Skip to main content

Table 5 Results of the association parameter α estimated from the TVCM and joint model fitted to data generated considering slight and gross misspecifications of the longitudinal trajectories (scenarios 7 and 8 of Table 1), σε∈(0.1,0.3,0.5) with CV ∈(3.1%,9.4%,15.6%) and the true α=0.3

From: Joint model robustness compared with the time-varying covariate Cox model to evaluate the association between a longitudinal marker and a time-to-event endpoint

  

Slight misspecification

Gross misspecification

σε

Model

Est

ESE

ASE

%Bias

CP

Est

ESE

ASE

%Bias

CP

0.1

TVCM(1x/week)

0.295

0.056

0.056

-1.7

95

0.297

0.055

0.056

-1.0

95

 

TVCM(4x/week)

0.309

0.054

0.055

3.0

96

0.310

0.053

0.054

3.3

96

 

joint model

0.297

0.209

0.060

-1.0

92

0.280

0.222

0.059

-6.7

91

0.3

TVCM(1x/week)

0.277

0.054

0.054

-7.7

93

0.280

0.053

0.054

-6.7

93

 

TVCM(4x/week)

0.292

0.051

0.053

-2.7

95

0.294

0.051

0.052

-2.0

95

 

joint model

0.284

0.291

0.067

-5.3

91

0.273*

0.290

0.059

-9.0

91

0.5

TVCM(1x/week)

0.249

0.050

0.051

-17

83

0.252

0.050

0.051

-16

84

 

TVCM(4x/week)

0.264

0.047

0.050

-12

89

0.266

0.047

0.049

-11

90

 

joint model

0.265

0.192

0.071

-12

91

0.265*

0.167

0.065

-12

91

  1. Mean of the maximum likelihood estimates (Est), empirical Monte Carlo standard error (ESE), asymptotic standard error (ASE), percentage bias (%Bias) and 95% coverage probabilities (CP) are shown. The star (*) indicates that one extreme outlier estimate was removed