Table 3.
Crude Cox regression analysis of survival from the beginning of best supportive care.
| Variable | Univariate Analysis | Multivariate Analysis | ||||
|---|---|---|---|---|---|---|
| HR | 95% CI | p-value | HR | 95% CI | p-value | |
| Anticancer therapy (none vs. used) | 1.40 | 1.26–1.55 | 0.0001 | 1.73 | 1.27–2.37 | 0.001 |
| Performance status (0 → 4) | 1.70 | 1.48–1.94 | 0.0001 | 1.66 | 1.44–1.93 | 0.0001 |
| Disseminated intravascular coagulation (absent vs. present) | 4.85 | 2.72–8.66 | 0.0001 | 3.74 | 2.01–6.96 | 0.0001 |
| Age (low → high) | 0.96 | 0.95–0.98 | 0.0001 | 0.97 | 0.95–0.98 | 0.0001 |
| White blood cell (×103) (low → high) | 1.05 | 1.03–1.07 | 0.0001 | 1.02 | 1.01–1.04 | 0.007 |
| Lactate dehydrogenase (×103) (low → high) | 1.01 | 1.003–1.02 | 0.007 | 1.01 | 1.00–1.02 | 0.011 |
| Dementia (absent vs. present) | 0.61 | 0.38–0.96 | 0.034 | 0.66 | 0.40–1.08 | 0.10 |
| Alanine transaminase (low → high) | 1.002 | 1.001–1.004 | 0.006 | 1.001 | 0.999–1.003 | 0.33 |
| Brain metastasis (absent vs. present) | 1.53 | 1.12–2.09 | 0.007 | 0.99 | 0.72–1.36 | 0.94 |
| Blood urea nitrogen (low → high) | 1.01 | 0.999–1.02 | 0.092 | 0.998 | 0.99–1.01 | 0.69 |
| Stage (1 → 4) | 1.30 | 1.14–1.48 | 0.0001 | NI | ||
| Hemoglobin (low → high) | 0.84 | 0.79–0.89 | 0.0001 | NI | ||
| Albumin (low → high) | 0.40 | 0.32–0.49 | 0.0001 | NI | ||
| C-reactive protein (low → high) | 1.04 | 1.03–1.06 | 0.0001 | NI | ||
| Patient’s wish about best supportive care (absent vs. present) | 0.45 | 0.32–0.62 | 0.0001 | NI | ||
| Doctor’s decision about best supportive care (absent vs. present) | 2.13 | 1.48–3.09 | 0.0001 | NI | ||
| Transfer to hospice or home (absent vs. present) | 0.44 | 0.37–0.53 | 0.0001 | NI | ||
| Liver dysfunction (absent vs. present) | 1.46 | 0.96–2.21 | 0.074 | NI | ||
Abbreviations: HR, Hazard ratio; C.I., confidence interval; Variables with a p-value < 0.10 on univariate analysis were entered into multivariate Cox analysis using.
the simultaneous method. NI, not included in the best multivariate Cox regression model.