Model-based estimation of measures of association for time-to-event outcomes

Table 1 Event times generated according to a Cox-exponential model with a confounder X and an exposure status Z

CGPM was adopted using the Cox proportional hazard regression model. PV model was adopted with identity link. Pseudo times were fixed at quantiles 5%, 25%, 50%, 75%, 95% of the failure time distribution. The Z covariate was inserted in the PV model with an interaction with the baseline risk function modelled with a restricted cubic spline with 3 knots (Z td) giving an estimated risk difference varying through follow-up time, RD(t), and without time-dependence to estimate a constant RD through follow-up time between the first and the last pseudo-times (Z const).

ISSN: 1471-2288