Candida albicans and non-albicans bloodstream infection in adult intensive care units: A comparison of risk factors and mortality

Candidemia is an important bloodstream infection which is seen in intensive care unit with high mortality rate. Early diagnosis and treatment is difficult. For this reason, it is very important to identify patients who are at high risk so that preemptive and/or prophylactic treatment can be initiated. The species and incidence of candidemia may vary in years from country to country, even within the same country between hospitals. While the most common strain is C. albicans, in recent years the incidence of candidemia due to non-albicans species has been on the rise. This study was done retrospectively in the patients with candidemia hospitalized in the adult intensive care units of Baskent University Faculty of Medicine Ankara Hospital between 1 January 2011-31 August 2015 dates. Epidemiological and clinical data of patients enrolled in the study were noted retrospectively on structured form. Statistical analysis was done using T-test, Mann Whitney U test, Chi-square, and Fisher's Exact analysis with p<0.05 being accepted as significant. There were a total of 106 patients with candidemia enrolled to the study: 58 (55%) C. albicans, 48 (45%) non-albicans Candida species were isolated. Non-albicans species were isolated as C. glabrata 19% (n = 20), C. tropicalis 11% (n = 12), C. parapsilosis 7% (n = 7), Candida spp 4% (n = 4), C. lusitaniae 2% (n = 2). One case of each C. famata, C. guilliermondii, and C. dubliniensis were also isolated. While previous use of azoles was found to be a risk factor for development of candidemia due to non-albicans species, previous broad spectrum antibiotic use had no affect on the risk of both non-albicans and C. albicans candidemia development. In this study, of the patients who developed either C. albicans or non-albicans candidemia 99% had previous history of antibiotic use, 87% had central venous lines placed, 92% had urinary catheters, 68% were on total parental nutrition, 44% were on mechanical ventilation, and 23% had undergone abdominal surgery. However, none of these parameters were found to be a greater risk factor in candidemia due to non-albicans species and C. albicans. When patients with candidemia were subjected to evaluation in terms of the underlying disease (diabetes mellitus, chronic obstructive pulmonary disease, hematologic malignancy, solid organ malignancy, solid organ transplant, notropenia, renal failure, hemodialysis) no significant difference in risk was found between candidemia due to non-albicans species and C. albicans development. Because of candidemia detection after exitus in some patients antifungal susceptibility tests have not been studied at all. Antifungal susceptibility tests were studied in only 82 (77%) of 106 strains of candidemia. Flucytosine, fluconazole, voriconazole and itraconazole were used in all of the strains of antifungal susceptibility test; caspofungin and amphotericin B were used in the certain Candida strains. Because of three different antifungal susceptibility methods used during the study period statistical comparisons could not be made. Antifungal susceptibility tests were studied in 48 of the fifty-eight C. albicans strains (83%), in 12 of 20 C. glabrata strains (60%), in seven of 12 C. tropicalis strains (58%), in six of the seven C. parapsilosis strains (86%). In our study we also isolated Candida spp (n = 4), C. lusitaniae (n = 2) and one strain is isolated each from C. famata, C. guilliermondii and C. dubliniensis; because of small numbers, susceptibility of these strains has not been evaluated statistically. 15 of totally 106 strains of Candida, caspofungin susceptibility were studied; 39 of totally 106 strains of Candida, amphotericin B susceptibility were studied; and found to be susceptible at all. Fluconazole, voriconazole, itraconazole, caspofungin and amphotericin B were found susceptible in all strains of C. albicans and C. parapsilosis with antifungal susceptibility tests. In C. glabrata; susceptible 42%, dose-dependent susceptible 33%, resistant 25% to fluconazole; susceptible 50%, 25% dose-dependent susceptible, resistant 25% to voriconazole; 50% susceptible, 25% were found dose-dependent susceptible to itraconazole. In C. tropicalis; susceptible 100% to voriconazole, susceptible 86% to fluconazole and 14% dose-dependent susceptible to fluconazole; susceptible 86% to itraconazole, 14% were found to be dose-dependent to itraconazole. Mortality occurred in 69% of the patients with candidemia in our intensive care units. Mortality rates were 53% in non-albicans and 47% in C. albicans patients with candidemia. Mortality was seen more in non-albicans species. However, this difference was not found statistically significant. It is of utmost importance to keep in mind not only that patients who are in the intensive care unit for long periods of time, receiving broad spectrum antibiotic therapy, and undergoing invasive therapeutic procedures are susceptible to development of candidemia, but also to acknowledge the epidemiologic data specific to every intensive care unit so that early diagnosis and antifungal treatment can be initiated and ultimately reduce the morbidity and mortality of candidemia. Key words: Intensive care unit, candidemia, Candida species, risk factors.