An estimated 1.4 billion antibiotic prescriptions were filled in the US from 2000-2010. Despite a national antibiotic education campaign, approximately 30% of antibiotic prescriptions are unnecessary. As a result of the over-prescription of antibiotics, there is a current and pressing public health need for improved antibiotic stewardship. Compounding this problem are new modalities of healthcare like telemedicine that, while meeting patient demand for convenience, have been shown to have higher antibiotic prescription rates. As a result, there is a pressing need for a tool that these remote providers can use to distinguish between bacterial infections, where antibiotics can be effective, and viral infections where antibiotics are unneeded. In this research, we propose the development of Barcode Embedded Rapid Diagnostics (BERDs) as a point-of-care test to measure a host biomarker signature in response to an infection in blood. These tests have novel spatial patterning of reagents that allows them to automatically and quantitatively be analyzed by a mobile phone application. Once the biomarkers have been analyzed on the test, a clinical decision support engine developed in this proposal will provide a confident diagnosis. We have assembled an interdisciplinary consortium that consists of biomedical engineering, point-of-care device, and software development expertise through PragmaDx, Inc., and biomarker assay development in the Department of Chemistry at Vanderbilt University. To meet our shared goals, our specific aims will: optimize manufacturing conditions of BERDs, develop and evaluate a machine learning clinical decision support engine, and rigorously compare the performance of BERDs to standard clinical laboratory methods using de-identified patient specimens. A successful outcome of this Phase I STTR will be the development of a diagnostic test and clinical decision support engine that is capable of measuring multiple biomarkers in blood to differentiate the etiology of an infection at the point-of-care, with accuracy comparable to standard clinical laboratory assays. Upon completion of this effective demonstration, the platform will be ready for a large-scale clinical evaluation that will be the subject of a Phase II application.
An estimated 30% of antibiotic prescriptions are unnecessary. As a result, antibiotic resistance has developed into a major public health concern. Diagnostic tools that can distinguish between varying etiologies of infection are necessary to curb this growing problem.
