TABLE 4.
Hazard Ratios of Treatment Group and Other Patient Characteristics on All-Cause Mortality, Based on a Multivariable Cox Model*
Hazard ratio | P value | 95% CI | |
---|---|---|---|
Age (year) | 1.06 | <.001 | (1.05, 1.08) |
Female | 0.80 | .065 | (0.63, 1.01) |
Neurosurgery clipping (compared to coiling) | 1.35 | .009 | (1.08, 1.68) |
WFNS grade (compared to grade 1) | |||
Grade 2 | 1.62 | <.001 | (1.25, 2.08) |
Grade 3 | 1.47 | .070 | (0.97, 2.22) |
Grades 4-6† | 5.69 | <.001 | (3.74, 8.66) |
Maximum target aneurysm lumen size (mm) (compared to < = 5) | |||
6-10 | 1.30 | .028 | (1.03, 1.65) |
> = 11 | 1.71 | .005 | (1.17, 2.49) |
Number of aneurysms detected (compared to 1) | |||
2 | 1.27 | .099 | (0.96, 1.67) |
3 | 1.43 | .171 | (0.86, 2.38) |
4 | 1.75 | .101 | (0.90, 3.43) |
WFNS = World Federation of Neurological Surgeons.
*No interaction between the treatment and other covariates was detected; the PH assumption was examined and confirmed using Schoenfeld residuals.
†The hazard ratios for WFNS grade 4 to 6 were close to each other without statistically significant differences. Due to small sample size for patients with WFNS grade 4 to 6, these 3 levels were combined in the Cox model.