A predictive bacterial infection score according to eosinophil level: An observational study

Introduction. - Several studies have shown that eosinopenia less than 0.04 g/L is a marker of bacterial infection in the presence of unexplained inflammatory syndrome. The aim of our study was to test this hypothesis and to propose a predictive score for bacterial infection (score CIBLE, C reactive protein, bacterial infections, levels of leucocytes and eosinophils). Patients and methods. - This was a single-center observational study of patients admitted to an internal medicine department in the year 2015 and presenting with an inflammatory biological syndrome. Patients were divided into 2 groups: bacterial infections (group 1) and nonbacterial inflammatory diseases (group 2). Results. - One hundred and ninety patients were included: 92 men (48.4 %) and 98 women (51.6 %). Mean age was 73.5 +/- 18.2 years [19-104]. Group 1 consisted of 124 patients (65.2 %) and group 2 of 66 patients (34.8 %). ROC analysis confirmed a cut-off level for eosinophils count at 0.04 as discriminant to predict bacterial infection. In a multivariate analysis, the eosinophil/neutrophil ratio, a history of COPD, the C reactive protein concentration, patient age and initial temperature were the most significant variables. They were used to build the CIBLE score. For a score higher than or equal to 87, the probability of a bacterial infection is at least 70 %. Conclusion. - The CIBLE score appears to be a relevant and inexpensive tool to establish a probability for bacterial infection. (C) 2017 Societe Nationale Francaise de Medecine Interne (SNFMI). Published by Elsevier Masson SAS. All rights reserved.