Authors: Chloé Szurewsky, Guosheng Yin, Gwénaël Le Teuff
Pre-print avaible [here] (https://doi.org/10.48550/arXiv.2506.18608)
Tests_SA.R
This file contains the functions required to run the simulations :
- OSLRT and mOSLRT are the classical one-sample log-rank test (Finkelstein et al, 2003) and its modified version (Wu, 2014) ;
- Score_EE, Score_ME, Score_DE and Score_CH are the developed score tests for, respectively, an early, a middle, a delayed effect and crossing hazards ;
- rmst_single returns the estimated value of the RMST and test_SA the test based on the RMST for single-arm trials;
- maxcombo1 is the combination test presented in the article that combines the mOSLRT, early effect score test and delayed effect score tests.
Simulations.R
This file runs the simulation study and gives the type I error or power as shown in Figure 2, and in Figure B3 in appendix by changing the censoring rate, in Figure B4 and B5 in appendix with other hazard ratio.
Simulations_parameter_misspe.R
This file runs the simulations to study the impact of the variability on the exponential parameter of the historical control group (section 6.1) and returns the relative difference in terms of type I error and power as shown in Figure 4 and Figure B11 in appendix with other censoring rates.
Simulations_dsitribution_misspe.R
This file runs the simulations to study the impact of the model misspecification of the survival distribution of the external control curve (section 6.2) and returns the relative difference in terms of type I error and power as shown in Figure 5 and Figure B13 in appendix with other censoring rates.
For questions or remarks about the code, please contact C.Szurewsky (chloe.szurewsky@gustaveroussy.fr) or G. Le Teuff (gwenael.leteuff@gustaveroussy.fr).