This article is more than 5 years old. Information may no longer be current.
Three-host protein-based assay can distinguish bacterial, viral infections in young children
A recent double blind validation study determined that using a three-host protein-based assay to distinguish between bacterial and viral infections in young children with lower respiratory tract infection or fever can reduce antibiotic misuse.
“It is often not possible to differentiate between bacterial and nonbacterial disease on the basis of clinical judgement alone … consequently, antibiotics are prescribed almost twice as often as required in children with acute respiratory tract infections in the [United States],” Chantal B. van Houten, MD, from the division of pediatric immunology and infectious diseases at the University Medical Center Utrecht in the Netherlands, and colleagues wrote. “These data create an incentive to introduce diagnostic tools that can aid in distinguishing between bacterial and nonbacterial causes of infection.”
The researchers sought to confirm the accuracy of ImmunoXpert (MeMed) — a new host-response–based diagnostic assay that combines tumor necrosis factor-related apoptosis-inducing ligand (TRAIL), interferon gamma induced protein-10 (IP-10), and C-reactive protein (CRP) — in differentiating between bacterial and viral infections in young children.
In a prospective, double blind study, van Houten and colleagues enrolled children aged 2 to 60 months with lower respiratory tract infection or fever without source at four hospitals in the Netherlands and two hospitals in Israel. Using all clinical and laboratory information available, a panel of three experienced pediatricians masked to the assay results decided on a reference standard diagnosis (bacterial or viral infection) for all patients. The researchers externally validated the sensitivity and specificity of the ImmunoXpert assay in distinguishing between bacterial and viral disease and compared its diagnostic accuracy with commonly used CRP and procalcitonin testing.
Of 577 children assessed between Oct. 16, 2013 and March 1, 2015, most of the panel diagnosed 71 cases as bacterial infections and 435 as viral infections. The panel inconclusively diagnosed another 71 patients. In terms of diagnostic performance, the assay differentiated bacterial from viral infections with 86.7% sensitivity (95% CI, 74.8-93.1), 91.1% specificity (95% CI, 87.9-93.6), 60.5% positive predictive value (95% CI, 49.9-70.1) and 97.8% negative predictive value (95% CI, 95.6-98.9). In the cases with unanimous panel diagnosis (n = 354), the results showed 87.8% sensitivity (95% CI, 74.5-94.7), 93% specificity (95% CI, 89.6-95.3), 62.1% positive predictive value (95% CI, 49.2-73.4) and 98.3% negative predictive value (95% CI, 96.1-99.3).
In an accompanying editorial, Susanna Esposito, MD, and Nicola Principi, MD, both from the pediatric highly intensive care unit at the University of Milan Medical School, said that although the combined TRAIL, IP-10 and CRP assay improves the identification of patients with viral infections, there are limitations that preclude its use in routine clinical practice — including the fact that the test requires advanced laboratory techniques that cannot be used outside of the hospital. Another limitation is that upper respiratory secretions in children can lead to incorrect diagnosis of a lower respiratory infection, and that “bacteria and viruses can simply be carried and could have no association with the cause of a disease.”
They determined that future studies should confirm the results of three-host protein-based assays in larger populations. – by Savannah Demko
Disclosures: Van Houten reports funding by MeMed. Esposito and Principi report no relevant financial disclosures.
Perspective
Back to Top
James H. Brien
This sounds like something that certainly would be worthwhile. However, speaking as just one practicing infectious diseases specialist, I have seen similar studies come and go with the promise of potentially changing the practice such that unnecessary antimicrobial therapy could be avoided, only to find that in practice, it either becomes too expensive, too long getting results, or just needing more reassurance with the results, that it never really gets any traction. That is not to say that we should not continue to pursue clinical scoring systems, special differentiating assays, or a combination of these to try to limit diagnostic uncertainty, and therefore unnecessary antimicrobial prescribing. In that vein, this particular assay certainly holds promise, but an old skeptic, like me, will need to see it in a real-time, practical clinical application, along with the cost-effectiveness before jumping on board. But, I will jump on board if the promised benefit is met.