Procalcitonin adds to diagnosis, but does not reduce initial antibiotics in febrile neutropenic children.

Abstract

Introduction: The immune system is suppressed during chemotherapy. This makes diagnosis of severe life-threatening infections more difficult and it also intensifies the clinical course of such infections. Hence, empirical use of broad-spectrum antibiotics is mandatory. We investigated if procalcitonin (PCT) measurement may improve diagnostic accuracy.

Material and methods: In a prospective observational study, we included all admissions due to febrile episodes in a cohort of children below 16 years of age. PCT and C-reactive protein (CRP) were analyzed using LUMI test and VITROS CRP slides, respectively.

Results: We recorded 230 febrile episodes in 85 children. Severe systemic infection was found in 61 (27%) of these episodes. PCT performed better than CRP (p value ≤ 0.01). The discriminative power of PCT was significant already from admission. For CRP, discriminative power was significant after 48 hours. The cut-offs for PCT and CRP were 0.4 ng/ml and 336 nmol/ml to achieve sensitivities of 93%. The specificities for PCT and CRP were 45% and 22%, respectively. Severely infected patients were not found, either by PCT or by CRP in four (7%) cases. PCT levels rose in response to infection in the neutropenic population.

Conclusion: PCT measurement considerably improves biochemical information; however, the sensitivity is too low to safely alter the recommended administration of empirical antibiotics at admission.