A novel assay for distinguishing bacterial from viral infections significantly outperformed routine laboratory parameters and biomarkers in a study using serum remnants from children suspected to have acute infection.
The assay integrates measurements of 3 blood-borne host proteins: a novel viral-induced host protein, tumor necrosis factor-related apoptosis-inducing ligand (TRAIL); interferon y-induced protein-10 (IP-10); and the traditional bacterial-induced C-reactive protein (CRP). The serum samples were collected at 5 pediatric emergency departments and 2 wards (in Switzerland and Israel) from children aged 3 months to 18 years with fever of at least 100.4°F and symptoms lasting 7 days or less.
The study group included 361 children, of whom 229 were diagnosed with a viral infection, 68 with a bacterial infection, and 54 with infection of an indeterminate etiology. Of this study group, the assay classified 209 patients (58%) with a viral outcome, 99 patients (27%) with a bacterial outcome, and the remaining 53 patients (15%) with an equivocal outcome. This represents a sensitivity of 93.8% (the assay generated 4 false negatives for bacterial infection) and a specificity of 89.9% (it generated 21 false positives). The study group included 6 patients with serious bacterial infection and bacteremia, all of whom the assay correctly classified.
The assay outperformed white blood cell count and demonstrated significantly improved sensitivity and comparable specificity to absolute neutrophil count. It also showed comparable sensitivity and significantly improved specificity compared with the biomarker CRP alone and significantly improved sensitivity and specificity compared with procalcitonin. In addition, this test panel performed consistently across a wide range of patient subgroups, including various ages and clinical syndromes, regardless of time from symptom onset and receipt of antibiotics up to 48 hours before testing (Srugo I, et al. Pediatrics. 2017;140:e20163453).
Thoughts from Dr. Burke
The perfect test(s) for differentiating fever caused by viral versus bacterial infection is the holy grail of pediatrics. The search has included white blood cell count, absolute neutrophil count, CRP, and procalcitonin—each new candidate with better sensitivity and specificity than the last. This panel of 1 old (CRP) and 2 new (TRAIL, IP-10) proteins outperforms its predecessors in sensitivity and specificity. It has an additional real-life advantage of remaining accurate even in children who have been previously treated with antibiotics. Watch for further development with more rapid turnaround and increased availability of this novel diagnostic tool. This may be the one!