PROJECT SUMMARY/ABSTRACT Antimicrobial resistance (AMR) is a rapidly growing threat to global health that is primarily driven by the overuse and misuse of antimicrobials. The applicant’s preliminary work in Uganda confirmed what has been noted in other resource-limited settings - children who present with febrile illnesses, particularly those with associated respiratory symptoms, are frequently treated with antibiotic therapy despite most of them likely having self-limited viral infections that do not require such treatment. This inappropriate dispensation of antibiotics is in part due to the diagnostic uncertainty inherent in the reliance on clinical symptoms for diagnosis and management of acute respiratory infections (ARI). Therefore, there is an urgent need for novel diagnostic tools that can distinguish children with bacterial ARI from those with non-bacterial infections, thus reducing antibiotic overuse for children presenting with this clinical syndrome. The objective of the scientific aims proposed in this K23 application is to develop a diagnostic model that accurately predicts bacterial infection in children with febrile ARI in Uganda. This work will facilitate the long-term goal of the applicant, which is to become an independent investigator with expertise in diagnostic evaluation and implementation working to improve quality of care and antimicrobial stewardship in both domestic and global contexts. Specifically, under the mentorship of experts in the fields of emerging molecular technologies, diagnostic evaluation, and prediction modeling at the University of North Carolina at Chapel Hill, she will (1) determine the etiology of febrile ARI in a cohort of previously and prospectively enrolled children, enhancing the diagnostic assessment with PCR-based and metagenomic next generation sequencing tools, (2) validate a host immune response gene expression assay, incorporating genes previously identified to be differentially expressed in viral and bacterial ARI, that minimizes the number of genes included to maximize feasibility in resource-constrained settings, and (3) develop novel clinical algorithms that combine host-based classifiers with other clinical and laboratory data to accurately predict bacterial ARI. Through pursuit of the scientific aims, the associated experiential and hands-on laboratory and biostatistical training, and focused didactic coursework, the applicant will address the following gaps in her knowledge: (1) molecular technologies and their role in infectious disease diagnosis and diagnostic development, (2) clinical prediction modeling, and (3) diagnostic tool evaluation and implementation. By the conclusion of the K23 award period, the applicant will have generated key results to inform a competitive R01 application. She will also be well-positioned to transition into an independent physician scientist and leading expert in diagnostic evaluation and antimicrobial stewardship in resource- constrained settings.