Skip to main content

Table 3 (A) The binary logistic regression analysis of the patient sample revealed that the infusion of colloids within the first 48 h after trauma is an independent predictor for the development of SIRS and sepsis. Hosmer–Lemeshow test, P < 0.001 for SIRS and P < 0.001 for sepsis. (B) ROC curve of the patient sample

From: Resuscitation with polymeric plasma substitutes is permissive for systemic inflammatory response syndrome and sepsis in multiply injured patients: a retrospective cohort study

Binary logistic regression Wald Odds P value
(A) The binary logistic regression analysis of the patient sample
SIRS (colloids) 174.229 3.325 <0.001
Sepsis (colloids) 108.989 8.984 <0.001
SIRS (EC) 39.242 1.955 <0.001
Sepsis (EC) 69.910 1.848 <0.001
SIRS (platelets) 6.303 0.972 0.012
Sepsis (platelets) 0.005 0.998 0.944
SIRS (FFP) 4.335 0.942 0.037
Sepsis (FFP) 10.447 1.217 0.001
ROC AUC P value
(B) Predictive quality depicted by AUC of the corresponding ROC
SIRS (colloids) 0.693 <0.001
Sepsis (colloids) 0.669 <0.001
SIRS (EC) 0.539 <0.001
Sepsis (EC) 0.821 <0.001
SIRS (platelets) 0.501 <0.001
Sepsis (platelets) 0.677 <0.001
SIRS (FFP) 0.512 <0.001
Sepsis (FFP) 0.807 <0.001
  1. AUC area under the curve, FFP fresh frozen plasma
  2. Significant if P < 0.05