Abstract: This study  aimed to evaluate risk factors for 30-day mortality among hospitalised  patients with Pseudomonas aeruginosa bacteraemia, a highly fatal  condition. A retrospective study was conducted between 1 January 2009  and 31 October 2015 in 25 centres (9 countries) including 2396 patients.  Univariable and multivariable analyses of risk factors were conducted  for the entire cohort and for patients surviving ≥48 h. A propensity  score for predictors of appropriate empirical therapy was introduced  into the analysis. Of the 2396 patients, 636 (26.5%) died within 30  days. Significant predictors (odds ratio and 95% confidence interval) of  mortality in the multivariable analysis included patient-related  factors: age (1.02, 1.01-1.03); female sex (1.34, 1.03-1.77); bedridden  functional capacity (1.99, 1.24-3.21); recent hospitalisation (1.43,  1.07-1.92); concomitant corticosteroids (1.33, 1.02-1.73); and Charlson  comorbidity index (1.05, 1.01-1.93). Infection-related factors were  multidrug-resistant Pseudomonas (1.52, 1.15-2.1), non-urinary source  (2.44, 1.54-3.85) and Sequential Organ Failure Assessment (SOFA) score  (1.27, 1.18-1.36). Inappropriate empirical therapy was not associated  with increased mortality (0.81, 0.49-1.33). Among 2135 patients  surviving ≥48 h, hospital-acquired infection (1.59, 1.21-2.09), baseline  endotracheal tube (1.63, 1.13-2.36) and ICU admission (1.53, 1.02-2.28)  were additional risk factors. Risk factors for mortality among patients  with P. aeruginosa were mostly irreversible. Early appropriate  empirical therapy was not associated with reduced mortality. Further  research should be conducted to explore subgroups that may not benefit  from broad-spectrum antipseudomonal empirical therapy. Efforts should  focus on prevention of infection, mainly hospital-acquired infection and  multidrug-resistant pseudomonal infection.

Keywords: Bacteraemia; Mortality; Pseudomonas; Risk factors.