Table 2.
Results in term of univariable analysis and final multivariable model.
| Variable | Transition Univariable Selection | Final Multivariable Model | ||
|---|---|---|---|---|
| Transition | Hazard Ratio (95% CI) | p-Value | ||
| Sex | 16 | None | ||
| Age > 50 | 16 | None | ||
| Age > 60 | 6, 11, 12, 13, 14, 16, 17 | 13 | 0.14 (0.04, 0.48) | 0.002 |
| 14 | 1.55 (0.99, 2.41) | 0.054 | ||
| 16 | 9.84 (4.41, 21.97) | <0.001 | ||
| Age > 70 | 9 | 9 | 7.5 (2.47, 22.76) | <0.001 |
| BMI > 25 | 6, 7, 9 | 6 | 1.8 (1.22, 2.66) | 0.003 |
| BMI > 30 | 6, 13 | 6 | 0.67 (0.48, 0.93) | 0.016 |
| SAPS II > 25 | 1, 6, 7, 10, 11, 12, 13, 16 | 1 | 0.66 (0.51, 0.85) | 0.001 |
| 6 | 0.58 (0.41, 0.82) | 0.002 | ||
| 7 | 1.66 (1.11, 2.47) | 0.014 | ||
| 13 | 5.86 (1.24, 27.64) | 0.026 | ||
| 16 | 0.23 (0.08, 0.61) | 0.003 | ||
| SAPS II > 33 | 7, 13 | 13 | 0.4 (0.16, 1.03) | 0.058 |
| SAPS II > 44 | 1, 9, 14 | 14 | 1.6 (1.06, 2.41) | 0.025 |
| Charlson > 0 | 1, 6, 10, 14 | 1 | 0.75 (0.59, 0.95) | 0.017 |
| 6 | 0.64 (0.49, 0.84) | 0.001 | ||
| 14 | 2.23 (1.34, 3.70) | 0.002 | ||
| Charlson > 2 | 9, 10 | None | ||
| Number of days in hospital before ICU > 2 | 6, 7, 11, 12, 14, 16, 17 | None | ||
| Number of days from first symptoms to ICU> 10 | 7, 16, 17 | None | ||
| SOFA > 4 | 7 | 7 | 2.39 (1.71, 3.35) | <0.001 |
| SOFA > 5 | 1, 7, 17 | None | ||
| SOFA > 8 | 1, 11, 14, 17 | 11 | 0.19 (0.04, 0.78) | 0.022 |
| 14 | 1.84 (1.21, 2.81) | 0.004 | ||
| Leucocytes > 6000 (×109 per L) | 12, 14, 17 | None | ||
| Leucocytes > 10,000 (×109 per L) | 10, 14, 17 | 14 | 0.57 (0.36, 0.89) | 0.014 |
| CRP > 150 | 6, 12 | 12 | 0.53 (0.34, 0.81) | 0.004 |
| Lymphocytes > 1000 (×109 per L) | 1, 6, 10, 12, 14 | 6 | 1.49 (1.13, 1.96) | 0.005 |
| Temperature > 39 °C | 7 | 7 | 1.98 (1.41, 2.77) | <0.001 |
| Corticosteroids | 6, 7, 11, 14 | 7 | 0.59 (0.39, 0.90) | 0.016 |
| Ritonavir/lopinavir | 1, 12 | None | ||
| Hydroxychloroquine | 1, 12, 17 | None | ||
| Tocilizumab/anakinra | 1, 11, 12 | 12 | 1.8 (1.02, 3.17) | 0.043 |
| Heparin (therapeutic) | 6 | 6 | 0.58 (0.42, 0.81) | 0.001 |
Legend: In the second column, the transitions in which each variable was selected as a risk factor (p-value < 0.2 in univariate analysis) are shown. Transition coding: (1) from ICU discharge to hospital discharge, (2) from ICU discharge to ICU non-invasive (no covariate tested), (3) from ICU discharge to ICU invasive (no covariate tested), (4) from ICU discharge to death (no covariate tested), (5) from ICU non-invasive to hospital discharge (no covariate tested), (6) from ICU non-invasive to ICU discharge, (7) from ICU non-invasive to ICU invasive, (8) from ICU non-invasive to ECMO (no covariate tested), (9) from ICU non-invasive to death, (10) from ICU invasive to hospital discharge, (11) from ICU invasive to ICU discharge, (12) from ICU invasive to ICU non-invasive, (13) from ICU invasive to ECMO, (14) from ICU invasive to death, (15) from ECMO to ICU non-invasive (no covariate tested), (16) from ECMO to ICU invasive, and (17) from ECMO to death. ICU: intensive care unit. ECMO: extracorporeal membrane oxygenation. ICU non-invasive: in ICU without mechanical ventilation. ICU invasive: in ICU with mechanical ventilation. BMI: body mass index. SAPS: Simplified Acute Physiology Score. SOFA: sequential organ failure assessment score. CRP: C-reactive protein.