Premium
Outcome and potentially modifiable risk factors for candidemia in critically ill burns patients: A matched cohort study
Author(s) -
Dudoig Emmanuel,
Alanio Alexandre,
Anstey James,
Depret François,
Coutrot Maxime,
Fratani Alexandre,
Jully Marion,
Cupaciu Alexandru,
Chaussard Maïté,
Oueslati Haikel,
Ferry Axelle,
Benyamina Mourad,
Tymowski Christian,
Boccara David,
Serror Kevin,
Chaouat Marc,
Mimoun Maurice,
Lafaurie Matthieu,
Denis Blandine,
GitsMuselli Maud,
Bretagne Stephane,
Mebazaa Alexandre,
Legrand Matthieu,
Soussi Sabri
Publication year - 2019
Publication title -
mycoses
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.13
H-Index - 69
eISSN - 1439-0507
pISSN - 0933-7407
DOI - 10.1111/myc.12872
Subject(s) - medicine , retrospective cohort study , septic shock , sepsis , univariate analysis , fluconazole , cohort study , cohort , colonisation , intensive care medicine , antifungal , multivariate analysis , dermatology , colonization , microbiology and biotechnology , biology
Summary Background Patients with extensive burns are at risk of developing candidemia. Objectives To identify potentially modifiable risk factors and outcomes of candidemia in critically ill burns patients. Patients and Methods Retrospective matched cohort study including adult burns patients. Patients who developed candidemia were matched with burns patients with Candida spp colonisation and sepsis or septic shock without candidemia in a ratio of 1:3 (same severity scores and colonisation index). Univariate and multiple regression analyses were performed. Results Of 130 severely burned patients with Candida spp colonisation and at least one episode of sepsis or septic shock, 14 were diagnosed with candidemia. In the candidemia group, patients had a median ( IQR ) total burns surface area ( TBSA ) of 57 (38‐68)%, SAPSII of 43 (36‐58) and ABSI of 11 (8‐13). Multiple regression analysis showed that only duration of prior antibiotic therapy was independently associated with candidemia. ICU mortality was higher in the candidemia group (71% vs 35% [ P = 0.02]). The log‐rank test for 28‐day mortality comparing patients with candidemia treated with an empirical strategy vs a curative strategy did not reach significance ( P = 0.056). Conclusions Burns patients having received recent antibiotherapy have a higher risk of candidemia. Antifungal strategies did not influence outcome in this series.