In a May Becker’s Hospital Review webinar sponsored by Consensus, panelists discussed how these technologies are making important data more accessible and usable:
Vishwanath Anantraman, chief technology officer, Mayo Clinic (Rochester, Minn.)
Terry Jeffery II, vice president for technology, Ametras USA
John Nebergall, chief operations officer, Consensus Cloud Solutions
Four key takeaways were:
- Unstructured data is ubiquitous and problematic. Unstructured data makes up a big part of health data, estimated by Mr. Nebergall to represent approximately 80 percent of the clinical data within healthcare organizations. With unstructured data, it’s hard to search and even more difficult to transform that data into usable information. At the Mayo Clinic, and at many other healthcare organizations, medical records are still faxed. “We have a lot of challenges in processing the data and making it more easily usable for the physician, who is trying to review a history of five or 10 years across many providers,” Mr. Anantraman said.
- The reasons for the large quantities of unstructured data are many. Lack of high-quality provider directories is one issue. Fax machines remain in use in healthcare because all one needs is a phone number to send data, Mr. Anantraman said. EHRs, on the other hand, require much more interoperability in order to share information and often are not or cannot be equipped to do so. “The financial system has SWIFT codes and routing numbers, and that equivalent doesn’t exist in healthcare right now,” Mr. Nebergall said.
- New applications of AI and natural language are being used to successfully structure data. “The practical application of the AI is really quite simple,” Mr. Jeffery said. Imagine a fax with two patients’ data included. It’s easy for a human to discern that it’s two different people. It is possible to “have a system be able to make that determination, create a logical split and create two unique documents,” he said.
- To get started using the technology, define specific use cases. Billing optimization is one example. “Coding for comorbidities that may not have been captured in structured data and therefore did not show up in the bill is a potential revenue loss for the organization,” Mr. Anantraman said. Technology can be used to identify comorbidities, bill for them and generate increased — and appropriate — revenue.
COVID test result reporting is another example. An organization running 10 days behind in reporting — a critical lag during the pandemic — was able to reduce that time to seconds, which not only saves time and resources but provides immediate value.
Using these technologies to begin to structure data is not a small undertaking. But with the right use cases and an experienced partner like Consensus, organizations can provide value and ROI, reduce the burden of manual labor and bring to the fore the possibility of interoperability with a solution that isn’t costly.
To watch the webinar on-demand, get access here. To register for upcoming webinars, click here
