Utility of Anaerobic and Fungal Blood Cultures in the Pediatric Oncologic Population
Madan Kumar, DO; Benjamin Hanisch MD
Background: In our institution, a febrile or ill appearing oncology patient will often be evaluated with aerobic, anaerobic, and fungal cultures. This is especially true in patients with persistent fevers without a clear etiology on empiric antimicrobial therapy. It is common for all three cultures to be repeated multiple times per admission.
Although this practice may seem sensible, there is to our knowledge little evidence to confirm its necessity in this population.
Methods: A record of all positive blood cultures originating from our institutions oncology ward was obtained from January 2010 to April 2017. Duplicate cultures (obtained on consecutive days with repeat organisms) were excluded. Each anaerobic and fungal culture was then evaluated for corollary positive aerobic cultures from the same time frame.
Results: A total of 10,950 blood cultures were evaluated for this study, including 2391 anaerobic cultures and 1980 fungal cultures. Forty-two unique anaerobic cultures (1.7%) were identified. The viridans group of Streptococcus was a large contributor with 9 unique cultures. Only 7 cultures of obligate anaerobes were observed: 4 cultures of Clostridial species, two Propionobacterium acnes, and one Peptostreptococcus species. 23 unique fungal cultures (1.2%) were identified. Notably most of these isolates (14) were identified as having one colony present and regarded as probable contaminants. Penicillium, Cladosporium, and unidentified dermatiaceous molds were present in greatest frequency.
Conclusion: Over a 7-year period of routinely obtaining anaerobic and fungal cultures for febrile oncology patients only 42 unique anaerobic and 23 unique fungal cultures were identified. Given the predominance of facultative anaerobes this may simply reflect the findings of increased blood sampling rather than added utility of the growth medium. Similarly, even among the limited unique fungal cultures the majority were of suspect validity given the presence of a single colony. These findings suggest judicious use of selective growth media in cases with higher clinical suspicion may be more useful than empiric evaluation.