Evaluation of new laboratory tests to discriminate bacterial from nonbacterial chronic obstructive pulmonary disease exacerbations

G-J M van de Geijn,S Denker,V Meuleman-van Waning,H G M Koeleman,E Birnie,G-J Braunstahl,T L Njo

DOI: https://doi.org/10.1111/ijlh.12550

Abstract:Introductions: Discriminating bacterial from nonbacterial acute exacerbations of chronic obstructive pulmonary disease (AECOPD) is difficult, causing antibiotics overuse and bacterial resistance. Sputum cultures are of limited use because results take time. In our hospital, only leukocyte concentration and CRP are laboratory parameters evaluated in AECOPD. We evaluated additional tests to discriminate bacterial vs. nonbacterial AECOPD: 5-part leukocyte differentiation (hematology analyzer), leukocyte differentiation using flow cytometry (Leukoflow, Cytodiff), Leuko64 kit, and procalcitonin. Methods: Retrospectively, patients were classified as bacterial or nonbacterial AECOPD. ROC analyses tested how the additional tests discriminate these groups. Results: Twenty-two AECOPD were classified as bacterial and 23 as nonbacterial. From the additional tests, basophil percentage (Cytodiff) has superior AUC (0.800). At a cutoff resulting in ≥90% sensitivity, neutrophil/lymphocyte ratio (AUC:0.755) and CD4-positive T cells (Leukoflow, AUC:0.747) have the highest specificity (57%). Both neutrophil mean volume and standard deviation (Cell Population Data, DxH800 hematology analyzer) had good combined sensitivity and specificity (AUC:0.846/0.804, 91% sensitivity, 69% specificity). Addition of leukocyte populations and procalcitonin to CRP in regression models (AUC: 0.907/0.876/0.890) increased specificity compared to CRP alone (71% or 73% vs. 39%). Conclusion: No additional test has sufficient accuracy on its own to predict bacterial AECOPD. Combining CRP with several parameters from the additional tests may improve this.