Consolidating a clinical imaging IT environment from multiple disparate systems can be challenging and costly for health systems. But it's worth it, according to Evan Jackson, executive director of IT services at Yale New Haven Health System (YNHHS).
"There are great rewards when you do it right," Jackson said.
He spoke during a presentation on Monday at the New York Medical Imaging Informatics Symposium in New York City.
In 2010, YNHHS included three hospitals: Yale-New Haven Hospital, Bridgeport Hospital, and Greenwich Hospital. Operational and technical ties between the hospitals were limited, however.
Evan Jackson, executive director of IT services.
"The IT folks were disaggregated; they knew of each other but did not collaborate on much of anything," Jackson said.
At that time, YNHHS was a major academic partner of Yale University School of Medicine, and there was increasing integration between Yale-New Haven Hospital and the medical school.
The legacy IT environment for the health system, however, was characterized by hundreds of applications with multiple versions and multiple vendors supporting the same function, Jackson said.
In addition, IT strategies, deployment, and support were set by the local clinical and operation leadership. The support teams were also separate and aligned to clinical leadership.
There were disparate IT strategies across the health system, with no shared vision or strategy, he said.
At the same time, the health system was facing sweeping change due to scale and scope pressures. There was a growing focus on clinical data sharing, care management, and risk management.
"There was a belief that real clinical and business integration was needed," Jackson said.
This vision was shared by the health system's physician partners with Yale School of Medicine and Yale Medical Group, he said.
A new chief information officer was brought in who consolidated the system's IT resources, and a new IT services leadership team was created.
The health system leadership decided that adopting an electronic medical record (EMR) from Epic Systems would serve as the tool for consolidating applications across the system.
"We essentially bought everything from Epic," Jackson said. "The commitment was if Epic made it, unless there was a really good reason ... we were going to use Epic."
Rollout of the Epic EMR has been fast and on schedule, although the timeline has been extended by four months for the Hospital of St. Raphael in New Haven, he said. Four inpatient locations representing more than 2,300 beds are live on Epic, as are 1,100 providers and 260 practices. Initial ambulatory rollout is expected to be completed in January 2014, adding 150 more providers.
Prior to implementing Epic systems, the health system had to determine how to handle the complex realm of third-party ancillary software. This would require integrating multiple third-party applications for similar functions or consolidating to a common third-party application.
An aggressive planning process was launched, Jackson said. This considered several clinical and operational mandates, including a growing focus on consistent clinical workflow, heightened demand for clinical data sharing, and a need for full and effective EMR integration.
Seamless access to data was expected anytime, anywhere, with demand for improved support and user satisfaction and continued pressure to maintain "leading edge" activities, he said.
As for financial mandates, capital was available under the Epic EMR purchasing budget to drive some of these consolidations, but financial resources were finite. There also was an expectation that IT services could drive greater cost efficiency.
Technical considerations involved a need to improve application uptime and to manage IT with fewer personnel.
Senior-level leadership agreed with the consolidation mandate -- with a caveat. Consolidation could be performed, but it had to be paid for out of the Epic project funds. There also had to be a consensus that consolidation could be fulfilled in very short order, Jackson said.
"Everything that we chose to do had to be done consistent with the Epic timeline," he said.
In clinical imaging, planners identified the critical applications, assessed the installed base, and determined the priority of the consolidation project.
"We had a really tight window, which took what normally would be a one-year planning process and condensed it into about three months," Jackson said.
Radiology and cardiovascular workgroups were established that included a cross-section of physicians, nurses, and IT staff.
After a review of applications in radiology, planners determined that PACS, voice recognition, and mammography documentation were required to have a common application and version level prior to Epic deployment.
Critical results reporting and radiology decision support were given a strategic recommendation rating, meaning it was recommended for all sites but the timing was not critical for rollout of the Epic system. Image sharing, peer review, and enterprise 3D visualization were rated as optional deployments, with standard vendor direction and deployment remaining a local decision.
"Really what we were trying to do is drive a consensus opinion about which solution sets and which vendors fit in each category to avoid any further disaggregation as we went down the road," he said.
In some categories, there were four different vendors among the various sites, he noted.
The accelerated planning process involved weekly meetings, collaborative discussions with clinical users, internally hosted product reviews and demonstrations, and vendor involvement when needed.
The health system sought a rapid consensus on direction (or the lack of a clear obvious path) for integration, requirements to consolidate, and the priority to change pre- versus post-Epic implementation.
After consolidation decisions were made, IT services leadership assessed the value of the various system consolidations and the feasibility of implementing Epic without consolidation. The Epic project steering committee provided the final vote on direction and allocated funding.
Progress and next steps
It's been a tale of two clinical disciplines for Yale. In radiology, the consolidation direction was clearly identified and funding was secured through the Epic project. Consolidation has largely occurred with or in advance of the Epic installation, Jackson said.
In cardiology, however, no clear consensus was achieved for the consolidation. A limited effort to leverage a promising vendor failed after a first expansion attempt. The planning process to set direction for consolidation is continuing. Consolidation is still the goal but funding will determine the pace.
As much as possible, the health system made sure common applications and common versions of its applications were in place at each stage of Epic deployment, Jackson said.
In radiology, there are some remaining technical integration efforts planned for the coming year. Product optimization is a priority, and the health system would like to resume more aggressive adoption of new technology for high-end needs.
"We've got some new technology transitions that we're working on, like a [vendor-neutral archive (VNA)]," he said. "A VNA, vendor-neutral image viewing for integration into the EMR ... all of those things give us a lot more flexibility."
Reporting and analytics specifically associated with radiology is also a priority, Jackson added.
Integrating multiple disparate systems is a challenge, Jackson said. At Yale, adopting the Epic EMR really drove the consolidation process. Technology for clinical imaging is neither easy nor inexpensive to consolidate, but addressing it effectively is critical, he concluded.
Jackson noted that vendors are still maturing, so there may be some product sacrifices that have to be made during consolidation.
"We weren't going to be able to fit every desire into a consolidation strategy," he said.
It's important to engage physicians and clinical leadership involved in setting the direction when consolidating IT systems.
"There's nothing harder than implementing a system where the people who ultimately are going to use it are not fully bought into where you're going," he said.
Particularly in academic institutions, consolidation will affect "leading edge" IT adoption.
"I had to be very overt with folks that we're just not going to be able to address that over the next couple years," Jackson said.