Your medical history—and the histories of patients with similar conditions—can help you.
That's the foundation of a RENCI project to build an easy-to-use data analysis and visual dashboard to help doctors quickly determine the best treatment options for their patients.
The project teams Chris Bizon, RENCI senior research scientist, Ketan Mane, RENCI senior research informatics developer, and Charles Schmitt, RENCI's director of informatics, with Dr. Kenneth Gersing, a psychiatrist and medical director of clinical informatics in the psychiatry department at Duke University Health Center, and Bruce Bruchett, a statistician in the Duke psychiatry department.
Gersing is the mastermind behind MindLinc, a tool for managing, evaluating and improving behavioral health treatment used at Duke and in medical clinics across the U.S. MindLinc includes anonymous emergency medical record (EMR) data accumulated from psychiatric patients over 20 years. As with any large data set, quickly analyzing data and translating the analysis into better patient care poses many challenges.
"The EMR data is there. It represents patients from multiple sites and across time," said Mane. "We are developing a way to find important information in the data quickly and to view the results in a way that is easy and quick to understand at the point of care. Doctors don't have time to pore over pages of spreadsheets."
Using the RENCI dashboard, a clinician can compare one patient's medical history—for example, a 40-year-old male with depression—to the EMRs of other patients with similar conditions and histories and with similar demographic characteristics. From the dashboard on their computer screen, the clinician can examine graphs that show which medications have been most effective in treating similar patients.
The dashboard also presents graphs that map the effectiveness of different treatments over time and show the patient's medical history over time, including doctor and emergency room visits. By examining the visual data, the clinician can draw conclusions about what triggered an emergency room visit or what medications are most likely to remain effective over time.
The clinician also can change the parameters for comparisons as needed. If, for example, race has little relevance to a particular disease or to the effectiveness of a medication, the clinician can tell the system to disregard race when presenting the data. If the clinician suspects that time since the initial diagnosis might factor into the effectiveness of a treatment, that variable can be included in determining results.
"It's similar to when you buy a book on Amazon.com," said Gersing. "You buy a book and you see a note 'people who liked this also liked …' They are making predictions based on your past behavior.
"In our case, we're not trying to get you to buy a book. The bottom line is more effective treatments, lower costs, and healthier people," he said.
Gersing hooked up with RENCI through Ricardo Pietroban, vice chair of the department of surgery at Duke University Health Center, who knew Mane and his work and thought that RENCI's informatics expertise could enhance Gersing's efforts to use EMRs to improve medical decision-making.
"RENCI has been extremely helpful, not only by providing a visual interface but also the analytics," said Gersing. "To use the data, you first need to pick out the important characteristics from the background noise."
A prototype of the RENCI visual dashboard is almost complete and will be tested on the full MindLinc database later this year. The tool will then be tested in clinical settings in the psychiatric department at Duke Health Center and refined based on feedback from doctors and other clinicians.
Charles Schmitt, RENCI's director of informatics, sees the project as an example of how RENCI can help the medical community translate research data into better patient care.
"There are guidelines that doctors follow when determining how to treat a patient, but they get dated pretty quickly," said Schmitt. "Analytics and information visualization is an intuitive way to put data to use and have immediate results for the healthcare system."
Professor Nancy Ip: Pioneering New Paths in Neurodegenerative Therapy