Elena Magnani (Italy)
"Bufalini" Hospital Internal MedicineAuthor of 1 Presentation
PREDICTORS FOR CANDIDEMIA IN INTERNAL MEDICINE: A NEED FAR TO BE SATISFIED
Abstract
Background and Aims
Candidemia is a challenging clinical condition burdened by relevant mortality and morbidity. Identification of patients with high suspicion of candidemia might lead to more prompt diagnosis and therefore better outcome. Aim of this analysis is to evaluate the predictive value of some risk assessment models (RAMs) for candidemia recently developed for the specific setting of Internal Medicine (IM), referring to the data collected through a registry promoted by the Italian Scientific Society of IM FADOI.
Methods
The characteristics of patients enrolled in a national registry promoted by FADOI were matched with the RAMs proposed by Falcone et al. (Eur J Intern Med 2017; 41: 33-38), Sozio et al. (Infection 2018; 46: 625-633) and Atamna et al. (Diagn Microbiol Infect Dis 2019; 95: 80-83).
Results
In the registry promoted by FADOI a total of 111 patients with candidemia confirmed by blood culture were enrolled. By analyzing the clinical characteristics of these patients, 29.7%, 37.8% and 1.8% of them would have been identified as at high risk of candidemia by applying the RAMs of Falcone et al., Sozio et al. and Atamna et al., respectively.
Conclusions
In the specific setting of IM, the identification of an effective predictive tool for early recognition of patients at high risk of candidemia remains an unmet issue. Timely selection of patients at high risk of candidemia probably needs research that uses larger cohorts of derivation and validation than those studied so far, or bundle strategies that integrate clinical variables and rapid diagnostic tests.