BACKGROUND/AIMS: Can repeated blood PCR assays in critically ill patients with culture-confirmed candiduria help in the early detection of candidemia? METHODOLOGY: Urine samples were obtained on the day of admission and if negative were repeated on the 5th ICU day. Upon laboratory confirmation of candiduria, preemptive antifungal therapy was commenced. Whole blood samples were tested simultaneously with conventional cultures and PCR repeated twice weekly for two weeks. RESULTS: In ICU patients with candiduria, preemptive antifungal therapy resulted in candiduria clearance and clinical improvement in 94% of them. Candida DNAemia (detected in 64 out of 136 repeated PCR assays) persisted for 1 or 2 weeks in a significant number of patients. Even after multiple positive PCR assays from each individual candiduric patient, the specificity and positive predictive value of the assay in the early detection of candidemia were low (negative predictive value was 100%). No differences were detected on demographics among patients with or without candiduria. The correlations (PCR with culture) with kappa test presented a poor agreement (K = 0.103, Kmin= -0679, Kmax = 0.103). CONCLUSIONS: PCR in the blood should not be used in early diagnosis of (but it could help in excluding) candidemia, in ICU patients with candiduria.
BACKGROUND/AIMS: Can repeated blood PCR assays in critically illpatients with culture-confirmed candiduria help in the early detection of candidemia? METHODOLOGY: Urine samples were obtained on the day of admission and if negative were repeated on the 5th ICU day. Upon laboratory confirmation of candiduria, preemptive antifungal therapy was commenced. Whole blood samples were tested simultaneously with conventional cultures and PCR repeated twice weekly for two weeks. RESULTS: In ICU patients with candiduria, preemptive antifungal therapy resulted in candiduria clearance and clinical improvement in 94% of them. Candida DNAemia (detected in 64 out of 136 repeated PCR assays) persisted for 1 or 2 weeks in a significant number of patients. Even after multiple positive PCR assays from each individual candiduric patient, the specificity and positive predictive value of the assay in the early detection of candidemia were low (negative predictive value was 100%). No differences were detected on demographics among patients with or without candiduria. The correlations (PCR with culture) with kappa test presented a poor agreement (K = 0.103, Kmin= -0679, Kmax = 0.103). CONCLUSIONS: PCR in the blood should not be used in early diagnosis of (but it could help in excluding) candidemia, in ICU patients with candiduria.