Methods: We conducted a retrospective study of 1131 patients at the University of Michigan Health System who were hospitalized for >72 hours and had diarrhea with stool testing for presence of toxigenic C. difficile. Potential BSIs were reviewed if they occurred after onset of CDI; standard criteria for our primary outcome of a true BSI vs. a contaminant were applied. Secondary outcomes included mortality, colectomy, readmission, and ICU admission within 30-days of CDI. Unadjusted and adjusted statistical models were employed.
Results: BSI occurred in 86 (7.6%) of 1131 patients. Enterococcus (16.3%) and Klebsiella (16.3%) species were the most common organisms, followed by E. coli (12.8%) and S. aureus (12.8%). Age and race were similar between groups. Patients with BSI were more likely to be male (OR [odds ratio] 1.74, 95% CI [confidence interval] 1.11-1.2.72, P =.016), on immunosuppression (OR 1.94, 95% CI 1.22-3.07, P =.005), have a higher Charlson weighted comorbidity score (OR 1.12, 95% CI 1.04-1.21, P =.004) and met modified systemic inflammatory response syndrome (SIRS) criteria (OR 2.03, 95% CI 1.17-3.53, P =.012). Of the patients with BSI, 36 (41.9%) had CDI and 50 (58.1%) did not have CDI. However, CDI was not associated with BSI (OR 0.69; 95% CI 0.44-1.08; P =.103). This lack of association remained true even after adjusting for age, sex, immunosuppression, weighted Charlson comorbidity score, and modified SIRS criteria (OR 0.64, 95% CI 0.38-1.08, P=.092).
Conclusion: In this cohort of hospitalized adults that underwent testing for CDI, presence of CDI was not associated with developing BSI.
R. J. Ulrich,
K. Santhosh, None
J. Mogle, None
V. B. Young, None
See more of: Poster Abstract Session