Is Colony Count Important in the Managements of Patients with Candiduria?

Introduction: Candiduria is a common problem encountered in hospitalized patients, especially in intensive care units. Detection of Candida spp. in a urinary culture could be a sign of colonisation or severe infection. Due to the lack of well-defined, standard criteria it is difficult to differentiate between colonisation and infection. In this study, the role of the number of colonies was investigated to differentiate between colonisation and infection in hospitalized patients with established candiduria according to urinary culture. Materials and Methods: Patients who had urine culture positive for yeasts between December 2013 and March 2015 in Ege University Hospital (n= 217) were prospectively followed up. The data were evaluated observationally. Patients’ age, sex, diagnosis, clinical signs, urinary catheterization, hospital stay, underlying illness, hospitalization in intensive care, presence of candidemia, use of antibiotics and antifungals were evaluated. Culture results were reported as number of colonies/mL through quantitative cultures of all urine samples included in the study. Candiduria patients receiving antifungal treatment targeting candiduria were classified as infected. The number of colonies was used both as scale type and categorical data classified into two groups through three cut-points (103, 104, 105 and higher) in the analyses. The colony counts of patients receiving treatment and without treatment were compared using t-test and ChiSquare test. The relationship between colony counts and treatment was examined according to the presence of a urinary catheher, hospitalization in an intensive care unit and the microbial agents identified. Logistic regression was used in multivariate analyses. Mean values are shown with their standard deviations. Results: Among the patients, 134 were female and 83 male with a mean age of 64.03 ± 17.85 (17-109). The mean duration of hospital stay was 14 ± 17 days (1-123 day). Urinary catheter was present in 139 (64.1%) of the patients and their mean catheterization duration was 13.12 ± 7.81 days (1-36). The most frequent agent was Candida albicans (n= 98) followed by Candida glabrata (n= 76), Candida tropicalis (n= 21), Candida kefyr (n= 9). Vaginitis in 32 patients and candidemia in 6 patiens were detected by concurrent cultures. Number of colonies was detected as 105 and above in 107 of the patients (49.3%), below 103 in 43 (19.8%). Among the patients, 127 (60.5%) received treatment. No statistically significant difference was found among patients receiving treatment and patients without treatment accoring to the presence or absence of a catheter, Candida albicans versus non-albicans Candida as the agent, the cut-points of the number of colonies. Candiduria related treatment rates of hospitalized patients with catheter were significantly higher than catheterized patients in ICUs. Conclusion: Despite being the first prospective study in Turkey, It has been concluded that the presence of a catheter of the number of colonies according to the Candida species cannot be used as indicators of infection.