Peripheral vs. Central Blood Cultures in Patients Admitted to a Pediatric Oncology Ward

Blood cultures are essential tools for management of febrile children admitted to an oncology ward. There is little information on how often to draw blood cultures and from what source during the course of a hospitalization. In this study we reviewed overall blood culture utilization and the clinical utility of peripheral blood cultures when those were obtained at the same time as central venous catheter cultures in patients hospitalized for fever and neutropenia. Patients and methods. The microbiology laboratory records of febrile patients admitted to our pediatric oncology service were reviewed during a 9-month period. On admission all patients had routine peripheral blood cultures (PBC) and central venous catheter blood cultures drawn. Subsequent cultures depended on the clinical judgment of individual practitioners. Inpatient blood cultures taken up to 60 days after the day of admission were included in the analysis. A blood culture set was defined as blood culture samples taken at the same time from one site or more. For the purpose of this study, if only one culture was collected in a day it was also considered to be a set. A set was considered positive if one culture or more was positive. The clinical utility of PBCs was evaluated in culture sets where both PBCs and central venous catheter blood cultures were obtained at the same time. Results. Eighty-nine patients were admitted 128 times for fever, representing ∼10% of all days of admission on the pediatric oncology service. Their mean age was 7 years (range, 3 months to 23 years) and their oncologic diagnoses were leukemia 54%, solid tumors 33%, lymphomas 7% and myelodysplastic or histiocytic syndromes 6%. There were 681 blood cultures obtained, representing 5.3 per admission. Sites of culture were as follows: double lumen catheters with each lumen counted separately (34%); ports (20%); other or unspecified central venous catheters (4%); peripheral cultures (34%); and site not identified (8%). Overall 7.3% of blood cultures were positive. When blood cultures were grouped into 359 sets (mean, 1.9 cultures per set), 9.7% of these sets were considered positive. The number of sets (percent positive) by day of admission were: Day 0, 129 (10%); Days 1 to 20, 161 (8%); Days 21 to 40, 49 (14%); Days 41 to 60, 20 (10%). One hundred seventy-six sets (176) included central and peripheral cultures; of those 21 (12%) were positive (88 on admission, 12%). In 2 (1%) cases the peripheral culture added information to the culture set; both of these were from the time of admission (Table 1). One patient grew Pseudomonas aeruginosa from a peripheral site, the other grew Staphylococcus aureus from 2 peripheral sites. In the first case the central venous catheter culture was sterile; in the later case the central specimen grew a different organism (Klebsiella pneumoniae). The remaining 5 of 6 sets (83%) with positive peripheral cultures and negative central cultures yielded coagulase-negative staphylococci. Two (33%) of these sets grew multiple strains of coagulase-negative staphylococci, indicating probable skin contaminants. Of the 7 positive sets from double lumen catheters where both lumina were cultured, 3 (43%) had an organism identified from only 1 lumen cultured (2 Escherichia coli, 1 coagulase-negative Staphylococcus with a Gram-positive coccobacillus).TABLE 1: Central vs. peripheral positive blood culture results Discussion. We examined blood culture utilization on a pediatric ward during a 9-month period. The percentage of positive blood cultures varied little during the 60-day follow-up period. Thus there was no significant overutilization of blood cultures when the decision to draw blood cultures was left to clinical judgment. This survey also demonstrated the importance of taking a culture from all catheter lumina, because greater than one-third of the double lumen catheters that were positive were positive from only one of the two lumina. On the other hand the utility of peripheral cultures was less apparent. Current guidelines recommend that peripheral cultures be drawn in parallel with central venous catheter cultures in patients with fever and neutropenia.1 This is to help distinguish between a central venous catheter infection and a bacteremia from another source. Peripheral and central cultures may both be positive as a result of significant catheter infection. In the absence of another source such culture results are classified as probable catheter-related sepsis.2 Definitive catheter-related sepsis is defined as: exit site infection with pus; refractory sepsis responsive only to catheter removal; quantitative catheter tip culture with the same organism recovered from the blood stream; or differential quantitative culture with a >10-fold colony count from the central venous catheter culture as compared with the peripheral blood culture.2 Differential quantitative culture methods do not require catheter removal but are costly, are not readily available and may lack sensitivity.3 A recently published method comparing the time of positivity of peripheral and central venous catheter blood cultures appears simpler in concept than quantitative culture technique but requires further validation.4 The utility of peripheral cultures done in parallel with central cultures without bacterial quantification is less clear. McCoig et al.5 reviewed 66 admissions of children with fever and neutropenia; 12 had positive cultures from central venous catheters only, 3 from central and peripheral sites and none from peripheral sites only. They concluded that peripheral cultures were not helpful in the management of these children. Lecciones et al.6 reviewed 155 episodes of catheter-associated fungemias; culture sets with positive central cultures and negative peripheral cultures had the same rate of dissemination and mortality as sets where both peripheral and central cultures were positive. In our study the peripheral sample added clinical information in 2 patients with long term central venous catheters (1% of blood culture sets). Recently Kaditis et al.7 demonstrated that when a standardized appropriate volume of blood was drawn from children with fever and neutropenia, the recovery rate of microorganisms in the blood increased significantly. Benefits from excluding routine peripheral cultures in children in an oncology ward with central venous catheters and fever seem to outweigh their continued routine use. Advantages of not obtaining routine PBC include reduction of the recovery of potentially confusing bacterial skin contaminants that can contribute to unnecessary treatment; avoidance of phlebotomy-associated complications such as pain, infection or phlebitis; and reduction of cost (by 13% in this study). In addition nonquantitative peripheral cultures are seldom useful in identifying infected catheters that should be removed. Even if peripheral cultures did help to identify infected catheters, current recommendations do not include catheter removal in many instances of sepsis1, 2; thus treatment would not be modified. Routine use of peripheral blood cultures in patients with fever and central venous catheters admitted to a pediatric oncology ward is not supported by this study. Acknowledgments. We thank Robert Keleher and Gina Lee for aiding in the data extraction and Patrick Daniel for support. Thomas V. Adamkiewicz, M.D., F.R.C.P.(C) Adonis Lorenzana, M.D. John Doyle, M.D., F.R.C.P.(C) Suzan Richardson, M.D., F.R.C.P.(C) Hospital for Sick Children; Toronto, Canada