Changes in Central Line–Associated Bloodstream Infection (CLABSI) Rates Following Implementation of Levofloxacin Prophylaxis for Children and Adolescents With High-Risk Leukemia

Abstract

Background: Despite initiatives to reduce central line–associated bloodstream infection (CLABSI), children and adolescents with hematologic malignancies, as well as those with relapsed disease, remain at the greatest risk for infection. This single-institution project evaluated changes in CLABSI rates following implementation of antibacterial prophylaxis with levofloxacin for patients with high-risk hematologic malignancies. Methods: Positive blood culture events meeting National Health Safety Network surveillance criteria to be classified as CLABSIs from January 1, 2006, to December 31, 2019, were included. Data were organized into four time periods for comparison based on implementation of CLABSI-reduction interventions. Conditional Poisson regression models were used to evaluate the effect of time (intervention period) on CLABSI rates with post hoc Tukey pairwise comparisons between each of the four time periods. Results: From 2006 and 2019, 227 patients experienced 310 CLABSIs. Clinically important decreases in CLABSI rates from baseline (4.84 per 1,000 line days) occurred with implementation of Children's Hospital Association (CHA) bundles (3.29 per 1,000 line days); however, this difference was not significant ( p  =  .16). CLABSI rates decreased from baseline with the addition of formalized supportive cares (2.66 per 1,000 line days; incidence rate ratio [IRR]  =  0.60; p < .01), and with the use of antibacterial prophylaxis (1.66 per 1,000 line days; IRR  =  0.35; p < .01). Post hoc comparisons indicated decreased CLABSI rates with the use of antibacterial prophylaxis compared with CHA bundles alone (IRR  =  0.49; p  =  .011) and CHA bundles plus formalized supportive cares (IRR  =  0.58; p  =  .046). Discussion: Results demonstrate sustained success using a practice-based evidence approach to guide CLABSI-reduction interventions. Follow-up research, applying machine learning algorithms, may identify additional risk factors and inform future interventions.