661. Futility of Bacterial Bone Marrow Cultures: Experience over a 19 Year Period

Background Bone marrow biopsies are often performed on patients with unclear diagnoses and cultures may be ordered for both routine bacterial, mycobacterial and fungal pathogens. They are performed in semi-sterile conditions and involve needle penetration through the skin, posing an increased risk of skin contamination. These cultures also require a substantial amount of laboratory personnel time. Methods Cultures collected from 2001-2020 were surveyed in the lab electronic record. We assessed the culture type (fungal, bacterial, mycobacterial), and the presence of pathogens and contaminants. An organism was deemed a contaminant if it was consistent with skin flora or listed as a contaminant in the report given to the physician. Organisms for which the role in bone marrow disease is unclear were included as possible pathogens. For questionable non-contaminant organisms, clinical significance was determined based on if patient was treated for the organism. For all bone marrow cultures, growth of the same organism within 1 month of the bone marrow specimen was surveyed to determine whether the organism would have been found by alternative methods. Results Of 483 bacterial bone marrow cultures, there were 110 (23%) positives, of which 76 (69%) were deemed contaminants. Twenty (18%) of the 76 contaminants grew in the routine bacterial culture. However, 49 (65%) contaminants grew in the AFB culture, of which 10 also grew in the bacterial culture. For the 34 non-contaminant organisms, 26 were determined to be clinically significant. Nineteen of the 26 had a matching culture (usually blood) growing the organism within 1 month. The majority of pathogens were mycobacteria (18 of the 34). Fungal organisms represented 5 cultures and 11 were bacterial. Of the 11 bacterial organisms, 1 was a Helicobacter species (grown in special media), and 4 had a matching positive blood culture. Only 4 (1% of 483) bacterial non-contaminants grew in the routine bacterial culture. Given an unknown number of true negatives, we can only conclude a positive predictive value (PPV) of 0.16 for routine bacterial cultures. Including AFB and fungal cultures, the PPV increased to 0.30. Conclusion Our findings indicate that routine bacterial bone marrow culture is unlikely to yield a novel result and is likely a poor use of lab resources. Disclosures All Authors: No reported disclosures