Skip to main content
. 2019 Apr 15;199(8):980–986. doi: 10.1164/rccm.201807-1419OC

Table 2.

Comparison of Outcomes of Patients with Sepsis Response Signature 1 and 2 Phenotypes Given Either Hydrocortisone or Placebo as Study Drug 2

  SRS1
SRS2
P Value for Interaction
Hydrocortisone Placebo Hydrocortisone Placebo
n 27 35 31 24  
Kidney failure–free days, median (IQR), d 25 (1–28) 25 (9–28) 25 (4–28) 28 (25–28) 0.43*
28-d mortality, n/total (%) 9/27 (33) 13/35 (37) 13/31 (42) 2/24 (8) 0.02
ICU mortality, n/total (%) 7/27 (26) 9/35 (26) 11/31 (35) 2/24 (8) 0.08
Hospital mortality, n/total (%) 8/27 (30) 12/35 (34) 13/31 (42) 2/24 (8) 0.02
Kidney failure, n/total (%) 13/27 (48) 19/35 (54) 17/31 (55) 10/24 (42) 0.30
No. weaned from vasopressors for >24 h, n/total (%) 25/27 (93) 31/35 (89) 28/31 (90) 23/24 (96) 0.36
Time to shock reversal, median (IQR), h 30.6 (18.1–77.7) 43.8 (21.5–91.5) 58.9 (36.1–82.3) 89.5 (31.5–122.0) 0.60
Duration of mechanical ventilation, median (IQR), d 3.0 (2.0–12.0) 6.0 (2.0–11.5) 6.0 (2.0–14.5) 9.0 (6.0–20.0) 0.67*
Mean total SOFA score, median (IQR) 5.7 (3.6–9.0) 4.9 (3.6–7.2) 5.6 (3.7–8.3) 4.7 (3.5–6.3) 0.72§

Definition of abbreviations: IQR = interquartile range; SOFA = Sequential Organ Failure Assessment score; SRS = sepsis response signature.

Bold type indicates P < 0.05.

*

From the aligned rank transform test.

From logistic regression.

From Cox regression, treating deaths as never having the event of interest. Results were similar treating death as a competing risk.

§

From linear regression, applying a square root transform to the outcome.