Project 461493

Each second, five newborn babies die from suspected life-threatening infections somewhere in the world. Newborns and young babies are more vulnerable to severe infections than all other age groups. However, there is no fast and easy way to tell which microbes are involved, resulting in babies being both over- and under-treated with antibiotics that may or may not help them, contributing to antibiotic resistance and wasting limited resources. In this study, we will use advanced molecular diagnostic tests that are generally only available in a research setting and machine learning techniques to identify specific whole-blood signatures that could be used to determine which babies are infected by bacteria and require urgent antibiotics. Our ultimate goal is to develop a rapid, easy-to-use tool that can be used around the world to guide the diagnosis and treatment of babies with systemic infections. A better understanding of the types of microbes involved will improve our abilities to differentiate between babies requiring urgent antibiotics and those requiring other (non-antibiotic) treatments, saving the lives of newborn babies worldwide.