MultiCare has grown from 3 hospitals to 13, soon to be 18. Its interface team didn't grow anywhere near as fast. That gap between growth and integration capacity is the problem most growing healthcare organizations hit, and it's what eventually pushed MultiCare to stop building interfaces in house.
Every acquisition brings a legacy EHR with it. The data on that system has to go somewhere: converted into the destination EHR, or archived where it stays accessible for as long as the law requires. In Washington state, that's 26 years. Do that once and it is a project. Do it a dozen-plus times while running a health system and it is a capacity problem no hiring plan solves.
MultiCare's answer was to stop managing this project by project and hand execution to one accountable partner instead. ELLKAY supports data archiving across 486 healthcare applications and clinical archiving and migration across 293 EHRs.
Why Does Acquisition Growth Break Healthcare Data Integration?
Growth multiplies data sources faster than it multiplies the people who manage them. MultiCare had over 50 siloed data sources before consolidating, spanning pharmacy, imaging, radiology and every other system that collects its own information in its own format.
“There are so many different data sources, whether it's pharmaceutical or imaging or radiology data,” Shepard told Healthcare Innovation. “You have different systems that collect different things, and those systems aren't necessarily meant to communicate with one another.”
The staffing math is what forces the decision. MultiCare had six people on its integration team in 2001. In 2021, after growing from three hospitals to eighteen, it still had six people on that team.
{{quote-1}}
The work itself does not shrink. Every affiliation brings archival, then data conversion off the legacy system, then master patient index reconciliation so the people in the old system are correctly matched to the people in the new one. Then come the third-party integrations: lab, pharmacy, radiology, transfer center, admissions and discharges. Each one has to connect into the destination EHR and keep working.
As Shepard put it, before the platform approach the work “was very manual, and it was done by heroics.”
What Is the Difference Between EHR Data Conversion and Data Archival?
These are two different jobs and most organizations facing an EHR transition need both.
Data conversion is migration. It moves data from the legacy system into the destination EHR so it lives inside the clinician's primary workflow, including the demographic and clinical records that have to be correctly matched to the right patient on arrival.
Data archival is retention. Some data does not need to live in the new EHR but still has to be accessible, usually because retention law requires it. Washington state requires healthcare organizations to retain records for 26 years. An archive keeps that data reachable without carrying it into the destination system.
“That's what we call data conversion vs. data archival,” Shepard said. “Data archival is one aspect. In the state of Washington, for example, we are required by law to retain data for 26 years.”
Getting the split wrong is expensive in both directions. Converting everything inflates the destination EHR with data nobody uses. Archiving too much strands records clinicians need at the point of care. ELLKAY supports both paths, with clinical archiving and migration across 293 EHRs and archiving capability across 486 healthcare applications in total.
What Does Epic Community Connect Require for Data Integration?
MultiCare runs Epic as its primary EHR and extends it to other organizations as an Epic Community Connect partner, offering the platform to organizations that do not have the infrastructure or resources to deploy it themselves. More than 40 organizations are Epic Community Connect customers of MultiCare.
Community Connect makes the integration problem someone else's data as well as your own. Epic's principle is one patient, one record, and honoring that across affiliated organizations means every partner's legacy data has to be handled before it can join the environment.
“In that scenario, you need to bring those legacy systems either into archival solutions, which we use ELLKAY for, or into the Epic data environment itself,” Shepard said.
Alongside that sits the third-party integration work, which Shepard describes as the larger share. Lab systems, pharmacy, radiology, transfer center management, admissions, discharges and transfers all have to connect into Epic and keep running as the organization grows.
When Should a Healthcare Organization Outsource Integration Work?
MultiCare did not begin with a platform decision. It began with archival, ran on that for three and a half to four years, then tested integration on a single site before committing.
“We had the opportunity to utilize them as our integration engine for one of our sites, Forks Community Hospital,” Shepard said. “We'd never used them for integration or for interoperability, and that's where they kind of blew our socks off.”
At the end of January 2026, MultiCare moved its entire interface and integration function to ELLKAY. Several MultiCare staff were rebadged as ELLKAY employees, including people who had been with the organization for more than twenty years.
{{quote-2}}
The scale difference is what made it work. MultiCare's six-person team became access to an integration organization of more than 900 people.
“You don't do that with a vendor; you do that with a partner,” Shepard said.
The sequence matters for anyone evaluating the same decision. Archival first, a single-site integration trial second, full transition third, across roughly four years. Every step was reversible until the last one.
What Does a Data Platform Approach Enable Once Integration Is Handled?
With integration handled, MultiCare's data work moved to what the data can do. The organization runs an Azure-based enterprise data warehouse on Microsoft Fabric, reporting through Power BI, and is now looking at AI enablement on top of it.
“Soon we'll be looking at AI enablement in the native capabilities there, and allowing people to just chat with the data,” Shepard said.
That sequence is the point. The analytics and AI layer only works when the data underneath it is complete, correctly matched and accessible, which is what conversion, archival and integration deliver. Identity resolution in particular determines whether records from a legacy system are correctly attached to the right person after migration.
ELLKAY’s patient matching solution delivers identity resolution at scale:
{{stat}}
MultiCare's growth is not finished. Shepard describes acquisitions, affiliations and joint ventures planned years ahead, with the goal of remaining locally governed while continuing to expand.
“I have the partner, and they have the scope and the scale to be able to meet our needs,” Shepard said.
Nick Shepard's comments are drawn from an interview with Healthcare Innovation, published September 18, 2026. Read the full interview at Healthcare Innovation.
ELLKAY orchestrates healthcare data end-to-end so the right information reaches the right place at the right time. See how EHR data archiving preserves legacy records, how EHR migration moves data into a new system, or read how MultiCare consolidated 50 data sources into one foundation.
“We've had some really good talent at MultiCare that enabled us to do this work. As we grew from three hospitals to 18 hospitals, we recognized the scope and scale was not there for us to continue to bring that talent pool in house.”
“I no longer have an interface team at MultiCare. ELLKAY, as my partner, is my interface or integration team, my data orchestration team.”
