Candidemia, a leading cause of blood stream infections, causes significant mortality. The impact of fluconazole resistance on mortality remains poorly defined.
Active population-based surveillance for candidemia was conducted in Atlanta and Baltimore. Case-isolates were sent to CDC for speciation and susceptibility testing. Independent risk factors associated with fluconazole resistance and all-cause 2-30-day mortality were identified using logistic regression in patients with incident candidemia.
3,553 cases of candidemia were identified between 2008 and 2013, with an average yearly incidence of 18.9 cases/100,000. 21.1% died between 2-30 days. Overall, 6.9% of case-isolates were resistant to fluconazole, 26.6% of cases with fluconazole resistance died between 2-30 days. Black race (adjusted odds ratio: 1.5; 95% confidence interval: [1.0-2.1]), prior azole exposure (3.9; [2.7-5.8]), HIV infection (2.8; [1.7-4.8]), hematological malignancy (3.6; [2.0-6.4]), and solid organ transplant (3.0; [1.4-6.6]) were associated with fluconazole resistance. Chronic liver disease (0.4; [0.2-0.8]) was associated with decreased fluconazole resistance. Mortality risk factors included: age ≥65 years (1.7; [1.4-2.1]), malignancy (1.7; [1.3-2.2]), chronic liver disease (1.7; [1.3-2.4]), HIV infection (1.6; [1.1-2.5]), hospital onset infection (1.9; [1.4-2.5]), ICU admission before candidemia (3.6; [2.7-4.7]), treatment with amphotericin B (2.8; [1.7-4.5]), and infection with C. tropicalis(1.7; [1.2-2.4]). Any azole exposure (0.5; [0.4-0.6]) and the removal of a central venous catheter (CVC) (0.4; [0.3-0.5]) were associated with lower odds of mortality in this model. No association was found between fluconazole resistance and mortality.
Fluconazole resistance was increased in conditions associated with prior azole exposure. Chronic liver disease was associated with decreased resistance, possibly secondary to avoidance of azoles in this patient population. Following treatment guidelines, specifically removal of CVCs, may improve candidemia outcomes. However, further investigation is needed to evaluate impact of health status on the association between CVC removal and mortality.
A. Cleveland, None
R. Hollick, None
S. Lockhart, None
G. M. Lyon III, None
T. Chiller, None
L. Harrison, None
M. Farley, None