ledoux Sofeu and Virginie Rondeau (2020). How to use frailtypack for validating failure-time surrogate endpoints using individual patient data from meta-analyses of randomized controlled trials. PLOS ONE, 15, 1-25.

Shiny

A shiny application of frailtypack is available, allowing modelisation and prediction of several models presents in frailtypack, at https://frailtypack-pkg.shinyapps.io/shiny_frailtypack. This application can be run in a local mode in R thanks to the function runShiny().

Note

If you want to take in charge frailtypack as a project, but work on a macOSX with a version of R >= 3.6, a problem must happen between openmp and clang. clang does not seem to support option '-fopenmp'. To avoid this, you should edit your personal Makevars (~/.R/Makevars), admitting you use gfortran, by relocating CC=clang by CC=gfortran