Skip to main content

Table 2 Performance for scenario I with true intervention effect of \({\beta }_{t}=-0.264\)

From: Comparison of statistical models for estimating intervention effects based on time-to-recurrent-event in stepped wedge cluster randomized trial using open cohort design

Event Generation model Number of steps (clusters)    
AG model PWP-TT model PWP-GT model
Bias MSE CP Bias MSE CP Bias MSE CP
Poisson 2 0.0267 0.0034 0.937 0.0006 0.0032 0.959 0.0393 0.0047 0.868
4 0.0259 0.0036 0.941 0.0013 0.0036 0.948 0.0386 0.0051 0.894
5 0.0266 0.0040 0.933 0.0021 0.0040 0.940 0.0380 0.0054 0.899
8 0.0255 0.0041 0.943 0.0011 0.0041 0.938 0.0397 0.0056 0.896
10 0.0251 0.0042 0.936 0.0016 0.0042 0.943 0.0376 0.0057 0.897
20 0.0259 0.0042 0.938 0.0020 0.0042 0.949 0.0363 0.0057 0.912
Mixed-Poisson 2 0.0283 0.0035 0.930 0.0085 0.0032 0.947 0.0572 0.0065 0.785
4 0.0284 0.0039 0.923 0.0099 0.0037 0.936 0.0563 0.0068 0.826
5 0.0290 0.0041 0.927 0.0106 0.0039 0.938 0.0543 0.0067 0.844
8 0.0282 0.0043 0.932 0.0097 0.0041 0.943 0.0547 0.0070 0.846
10 0.0254 0.0042 0.937 0.0075 0.0042 0.947 0.0534 0.0071 0.845
20 0.0281 0.0046 0.920 0.0097 0.0045 0.938 0.0530 0.0073 0.856
Weibull (constant) 2 0.0462 0.0052 0.870 0.0424 0.0050 0.882 0.0016 0.0032 0.944
4 0.0546 0.0065 0.848 0.0509 0.0062 0.867 0.0003 0.0037 0.945
5 0.0567 0.0067 0.858 0.0529 0.0065 0.864 0.0032 0.0037 0.949
8 0.0563 0.0071 0.849 0.0522 0.0068 0.863 0.0016 0.0043 0.934
10 0.0566 0.0072 0.842 0.0522 0.0069 0.860 0.0019 0.0042 0.945
20 0.0553 0.0070 0.869 0.0496 0.0067 0.884 0.0004 0.0044 0.949
Weibull (change) 2 0.1229 0.0162 0.100 0.0253 0.0026 0.897 -0.0014 0.0019 0.950
4 0.1221 0.0162 0.157 0.0299 0.0032 0.892 -0.0019 0.0021 0.963
5 0.1247 0.0168 0.162 0.0328 0.0033 0.896 -0.0004 0.0022 0.955
8 0.1240 0.0167 0.171 0.0330 0.0035 0.902 -0.0014 0.0024 0.954
10 0.1250 0.0171 0.191 0.0348 0.0038 0.888 0.0010 0.0025 0.956
20 0.1237 0.0169 0.222 0.0334 0.0038 0.901 -0.0012 0.0026 0.949
  1. AG Andersen-Gill, PWP-TT Prentice-Williams-Peterson Total-Time, PWP-GT Prentice-Williams-Peterson Gap-Time, MSE Mean square error, CP Coverage probability