Utility of Routine Blood Cultures for Inpatient Hematology/Oncology Patients Receiving Antimicrobials

Abstract

BackgroundBlood cultures are drawn regularly on hematology/oncology patients due to the concern for infectious complications. However, this practice is costly in many ways, and there are limited data to support this practice. We aimed to investigate the frequency with which blood cultures are drawn in these patients, and moreover, how frequently they provide meaningful data. Materials and MethodsWe performed a single-center retrospective review study in hematology/oncology patients admitted to our hospital. We reviewed 1,437 blood cultures from 220 unique patients between July 1, 2013 and June 30, 2014. We reviewed the proportion and characteristics of the positive blood cultures, as well as organisms isolated. ResultsOf all 1,437 blood cultures drawn during the study period, 111 (8%) of the blood cultures grew a clinically meaningful organism. Gram-positive organisms were more likely than gram-negative organisms, with coagulase-negative Staphylococcus and vancomycin-resistant Enterococcus being most common. In patients who were receiving broad-spectrum antimicrobials, only 13 (4%) of 358 blood cultures collected grew a clinically meaningful organism, all of which were performed for clearance. None of these 13 positive blood cultures represented a new infectious organism. No patient receiving broad-spectrum antibiotics with blood cultures drawn in response to a fever had a positive culture. ConclusionsIn hospitalized cancer patients receiving broad-spectrum antimicrobials, the likelihood of growing a new clinically significant organism from a blood culture was extremely low. We speculate that the practice of repeating blood cultures in hospitalized cancer patients on broad-spectrum antimicrobials, particularly in response to fever alone, should be eliminated.