[Retrospective analysis of fluconazole efficacy in Candida-colonized, non-neutropenic, surgical patients in long-term intensive care]

Abstract

Background and objective: The early clinical diagnosis of invasive candidiasis is difficult. Fluconazole, which has been available since the early 1990s, is a relatively atoxic intravenously applicable antimycotic agent. For this reason it has been widely used - possibly too much. The aim of this study was the retrospective critical evaluation of the efficacy of systemic antifungal chemotherapy in non-neutropenic, Candida-colonized, surgical patients in long-term intensive care.

Patients and methods: 69 patients (54 men and 15 women, aged 55.8 [range 18-87] years) of 364 patients of the anaesthesiological intensive care unit (ICU) of the University Hospital of Heidelberg in 1991 and 1992 were selected for the study. None of the 69 patients was suffering from proven invasive candidiasis according to the gold-standard criteria of positive histology, blood culture, or isolation from a sterile compartment. However, 35 of the 69 patients were systemically treated with fluconazole (on average 295 mg per day for 10.2 days intravenously). 34 patients did not receive any antifungal therapy. Retrospectively we analysed the course of the disease in both groups of patients. Furthermore, 173 serum samples of these patients were available for investigations by Western blot for anti-Candida antibodies of the immune globulin classes M and G.

Results: Both groups, antimycotically treated and untreated patients, had similar characteristics at base-line: age, sex, underlying disease, severity of the disease (APACHE II Score), and also mortality (approximately 20 % in both groups). Only times in the ICU and on mechanical ventilation were significantly enhanced in fluconazole treated patients (p values 0.0004 each). Before therapy, the fluconazole patients had significantly more often yeasts in primarily non-sterile compartments (chi (2) test 0.05). The yeasts were partly eradicated by fluconazole (32/54, 59.3 %). Anti-Candida antibodies significantly correlated with higher age (anti 47 kDa antigen, p = 0.02), but not with other, clinically, diagnostically or prognostically relevant parameters.

Conclusion: Fluconazole in non-neutropenic, Candida-colonized, surgical patients in long-term ICU care neither improved the clinical course nor the mortality rate among these patients. These observations indicate that there was a trend of overestimating the clinical significance of Candida in this group of patients. Fluconazole therapy may be significantly reduced in such patients.