Routine Use of Biomarkers to Rationalize Antibiotic Use During Febrile Episodes in Pediatric Bone Marrow Transplantation Units.

ABSTRACT

BACKGROUND: Children frequently develop fever after hematopoietic stem cell transplant (HSCT). Although the etiology of many febrile episodes (FEs) is not an infection, patients often receive broad-spectrum antibiotics in response. METHODS: To improve the judicious use of antibiotics in pediatric HSCT patients, we performed a prospective cohort study of children receiving an HSCT in Clínica Imbanaco (Cali, Colombia) between September 2016 and December 2019. We assessed all FEs occurring during 3 periods (infusion, neutropenic and engraftment). We measured procalcitonin and C-reactive protein (CRP) sequentially during each FE and compared levels among patients with fever due to significant infection (FSI) versus fever not attributable to infection (FNI) in each transplant period. RESULTS: There were 166 FEs in 95 patients. FSI accounted for 12%, 42% and 42% of FE during infusion, neutropenic and engraftment periods, respectively. CRP had better discriminatory capacity for FSI versus FNI in the infusion period [area under the curve (AUC) 0.80 (95% confidence interval [CI], 0.62-0.96) for a CRP level of 50 mg/L]. Neither biomarker performed well in the neutropenic period. During the engraftment period, a CRP of 65 mg/L had an AUC of 0.81 (95% CI, 0.65-0.96), while a procalcitonin level of 0.25 ng/mL had an AUC of 0.83 (95% CI, 0.63-1.0). In contrast to procalcitonin, the CRP's pattern of change throughout the first 3 days of fever in each transplant period was different in FSI compared with FNI. CONCLUSION: Sequential measurement of biomarkers, especially CRP, may allow clinicians to more appropriately manage antibiotic use in pediatric HSCT units.