Modifying the EHR to ease research

When Dr. Demetrius Maraganore left the Mayo Clinic in 2009 to take over as chairman of NorthShore University HealthSystem's neurology department, he brought more than a decade's worth of data from his National Institutes of Health-funded study of Parkinson's disease treatment. He hoped to continue his work in evaluating quality-improvement measures.

At Mayo, research team members hand-wrote patient information on 72-page forms during doctor visits; phone interviewers hand-wrote information on a 25-page evaluation form to assess risk factors. This required hiring a large team of study coordinators, data-entry clerks and others to help collect information. He expected that would be necessary at NorthShore too.

But his new colleagues suggested an alternative approach—configuring NorthShore's electronic health-record system to make it easier to gather needed information for research purposes during the patient-care process. The goal was to meld the historically separate clinical and research data systems and slash the added expense of duplicate data entry.

The Food and Drug Administration has said single-point data capture could produce more complete, accurate and timely data, speed discovery and dissemination of therapeutic advances, and boost the number of doctors willing to participate in clinical research.

“Every transaction in healthcare is happening in the (EHR),” Maraganore said. “If we can harness that technology to capture clinical information of the highest quality, that obviously would be very appealing.”

So Maraganore and the staff got to work to see if NorthShore's EHR system could collect patient data that would be useful for both clinicians and researchers. By June 2010, they had developed their first structured clinical document support (SCDS) toolkit, an application programming interface they built into the Epic Systems EHR. The system allowed clinicians to better track care for their Parkinson's patients and enabled researchers to collect patient data to track patterns related to the disease.

“That was kind of a eureka moment for us,” Maraganore said. “We realized that this same tool could be deployed for other neurological disorders.”

He and his colleagues expanded the project to develop toolkits for nine additional neurological disorders: brain tumors, epilepsy, migraine, memory disorders, mild traumatic brain injury, multiple sclerosis, neuropathy, restless leg syndrome and stroke. They added another project to assess risk factors for the onset of Alzheimer's disease.

The project involved members of NorthShore's departments of neurology and health information technology and its Center for Biomedical Research Informatics. The team worked to create an EHR system that would standardize office visits to capture patient data that could be used to develop patient safety and quality metrics. The toolkits are designed to capture up to 1,000 fields of data per office visit.

The team sought to create a data system that would be part of the routine clinical workflow. It was designed to limit the time for a medical assistant or nurse to collect data from a patient to 15 minutes; neurologists could take up to an hour to make their assessment.

Researchers used the toolkits to enroll patients in a DNA biobanking initiative. Based on the patient information entered, the system prompts the physician during the office visit to try to enroll the patient. An electronic notice is sent to the research team letting them know whether the patient has consented to participate in the trial.

As of April, eight SCDS toolkits have been used several thousand times during routine office visits, according to a study by Maraganore and colleagues published online in September in the journal Neurology Clinical Practice.

Maraganore couldn't give a cost-savings estimate for the initiative. But he said the project has improved the quality of care and data quality, and may reduce costs. Quarterly quality- improvement reports are generated to help identify opportunities to improve patient safety and outcomes.

NorthShore has expanded the project by offering the toolkits to neurology departments at other systems to share information, broaden research databases and encourage more research collaboration. “The work we've done provides a roadmap for how other neurology practice groups can … make clinically meaningful use of their (EHR),” Maraganore said.

James Hendler, director of the Rensselaer Institute for Data Exploration and Applications, said initiatives such as the one at NorthShore are the tip of the iceberg when it comes to the potential of EHR systems to ease research and improve quality of care. But there are lingering privacy issues that will require legislative remedies. “The potential for savings and to improve healthcare quality is huge,” he said.