Teamwork called necessary for EMR clinical research success

Washington -- Physicians and researchers say they have the potential to make enormous strides in medical care by marshaling the unparalleled amount of clinical information tucked into tens of millions of electronic medical records. But that promise hinges on private stakeholders and the government getting together on a way to share the data.

That was one of the main themes of a Jan. 26 seminar sponsored by the journal Health Affairs that explored the push to implement a "rapid learning" health system. By allowing physicians quickly to see real-world data on how well certain medical procedures and products work, such a system vastly could expand the medical community's understanding of how best to treat even the worst diseases, seminar participants said.

Until the educational potential of millions of EMRs can be harnessed, patients will continue to suffer from major gaps in clinical knowledge, said Lynn Etheredge, a consultant with George Washington University's Health Insurance Reform Project in Washington, D.C. He is author of the lead report in a special edition of Health Affairs that corresponded with the event.

Even physicians who are as diligent as possible about keeping up with the more traditional studies will not be able to keep up with medical technology advances in the field, Etheredge said. "The rate of growth of technology is going faster than the knowledge of how to use it."

A rapid learning health system would provide the opportunity to compare the benefits and risks of treatment options much more quickly and inexpensively than the current system, Etheredge said.

A typical randomized controlled trial for a medical procedure or technology can run as long as a decade and cost as much as $300 million, said David Eddy, MD, the medical director and founder of Archimedes Inc. His San Francisco-based firm aims to improve health care quality and efficiency through complex mathematical modeling of data.

"You wait 10 years, you spend that money, and you have one answer to one question," he said.

Relying on clinical trials instead of vast patient record databases to determine which treatment options are the most promising also fails to address the needs of entire groups of patients, Etheredge said. Typical studies use younger, relatively healthy subjects and largely exclude the older, sicker Medicare and Medicaid patients who could benefit from an entirely different treatment plan.

Physicians participating in the seminar stressed that clinical trials, especially those testing unapproved drugs, always will have a place. But finding out what works best once medical products or technologies start getting widespread use is an area in which rapid learning can fill an important niche, they said.

The government's role

The federal government is keenly aware that EMRs could be one of the most valuable physician resources if the information is shared appropriately, said Agency for Healthcare Research and Quality Director Carolyn M. Clancy, MD, who addressed the seminar. "Everyday experience is the best educational tool," she said.

In recognition of this, President Bush included $15 million in his fiscal 2008 budget request to help large medical centers share their clinical databases over a network. Dr. Clancy said the administration also will continue to drive a public-private effort to develop EMR standards so that information can be shared more widely while maintaining patients' privacy.

Universal interoperability standards will be necessary for developing a rapid-learning health system as data become more abundant and less expensive to gather, said Paul Wallace, MD, medical director of Kaiser Permanente's health and productivity management programs.

"We're going to have to address the Babel issue," he said.

Progress so far

For some physicians and researchers, the future is already here. Rapid learning is already being put into place in a number of public and private health arenas with promising initial results.

The Veterans Health Administration has crunched the electronic medical record data for hundreds of thousands of patients with diabetes to come up with potential clinical insights, said Joel Kupersmith, MD, the VHA's chief research and development officer.

The administration found a high rate of mental illness among diabetics, for instance, and is exploring how the weight gain associated with newer psychotropic drugs might be contributing to diabetes.

On the private-sector side, 40 million patients enrolled in 14 HMOs are providing the clinical backbone of a rapid-learning system focusing on cancer research and vaccine safety, Etheredge said.

In both cases, the programs' architects are heavily involving physicians and clinical researchers in the process from the earliest stages in an effort to ensure success.