Fattori di rischio ed outcome delle sepsi da C. parapsilosis

Candidemia is a major cause of morbidity and mortality in hospitalized patients. Over the years an increase detection of non-albicans species, in particular by C. parapsilosis. Aim of this study was to evaluate Candida parapsilosis bloodstream infection incidence, its predictors and the 30 days overall mortality. During the study period (2008-2016)1240 episodes were identified, 528 in the retrospective (2008-2011) and 712 in the prospective. An increased incidence of cases of both C. albicans (1.02/10000 hospital days - p = 0.004) C. parapsilosis (1.04/10000 hospital days - p < 0.001). Complete clinical data were avaible for 694 episodes of the prospective phase, cases were mainly in medical ward. C. parapsilosis increasing at the same time with the activation of a hospital service for PICC positioning, this led us to presume a potential correlation between C. parapsilosis and the device. This hypothesis wasn’t confirmed by the univariable analysis, while, at the multivariable, time to infection (p = 0.004) and previous treatment with echinocandins (p < 0.0001) results as predictors of the infection. 30 days overall mortality was 35.7%. From a microbiological point of view every C. parapsilosis strain resulted sensu strictu and a certain homogeneity was found in the strains tested for possible clusters between 2014 and 2016. Moreover, a high prevalence of resistance to fluconazole was found in this species (32.4%). In conclusion in our center we found an increasing incidence of C. parapsilosis probably due to a widespread use of echinocandins and to the spread of a clone, although the data in this regard are limited. The fact that patients with candidemia, regardless of the species, have very similar characteristics did not allow us to find other predictors. Further studies and preventive measures are needed in order to better understand and control this clinical problem.