objective. To assess the burden of bloodstream infections (BSIs) among pediatric hematology-oncology (PHO) inpatients, to propose a comprehensive, all-BSI tracking approach, and to discuss how such an approach helps better inform within-center and across-center differences in CLABSI rate.design. Prospective cohort study.setting. US multicenter, quality-improvement, BSI prevention network.participants. PHO centers across the United States who agreed to follow a standardized central-line-maintenance care bundle and track all BSI events and central-line days every month.methods. Infections were categorized as CLABSI (stratified by mucosal barrier injury-related, laboratory-confirmed BSI [MBI-LCBI] versus non-MBI-LCBI) and secondary BSI, using National Healthcare Safety Network (NHSN) definitions. Single positive blood cultures (SPBCs) with NHSN defined common commensals were also tracked.results. Between 2013 and 2015, 34 PHO centers reported 1,110 BSIs. Among them, 708 (63.8%) were CLABSIs, 170 (15.3%) were secondary BSIs, and 232 (20.9%) were SPBCs. Most SPBCs (75%) occurred in patients with profound neutropenia; 22% of SPBCs were viridans group streptococci. Among the CLABSIs, 51% were MBI-LCBI. Excluding SPBCs, CLABSI rates were higher (88% vs 77%) and secondary BSI rates were lower (12% vs 23%) after the NHSN updated the definition of secondary BSI (P < .001). Preliminary analyses showed across-center differences in CLABSI versus secondary BSI and between SPBC and CLABSI versus non-CLABSI rates.conclusions. Tracking all BSIs, not just CLABSIs in PHO patients, is a patient-centered, clinically relevant approach that could help better assess across-center and within-center differences in infection rates, including CLABSI. This approach enables informed decision making by healthcare providers, payors, and the public. 2017;38:690-696 Bloodstream infections (BSIs) cause morbidity and mortality and are especially concerning in pediatric hematology-oncology (PHO) patients. 1 Most oncology patients and some hematology patients have a long-term central line that facilitates their treatment, and these devices can be a source of centralline-associated BSI (CLABSI). Other sources of BSI in immunosuppressed PHO patients include disrupted skin or mucosal surfaces, such as the upper or lower gastrointestinal tract, and deep-seated infections. Frequently, an obvious source of BSI cannot be found. CLABSIs in inpatients are considered healthcare-associated infections that are preventable and must be reported 2 to payors and local, state, and federal agencies, with implications for reimbursement. 3,4 As a result, BSI surveillance in PHO inpatients has disproportionately focused on CLABSIs; the Centers for Disease Control and Prevention's National Healthcare Safety Network (NHSN) surveillance definition of CLABSI is widely used for this purpose. 5 A BSI from an alternate source of infection (other than the central line), defined by the NHSN as "secondary BSI," has not received similar scrutiny. Additionally, the NHSN definitio...