Table A1.
Univariate Cox Proportional Hazards Model of Missing v Determined Marker Data
Marker Data by End Point | Patients |
Analysis |
|||
---|---|---|---|---|---|
No. of Patients | No. Who Experienced Failure | HR* | 95% CI | P† | |
Overall mortality | |||||
Mssing | 1,043 | 565 | 0.98 | 0.85 to 1.14 | .82 |
Determined | 478 | 261 | |||
Cause-specific mortality | |||||
Missing | 1,043 | 146 | 1.21 | 0.92 to 1.58 | .17 |
Determined | 478 | 82 | |||
Distant metastasis | |||||
Missing | 1,043 | 189 | 1.12 | 0.92 to 1.49 | .20 |
Determined | 478 | 102 |
Abbreviation: HR, hazard ratio.
A hazard ratio of 1 indicated no difference between the two subgroups. The marker data indicator was coded as follows: 0, missing; 1, determined.
P value from χ2 test by using the Cox proportional hazards model.