Please share a little about your background and your research experiences.
Andrew King, PhD, is an informatics postdoc in the Department of Critical Care Medicine at the University of Pittsburgh. Through his graduate training in the Department of Biomedical Informatics, Dr. King has experience applying machine learning (ML) methods to electronic health record (EHR) data and developing clinical decision support systems for intensive care unit (ICU) clinicians.
Derek Angus, MD, MPH, FRCP, is a professor in the Department of Critical Care Medicine, Associate Vice Chancellor for Health Care Innovation at the University of Pittsburgh, and also serves as UPMC’s chief health care innovation officer. Dr. Angus’ research interests are learning healthcare systems, organization and delivery of critical care services, sepsis, precision medicine, and causal inference methods.
Jeremy Kahn, MD, MS, is a professor in the Department of Critical Care Medicine and Health Policy & Management. Dr. Kahn’s interests are in ICU organization and management, clinical decision making, translation of evidence into practice, quality measurement, and socioeconomic determinants of outcomes in critical care.
Gregory Cooper, MD, PhD is a professor in the Department of Biomedical Informatics and Intelligent Systems. Dr. Cooper’s interests are in the development and application of Bayesian statistics and artificial intelligence (AI) methods to biomedical informatics research problems, including causal discovery, computer-aided medical diagnosis and prediction, patient safety, and biosurveillance.
Christopher Seymour, MD, is an associate professor in the Department of Critical Care Medicine and Emergency Medicine. Dr. Seymour’s interests are in early recognition and treatment of sepsis, phenotypes of critical illness, and prehospital recognition and management of critical illness.
Jennifer Seaman, PhD, RN, is an assistant professor in the Department of Acute & Tertiary Care. Dr. Seaman’s research focuses on enhancing clinician-family communication and the implementation of routine, timely interdisciplinary family meetings in the ICU.
What led you to the PHDA?
For us, the PHDA fills a critical gap between research projects (which typically have a multi-year timeline for discovering and publishing generalizable knowledge) and commercial projects (which are much more product-focused). When faced with picking between these two types of projects, the PHDA allowed us to say “yes” to generating generalizable knowledge, “yes” to developing a commercial product, and, most importantly, “yes” to improving care for patients at UPMC while working towards the other endpoints.
Walk us through your project.
Healthcare, particularly in the ICU, is a team sport. It takes providers from a dozen different specialties to support a patient while recovering from multiple organ dysfunction. Every day, the patient’s care team makes hundreds of decisions about how to best support a patient while they recover. If any of those decisions are wrong or delayed, the patient’s recovery becomes more difficult. Because it is impossible for anyone on the care team to keep track of all the complexities of ICU patient care, we are creating a virtual assistant to help the team. Aviva (a voice-interactive virtual assistant) is an extra pair of eyes, or rather ears, to help an ICU care team keep a patient’s recovery on track. Aviva fuses information extracted from care team discussions of a patient with information extracted from the patient’s EHR to compose a complete and timely record of the patient’s trajectory. Afforded with more situational awareness than tools based only on EHR-data, Aviva is uniquely positioned to support the team by ensuring that every patient receives evidence-based medical care.
In what ways has UPMC played a role lending clinical expertise and sharing data?
UPMC is supportive of trying new things that have the potential to benefit patients. Our team consists of three leaders in the care of critically ill patients and benefits from input by countless nurses, physicians, and other specialists. UPMC puts their expertise, data, and funding support behind innovation in healthcare.
When you look at Pittsburgh as a region, what role do you see the PHDA playing? What do you foresee the future of innovation looking like here?
Pittsburgh’s health innovation scene is a well-stocked refrigerator. The shelves are full of the necessary ingredients: machine learning talent, clinical experience, human factors expertise, EHR data access, and a clinical environment to conduct randomized trials. The PHDA plays a role in pairing the ingredients and giving them time to marinate (via seed funding). A third role, which does not fit into this metaphor, is offering commercialization and acceleration experience.