Differential quantitative blood cultures for the diagnosis of catheter-related bloodstream infections associated with short- and long-term catheters: a prospective study.

We prospectively evaluated the value of two levels of differential quantitative blood culture (DQBC) ratio (> or =2:1 or > or =5:1) in diagnosing catheter-related bloodstream infections (CRBSIs) in patients with malignancy that have short-term and long-term central venous catheters (CVCs) (<30 and > or =30 days of placement). Diagnosis of CRBSIs was based on results of semiquantitative cultures of removed catheters. For short-term CVCs a 5:1 or greater DQBC ratio had an 18% sensitivity, 67% specificity, 18% positive predictive value (PPV) and 67% negative predictive value (NPV), whereas a 2:1 or greater cut-off point for the DQBC was associated with a 45% sensitivity, 48% specificity, 26% PPV, and 68% NPV for the diagnosis of CRBSIs. For long-term CVCs, a cut-off point of 5:1 or greater of the DQBC had an 81% sensitivity, 40% specificity, 36% PPV, and 83% NPV for the diagnosis of CRBSIa, whereas a 2:1 or greater cut-off point had a 90% sensitivity, 30% specificity, 35% PPV, and 88% NPV for the diagnosis of CRBSIs. In conclusion, DQBCs are not diagnostic of CRBSIs for short-term CVCs. In long-term CVCs, DQBCs at 2:1 or greater or 5:1 or greater are sensitive but associated with low specificity and positive predictive value.