Skip to main content
. 2016 Jun 13;6:52. doi: 10.1186/s13613-016-0156-2

Table 6.

Univariable and multivariable logistic regression analyses of factors associated with ICU mortality in ARDS patients

n Death
n (%)
Univariable analysis Multivariable analysis
OR (95 % CI) p aOR (95 % CI) p
Age (years) 1.02 (1.01–1.03) <0.0001 1.02 (1.00–1.03) 0.029
Year of inclusion 0.89 (0.82–0.95) <0.001
Liver cirrhosis
 Yes 44 31 (70.5) 2.69 (1.37–5.31) 0.004 2.62 (1.24–5.54) 0.012
 No 379 178 (47.0) 1 1
Immunosuppression
 Yes 100 58 (58.0) 1.57 (1.00–2.47) 0.050 1.83 (1.08–3.11) 0.024
 No 323 151 (46.7) 1 1
PaO2/FiO2 ratio (mmHg) 0.99 (0.99–0.99) <0.0001 0.99 (0.99–0.99) <0.0001
SAPS II 1.03 (1.02–1.04) <0.0001 1.02 (1.00–1.03) 0.018
LODS 1.19 (1.13–1.25) <0.0001 1.12 (1.05–1.20) <0.001
Antifungal treatmenta
 Yes 17 12 (70.6) 2.55 (0.88–7.36) 0.084
 No 406 197 (48.5) 1
Blot et al. algorithm[16]
 No Aspergillus spp. colonization 388 188 (48.5) 1 1
 Aspergillus spp. colonization 18 6 (33.3) 0.53 (0.20–1.45) 0.22 0.64 (0.21–1.99) 0.44
 Putative or proven IPA 17 15 (88.2) 7.98 (1.80–35.36) 0.006 9.58 (1.97–46.52) 0.005

IPA invasive pulmonary aspergillosis

aAs prescribed for a suspicion of invasive pulmonary aspergillosis; the Hosmer–Lemeshow goodness of fit test showed good calibration of the model (p = 0.28); the area under the curve of the model is 0.78 (0.73–0.82); OR (95 % CI), odds ratio (95 % confidence interval); aOR, adjusted odds ratio