Sepsis is a major cause of organ failure and death in both children and adults around the world for which there are no specific therapies other than antibiotics and supportive care. A major reason for the lack of targeted therapies is our incomplete understanding of the genetic, environmental, microbial, socioeconomic and biologic factors that underly risk of and outcomes from human sepsis and that drive heterogeneity within this clinical syndrome. We propose to create a unique resource, the Sepsis ClinicAl Resource And Biorepository (SCARAB), a two center prospective study in adults and children with sepsis and critically ill controls that is designed to be both comprehensive and scalable to large multicenter implementation. To create this resource, we will leverage our experience with large scale prospective observational studies in the ICU, our novel method of non-invasively sampling the distal airspace in mechanically ventilated patients, our world class bioinformatics expertise, state-of-the-art phenotyping methodologies, and unique institutional resources. SCARAB will prospectively enroll subjects from pediatric and adult ICUs at VUMC and Meharry Medical College, an institution that cares for a primarily minority population. Subjects will be enrolled early in their illness and serial biologic specimens of blood, DNA, respiratory micro-droplets (from Heat Moisture Exchange (HME) filters), tracheal aspirates and urine will be collected. To study the end organ most commonly injured in sepsis, we will establish best practices for collection, processing and storage of lung fluid collected from HME filters, a novel method for sampling the distal airspace in ventilated patients. Working with our collaborators in Biomedical Informatics, we will develop automated phenotyping algorithms to identify patients with sepsis and organ failures including ARDS, AKI and delirium. We will use cutting edge natural language processing methods such as concept mapping to conduct detailed sub-phenotyping for clinically important variables that are challenging to phenotype on a large scale, such as hyper- and hypoactive forms of delirium. SCARAB will also be linked to unique VUMC resources such as MicroVU which retains clinical isolates of all microbial specimens at VUMC allowing large studies of host-pathogen interactions using contemporary multi-omics approaches. SCARAB clinical data and biospecimens will be accessible to the broader research community through an application on our web based study portal. Proposals will be reviewed by a Resource Utilization Committee comprising study investigators and national experts in sepsis and organ dysfunction. Data will be de-identified using the most rigorous security standards. SCARAB is also designed to be readily accessible to early career investigators by providing statistical support, customized phenotyping and access to Studios for expert engagement for study design. Bringing together all of these innovative elements will allow for mechanistic studies, development of predictive biomarkers, understanding of pre-hospital factors and clinical outcome studies in patients with sepsis and appropriate controls at a scale not previously possible.

Project Narrative Here we propose to create a unique resource, the Sepsis ClinicAl Resource And Biorepository (SCARAB) registry comprising prospectively enrolled sepsis patients and critically ill control patients, their complete, de-identified medical records, automated phenotyping algorithms, and a complement of serial biological specimens to be used by the sepsis research community.