Cargando…

2624. Viral Pneumonia in Children: Facing the Challenge Using the Host Response

BACKGROUND: Diagnosing viral pneumonia in children is challenging. Chest radiographic imaging and clinical findings cannot reliably distinguish viral from bacterial pneumonia. Furthermore, pathogen-based diagnosis is limited by inaccessible site of infection and high asymptomatic detection rates. Th...

Descripción completa

Detalles Bibliográficos
Autores principales: Mastboim, Niv Samuel, Gottlieb, Tanya, Srugo, Isaac, Gervaix, Alain, Paz, Meital, Navon, Roy, Boico, Olga, Bamberger, Ellen, Klein, Adi, Stein, Michal, Chistyakov, Irina, Oved, Kfir, Eden, Eran, Shani, Liran
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Oxford University Press 2019
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6809646/
http://dx.doi.org/10.1093/ofid/ofz360.2302
Descripción
Sumario:BACKGROUND: Diagnosing viral pneumonia in children is challenging. Chest radiographic imaging and clinical findings cannot reliably distinguish viral from bacterial pneumonia. Furthermore, pathogen-based diagnosis is limited by inaccessible site of infection and high asymptomatic detection rates. The objectives of this analysis were twofold: first, to establish pneumonia etiology by applying a rigorous expert panel process, and second, to evaluate whether a novel host-immune signature that integrates viral induced proteins TRAIL and IP-10 together with bacterial CRP, can accurately differentiate viral from bacterial pneumonia. METHODS: This analysis included 1025 febrile children enrolled in two multi-center clinical studies that evaluated the host-immune signature performance: ‘Curiosity’ study (Oved et al., PLoS One 2015) and ‘Pathfinder’ study (Srugo et al., Pediatrics 2017). Pneumonia etiology – viral or bacterial – was determined by a panel of 3 independent experts, after reviewing patients’ clinical, laboratory, microbiological, and radiological data. Only cases with majority panel assignment were included. The host-signature generated one of the three results: viral, equivocal or bacterial, based on predetermined cut-offs. RESULTS: A total of 709 children were eligible for analysis and had an expert panel etiology determination. Of them, 114 were diagnosed with pneumonia: 51 assigned viral and 63 assigned bacterial (Figure 1). The signature separated viral from bacterial pneumonia with a sensitivity of 94% (95% CI: 85%–99%) and specificity of 95% (85%–99%) with 14% equivocal test results. Out of the 51 children diagnosed with viral pneumonia by the expert panel, 40 (78%) were given antibiotics, and 43 (83%) underwent chest x-ray evaluation. The signature correctly classified 42 of these 51 viral children, indicating its potential to reduce antibiotic overuse rates by 4.4-fold (from 78% to 18%; P < 0.001) and chest x-ray examination by 4.8-fold (from 83% to 18%; P < 0.001). CONCLUSION: The TRAIL/IP-10/CRP signature exhibits high accuracy for diagnosing viral pneumonia in children. The signature’s potential to safely decrease unnecessary antibiotics and chest radiographic imaging should be examined in future utility studies. [Image: see text] DISCLOSURES: All authors: No reported disclosures.