Assistant Professor, Johns Hopkins University
Suchi Saria is an assistant professor of Computer Science with joint appointments in Applied Math & Statistics and Health Policy at Johns Hopkins University. Her interests are in statistical machine learning and its applications to domains where one has to draw inferences from observing a complex, real-world system evolve over time. Her research focuses on the methodological questions underlying the goal of individualizing prognosis in healthcare from the disparate high-dimensional (clinical) data streams that are routinely collected. Her work has led to new ideas for characterizing disease subtypes using longitudinal data, latent variable models for individualizing prognosis, and methods for scaling up inference.
Assistant Professor, Johns Hopkins University School of Medicine
Nishi Rawat is a physician and healthcare services researcher, double-board certified in Critical Care and Emergency Medicine and faculty at Johns Hopkins University School of Medicine. Her research is focused on improving quality of health care and reducing costs with an emphasis on the application of information technology to evaluate performance, minimize waste and eliminate redundancies. She is an NIH-sponsored Principal Investigator to apply sensor technology to improve provider compliance with evidence-based interventions, and has been awarded several grants to commercialize emerging technologies. She is also a key co-investigator for a national quality improvement collaborative to improve care for mechanically ventilated patients, funded by the Agency for Healthcare Research and Quality. Nishi has won several awards at national research meetings for her research, and a leadership award from Johns Hopkins for her quality improvement activities.
13 - An 85% prediction model? Advances in Sepsis Prediction at Johns Hopkins (Case Study)
Suchi Saria, PhD (Assistant Professor, Johns Hopkins University), Nishi Rawat, MD (Assistant Professor, Johns Hopkins University School of Medicine
Led by Professor Suchi Saria, a team of data scientists at Johns Hopkins University have made substantial advances in accurately predicting the patients most likely to experience septic shock. On the back end of a six-year study period, the research team drew from a 16,000 patient data set. Eighty-five percent of the time, their prediction model successfully predicted septic shock. And equally important, Dr. Saria’s work did not require additional screening.
Join Dr. Saria and several from their team as they share lessons learned across the six-year study including both clinical implementation and IT development challenges that they faced. In its last stage of testing in collaboration with clinical teams, these tools may soon become available to others. Join us in learning more.