Skip to main content

Table 1 Summary of the simulation protocol comprising main parameter values, marker and survival time distributions used for each of the simulation scenarios

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

Scenario

β0

β1

β2

bi

Σ11

Σ22

Σ33

h0(t)

a) LOCF and measurement error impact

1

3.2

0

0

N(0,Σ11)

0

0

0

Weibull (0.1,1.4)

2

3.2

−0.07

0

N2(0,Σ)

1.44

0.04

0

Weibull (0.1,1.4)

b1) Marker distribution

3

3.2

−0.07

0

BM ∗

1.44

0

0

Weibull (0.1,1.4)

4

3.2

−0.07

0

χ2(0.72)

1.44

0

0

Weibull (0.1,1.4)

5

3.2

−0.07

0

Γ(0.5,1.7)

1.44

0

0

Weibull (0.1,1.4)

6

3.2

−0.07

0

N(0,Σ11)

1.44

0

0

Weibull (0.1,1.4)

b2) Marker profile

7

3.2

−0.07

0.004

N3(0,Σ)

1.44

0.6

0.09

Weibull (0.1,1.4)

8

3.2

−0.16

0.01

N3(0,Σ)

1.44

0.6

0.09

Weibull (0.1,1.4)

b3) Baseline hazard

9

3.2

−0.07

0

N2(0,1.44)

1.44

0.04

0

h0(t)nonm

  1. BM ∗ denotes a bimodal mixture distribution 0.65∗N(8,1.44)+0.35∗N(15,1.44)
  2. h0(t)nonm=νκtκ−1/(c+tκ), where ν=1,κ=2,c=10