The Usefulness Of Biomarkers In Differentiating Bacterial And Viral Infections - A Systematic Review
DOI:
https://doi.org/10.53555/AJBR.v28i4S.9046Keywords:
“Viral infection”, “Bacterial infection”, “Biomarker”, “Differential diagnosis”.Abstract
Infection is one of the most common reasons why patients seek medical attention. Differentiating bacterial and viral infections remains a key diagnostic challenge with significant therapeutic implications. Classic acute-phase markers and morphological parameters (such as leukocytosis), although commonly used, are often insufficient to reliably distinguish between etiologies. In an era of increasing antibiotic resistance and the emergence of multidrug-resistant bacterial strains, making decisions about symptomatic treatment or initiating antibiotic therapy is crucial. Overuse of antibiotics also often leads to higher treatment costs and more complications. Over 60 scientific papers were used to prepare this systematic review. After establishing precise inclusion and exclusion criteria, 44 original publications from 2012-2025 were selected, assessing the diagnostic value of individual biomarkers and host-response panels (proteins, cytokines, interferon proteins, transcriptional signatures) in differentiating bacterial and viral infections. The results indicate that the best diagnostic accuracy is achieved by combined panels, particularly those that include both bacterial and viral markers (e.g., TRAIL + IP-10 + C-reactive protein (CRP)), as well as interferon-induced proteins, such as MxA. The sensitivity and specificity of these panels have been shown to exceed 90%. Furthermore, transcriptional signatures and machine learning-based approaches represent a promising direction for diagnostic development that, until recently, was unattainable in medicine.
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Kaja Moc, Wiktoria Grzelak, Julia Kluczniok, Jacek Kramek, Magdalena Matzner, Patryk Marchwiany, Joanna Przeniosło, Anna Rutkowska, Anna Sadowska, Wiktoria Stenka (Author)

This work is licensed under a Creative Commons Attribution 4.0 International License.