Diagnostic usefulness of differential time to positivity in neutropenic cancer patients with suspected catheter-related candidemia.

Methods for distinguishing catheter-related candidemia (CRC) from non-CRC before catheter removal remain limited. We thus evaluated the diagnostic performance of differential time to positivity (DTP) to diagnose CRC in neutropenic cancer patients with suspected CRC. Of the 35 patients enrolled, 15 (43%) with CRC (six definite and nine probable) and 17 (49%) with non-CRC were finally analyzed. Based on the receiver operating characteristic curve, the optimal cutoff value of DTP for diagnosing CRC was >= 1.45 hours with the sensitivity 80% (95% confidence interval [CI], 51-95) and specificity 100% (95% CI, 80-100), respectively.