Sensitivity of peripheral blood smear review for the diagnosis of Candida fungemia.

Abstract

Case reports have described detection of candidemia by examination of peripheral blood smears. It is unclear whether this method has wider applicability for early detection of fungemia. To determine the sensitivity of smear review for detecting candidemia. Normal and cytopenic blood was spiked with increasing concentrations of yeast. Smears were prepared and reviewed by a pathologist and by technical staff. Staff members blinded to the purpose of the study first performed a routine slide review and then a targeted review for yeast. The pathologist detected isolated yeast forms at a concentration of 1 to 5 x 10(5) colony-forming units (CFU)/mL. When blinded to the purpose of the study, technical staff could detect Candida in most samples when the yeast concentration was 1 to 5 x 10(7) CFU/mL, but found it in only a small fraction of samples with lower concentrations. When asked to examine the smears specifically for yeast, they could detect it in most samples containing 1 to 5 x 10(6) CFU/mL. Detection of candidemia by peripheral blood smear examination requires a yeast concentration of 1 to 5 x 10(5) CFU/mL or greater. This degree of fungemia is unusual; therefore, detection of candidemia by blood smear review will not be possible in most cases. Sensitivity of smear review for yeast detection is greatly increased if the microscopist is specifically directed to look for the presence of yeast.