Table 2. Clinical outcomes in the two molecular phenotypes.
Hypoinflammatory | Hyperinflammatory | P-value | ||||||
---|---|---|---|---|---|---|---|---|
N | ICU free days | Mortality | N | ICU free days | Mortality | ICU free days | Mortality | |
VALID * | 804 | 21 (11 -24) | 133 (17%) | 336 | 7 (0 - 22) | 143 (43%) | < 0.0001 | < 0.0001 |
EARLI * | 530 | 23 (17 - 25) | 107 (20%) | 288 | 12 (0 - 23) | 129 (45%) | < 0.0001 | < 0.0001 |
VALID* ARDS excluded | 532 | 23 (17 - 25) | 70 (13%) | 236 | 16 (0 - 23) | 80 (34%) | < 0.0001 | < 0.0001 |
EARLI* ARDS excluded | 346 | 24 (21 - 25) | 48 (14%) | 226 | 22 (0 - 24) | 77 (34%) | < 0.0001 | < 0.0001 |
PROWESS-SHOCK ** | 1142 | 13 (0 - 20) | 233 (20%) | 538 | 0 (0 - 14) | 192 (36%) | < 0.0001 | <0.0001 |
VASST ** | 455 | 11 (0 - 20) | 127 (28%) | 323 | 0 (0 - 12) | 163 (51%) | < 0.0001 | <0.0001 |
In-hospital mortality;
Mortality at Day 28.
Intensive care unit (ICU) free days were censored at Day 28, such that patients that died before Day 28 were assigned zero ICU-free days. P-values for ICU free days were generated using the Wilcoxon-rank test and for mortality using the Chi-square test.