Abstract:
Objective: To evaluate the comparative epidemiology of hospital-onset bloodstream infection (HOBSI) and central line-associated bloodstream infection (CLABSI).
Design and setting: Retrospective observational study of HOBSI and CLABSI across a three-hospital healthcare system from 01/01/2017 to 12/31/2021.
Methods: HOBSIs were identified as any non-commensal positive blood culture event on or after hospital day 3. CLABSIs were identified based on National Healthcare Safety Network (NHSN) criteria. We performed a time-series analysis to assess comparative temporal trends among HOBSI and CLABSI incidence. Using univariable and multivariable regression analyses, we compared demographics, risk factors, and outcomes between non-CLABSI HOBSI and CLABSI, as HOBSI and CLABSI are not exclusive entities.
Results: HOBSI incidence increased over the study period (IRR 1.006 HOBSI/1,000 patient days; 95% CI 1.001-1.012; P = .03), while no change in CLABSI incidence was observed (IRR .997 CLABSIs/1,000 central line days, 95% CI .992-1.002, P = .22). Differing demographic, microbiologic, and risk factor profiles were observed between CLABSIs and non-CLABSI HOBSIs. Multivariable analysis found lower odds of mortality among patients with CLABSIs when adjusted for covariates that approximate severity of illness (OR .27; 95% CI .11-.64; P < .01).
Conclusions: HOBSI incidence increased over the study period without a concurrent increase in CLABSI in our study population. Furthermore, risk factor and outcome profiles varied between CLABSI and non-CLABSI HOBSI, which suggest that these metrics differ in important ways worth considering if HOBSI is adopted as a quality metric.
Reference:Krishnan J, Gettler EB, Campbell M, Kalu IC, Seidelman J, Smith B, Lewis S. Comparative epidemiology of hospital-onset bloodstream infections (HOBSIs) and central line-associated bloodstream infections (CLABSIs) across a three-hospital health system. Infect Control Hosp Epidemiol. 2024 Mar 20:1-7. doi: 10.1017/ice.2024.38. Epub ahead of print. PMID: 38505940.