At Yale New Haven Health in 2010, the health system had five different electronic health record systems and variance in its common revenue cycle platform implementations.
Like many health systems, Yale New Haven Health was in the process of consolidating on a single electronic health record and revenue cycle platform. It needed to reduce the budgetary waste of maintaining multiple applications, but also understood the critical need for access to data used in clinical care, compliance, billing and audits.
While the strategy for the future was to have a single electronic health record and revenue cycle platform, cleaning up all the legacy data was a significant challenge. Many health systems are still struggling with these issues today.
“Our first step was simply to keep the legacy systems available in a read-only format so the important information could be readily accessed,” said Lisa Stump, senior vice president and CIO at Yale New Haven Health.
“On the revenue cycle side, we had to maintain limited access to those who needed it in order to continue processing transactions,” she explained. “Running out the old accounts receivable is a very real need, but there really wasn’t a long-term option to maintain ready access to incredibly important medical information that was in the legacy systems.”
Yale New Haven Health came upon the MediQuant DataArk product, which created an easy user interface that the health system could integrate directly into the electronic health record for access to the clinical information, Stump explained.
“The product also simultaneously allowed transactions to be performed on the revenue cycle side on the back end,” she said. “The vendor provided us with a strong project management team that helped work through all those challenges, proving to be an exceptional team to work with us through a complex process.”
Managing legacy data may not be the most exciting topic in healthcare or IT, but this is a bread-and-butter, block-and-tackle piece of work that health systems must do, and many organizations struggle with it, Stump said.
“All the free text notes aren’t necessary, but the discrete data, like allergy information problem lists and medication lists, are all brought over. Along with lab and imaging results.”
Lisa Stump, Yale New Haven Health
“Our decision to use the MediQuant product served us well in transition and is still serving us well today,” she said. “And the cost savings were real in allowing us to shut down those legacy systems.”
Yale New Haven Health first came across the MediQuant product at a HIMSS conference.
“I understood the problems we faced, but I had yet to see anything that offered a solution,” Stump said. “A colleague on my team had just come from a booth at the conference and told me to go see the demonstration. That first demonstration met my first gut-check that it could do what it claimed.
“We brought the team out and worked through the plan in detail, had our end users – the clinicians who would access the data on the front end, my technical team, our finance team, the medical records and health information management team – all kicked the tires, which convinced us it was worth the try.”
There are many systems integration tools on the health IT marketplace. Some of the vendors of these tools include Corepoint Health, Datica, Health Catalyst, InterSystems, Jitterbit, MicroHealth and Orion Health.
MEETING THE CHALLENGE
Taking different data structures out of different legacy systems and trying to put them into a single common format that makes sense across a broad base of users created multiple challenges.
“We had created a single electronic health record for the go-forward strategy, that’s Epic for us, and using Mediquant, we created that single, continuous record, even back into the legacy world,” Stump explained. “So if you are in a patient’s EHR today, you can click one button, Pre-Epic EMR, to view important legacy data and also know whether that data had existed in Cerner or Meditech or any of our other legacy systems, who entered it and when.”
So the clinicians can make the right interpretation about that historical data relative to the patient’s current state, she added.
“We asked ourselves if we needed the active archive or if just keeping the data available in some legacy server somewhere would be adequate,” Stump recalled. “That’s where clinical data drove the answer – not losing that patient history. We had experienced a real example in our institution where challenges in accessing that historical data interfered with the best outcome for a patient.”
The need was very real, she said. So after being convinced this product could provide the active access for the best care, staff moved forward, she added.
Integration with other systems was key. Yale New Haven Health made it viewable from within the Epic medical record, spending significant time focused on the importance of making the vital information easily accessible in the clinicians’ workflow.
The archived clinical data is used by physicians, nurses and full care teams to see that view into the patient’s past medical data. A busy physician or nurse going out to a separate system, even if it has a single sign-on capability, is not a reasonable expectation, Stump contended. Making that legacy data, the archives, completely viewable and accessible from within their natural workflow within the Epic EHR, was a key selling point.
“Also, maintaining the important pieces of the legal medical record was imperative so we needed to make sure all that was kept intact,” she said. “That has proven to be very useful and valuable.”
To be clear, Yale New Haven Health did not take all the data that existed in the legacy systems. It did not believe that was necessary. The governance board was involved, but the same stakeholder groups (clinicians, the finance team, the medical records team, etc.) worked with IT to identify which data elements were important and that is what was transferred into MediQuant’s DataArk.
“For example, all the free text notes aren’t necessary, but the discrete data, like allergy information problem lists and medication lists, are all brought over,” she explained. “Along with lab and imaging results.”
For the finance team, with focus on the revenue cycle, transacting against the data was needed to manage accounts receivable and ongoing payment cycles. That data was more clean-cut based on that need and the ability for them to manage from a single platform for those old transactions was a big win, Stump said.
“As mentioned, this process is very complex and touches on many functions and users throughout the organization,” she said. “And no project of this scope transforming so much information is without its bumps. That’s why the partnership between the vendor and all of our teams is what makes a project like this successful.”
Stump was clear at the outset that she really wanted a vendor partnership, not just a vendor, and was met with a lot of enthusiasm and agreement around the approach.
“As part of that, during the transition, we worked with MediQuant to test and stage the data ahead of migrating it into the database,” she added.
Yale New Haven Health focused on two goals: cost savings and accessibility for clinicians.
“Financial savings are important,” Stump stated. “Every health system and IT division is looking to reduce its cost of ownership. In our case, legacy costs before the transition reached $2,980,000 annually and the post-project costs run $326,000, so we achieve more than $2.6 million in annual cost savings. Additional savings include the net hardware refresh and data center costs.”
But Stump is a clinician by background, a pharmacist, so ease of accessibility to important information for patient care was a key driver in the strategy.
“That proves out in patient care every day,” she reported. “It’s been an incredible success and it has brought real cost savings. We ended up with what I think is the best of both worlds.”
ADVICE FOR OTHERS
“First, ask the right questions,” Stump advised. “Think about what type of data you are trying to manage and how it will be used. Does it need to be accessed within another workflow? Is it needed with some degree of frequency? If so, this active archive model is very well suited. We certainly don’t need an active archive for all the data we store but seeking a balance of need for a specific data set is important.”
For the select data types that meet those key questions, active archiving is very important, she added.
“Second, and perhaps the single best piece of advice I can offer, is to include all stakeholders in the process: clinicians, finance, medical records and tech – all teams need to work in partnership,” she stated. “And MediQuant was at the table with us – with a project manager and technical support team that worked side by side with our teams.”