In healthcare, the movement of data can be critical to patient outcomes. Accurate, up-to-date patient information has to be accessible both within a healthcare organization and to partner organizations. For years, requesting medical records has been a cumbersome, manual process that can take weeks, even after the shift from paper documents to electronic medical records (EMRs).
“They definitely succeeded in digitizing the record, but sharing the record is still very challenging,” says George Rosello, director, digital transformation at Jackson Health System. “Ultimately, faxing is still the easiest and cheapest way for people to do it.”
The challenge is especially acute for an organization like Jackson Health System (JHS), one of the largest public health systems in the country. JHS is a nonprofit academic medical system in Miami, Fla., which runs six hospitals, including Jackson Memorial Hospital (the safety net hospital for Miami-Dade County), a network of urgent care centers, multiple primary and specialty care centers, long-term care nursing facilities, corrections health services, trauma centers, and the Miami Transplant Institute. It also works with a multitude of referring organizations, many of them private practices.
“When outside organizations send a patient to Jackson, they end up with an incomplete medical record on their side because that instance of care took place within Jackson. We documented it within our EMR, but that’s not something that exists in their EMR, the home EMR of that patient,” Rosello says.
To resolve that, a referring organization typically contacts JHS’s medical records department and requests a medical records transfer. Like many large hospital systems, JHS is sees high volumes of medical record requests and can sometimes take weeks to fulfill a request.
“It’s a very manual process and usually ends with somebody faxing or mailing things that need to be sent over,” Rosello adds. “That process is very cumbersome and very people-driven.”
Automate and simplify
JHS typically electronically faxes medical records to referring organizations. Because the documents are printed from the JHS health record, they have different identifiers than the referring organizations use. Those referring organizations are often small practices, and so are frequently limited in their technology tools and must manually perform a patient match.
To address this data challenge, Rosello and his team have created the Medical Record Surveillance and Record Sharing project, an in-house solution that scans medical records in real-time, matches them to partner organizations, and automates the medical record transfer. The project has earned JHS a CIO 100 Award in IT Excellence.
“We wanted to really simplify that process,” Rosello says. “We built an automated process that sits between Jackson and these other organizations. To start, we partnered with an organization that was sending us patients but facing some of these frustrations.”
The result is a transformed process as much as a technology solution. It starts with referring organizations securely sending a daily flat file that contains their patient census and some back demographics. On JHS’s end, the surveillance platform constantly monitors all medical records as they are created—including unstructured notes—and sends them through an integration engine. The engine runs a patient matching algorithm to match patients against the census of the referring partner.
Once a match is identified, the surveillance platform sends an electronic document in HL7 format, an international standard for transferring medical information, to the integration engine for processing in real-time. The engine extracts the results from the message and reconstructs it using easy-to-read templates stored as objects in the JHS data model. It then updates the medical record identifiers in the object to reflect the identifiers of the referring organization so they can easily file the record in their system.
Meeting providers where they are
As it is being processed in the JHS system, the medical record is sent to the referring organization by the method of their choice in real time. It currently supports files, direct messages, and a custom process to send an electronic or physical fax.
“If we get a referral from an outside organization, we’re using their medical record number in the report that we send back to them, so they know exactly who it’s for,” Rosello says.
“They’re getting it in real-time because we fax it as we find the documents being created,” he adds. “If I’m a provider and I just signed off on that document, however long it takes to fax something, is how long it’ll take for it to get there. We joke that it’s one of the most high-tech low-tech solutions we’ve ever built because it still uses an electronic fax.”
Initially, Rosello’s team just planned to send records via electronic messaging or secure FTP, but ultimately, they had to meet referring organizations where they are.
“Most of these organizations prefer the fax,” Rosello explains. “Healthcare has not gotten over the fax yet. It’s still one of the most widely accepted and easiest points of access that is acceptable to send patient information for organizations. There’s lots of capabilities that we have, but some of these groups are smaller. Some of them are just private practices that have a large referral base of patients, and they don’t have the infrastructure to support an HL7 interface, they don’t have an interface team, they may not even have a full-fledged EMR.”
Addressing technical challenges
Rosello notes that designing the transformation of the HL7 message to an elegant and easy-to-read document and developing the patient matching algorithm was one of the biggest challenges of the project.
The HL7 transformation involved over a dozen custom-written libraries to handle the variety and formatting of the HL7 message. The team ran nearly six months of medical records as a test to ensure they had accounted for every conceivable permutation of HL7 and that the document formatting was compatible.
For the patient matching algorithm, Rosello’s team simulated hundreds of thousands of variations in all of JHS’s data elements to select a reliable combination and threshold for matching accuracy. They also included fail safes in cases where matching was ambiguous, so JHS doesn’t overshare if the algorithm isn’t certain.
Reaping the benefits
The project, which was started in 2020 and is now fully deployed, has produced measurable benefits to the organization. Rosello points to three:
- Productivity savings for medical records staff. By automating this process, the project saves the medical record staff more than a dozen hours per week.
- Time savings and increased efficiency for referring organizations. Referring organizations now receive records with the correct identifiers to make them easy to file, and JHS also now sends all records pertaining to the patient, which reduces or even eliminates gaps in the medical record.
- Increased referrals and revenue for JHS. Organizations are now more eager to work with JHS and Rosello says there has been a dramatic increase in referrals, including new organizations that started working with JHS because they heard about the solution.
“I think the biggest takeaway is incorporating good design in healthcare, not depending on vendors to design things for you, but rather taking responsibility for your own internal design of how you want solutions and how you want your technology landscape to work,” Rosello says. “As we continue down this journey in healthcare digital transformation, I think one of the biggest components is going to be good design and making sure that you own your design; you’re not outsourcing that to other players that have different objectives than you.”
