Center-level variability in broad-spectrum antibiotic prescribing for children undergoing hematopoietic cell transplantion for acute leukemia.

• Authors: Caitlin Elgarten, Children’s Hospital of Philadelphia; Staci Arnold, Emory University; Yimei Li, Children’s Hospital of Philadelphia; Y. Vera Huang, Children’s Hospital of Philadelphia; Jeffrey S. Gerber, Children’s Hospital of Philadelphia; Wael Saber, Center for International Blood and Marrow Transplant Research; Richard Aplenc, Children’s Hospital of Philadelphia; Brian T. Fisher, Children’s Hospital of Philadelphia
• Journal Title: Open Forum Infectious Diseases
• Volume: Volume 4, Number suppl_1
• Publisher: Oxford University Press (OUP) | 2017-10-04, Pages S277-S278
• Type of Work: Article | Final Publisher PDF
• Publisher DOI: http://doi.org/10.1093/ofid/ofx163.623
• Subjects: Health Sciences, Oncology; Health Sciences, Epidemiology; Biology, Biostatistics

Abstract

Background: Antibiotic exposure after allogeneic hematopoietic cell transplant (HCT) is common. Exposure to specific classes of antibiotics after HCT has been associated with mortality, relapse and graft-vs.-host disease. Exploring differences in antibiotic utilization across hospitals could provide opportunities for comparative effectiveness studies and quality improvement interventions. Methods: We conducted a retrospective cohort study of patients undergoing HCT for acute leukemia using a dataset merged from two sources: the Pediatric Health Information System and the Center for International Blood and Marrow Transplant Research. Medication use data were obtained from the day of transplant through engraftment. Hospital antibiotic utilization rates were reported as antibiotic days/1000 neutropenic days. Adjusted rates were calculated using a poisson regression controlling for age, sex, race, graft characteristics and days of ICU-level care. Results: After adjustment, hospital rates of anti-pseudomonal antibiotic use varied from 410 to 1037 antibiotic days/1000 neutropenic days (Figure 1A) and for Gram-positive antibiotic use from 109 to 771 antibiotic days/1000 neutropenic days (Figure 1B). As shown in Figure 1, within anti-pseudomonal antibiotics, there was variation by hospital in the use of Fourth and 5th generation cephalosporins, anti-pseudomonal penicillins and carbapenems; variation in Gram-positive exposure was driven by vancomycin. Gram-positive antibiotic use was moderately associated with days of ICU-level of care (spearman correlation coefficient = .55) but anti-pseudomonal antibiotic use was not (Figure 2). There was no association between days of antibiotic exposure and 30-day mortality. Conclusion: Among a homogenous population of children undergoing transplantation for acute leukemia, both the volume and spectrum of antibiotic exposure in the immediate post-transplant period varied widely. These data present an opportunity for hospitals to benchmark their antibiotic utilization practices and can be further leveraged to assess the clinical impact of differential antibiotic exposure.