| W/P/W | W/P/P | W/W/W |
---|
α0 | Bias | SD | RMSE | Power | Bias | SD | RMSE | Power | Bias | SD | RMSE | Power |
---|
0 | −0.0142 | 0.292 | 0.298 | 0.358 | −0.0087 | 0.287 | 0.292 | 0.345 | 0.0022 | 0.275 | 0.275 | 0.359 |
0.3 | −0.0541 | 0.258 | 0.255 | 0.447 | −0.0366 | 0.256 | 0.248 | 0.416 | −0.0070 | 0.245 | 0.231 | 0.396 |
0.6 | −0.0736 | 0.245 | 0.248 | 0.505 | −0.0544 | 0.243 | 0.240 | 0.464 | −0.0111 | 0.233 | 0.218 | 0.418 |
1 | −0.0883 | 0.235 | 0.247 | 0.554 | −0.0702 | 0.235 | 0.239 | 0.518 | −0.0141 | 0.223 | 0.212 | 0.448 |
- aResults correspond to scenario S27 defined with Weibull survival distribution for the historical control data and piecewise exponential distribution for the new data with HR = 0.786 and analysed with ω = 0 either with a Weibull model (W/P/W) or with a piecewise exponential model (W/P/P). These results are compared to scenario S3, given as a benchmark and defined by commensurate historical and new control data which follow a Weibull distribution, and are analysed with a Weibull model (W/W/W)