Obama and Congress Inadvertently Created Obstacles to Data Sharing, Biden Acknowledges—Now What?

Share on facebook
Share on twitter
Share on linkedin
Share on email
Share on print

This article is part of The Cancer Letter's To The Moon series.

Seven years ago, when Congress sought to jumpstart the U.S. economy, few imagined that one aspect of the $800 billion stimulus program would turn electronic health records into the Tower of Babel.

Speaking at Memorial Sloan Kettering Cancer Center last week, Vice President Joe Biden took full responsibility for the major bioinformatics snafu triggered by the American Reinvestment and Recovery Act of 2009.

True, ARRA forced the U.S. medical establishment to digitize records in a hurry. Alas, today, fax machines and legions of data entry technicians are often required to transfer patient records from one hospital to another.

Biden said the Obama administration didn’t foresee the consequences of mandating the switch to EHRs without developing a standard infrastructure for aggregating data.

“When we wrote [ARRA], during the interregnum time between the time we were elected and the time we were sworn in, the president and I sat alone out on that 67th floor of a Chicago office building, putting together a new administration,” Biden said at the May 26 roundtable discussion at MSKCC.

“We had the leading economists in the country, from the Larry Summerses to the former heads of the Fed, talk about the need for a massive stimulus, but we decided to try to use that also to change the way we governed and to do it more effectively,” Biden said.

“So one of the things we did, everybody talked about electronic recordkeeping for docs and it being able to be easily transferred, and we came up with $35 billion in that legislation—before the Affordable Care Act—to accommodate doctors being able to transition to electronic records. We thought that we knew what we were doing, but what we didn’t realize is that we essentially put up five different firewalls.

“All $35 billion was spent, but the ability to communicate, aggregate that data, unlike the human genome, [there was] no agreed-to vocabulary—oh, you’ve got a broken femur, I’ve got a broken leg—the computer doesn’t pick it up.

“So the big debate has been, what do we do? Do we get docs and medical schools to start to do—which is very difficult—common definitions of common language so we can begin to be able to transfer those records and know what’s in the records?”

Biden’s remarks appear here.

With just 30 weeks remaining before the National Cancer Moonshot Task Force submits a final report to President Barack Obama, Biden is racing against the clock to solve this important issue that is compromising patient care and progress in cancer research.

“The task force you have assembled is focusing on streamlining cancer research, exploring how to best facilitate combination therapies, and how to reduce the barriers to data sharing and analysis,” MSKCC President Craig Thompson said to Biden at the meeting. “But we also need to continue to improve, to improve access to care, to make the latest diagnostic procedures—particularly DNA testing—more widely available, and to make cancer care more patient-centered.”

Clifford Hudis, the incoming CEO of the American Society of Clinical Oncology, likens the moonshot data-sharing debate to the technology revolution two decades ago.

“In the early 90s, everyone was looking for productivity gains and people were having a hard time describing it with respect to the PC movement,” Hudis said to The Cancer Letter. He currently serves as vice president for government relations, chief advocacy officer and chief of the Breast Medicine Service at MSKCC.

“But something quietly transformed in the workplace. The arguments about productivity gains have gone away. Now, no one doubts that there is speed and agility with the technology.

“We’re kind of in that valley of death with electronic records. Practitioners are spending an insane amount of time trying to figure out how to use these records. It is demoralizing. But out of that frustration will come creativity and progress. In the future, we won’t be sitting here. We will get to a better place.”

The moonshot, announced by President Barack Obama Jan. 12, aims to conduct a decade’s worth of cancer research over the next five years—primarily by breaking down data siloes and facilitating the creation of a central bioinformatics database for oncology (The Cancer Letter,Jan. 22).

The administration’s proposalestablishes a game plan for spending $1 billion in mandatory funds: the moonshot initiative will begin with $195 million in cancer research at NIH in fiscal 2016, according to the White House.

Though initial funding is relatively modest by comparison with the overall federal spending on biomedical research, the moonshot is shaping up as a broad-based research and public health initiative.

The administration’s budget proposal for the 2017 fiscal year would allocate $755 million in mandatory funds for new cancer-related research activities—$680 million for NIH and $75 million for FDA. The remaining $50 million is expected to fund Centers of Excellence in the Departments of Defense and Veterans Affairs.

“Fam. Hx.”

It’s plausible that in February 2009, lawmakers and the administration didn’t fully fathom that they were creating obstacles to interoperability.

Creating a common dictionary for all EHR systems would have been a daunting—perhaps impossible—task at the time. Instead, the federal government simply required the universal adoption of EHRs, ultimately forcing the medical field to confront the urgency of developing common standards.

This is not the first time such things have happened in the evolution of technology.

To be fair, in the late 1980s, the technology revolution did not begin with a comprehensive guidebook for all the uses of the Internet and IP addresses. The Internet materialized, and it continues to evolve.

In today’s Wild West of data, it shouldn’t be taken for granted that even EHRs produced by the same vendors are able to exchange data.

Some systems are unable to talk to one another because of sheer semantics: what one database labels as “Family History” may be synonymously recorded in another as “Fam. Hx.”

Experts say that these problems arose because of ARRA’s blueprint—or lack thereof—for electronic records: the “Meaningful Use” regulations. These regs are a part of the Health Information Technology for Economic and Clinical Health Act, an element of the ARRA.

Under the act, providers have to show that they are “meaningfully using” their certified EHRs by meeting measurement thresholds that range from recording patient information as structured data to exchanging summary care records. The Centers for Medicare and Medicaid is responsible for implementing the EHR incentive programs.

Meaningful Use does not provide a framework for how data should be aggregated, said Jonathan Hirsch, founder and president of Syapse, a company that integrates oncology data from EHRs with genomic data to enable precision medicine.

“At its core, Meaningful Use incentivized the use of technology by providers, not the achievement of any improvement in healthcare operations and outcomes through technology,” Hirsch said to The Cancer Letter. “While the government mandated the use of technology, they did not mandate any substantial data standards for what these clinical systems needed to store and open interoperability frameworks to make sure that data can flow freely.”

On June 2, Syapse launched the Oncology Precision Network, which enables interoperability between 79 hospitals and 800 clinics across 11 states, Hirsch said. Stanford Cancer Institute, Intermountain Healthcare and Providence Health and Services jointly announced the network, which currently has 100,000 patients.

“Syapse has succeeded where many others have failed, which is breaking down the silos to integrate clinical and molecular data in a software framework that provides value to physicians at point-of-care,” Hirsch said. “OPN is different in that its primary goal is to drive improvement of patient care today using aggregated cancer genomics data, rather than focusing on research or practice management.

“Through the network, we are leading the way in the clinical implementation of precision medicine, turning the Cancer Moonshot’s vision of improved cancer care into a reality for patients in the community.”

2016 Data Olympics

Syapse, along with other major players in oncology bioinformatics—ASCO’s CancerLinQ, M2Gen’s ORIEN, AACR’s Project GENIE, NCI’s Genomic Data Commons, and Flatiron Health—are vying to solve the interoperability problems, thereby assuming the leading role in Biden’s moonshot.

It is unclear whether the moonshot task force would eventually endorse individual initiatives. If it does, the winner or winners could dominate data-sharing standards on a global scale.

At the moment, Biden appears to be focused on getting these organizations to collaborate.

“I know MSK is part of Project GENIE and part of data-sharing coalitions. But I’ve met also with ORIEN, I’ve met with CancerLinQ,” Biden said at the MSKCC roundtable discussion. “You’re all doing the same thing! I find it curious.

“I don’t know if that’s the only way it that can be done. What happens if it was all aggregated in one place? I don’t know, so with that, why don’t I hush up? I’m going to leave it to you, [MSKCC President Craig Thompson], to call on whoever has any input here—I’d like to hear from everybody—and be more educated as to how we can move things more rapidly.”

Biden turned to Nat Turner, co-founder of Flatiron Health, who was on the panel at the MSKCC discussion: “Do we go to Flatiron? Have you solved the problem yet, Nat?”

The big problem is science isn’t moving fast enough, Turner said.

“There’s all this data out there, and we’ve had to raise hundreds of millions of dollars to liberate it, because we have teams of almost 300 people, engineers from places like Google who still can’t figure out how to make sense of all this data,” Turner said. “But what the Recovery Act did is it got all of this stuff in electronic format, this was a huge step forward.

“We are making strides as an industry, where we have armies of people and machines that are learning what these PDFs and scanned images say. When it says, ‘Mrs. Smith has this genetic mutation, she responded to this therapy or didn’t respond to this therapy; here are the side effects that occurred,’ that’s actually happening, and it’s super exciting.”

On May 25, Flatiron announced a research collaboration agreement with FDA to determine how real-world evidence derived from de-identified, HIPAA-compliant patient data captured outside of clinical trials can provide new insights into the safety and effectiveness of emerging anti-cancer therapies such as immunotherapeutic agents.

“We, too, want to know why only 20 percent of patients are responding,” Turner said at the MSKCC event. “No one institution, including the one we’re sitting in today, has enough patient data to figure that out. So an aggregated data set is what you need.

“We just looked through our database and found tens of thousands of patients who received those drugs, and we have linked data from genomics, and other sorts of tests that we think might lend some sort of signal as to why are those 20 percent of people not responding and can we target them.

“So I think the good news is that industry—in large part a lot due to your efforts—are very open to data-sharing. I think we’re actually past that. In years past institutions treat their data like intellectual property even though the patients own it. We’re largely past that.”

Katie McKinney contributed to this story.


Associate Editor