Table 4.
Comparing distributed Cox regression, pooled Cox regression, and meta-analyses with real data, under two partitions. Results presented are log hazard ratio (logHR) and estimated standard error (ese)
Partition 1 | Partition 2 | ||||
---|---|---|---|---|---|
logHR | ese | logHR | ese | ||
Unstratified pooled |
![]() |
0.1007 |
![]() |
0.1007 | |
Unstratified distributed |
![]() |
0.1043 |
![]() |
0.1063 | |
Stratified pooled |
![]() |
0.1007 |
![]() |
0.1240 | |
Stratified distributed |
![]() |
0.1045 |
![]() |
0.1272 | |
Univariate meta-analysis |
![]() |
0.1012 |
![]() |
0.1302 | |
Multivariate meta-analysis |
![]() |
0.1011 |
![]() |
0.1210 |