Association of an emergency department febrile neutropenia intervention protocol with time to initial antibiotic treatment.

Abstract

Earlier initial antibiotic treatment for febrile neutropenia is associated with improved clinical outcomes. This study was conducted to evaluate the association of an emergency department (ED) intervention protocol with time to initial antibiotic treatment for febrile neutropenia patients. We conducted a cohort study of adult ED febrile neutropenia patients before and after implementation of an intervention protocol. Analyses included comparison of means and medians, Kaplan-Meier estimates, multivariable regression analyses, interrupted time-series analyses, and causal mediation analyses. The intervention protocol included specific triage and process-of-care actions to reduce the primary outcome of time to initial antibiotic treatment. There were 69 patients in the 12-month preintervention period and 52 patients in the 8-month postintervention period. The mean (±SD) times to initial antibiotics were 197.6 (±85.4)min for the preintervention group and 97.7 (±51.0)min for the postintervention group (difference of 99.9min with 95% confidence interval [CI]= 73.5 to 126.4, p<0.001). The patients' probability for receiving initial antibiotics within 90min was severalfold greater (adjusted risk ratio= 10.31, 95% CI= 4.99 to 21.30, p<0.001) for the postintervention group versus preintervention group. ED length of stay, hospital length of stay, 30-day readmissions, and 30-day all-cause mortality were not different between the study groups. The association of the intervention protocol with time to initial antibiotics appeared to be mediated through times to treatment room placement, report of absolute neutrophil count, and initial antibiotic order. The intervention protocol was associated with a significant reduction in time to initial antibiotics for ED patients with febrile neutropenia. This association appears to be facilitated through specific intermediate process-of-care variables. A larger multicenter study is needed to assess the potential effects of an ED febrile neutropenia protocol on patient-centered clinical outcomes and resource utilization.