How Precision Medicine is Unlocking EHRs' Potential

The prospect of incorporating human genomic data into electronic health records is closer than you think. Extensions to one system promise to improve patient safety, research, and clinical outcomes for 10 neurological disorders, research shows.

The healthcare IT industry has been vigorously touting President Obama's precision medicine initiative all year, ever since he announced it in the State of the Union address.

No doubt, IT's enthusiasm for precision medicine is fueled by visions of extracting even more IT-oriented revenue from payers and providers. The assumption is that great medical breakthroughs are just around the corner.

Some restraint is in order. With the continuing struggle to simply stand up electronic medical records systems and get them exchanging information, the prospect of incorporating human genomic data into the EHR falls into the category of bright, shiny objects of the moment.

But in September, IL-based NorthShore University HealthSystem published findings about extensions to the Epic EHR which promise to improve patient safety, research and clinical outcomes for 10 neurological disorders, brain health risk assessments, and interventions for Alzheimer's disease. Precision medicine plays a role, and more than 2,000 NorthShore patients have already signed up to participate in a DNA biobank to support the effort.

The study was posted in the September 24 issue of Neurology Clinical Practice, a journal of the American Academy of Neurology. The initial project received funding from the Agency for Healthcare Research and Quality.

Most impressively, seven healthcare systems in addition to NorthShore have joined together to form a Neurology Practice-Based Research Network (NPBRN), to share de-identified data and best practices and to help move the state of the art forward that much faster: Dartmouth, University of Pennsylvania, Wake Forest Baptist Health System, Medical University of South Carolina, Ochsner Health System, University of Arkansas, and the University of Nebraska.

Precision Medicine at the Point of Care
Leading all this enthusiastically is Demetrius "Jim" Maraganore, MD, medical director of NorthShore's Neurological Institute and director of the Center for Brain Health. "It's creating precision medicine at the point of care, through the EMR," he says.

Until now, like many other physicians, most neurologists have expressed frustration with EHRs, which seem designed to serve a broad spectrum of healthcare interests, including the meaningful use program, but not to add value for the practice of neurology, Maraganore says.

The problem was such that "on the one hand, the EMR is where medicine now takes place, and on the other hand, it really isn't designed with neurology in mind," he says.

Out of the box, today's EHRs do not allow neurologists to capture neurology-specific standardized information about their patients or to learn about their outcomes, Maraganore says.

Believing that the EHR represents a "Gutenberg moment" for all of healthcare, he decided to create neurology-specific EHR extensions for Epic. "We've been on a quality journey over the last five years that I think has succeeded in accomplishing just that."

The initiative started with Maraganore and his team meeting with neurologists to imagine "the perfect office visit" for a patient with the disorder they specialized in—"Parkinson's, stroke, multiple sclerosis, headache, brain tumors, or epilepsy."

He asked them questions:

• If you could spend as much time as you needed, what would you want to learn from that first visit with the patient?
• How would you define the patient's disorder? What are the outcomes that matter most to your patients and to you?
• What are the best ways of measuring those outcomes in the office, and do you have permission, parenthetically, to use those outcome measures?
• And then what factors do you know to independently associate or influence those outcomes, and how would you capture those independent variables also?

Those meetings yielded consensus to focus on conditions which were progressive, variable, and unpredictable. "That's where we have the most opportunity to learn and to improve," Maraganore says.

With the modifications to the Epic system, "neurologists can write very comprehensive, perfectly spelled, perfectly paginated, perfectly structured progress notes. And every click of the mouse would also capture data that could be analyzed to improve quality and to conduct practice-based research and to make discoveries."

Patient-approved and Carefully Vetted
Before such discoveries, NorthShore had to convince patients to enroll in the precision medicine portion, to give blood samples for extraction and storage. "As of today, 97% of our patients consent at the point of care," Maraganore says. "Of those, 88% say, 'you can use this not only for your own research, but provided you protect my identity, you can share it with any investigator doing any genomic research at any institution.'"

NorthShore's goal is to enroll 1,000 such patients in each of its 11 cohort studies. It's off to a good start: So far, it has DNA from 700 migraine patients, more than 500 Parkinson's patients, even more than 300 patients with restless leg syndrome, and so on.

It's reassuring to me that these EHR extensions have been carefully vetted with American Academy of Neurology quality guidelines and parameters, as well as common data elements from the National Institute of Neurlogical Disease and Stroke, and guidelines from subspecialty societies.

"We also appreciate the expertise of our NPBRN sites, so when they identify quirkiness in our tools… we invite them and encourage them to give us the feedback," Maraganore says. "If there's consensus across the NPBRN to make the changes, we make the changes, and then we disseminate the upgrades to all the participating sites."

Early learnings from the effort include identifying cognitive impairments in early Parkinson's patients, and alerting clinicians to check in with such patients about any concerns they or those close to them have about their safety while driving a motor vehicle.

"What we basically validated is the need for a history and a physical that both give a valuable perspective on the patient's illness and that are not necessarily measuring the same things. You really need to complement them both to understand what's wrong with your patient."

Can the efforts of neurologists also show the rest of healthcare how to make lemonade out of EHR lemons? It will require leadership such as NorthShore and its network partners are showing. Cooperation will yield faster progress than each academic center of excellence simply competing the old-fashioned way, not sharing information enough to move an entire profession forward.

The effort will also require NPBRN and others to extend beyond the Epic EHR. "Only about 40% of neurology practices are in Epic," Maraganore says. "We know that there is a need to extend our project from the Epic EMR. If we're successful in demonstrating that we can disseminate structured tools within the Epic platform to other sites, we can then go to neurology practices on the Cerner or Allscripts systems, or any of the other popularly-used platforms."

"[Then we can] say 'we will share with you our fields, forms and workflows. If you can build them into our EMR system, we would be happy to share with our content.'" He also points out that for deep analysis, clinical informaticists extract the data from the EHR and present it via a data warehouse, further suggesting EHR-agnostic platforms to come.

Judging by the patient biobanking participation rates in this case, it's also clear that today's slow pace of clinical trials could be supplanted in a few years by patient-driven, data-driven healthcare research, much as we are now seeing in other professions. I find Maraganore's recently published findings a reason for optimism.