When it comes to pediatric heart transplant, any delay in the process — from organ procurement to transplantation — puts a child at risk. That’s why two UVA Health researchers are teaming up to create an analytics tool to streamline and hasten the decision-making process for organ acceptance. They are in the process of evaluating the last decade of pediatric heart transplants in the U.S., with the goal of creating a predictive dashboard to help pediatric cardiologists find the right donor for the right transplant candidate at the right time.
Developing a Data-Driven Tool to Analyze Organ Criteria
According to UVA Children’s Heart Center transplant cardiologist Michael McCulloch, MD, doctors must evaluate over 100 variables to determine whether a donor heart is a suitable match for their patient. Some of these considerations include:
- Time since the patient became an organ donor
- Whether the donor received CPR and for how long
- Current medications administered to the donor
- What the heart looks like on an ultrasound
The typical decision turnaround time is less than 30 minutes. Still, those are critical moments, according to McCulloch.
“I found myself wishing that I had some type of analytic tool that could summarize all of these variables and give me reports, so that I could quickly understand more about the heart in front of me," he says.
McCulloch believed it was possible to create a more systematized analysis so the acceptance criteria could be more comprehensive, data-driven, and faster. But he needed a partner.
Cross-Collaboration: Data Science & Medicine
Through seed grants provided by the UVA Center for Engineering in Medicine, iTHRIV and the Jefferson Trust, McCulloch set out to find a top-notch data scientist at UVA with whom he could start this project. He chose Michael Porter, PhD, an associate professor in the Schools of Engineering and Data Science. Porter's research focuses on finding meaning and relevance in large datasets by tracking down patterns — patterns that can’t easily be recognized without using data science tools.
Porter has used data science and modeling to help predict crime, terrorism, and even traffic incidents. Now he’s hoping to use the same type of modeling and predictive analysis to help identify viable, healthy, donated pediatric hearts.
Leveraging 10 Years of Transplant Data
The first step for Porter and McCulloch was to find commonalities in what makes for a successful organ match. They did this by analyzing 10 years of patient data from the United Network of Organ Sharing, which includes the history of more than 20,000 pediatric heart patients in the U.S. By documenting and analyzing characteristics of hearts accepted in the past, they believe they can better predict transplant success.
“We’re looking at data that’s never been looked at before,” says Porter. “For the first time, we will be able to see what happens from the time a heart is considered available, declaration of death by neurologic criteria, to the time the heart is accepted or declined — not just with one or two patients, but with all pediatric donors and waitlisted candidates. This baseline data will start to paint a picture.”
Before determining which variables are relevant for post-transplant outcomes, the team is first focusing on determining acceptance practices across the country. Knowing which donor and candidate variables doctors are currently using to make their decisions can help inform attempts to mathematically predict which variables should be used.
McCulloch adds, “We’re also looking at the organ procurement organizations, the groups that manage the donor heart once a patient is declared dead by neurologic criteria, and then seeing if the outcomes are different. This could suggest that differences in management strategies or approaches, maybe even just on the front end, could influence outcomes.”
To help address these important research questions, two UVA professors have joined the team. Peter Alonzi is a data scientist and expert in research computing, and Sara Riggs, PhD, is a cognitive systems engineer who specializes in developing displays and interfaces to aid complex decision-making.
One of the next steps for the team is to discover the relevance of the potential predictors. They need to define what traits, or sets of traits, accepted hearts had in common, contrasted with the common traits of hearts declined for transplant. The goal is to create algorithms that can narrow down meaningful combinations and relationships.
“There is a need for models and analytics that just don’t exist right now, and our conversations are driving these developments,” Porter says. “This is so new. No one’s done what we’re doing yet, and that’s very exciting.”
On the Horizon: Organ Viability Scorecard to Guide Pediatric Cardiologists
The team has finished collecting and combining the millions of datapoints, and the first thing they hope to produce for clinical testing is a scorecard. This will be a starting point for physicians when evaluating new donor hearts, providing a broader look at the parameters that spell transplant success, with answers to questions like:
- What were the measurement ranges for the donor and candidate variables of the hearts that were accepted?
- How does that compare to the declined organs?
“Imagine going from having just the data from your practice to use for decision-making to having the analyzed data of all pediatric patients from the past 10 years,” Porter says. “That’s a lot more confidence for that early morning call about a potential donor heart.”
The scorecard would be used as more of a guide to assist doctors in their decision-making, not as a fixed protocol. Eventually, the team intends to put the data in a cloud app where everyone can both use the tool and contribute their data, allowing the scorecard to continually be refined. With enough refinement, the scorecard could eventually become a predictive tool. “We hope that one day this modeling tool will tell us — all pediatric cardiologists — whether to take actions like: accept the heart right now, wait 6 hours to see what its condition is, or pass completely, and more," he says. “Eventually, the scorecard and predictive modeling techniques could be extended to other types of organ transplants.”