Synthetic Patient Records Help Deliver Real Health OutcomesFebruary 2019
Topics: Health, Healthcare IT, Health Innovation, Software (General)
At past health IT developers' conferences, there was limited focus on the methods of data analytics: showing actual methods, actual data, and actual results. That's because analytics are run primarily on real data—but real patient data sets cannot be shared due to privacy laws and ethical issues. This limited the community's ability to share information about how developers generated the analytics. And it limited the ability of innovators to use big data and data analytics to advance the quality of healthcare.
That all changed at a major conference in the summer of 2018.
"All of a sudden, people were opening up and showing their data analytics, including big data analytics vendors," recalls MITRE's Jay Walonoski. "And every single one of them, when they gave a presentation, was using Synthea—and giving us a shout-out. That's when we knew—this is becoming big."
That's because Synthea provides a public, open synthetic patient data set that enables developers to demonstrate analytics methods and the results of the analytics.
Due to privacy regulations, appropriately, it's hard to obtain access to patient data for health IT development. This makes it difficult for analysts and developers to use electronic health record (EHR) systems to conduct data analytics or develop clinical decision-support tools to improve medical outcomes. To resolve this issue, our researchers developed a way to create mock patients with realistic health issues to generate the data needed. They dubbed their solution "Synthea." (Read more about the creation of Synthea and Jay Walonoski.)
MITRE offers this open-source software solution without charge to the health IT community as part of our mission to solve problems for a safer world. "And it succeeded more than we ever really anticipated," Walonoski says.
Synthea started as a demonstration project, SyntheticMass, which models the health information of more than one million Massachusetts residents. Now, the federal government and industry are using Synthea to address healthcare issues.
Community Helps Build Synthea's Capabilities
"We had many scenarios in the past where we needed a lot of patient test data for our work," recalls Bo Dagnall, chief technologist and strategist in Perspecta's healthcare group. "We talked to private companies, but their data was either too expensive or not realistic. So, no progress. Then I learned about Synthea, in late 2017.
"We took a subset of the SyntheticMass records, about 100,000 of them, and did a bulk ingest into the interoperability platform we are building to create a base population we could use for our testing and demos," he explains. "That was awesome."
Dagnall also chairs the synthetic data working group of the Open Source Electronic Health Record Alliance (OSEHRA), which maintains the open-source Veterans Information Systems and Technology Architecture (VistA) EHR platform for the Department of Veterans Affairs.
The working group is developing tools to use Synthea more effectively. One tool allows developers to build a user interface on top of Synthea so they can specify how many mock patients with which specific health issues to generate.
"So, now you can use the VistA solution—which is a robust platform for assisting doctors in the delivery of care—with synthetic data populating the database."
The group is also creating new disease modules in Synthea. They started by generating records for mock patients suffering from anemia. Medical students at the University of Michigan, which participates in the group, are learning how to work with medical data and software in the delivery of healthcare, using Synthea data on demand.
"I think word is starting to get out," Dagnall says. "People are contacting me to understand what the working group does and about Synthea. Obviously, there's more work to do, but it's a valuable tool for testing, analytics, software development, and demonstrations.
"We're big fans."
Next Stop: SyntheticUSA?
MITRE has its own big ambitions for Synthea. "Our goal is to go to the next level and model the whole country, starting this year," Walonoski says.
But it's more than just scaling up and software testing, he adds. "It's also, 'what if we wanted to run a simulation about how we might change the nation's healthcare system?' Instead of just hypothesizing about what might work or trying out experiments in the real world and seeing how it affects people's real lives, we can try different experiments in Synthea and look at the results. That's where we want to go."
—by Jim Chido
Explore more at MITRE Focal Point: Health.