While radiology has been a leader among all specialties in adapting to the digital age of medicine, a nationally recognized leader in the field argues that the second generation of the digital age is among us, and with it some conceptual changes.
James Thrall, M.D., of the department of radiology at Massachusetts General Hospital, writes in the March issue of the Journal of the American College of Radiology that the first generation of the digital age should be referred to as the "data-in" era, in the sense that image data, demographics, patient metadata and administrative data have been converted into digital form and stored in huge digital data bases.
One of the overarching concepts of the next generation of the digital age, Thrall says, will be making a transition from "data-in" to "information out."
Data stored in radiology information systems, PACS and electronic medical record systems are "dumb data," according to Thrall, in the sense that it's up to users to integrate the data and extra value from them.
"The focus of the next 20 years will be turning dumb data from large and disparate data sources into knowledge and also using the ability to rapidly mobilize and analyze data to improve the efficiency of our work processes," he says. "The catchphrase 'big data' is now being used in connection with the potential for data mining and knowledge extraction from the large data sets that many enterprises have accumulated."
An example of data mining and knowledge creation, according to Thrall, is the fusion of decision support with computerized physician order entry.
"Providing appropriateness criteria in the context of the ordering process creates the paradigm of just-in-time knowledge delivery at the point of care," he writes. He adds that the power of the paradigm lies in the fact that hundreds of different procedures are offered in radiology departments, and that there are thousands of possibilities for performing them. It's impossible, he says, for a single physician to keep track of all the possible permutations.
At Massachusetts General Hospital the response has been different big data innovations like the Querative Patient Inference Dossier (QPID) knowledge extraction program. Developed over the last five years by Michael Zallis, an associate professor at Harvard Medical School, and Mitch Harris, a research scientist at MGH, the system is specifically designed to mine data from EMRs.
At MGH, the software is used in the emergency room to simultaneously display on a screen more than 50 EMR data elements so that an attending physicians doesn't have to waste time sifting through databases looking for relevant information. This innovation has actually transitioned into a commercial venture with the recent formation of QPID, Inc., according to the Boston Globe.
Other big data applications are being developed around the country, as well, Thrall points out, including automated alerts for important findings and stat cases, dashboards for aggregation and display of key business intelligence data, and real-time issues reporting for quality monitoring.
By recalibrating their thinking from "data in" to "knowledge out," Thrall says, radiologists will realize "enormous" benefits in terms of safety, quality, costs, and stakeholder satisfaction.