Project Spotlight: CHP VIEW

CHP VIEW aims to improve outcomes of hospitalized children by predicting critical deterioration outside of the pediatric intensive care unit (PICU). Inpatient critical deterioration is a severe worsening of a hospitalized child’s condition that can cause lifelong disability or death. These fast, unexpected changes can lead to transfer to PICUs, where a patient might receive intubation, cardiopulmonary resuscitation, transfusion, or serious outcomes that can increase the cost per admission by approximately $100,000. Every year in the U.S., about 500,000 children are admitted into the PICU, and approximately 1 in 3 will sustain long-term neurological defects that contribute to roughly $3 billion in healthcare expenditures. We connected with researcher Christopher Horvat, MD, MHA, assistant professor of critical care medicine at the University of Pittsburgh, to learn more about the project.

Please share a little about your background and your research experiences.

I am passionate about leveraging real-world data to support both quality improvement and research projects. I’m a board-certified pediatrician, pediatric intensivist, and clinical informaticist. I also have formal training in health administration, having earned a master’s degree in health administration from the University of Pittsburgh in 2017, as well as some additional training through Health Level 7, Cerner, and the College of Health Information Management Executives. My research focuses on leveraging data and other byproducts of care, such as biospecimens, to curb some of the uncertainties that envelop medical decision making, through a paradigm that is called the Learning Health System by the National Academy of Medicine.

Other members of the team include Srinivasan Suresh, MD; Gabriella Butler, MSN, RN; Denee Marasco, MBA; Harry Hochheiser, PhD; and Sarah Rubin, MD, MSCI.  Together they have backgrounds and expertise in pediatric care, clinical analytics, and informatics.

What led you to the PHDA?

The PHDA appeared to be perfectly aligned with our project’s core objectives, which involve enmeshing real-world health data, machine learning, and clinical investigation to improve outcomes among hospitalized children.

Walk us through your project.

We have a talented, multidisciplinary team that includes nurses, physicians, nurse informaticists, clinical informaticists, biomedical informaticists, systems architects, and systems analysts who have come together to create a surveillance system to identify hospitalized children who are high risk for suffering a life-threatening deterioration event. The project involves development of a machine learning model to predict inpatient deterioration, modifying existing information systems to extract the relevant data, and creating a custom user interface to present clinicians with the information over time. Our funding through the PHDA is allowing us to make some updates to the model, develop an alerting system in the hospital, and to develop an approach to package all features into an application or software bundle that will be commercialized.

In what ways has UPMC played a role lending clinical expertise and sharing data?

None of this work would have been possible if it were not for the engagement, resources, and talent provided by UPMC. UPMC partnership has been essential to access the real-world data, develop the information systems for running and deploying our model, and to educate frontline clinicians on the use of our surveillance system.

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?

This region is uniquely poised to develop data-driven solutions for some of the biggest problems facing healthcare. Real-world health data harbors innumerable insights that have the potential to optimize patient outcomes while curbing costs. The PHDA is well-positioned to help realize the National Academy of Medicine’s vision of a large-scale Learning Health System. I can imagine a future in Pittsburgh where there is a seamless, secure environment that allows for ready collaboration between clinicians, clinical researchers, and computer scientists using real-world data to support innovative solutions to clinical challenges.