THRIVING
October 2024
iTHRIV Pilot Grant Team Receives AHRQ Funding for Improving Pediatric Donor Heart Utilization with Predictive Analytics
Congratulations to researchers Michael McCulloch, MD (UVA Children’s Hospital, Pediatric Cardiology) and Michael Porter, PhD (UVA Schools of Data Science and Systems Engineering) who recently received notice of funding from the Agency for Healthcare Research and Quality (AHRQ) for $1.03 million dollars through an R21/R33 award! The team was funded by an iTHRIV Clinical Translational Research Pilot award in 2021 which first helped them define donor characteristics for pediatric heart transplants through their project titled “Impact Quantification of Donor Echocardiographic Data on Pediatric Heart Transplantation Recipient Outcomes.” The investigators then successfully competed for a Jefferson Trust grant that helped further their work.
Over 10% of infants and children with end-stage heart failure or inoperable congenital heart defects die waiting for a life-saving heart transplant, yet 40% of potential donor hearts are unused. When one becomes available, clinicians have minutes to evaluate over 100 donor, candidate and offer-specific variables and decide whether the heart is a suitable match. Complicating this crucial decision is a paucity of data-driven guidelines, leaving clinicians without clear understanding of what impact these variables have on recipient outcomes and the risks tied to declining an offer, prolonging a candidate’s time on the waitlist. Without tools to help clinicians quickly and confidently assess all of the complex considerations involved in organ decision-making, many suitable hearts may be unnecessarily rejected. Research is needed to address this critical gap, and provide data-driven predictions and an enhanced offer interface to empower clinicians with timely and confident decision-making that will increase donor utilization and waitlist mortality.
Michael McCulloch, MD
Michael Porter, PhD
In their iTHRIV pilot project, Dr. McCulloch, a Professor of Pediatrics/Pediatric Transplant Cardiologist at UVA Children’s Hospital Heart Center, and Dr. Porter, an Associate Professor in the UVA School of Data Science and of systems engineering in UVA’s School of Engineering and Applied Science, successfully cleaned and merged data from the United Network of Organ Sharing (UNOS), which was a herculean effort considering the extent of data within each database and the sheer number of separate and distinct databases that existed. The team also extracted the echocardiographic data for the 30,000+ donor offers provided to them from UNOS and entered the data into a REDCap database, which comprises the entirety of the pediatric donor data that the team elected to focus on in their initial work. Their recent analysis of UNOS data from 2010-2020 found prior donor offer refusals were the single strongest predictor or subsequent refusals, suggesting a potential decision cascade bias. Further, 60% of all discarded hearts were found to be completely normal on their final ultrasound. Together these findings underscore opportunities for improvement in pediatric donor heart acceptance practices.
The new AHRQ award will allow this experienced team of data scientists, engineers, behavioral scientists, and pediatric transplant cardiologists to produce the first predictive model capable of analyzing the vast amount of data and delivering real-time assessments of the likelihood of a successful transplant at the time a donor heart is offered. This research could shape future donor heart acceptance guidelines, optimize donor utilization, and save lives. The end result could advance the field of donor acceptance practices across all organ systems and demonstrate a proof-of-concept for predictive modeling research in transplant medicine.