Interpretation of tauBs in output of jointModelBayes

by DocEd   Last Updated October 09, 2019 15:19 PM

I am using the JMbayes package for R to fit joint models between a longitudinal and time-to-event outcome. The model output lists a variable for "tauBs" however I am uncertain as to what this refers. I'm am fairly new to a bayesian approach, so any help is appreciated!

See below for a reprex (drawn directly from the package documentation):

library(JMbayes)
lmeFit.aids <- lme(CD4 ~ obstime * drug, random = ~ obstime | patient, data = aids)
survFit.aids <- coxph(Surv(Time, death) ~ drug, data = aids.id, x = TRUE)
jointFit.aids <- jointModelBayes(lmeFit.aids, survFit.aids, timeVar = "obstime")
summary(jointFit.aids)

And output:

Call:
jointModelBayes(lmeObject = lmeFit.aids, survObject = survFit.aids, 
    timeVar = "obstime")

Data Descriptives:
Longitudinal Process        Event Process
Number of Observations: 1405    Number of Events: 188 (40.3%)
Number of subjects: 467

Joint Model Summary:
Longitudinal Process: Linear mixed-effects model
Event Process: Relative risk model with penalized-spline-approximated 
        baseline risk function
Parameterization: Time-dependent value 

 LPML      DIC      pD
 -Inf 11631.06 1155.18

Variance Components:
             StdDev    Corr
(Intercept)  4.5839  (Intr)
obstime      0.5724 -0.0284
Residual     1.7145        

Coefficients:
Longitudinal Process
                  Value Std.Err Std.Dev    2.5%   97.5%      P
(Intercept)      6.9712  0.0074  0.3048  6.3736  7.5726 <0.001
obstime         -0.2282  0.0010  0.0435 -0.3101 -0.1358 <0.001
drugddI          0.4871  0.0107  0.4354 -0.3784  1.3464  0.263
obstime:drugddI  0.0001  0.0015  0.0609 -0.1210  0.1183  0.990

Event Process
           Value Std.Err  Std.Dev    2.5%    97.5%      P
drugddI   0.3390  0.0077   0.1965 -0.0736   0.6991  0.092
Assoct   -0.2942  0.0016   0.0386 -0.3706  -0.2222 <0.001
tauBs   216.0732 18.0529 182.6835 19.3259 708.1347     NA

MCMC summary:
iterations: 20000 
adapt: 3000 
burn-in: 3000 
thinning: 10 
time: 2 min

Many thanks in advance!

Tags : bayesian


Related Questions



Help! Newton-Raphson explodes!

Updated April 22, 2017 09:19 AM



Hierarchical model for the amount of data I have

Updated May 13, 2019 10:19 AM