Infectious Diseases and Therapy (Dec 2023)

Developing a Tool for Differentiation Between Bacterial and Viral Respiratory Infections Using Myxovirus Resistance Protein A and C-Reactive Protein

  • Konstantina Iliopoulou,
  • Panagiotis Koufargyris,
  • Sarantia Doulou,
  • Elisavet Tasouli,
  • Sokratis Katopodis,
  • Stavroula-Porphyria Chachali,
  • Georgios Schinas,
  • Charalampos Karachalios,
  • Myrto Astriti,
  • Paraskevi Katsaounou,
  • George Chrysos,
  • Theodoros Seferlis,
  • Effrosyni Dimopoulou,
  • Myrto Kollia,
  • Garyphalia Poulakou,
  • Styliani Gerakari,
  • Ilias C. Papanikolaou,
  • Haralampos Milionis,
  • George N. Dalekos,
  • Vasiliki Tzavara,
  • Theano Kontopoulou,
  • Evangelos J. Giamarellos-Bourboulis

DOI
https://doi.org/10.1007/s40121-023-00901-2
Journal volume & issue
Vol. 13, no. 1
pp. 105 – 119

Abstract

Read online

Abstract Introduction The aim was to assess the performance of a blood assay combining measurements of MxA (myxovirus resistance protein A) and CRP (C-reactive protein) to differentiate viral from bacterial respiratory infections. Methods In a prospective study, MxA and CRP were measured in the blood by the AFIAS panel in adults admitted with respiratory infection. Patients were split into discovery and validation cohorts. Final diagnosis was adjudicated by a panel of experts. Microbiology-confirmed cases comprised the discovery cohort, and infections adjudicated as highly probable viral or bacterial comprised the validation cohort. Results A total of 537 patients were analyzed: 136 patients were adjudicated with definitive viral infections and 131 patients with definitive bacterial infections. Using logistic regression analysis, an equation was developed to calculate the probability for bacterial infection using the absolute value of MxA and CRP. Calculated probability ≥ 0.5 and/or MxA to CRP ratio less than 2 applied as the diagnostic rule for bacterial infections. This rule provided 91.6% sensitivity and 90.4% negative predictive value for the diagnosis of bacterial infections. This diagnostic sensitivity was confirmed in the validation cohort. A MxA/CRP ratio less than 0.15 was associated with unfavorable outcome. Conclusion The calculation of the probability for bacterial infection using MxA and CRP may efficiently discriminate between viral and bacterial respiratory infections.

Keywords