Collection of blood through a central venous catheter for the diagnosis of bacteremia is a debated topic. Quantitative cultures of organisms from blood collected through central venous catheters were found to be highly sensitive, specific, and predictive of bacteremia, especially when a cutoff point of 15 colonies of skin organisms was used.Drawing of blood through a central venous catheter (CVC) for the diagnosis of bacteremia is highly debated (2,4,(11)(12)(13)15) due to the possibility of culturing blood contaminated by organisms adhering to CVC lumen. Although quantitative blood cultures (QBC) collected simultaneously through a CVC and peripheral venipuncture (PV) have been used for the diagnosis of catheter-related bloodstream infections (3,6,14), the usefulness of QBC collected through CVC for the diagnosis of bacteremia from any source has not been thoroughly investigated.Between September 1999 and May 2002, we followed up adult cancer patients who required new insertions of a peripherally inserted or subclavian silicone CVC at the M. D. Anderson Cancer Center in order to evaluate the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of QBC collected through CVC for the diagnosis of bacteremia. Maximal sterile barrier precautions were followed during catheter insertions (9). Patients were followed up until catheter removal or up to 100 days, whichever occurred first. At the onset of fever or the suspicion of catheter-related bloodstream infections, simultaneous QBC collected through CVC and PV were obtained. When the CVC remained in place beyond 100 days, blood cultures were performed, even in the absence of signs and symptoms of infection (SSI). Microorganisms were identified according to standard methods (8). An initial 10 ml of CVC-collected blood was discarded to avoid contact with previously infused drugs with antimicrobial activity. Subsequently, two 20-ml blood samples were drawn for QBC, one through the CVC and the other through PV, and a 10-ml portion of each was cultured aerobically (Bactec 9240, Bactec Plus Aerobic/F; BD Diagnostic Systems, Sparks, MD).At our institution, anaerobes constitute less than 0.1% of positive cultures; most of these are found to be possible contaminants. The remaining 10 ml was placed into isolator tubes (Isolator 10; Wampole, Cranbury, NJ) to be cultured quantitatively (lysis centrifugation method) (5). We defined bacteremia according to guidelines from the Centers for Disease Control and Prevention (CDC) (7). Possible contaminating microorganisms included skin microorganisms such as coagulase-negative staphylococci (CNS), diphtheroids, and Bacillus spp., whereas pathogenic microorganisms included Staphylococcus aureus, alpha-hemolytic Streptococcus spp., gram-negative bacilli, and Candida spp. A QBC collected through CVC was considered a true positive if (i) a pathogen was isolated in the presence of SSI such as fever, hypotension, and rigors or when the same pathogen was isolated from a peripheral blood culture within 48...