Desktop version

Home arrow Economics

  • Increase font
  • Decrease font


<<   CONTENTS   >>

Two Time-to-Event Endpoints: A Two-Stage Approach

The analysis for the surrogacy setting of two time-to-event endpoints was discussed in Chapters 5 and 12, where a joint model and a two-stage approach

FIGURE 14.2

Age-Related Macular Degeneration Trial. Default output for the reduced fixed- effects model.

was used to estimate individual- and trial-level surrogacy. The two-stage approach is implemented in the Surrogate Shiny App. We use the ovarian cancer study (see Section 2.2.3) for illustration. The specification of the data is shown in Figure 14.3. In the same screen, we select the tab Two-stage (Survival- Survival) to perform the analysis, to select the censoring indicators and to determine the model's type for the second stage (weighted or unweighted). The output is presented in Figure 14.4. The estimated trial-level surrogacy is equal to R(ria = 0.9184 (0.8674, 0.9695). For the ovarian cancer dataset, the two-stage model discussed above is identical to the model specified in the function TwoStageSurvSurv:

Sur<-TwoStageSurvSurv(Dataset=ovarian, Surr=pfs,

SurrCens=PfsInd,True=surv,TrueCens=SurvInd,

Treat=Treat,Trial.ID=Center)

 
<<   CONTENTS   >>

Related topics