Skip to main content

Table 3 Results of the association parameter α obtained from the baseline Cox model, the TVCM and the joint model fitted to data generated considering the linear marker trajectory (scenario 2 of Table 1) with α(0,0.3,0.6) and σε(0.1,0.3,0.5) with CV (3.1%,9.4%,15.6%). Mean of the maximum likelihood estimates (Est), asymptotic standard error (ASE), bias, percentage bias (%Bias) and 95% coverage probabilities (CP) are shown

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

  α=0α=0.3α=0.6
σεModelEstASEBiasCPEstASE%BiasCPEstASE%BiasCP
 Cox.baseline0.0000.0600.000950.2540.056-15860.5450.059-9.283
0.1TVCM(1x/week)-0.0030.060-0.003960.2930.058-2.3950.5820.061-3.093
 TVCM(4x/week)-0.0010.060-0.001950.2980.059-0.7940.5940.062-1.094
 joint model-0.0030.059-0.003960.3010.0580.3950.6040.0610.794
 Cox.baseline0.0000.0580.000960.2400.054-20790.5080.057-1562
0.3TVCM(1x/week)-0.0030.058-0.003960.2750.057-8.3920.5410.059-9.883
 TVCM(4x/week)-0.0010.058-0.001950.2790.057-7.0920.5500.059-8.386
 joint model-0.0030.060-0.003950.3010.0590.3950.6040.0640.794
 Cox.baseline0.0000.0550.000960.2160.051-28610.4490.053-2523
0.5TVCM(1x/week)-0.0030.055-0.003960.2440.053-19800.4740.055-2135
 TVCM(4x/week)-0.0020.055-0.002950.2470.054-18840.4800.055-2040
 joint model-0.0040.062-0.004950.3010.0620.3950.6060.0691.094