ABSTRACT
Candida bloodstream infection (CBI) is associated with high mortality. The aim of this study was to compare the utility of the combined use of the Pitt Bacteremia Score (PBS) and Charlson Comorbidity Index (CCI) or Chronic Disease Score (CDS) to predict mortality among patients with CBI. Thereby, all consecutive patients with CBI at our institution between 2010 and 2014 were included. The PBS was used to evaluate CBI severity and the CCI and CDS were used to assess comorbidities of patients with CBI. Logistic regression analysis was used to estimate odds ratios for 30-day mortality in models including the PBS and CCI or CDS. A total of 189 CBI episodes were identified. Logistic regression models including the PBS and either CCI or CDS showed that the combined use of a comorbidity score and a severity score significantly predicted 30-day mortality. The performance of the different models was similar. Aggregated scores of comorbidity (CCI and CDS) and disease severity (PBS) are useful for the prediction of 30-day mortality risk in patients with CBI. Their use may facilitate the analysis of risk factors for poorer outcome and the development of an index for CBI mortality.
