Methods: Patients with VGS bacteremia between 1/2009 and 12/2014 were identified using microbiology records. Twelve whose culture results had been deemed to be contaminants were excluded. NP and non-NP patients at the time of positive culture were matched 1:1 to controls on the basis of ANC ≤ or >1000/mm3, respectively, hospital ward, and length of stay. We extracted demographic data, ANC, medications, chemotherapy, radiation therapy, mucositis and other oral inflammatory conditions, and GI conditions (e.g., C. difficile infection, graft-versus-host disease, typhlitis). Data were analyzed using McNemar’s test, Wilcoxon signed-rank test, and conditional logistic regression modeling.
Results: Among 101 patients, 63 were NP and 38 non-NP at the time of VGS bacteremia. No clinical factors were associated with VGS in non-NP. Among NP, however, VGS bacteremia was associated with lower ANC (p<0.0001), radiation therapy (p=0.0164), cyclophosphamide (p=0.0067) fludarabine (p=0.0143), mucositis (p=-0.039) and other oral inflammatory conditions (p=0.0105). In multivariable analysis of NP, only lower ANC predicted status as a case (OR=0.006, 95% CI: 0.001-0.134; p=0.0012).
Conclusion: The well-described risk factors for VGS bacteremia in NP did not predict infection in non-NP, possibly due to sample size or to the heterogeneity of the latter group. NP who had lower nadirs were more vulnerable to VGS bacteremia. Other associations in NP (chemotherapy, radiation, oral conditions) are all related to neutropenia. Further analyses will explore antibiotic exposure. Identification of modifiable risk factors may enable prevention among patients with adequate granulocytes.
A. Dulanto Chiang,
C. D. Spalding, None
D. K. Henderson, None
T. N. Palmore, None