What if the healthcare community had immediate access to data about COVID-19 patient care to see what treatments work best? MITRE researchers are taking what they’ve learned for digital data sharing and cancer—and applying it to pandemic response.
The COVID-19 pandemic has highlighted the fragile and fragmented state of the U.S. healthcare data infrastructure.
When the new virus hit in March 2020, there was no protocol for treating it. Clinicians tried medicines and treatments that worked for other respiratory viruses, hoping to find solutions. How could they quickly determine what worked and share best practices? People were dying.
Without the capability to digitally exchange case data across the country, health systems relied on their traditional tools—sharing faxes, paper exchanges, and spreadsheets.
Numerous randomized clinical trials—the gold standard of high-quality clinical data—were being started around the world, but even preliminary results on which treatments worked were months away.
- What could the real-world evidence—based on data extracted from patients’ electronic health records (EHRs)—tell us about a treatment’s safety and efficacy?
- Could real-world evidence help clinicians make treatment decisions by providing deidentified data from millions of patients’ records?
- Could EHR data provide comparable insights into effective care as those provided by clinical trials?
These are questions MITRE’s been asking for more than 10 years as we’ve pursued a modern data infrastructure for healthcare. And our work in oncology data may hold some of the answers.
mCODE Shows the Way to Better Data Standards
One of the key issues with electronic health records is that different systems don’t often work together. Put simply: They need a common language.
MITRE’s years of work in data standards and health IT capabilities led to a major research initiative called mCODE™ (minimal Common Oncology Data Elements). mCODE provides both a common data language and an open-standard, nonproprietary data model for interconnectivity across EHR systems.
While mCODE focuses on cancer, it’s also a test case for the use of real-world evidence to improve healthcare treatments and research.
When the SARS-CoV-2 virus exploded across the U.S., health systems turned all their energy to treating COVID-19 patients. MITRE staff from a range of disciplines quickly pivoted to work with government and industry to help keep health systems from being overwhelmed and to provide scientific data to decision makers.
A few of the mCODE and COVID-19 Healthcare Coalition (C19HCC) team members put their experience with health IT to work on COVID-19 projects—for example, facilitating the sharing of deidentified patient data to investigate clinical therapies. What could EHR patient data tell us about the effectiveness of such treatments as remdesivir, famotidine, and convalescent plasma?
“We wanted to show the value of contributing real-world evidence to the conversations around treating COVID-19 patients,” says MITRE’s Dr. Ryan Luginbuhl, a physician who led a coalition of companies, including data analytics and EHR vendors, that worked with us to create consensus-based data queries for observational cohort studies.
“We were able to leverage mCODE and C19HCC relationships to look at the value of real-world evidence for COVID-19,” he adds. “These companies—normally competitors—have donated many hours of time to create common queries that each could use in their own systems to identify trends.
“The federated queries allowed us to compare results across systems. For example, our cohort study showed that famotidine did not reduce risk of death in hospitalized COVID-19 patients, and this result was consistent across our partner EHRs and healthcare analytics platforms.”
A Future Where Data Sharing Leads to Better Care
Through mCODE, CodeX™ (Common Oncology Data Elements eXtensions), C19HCC, and many other programs, we’ve seen the value of creating public-private partnerships to combat complex, national problems. We continue to collaborate with stakeholders to identify and attack the barriers to interoperability and discuss the incentives. For example, we’ve found that the technical barriers to exchanging patient data are secondary to the organization-focused issues.
In contrast to healthcare, think about how quickly and seamlessly financial organizations can exchange customer data around the globe. You can stop at an ATM in any country and access cash from your local bank.
This same seamless integration is achievable in healthcare with agreed-on, broadly adopted standards.
“The health IT infrastructure is far from modern, even though it affects the quality of the healthcare we get every day and the state of health across the country,” notes Andre Quina, one of the mCODE leaders. “You can’t scale paper.”
Imagine: What if a “learning health system” powered by open source data standards and information architectures existed in the spring of 2020? Clinicians and researchers could have accessed data from thousands of COVID-19 patients and analyzed it for evidence of what treatments proved most beneficial for various demographics.
It’s a future MITRE envisions—and is helping to achieve.
“We’re invested in healthcare because we see how the gaps can create a national threat, and we’re creating responses to the threat,” says Dr. Jay Schnitzer, MITRE vice president, chief technology officer, and chief medical officer.
“The pandemic highlighted gaps in access to information that might have made a difference,” he adds. “Through data standards and interoperability, we can be better prepared not only for the next pandemic—but for the daily struggles to keep our populations safe and healthy.”