Eosinopenia Predicts Prognosis in Severe Community-Acquired Pneumonia

Background Eosinopenia has been reported as a marker of severity in infections, but its prognostic value in patients with severe community-acquired pneumonia (CAP) admitted to the intensive care unit (ICU) is unclear. Methods We retrospectively analyzed 496 patients with severe CAP admitted to the ICU, stratified by presence of eosinopenia at admission. Clinical characteristics, laboratory data, microbiological etiology, treatment, and outcomes were compared. Multivariable Cox regression identified independent predictors of 30-day mortality. Predictive performance of eosinophil counts, severity scores (CRB-65, CURB-65, PSI), and a modified CURB-65 incorporating eosinopenia (CURB-65Eos) were assessed. Results Eosinopenia was detected in 33% of patients. There were no differences in age and comorbidities between eosinopenic and non eosinopenic patients. Compared to non-eosinopenic patients, these patients had lower leukocyte, neutrophil, and lymphocyte counts and more frequent viral or polymicrobial infections. They more often required invasive mechanical ventilation (58% vs. 45%, p = 0.009) and developed pleural effusions (30% vs. 19%, p = 0.008). In-hospital and 30-day mortality were higher in the eosinopenia group (21% vs. 13%, p = 0.036; 20% vs. 12%, p = 0.022). Eosinopenia independently predicted 30-day mortality (HR 1.98; 95% CI 1.23–3.16; p = 0.005). Eosinophil counts alone had poor predictive accuracy (AUC 0.562), while established severity scores performed moderately. CURB-65Eos showed a numerical but nonsignificant improvement over CURB-65. Conclusions Eosinopenia is common in patients with severe CAP admitted to the ICU and is strongly associated with increased severity and mortality independent of age and comorbidities. It may serve as simple and inexpensive biomarker for the early identification of high-risk patients and could help guide more intensive therapeutic interventions.