SepSense: Improving Prediction of Sepsis in Community Settings
Improving outcomes of severe infection and sepsis in low-resource community settings
Keywords: Infectious disease, Healthcare, Computation, Global health
Sepsis, the uncontrolled response to infection, is among the most prevalent mortality drivers, causing 1 in 5 of all deaths worldwide. Intervening early significantly improves survival, motivating the development of electronic health record (EHR)-based sepsis prediction algorithms. However, these are ineffective for the 80% of cases classified as community-acquired, beginning outside the hospital. SepSense aims to develop a predictive algorithm to accurately assess patients’ likelihood of developing sepsis from community-acquired infections and recommend interventions.
Skillset needs: Collaborating with patients, physicians, and administrators in the intensive care unit and infectious disease hospital settings and outpatient clinics, Partnering with hospital systems abroad and working on global health, Backend engineering (specifically developing machine learning algorithms and electronic health record integration), Establishing IRB protocols
Contact: Aravind Krishnan