Methods

A levofloxacin prophylaxis practice guideline was implemented in March 2019 at Texas Children’s Hospital (Houston, TX, USA), which provides levofloxacin prophylaxis to patients with AML or r-ALL during periods of chemotherapy-induced severe neutropenia. Patients who received a hematopoietic stem cell transplant (HSCT) were excluded. ^{Reference Davis, Stevens and Brackett5} Systematic monitoring of adherence and outcomes was performed to confirm continued benefits and monitor for unintended consequences.

A retrospective, longitudinal cohort study was conducted for four years after implementation of the levofloxacin prophylaxis practice guideline. A cohort of patients from the early period of implementation (March 1, 2019–February 28, 2021) was compared to a cohort of patients from the subsequent two years of implementation (March 1, 2021–February 28, 2023) to ensure sustained benefits without an increase in infections due to resistant organisms or CDI. Patients with AML and r-ALL were eligible for the practice guideline and included in this study. Electronic health records were reviewed for underlying cancer diagnosis, HSCT status, and adherence to the levofloxacin practice guideline. Primary outcomes were NHSN CLABSI and any BSI event, including present on admission and secondary bacteremia. If patients had multiple BSIs, all events were included. Any BSI event was further categorized by etiology, gram-negative rod bacteria (GNR) versus viridans group streptococci or oral flora (VGS). Two positive cultures were required to be classified as a VGS BSI event. Secondary outcomes included culture from a sterile site that was positive for MDRO, bacteremia due to a levofloxacin non-susceptible GNR organism, diarrhea associated with stool positive for C. difficile toxin, death, and death due to bacterial infection. CDI was determined by C. difficile toxin PCR only if patients had new-onset diarrhea and greater than three unformed stools in a 24-hour period, without an alternative diagnosis or laxatives. Surveillance cultures for colonization with MDRO were not included. The patient clinical course was reviewed until completion of intensive chemotherapy, HSCT, death, transfer of care, or end of the study period. All patients eligible for levofloxacin prophylaxis were included in the analysis to capture the “real-world” impact and outcomes. ^{Reference Davis, Stevens and Brackett5}

Primary outcomes were analyzed by risk ratio (RR) of patients with NHSN CLABSI, any BSI, and any BSI with GNR or VGS, respectively. Secondary outcomes analyzed were RR of patients with BSI due to MDRO, BSI due to levofloxacin non-susceptible GNR organism, CDI, death, and death due to a bacterial infection. The Mann-Whitney U test was used to compare the days of severe neutropenia, levofloxacin prophylaxis, and cefepime therapy. Descriptive statistics were performed on demographics and underlying diagnosis using Fisher’s exact test. STATA 18 software (College Station, Texas) was used for statistical analyses.