In July 2008, Synergy Health in West Bend, Wisc. officially joined up with Froedtert Health, a large system based out of Milwaukee.
Adding Synergy, comprised of St. Joseph’s Hospital and the West Bend Clinic, stretched the system’s area of influence 30 miles or so to the north. It helped grow Froedtert into its current permutation, which includes Eastern Wisconsin’s only academic medical center, two community hospitals, and about 30 clinics.
The move also began Froedtert down a long path towards electronic medical record (EMR) integration. Instead of the traditional practice of bringing two different vendor EMR systems together, Froedtert had to integrate two different versions of an Epic EMR. Easy enough, right?
On the contrary, according to those involved, including Froedtert vice president and CIO, Bob DeGrand, the integration was a somewhat arduous process, and an unprecedented one at that.
“As we got into it, we realized Epic had done [a conversion from] IC Chart [InteGreat from Med3000] before, they had done Cerner-to-Epic conversions, they had done McKesson-to-Epic conversions. They had done those before, and they do them well. They hadn’t done Epic-to-Epic before. That was the area where they were least experienced in. It was a lot more work to do,” says DeGrand, who assumed the position of CIO in January of 2009, a few months after the West Bend affiliation became official.
We realized Epic had done [a conversion from] IC Chart [InteGreat from Med3000] before, they had done Cerner-to-Epic conversions, they had done McKesson-to-Epic conversions. They had done those before, and they do them well. They hadn’t done Epic-to-Epic before.
Why Turn Two into One
The conversations about merging the two Epics began shortly after the West Bend organizations were brought on board, according to Michael Sura, executive director of the EMR program and project management office at Froedtert.
There were four primary considerations when leadership made the move to one single platform across the enterprise. They looked at the cost and resource side to this kind of move, as well as the risks related to bringing converted data into an actual production database. On the positive side, DeGrand says they looked at clinical and convenience values. Ultimately, the positives outweighed the negatives and the integrated EMR fit the system’s strategic plans, DeGrand says.
“For the benefit of our patients, more than anything, we needed to make sure we had one single EMR platform so all the caregivers could have that immediate access to all of the relevant clinical information,” DeGrand says.
It wasn’t until 2012 that the integration actually began. Sura, who headed the initiative along with project manager Katie McCarthy, says other organizational priorities took hold.
Once they were ready to integrate, before they could even begin, Sura says they had to determine which data elements would be moved and if those elements could be converted safely in an electronic format. They then, with assistance from Epic, conducted a risk assessment for each data element. Data elements were moved or not moved based on whether or not there was risk involved with a potential conversion. If data wasn’t moved, it was either going to be pulled manually or clinicians would have to refer to the old West Bend Epic EMR.
The next step, after those decisions were made, was to create a game plan and timeline for those data elements that would be converted electronically. From the get-go, the team knew there would be various technical roadblocks they’d have to encounter.
“The first thing you have to do is make sure you can get all the patient records over from the older system. Because we share a lot of patients, one of the first challenges was the number of duplicates we found among the different databases. [Our first challenge was] how do we merge these patients so they have a single patient record? Technically, the biggest challenge there was just the volume. We were looking at approximately 300,000 merges that needed to be done within a four-week period,” McCarthy says.
We were looking at approximately 300,000 merges that needed to be done within a four-week period
With the sheer volume of data Froedtert had to convert, not just to avoid duplicates, but overall, project leaders at Froedtert made the executive decision to just do three years worth of patient history. This caused backlash from some of the clinicians, who said they should have done more. However, to do more, would have added a lot of time to the project, Sura says, and adhering to the timeline was a huge focus for the organization, which also had to manually abstract the data elements that wouldn’t be moved electronically.
In fact, they both say, adhering to the timeline in general presented a whole issue in its own right. The Epic to Epic conversion was part of a larger project to get the entire health system on one EMR system. Before that, Froedtert underwent another conversion, this one a non Epic EMR to the Epic system.
“The timeline had a lot of layers and if one thing was missed, it influenced five or six teams,” McCarthy says.
Another challenge was just the overall differences in the two versions of Epic, McCarthy says. For instance, something as basic as the tables which list the state from where the patient lives was different. They had to go through and make sure they mapped clinical equivalents between the two EMR systems. To do this, McCarthy says they used both IT and clinical team members.
All those who were involved with the Epic-to-Epic conversion convey the importance of having one “leadership voice,” on an extensive project such as this. According to Sura, when something touches the administrative, IT, and clinical side, everyone is bound to have an opinion. If issues arose, and they did, there was a clear path to figuring out a decision.
Those on the West Bend/St. Joseph’s side had grown accustomed to using their version of Epic, which was a lot more customized. Pushing the more standardized version of Epic, he says, wasn’t well received. And while there are still some clinicians from West Bend/St. Joseph’s who lament the loss of their customized EMR, Froedtert was able to get its standardization message across and get them involved with the data conversion.
While the process is still ongoing (manual abstraction will take a few more months to complete, says Sura), already the organization is reaping the benefits of one integrated system across the enterprise. For one, the large amount of patients that move between the various sites can now rely on one record. The providers that treat them don’t have to log into two instances of Epic to get the full picture. More than that though, DeGrand says the integrated EMR improves patient care.
“The perspective we take is that the more a physician can do while in the EMR, if we define the EMR to be a patient care tool, if they can do more within that without having to go out to another system or other data that resides elsewhere, the time they spend with the patient is just that much more effective,” DeGrand says.
“Yes it’s a convenience factor for the physician, but it’s also a quality of care for him and the patient. The patient isn’t left waiting for the physician to get to some results or get to an image. If we pull everything into one EMR, it’s just better for patient care,”