Table 1 Coefficients and 95% confidence intervals from the FPM and Poisson models of the mortality rates with two time-scales (t1 as time since surgery, and t2 as time since relapse or metastasis (RM)) and a time-varying indicator for the second time-scale (Indicator for RM) for Rotterdam Breast Cancer patients. Note: the spline coefficients are not interpretable on their own but they are used for making smooth predictions of the mortality rates
From: Modelling multiple time-scales with flexible parametric survival models
FPM | Poisson | |
|---|---|---|
Indicator for RM | 3.1891 (2.9314, 3.4468) | 3.1937 (2.9362, 3.4512) |
Age spline 1 | -0.0578 (-0.1092, -0.0063) | -0.0578 (-0.1092, -0.0063) |
Age spline 2 | -0.0005 (-0.0009, 0) | -0.0005 (-0.0009, 0) |
Age spline 3 | 0.0006 (-0.0002, 0.0014) | 0.0006 (-0.0002, 0.0014) |
Age spline 4 | -0.0001 (-0.0008, 0.0005) | -0.0001 (-0.0008, 0.0005) |
Age spline 5 | -0.0001 (-0.0004, 0.0002) | -0.0001 (-0.0004, 0.0002) |
Indicator for hormonal therapy | 0.3224 (0.1439, 0.5009) | 0.3221 (0.1436, 0.5006) |
log(t1) spline 1 | 0.8853 (0.4266, 1.344) | 0.8802 (0.4226, 1.3379) |
log(t1) spline 2 | 0.7396 (0.3272, 1.1519) | 0.7364 (0.3243, 1.1486) |
log(t1) spline 3 | -1.0513 (-1.7461, -0.3565) | -1.0467 (-1.7412, -0.3521) |
t2 spline 1 | 0.4519 (0.1646, 0.7392) | 0.4465 (0.1594, 0.7336) |
t2 spline 2 | 0.0562 (-0.0555, 0.1679) | 0.0546 (-0.0571, 0.1662) |
t2 spline 3 | -0.0153 (-0.0711, 0.0404) | -0.0146 (-0.0703, 0.0411) |
Intercept | -4.0529 (-4.3845, -3.7213) | -4.0549 (-4.3864, -3.7234) |