Fighting Cancer with Standard Health Records
When Andre Quina was two years old, his parents brought home their first computer. By elementary school, he was programming.
"I never questioned it—I wanted to go into software," he says.
Quina couldn't know it then, but he was already on the trajectory to where he is now: leading MITRE's Standard Health Record for Oncology project. The initiative has the audacious goal of harnessing patient data to dramatically improve cancer treatment plans and outcomes.
"I understood the importance of technology and saw the disparity in healthcare software. And since MITRE focuses on the public good, I knew there was an opportunity to make a real impact."
The Healthcare Data Challenge
Quina says that compared to other domains, collecting data for healthcare is particularly challenging. Partially, that's because human physiology and the diseases affecting it are incredibly detailed and complex. But the way the healthcare industry transitioned from paper to electronic health records (EHRs) also played an important part. This process, which took place largely in the early 2000s, was disruptive. Most healthcare institutions essentially digitized their individual paper systems. The issue is greater than computing compatibility. From hospital to hospital, there's little standardization about what information to capture, terminology, and more.
This is a problem both for patient care and research because most different EHR systems "can't talk." That means patients often repeat the same information to multiple providers. It's wasteful and time consuming. Worse, it can be dangerous—patients may not remember or even recognize essential information.
The lack of standard data collection and processing also creates a tremendous missed opportunity for researchers. "Within cancer patients' health records and their doctors' notes there's a goldmine of knowledge about what treatments worked for which demographics," Quina explains.
"However, that information is often trapped in individual physicians' notes—unstructured text that's difficult to analyze. If doctors and researchers had access to that information, they could analyze results and more effectively understand, choose, and develop treatments.
"Our goal is to tap into and organize that data goldmine."
Standardizing the Health Record
Along with his co-principal investigator Dr. Brian Anderson, Quina is partnering with Monica Bertagnolli, chief of surgical oncology at Brigham and Women's Hospital and the Dana-Farber Cancer Institute in Boston. Bertagnolli and the MITRE team collaborated with other leading organizations, including the American Society of Clinical Oncology (ASCO) and its nonprofit subsidiary CancerLinQ, Intermountain Healthcare, the Clinical Information Interoperability Council, Health Level Seven, the U.S. Food and Drug Administration, and the Alliance for Clinical Trials in Oncology, among others.
MITRE and our partners began with the development of a standard health record (SHR) for breast cancer. "We've defined the information that must be collected to support evidence-based, data-driven, personalized breast cancer treatment," Quina says. "Components of the SHR Breast Cancer Draft Standard are now being evaluated in clinical settings across the country through our ICAREdata™ study."
Quina and Anderson took what they learned from that work to create the Minimal Common Oncology Data Elements (mCODE)™ standard, which applies to a wider variety of cancers. ASCO, CancerLinQ, and Intermountain Healthcare are working with MITRE in that effort.
Testing the Approach
In 2019, the research team will build their achievements, working with more partners to test these approaches. Quina and his team will pilot the mCODE with Intermountain and ASCO. They plan to demonstrate the end-to-end use and value of mCODE at the point of care and to cancer research. The pilot will use mCODE to provide reports back to doctors that show a set of treatment options for the patient categorized by how effective and safe those treatments were for a set of similar patients.
Through the ICAREdata study, the team is working with two metastatic breast cancer trials across five clinical sites to do a side-by-side comparison of select mCODE data elements and clinical trials data. The goal of the ICAREdata study is to demonstrate that mCODE can be used to help evaluate equivalent clinical trial endpoints across EHR “real-world” data and clinical trial data. If the data analysis approach proves a match for the trial findings, it has the potential to replace slow, expensive, and random clinical trials with machine-speed, low-cost discoveries across a broader research base.
"Compared with traditional clinical trials, the ability to use real-world data provides access to an exponentially greater level of rich and meaningful information," Quina says. "That would allow quicker and easier analysis. By harnessing the power of real-world data, we can solve real-world problems for a safer, healthier world."
—by Bill Eidson
Join our community of innovators, learners, knowledge-sharers, and risk-takers: View Job Openings.