In multivariate analysis, absence of underlying cirrhosis, α-FP level and Okuda stage were revealed as independent prognostic factors affecting survival (Table 5).
By multivariate analysis controlling for underlying severity of illness as measured by APACHE II score, those who received empiric therapy had a 1.6 fold higher risk of death compared to directed therapy, although not statistically significant (p = 0.45) (Table 5).
The individual risk factors for isolation of MDR A. baumannii that were identified by the multivariate analysis were male sex, underlying comorbidity of ischemic heart disease, mechanical ventilation, and antimicrobial drug treatment.
Before the analyses, we examined age, current alcohol use, current smoking, ES, and the lead exposure metrics using univariate descriptive statistics to check for accuracy of data entry, missing values, and assumptions underlying multivariate analysis.
34 In the French Registry, male gender is also associated with a worse prognosis in patients with idiopathic PAH in univariate and multivariate analysis, although the explanation underlying this observation is not known.
Descriptive statistics and the multivariate analysis have been used to explore underlying structures of the data collected.
In the multivariate analysis we identified as independent risk factors: underlying medical disease (2,07(1,17-3,66) 0.01), immunosuppression (3,22(1,79-5,79) < 0.001) and superficial surgical wound infection (6,65(1,70.006 0.006).
After controlling for confounding variables such as baseline Candida species, APACHE II score, underlying disease, and so on, multivariate analysis revealed that neither CVC24 nor CVC48 were statistically significant prognostic indicators for 28-day and 42-day survival.
We included the Pitt bacteremia score instead of shock and, the Charlson's weighted index of co-morbidity instead of underlying diseases to avoid data overlap in the multivariate analysis.
