Table 3.
Regression analysis for primary and secondary outcomes from symptom day 15 onwards
| Outcome Kd = 0.05 nM | Highest free IL-18 response coefficient (standard error) | Odds ratio per 50 pg/ml increase in highest free IL-18 (95% confidence interval) | Significance |
|---|---|---|---|
| Primary outcome | |||
| PaO2/FiO2 Ratio (mmHg) | – 0.377 (0.111) | p < 0.03 | |
| Secondary outcomes | |||
| Mortality | |||
| Death at 60 days from symptom onset (crude mortality) | 1.41 (1.1–2.0) | p < 0.03 | |
| Death at 60 days from symptom onset (hypoxaemic respiratory failure)✸ | 1.90 (1.3–3.1) | p < 0.01 | |
| Organ-support dependency | |||
| Any organ support✦ | 1.66 (1.1–2.9) | p < 0.05 | |
| Artificial mechanical ventilation only | 1.66 (1.00–2.9) | p = 0.05 | |
| Per additional organ supported✦ | 50.72 (23.0) | p < 0.05 | |
| Per additional organ supported (hypoxaemic respiratory failure)✦✸ | 63.67 (23.3) | p < 0.01 | |
| Biochemical inflammatory markers | |||
| C-reactive protein (CRP) | 0.065 (0.29) | p > 0.5 | |
| Lymphocyte (as % of white blood cells) | − 118.4 (109) | p > 0.2 | |
| Neutrophil/lymphocyte ratio | 2.67 (1.3) | p = 0.05 | |
| Neutrophil/lymphocyte ratio (hypoxaemic respiratory failure)✸ | 3.54 (1.4) | p < 0.03 | |
Results that reach the significance threshold are bolded
✸The “hypoxaemic respiratory failure group” excludes 4 patients whose calculated PaO2/FiO2 ratio from values recorded in the last 24 h of life did not fall below 300 mmHg at any time. The group is used to exclude those who died from causes other than hypoxaemic respiratory failure
✦“Per Additional Organ supported” refers to each new organ supported, out of: vasopressor support; artificial mechanical ventilation; renal replacement therapy