Skip to main content
. 2022 May 27;17(5):e0269005. doi: 10.1371/journal.pone.0269005

Table 4. Cox regression model for the outcome of the composite endpoint.

Measures COVID-19 Viral controls Bacterial controls
n = 191 n = 227 n = 96
Univariable Multivariable Univariable Multivariable Univariable Multivariable
HR (95% CI) p-value HR (95% CI) p-value HR (95% CI) p-value HR (95% CI) p-value HR (95% CI) p-value HR (95% CI) p-value
IL-6 6.602 (3.113–14.003) <0.001 6.602 (3.113–14.003) <0.001 3.630 (1.602–8.227) 0.002 - - 2.277 (1.200–4.321) 0.012 2.101 (1.042–4.239) 0.038
CRP 5.519 (2.187–13.932) <0.001 - - 7.651 (1.562–37.48) 0.012 7.651 (1.562–37.480) 0.012 2.484 (0.914–6.754) 0.075 3.091 (0.839–11.393) 0.090
PCT 2.796 (1.476–5.297) 0.002 - - 3.445 (1.414–8.394) 0.006 - - 1.862 (1.064–3.258) 0.029 - -
Ferritin 4.397 (1.632–11.845) 0.003 - - 0.927 (0.126–6.811) 0.941 - - 0.558 (0.215–1.449) 0.231 0.290 (0.082–1.022) 0.054
Leukocytes 5.066 (0.834–30.771) 0.078 - - 10.899 (0.027–4377.491) 0.435 - - 0.710 (0.099–5.116) 0.734 - -

p-values for comparison of HR were calculated using the log-rank test

values were logarithmized to approach a normal distribution, values for the multivariable model were selected using a backwards selection process

HR = hazard ratio, CI = confidence interval, IL-6 = interleukin-6, CRP = c-reactive protein, PCT = procalcitonin.