- Healthcare led all industries in mergers and acquisitions in 2015, with $688 billion worth of consolidations coming to a close, according to financial information firm Dealogic. And while analysts expect more of the same this year due to favorable business and regulatory conditions, the individual parties to M&A transactions face their own unique challenges in combining medical staffs and ensuring health IT interoperability.
“Every one of our clients and most of our prospects are either involved in a merger, looking at a merger, being acquired or acquiring somebody else. It really is amazing what’s going on,” said Guy Scalzi, chief strategy officer at clinical decision support (CDS) firm medCPU.
“I’ve read that we’re headed for maybe five or six health systems in the whole country — and from the inside, it kind of feels like that,” added Scalzi, a former CIO at NYU Medical Center, New York/Cornell Medical Center and New York Presbyterian. “We are definitely in the thick of it, and it’s certainly about the data.”
Consolidating organizations typically have different instances of software, even if they’re using the same core vendor platform. Additionally, the software may be installed in different ways, impeding progress toward the ultimate goals of providing better care less expensively.
“If they can’t consolidate the data and read it across the entire enterprise, they’re not going to get the full value of what they’re trying to do. They’re just a bunch of friends who are talking together,” Scalzi told HealthITInteroperability.
Allan Strongwater, MD, a former pediatric orthopedic surgeon and now senior vice president of medical informatics at medCPU, agrees. “The IT shop in a medical center is a major cost center. Merging IT departments achieves significant cost savings. However, if you don’t have the ability to merge data, then you can’t accomplish IT integration,” he explained.
Integration in progress
Inspira Health Network formed in November 2012 through the merger of South Jersey Healthcare and Underwood-Memorial Hospital. Those facilities still maintain their own separately governed medical staffs.
From a technology perspective, they’re united across campuses by a common computerized provider order entry system. Concurrently, an organizational priority to reduce risk and liability in obstetrics led to investigation of medCPU’s OB module, which had been implemented at a neighboring healthcare system.
Inspira implemented the OB module about 8 weeks ago, and is in the process of building and developing a separate module for sepsis, another high-priority area for clinical improvement.
The underlying CDS technology is based on clinical rules — best practices that differ from one healthcare entity to another — complemented by natural language processing to build a clinical database distinct from the EHR system.
“We’re building a purely clinical database focused around taking care of the patient, what we know about the patient’s health and past experience. So once we’ve got that database, we can do our decision support and work with patients and their physicians across any of the organizations, looking at the same data,” Scalzi continued. “So if a doctor or patient moves between hospitals, from our perspective, we’re looking at the same data — not a different configuration or different vendor.”
The other key factor is knowing the sources of data.
“Where are we going to collect data?” Strongwater asked. “Is the bulk of the information going to be coming through an HL7 interface? Are we going to do it through a screen reader? Are we going to be on a Citrix platform? The more information that we can collect, the more accurate our decision support will be.”
Scalzi observed, “Most medical centers and integrated delivery networks today want to get to one platform. That’s what we hear all the time. If they’re on Cerner, they want to buy Cerner as much as possible. Or Epic or Meditech or Allscripts.”
However, clean transitions are difficult to execute.
“I can tell you, at New York Presbyterian I had 14 full-time people dedicated to interfaces and keeping them running,” recalled Scalzi. “Interfaces don’t just go up and then run forever. There are constant issues. They also affect the performance of systems.”
He said the approach of capturing data through HL7 with a natural language processor reading unstructured data brings about inherent advantages. “You don’t have the issue of keeping all these interfaces running and talking to each other,” Scalzi explained.
Strongwater added, “Because it is independent of the EHR system, our clinical decision support does not appear as a burden on the EHR database. That contrasts with some of the systems that are a part of the EHR; as they function and become busy, they load the EHR, which slows clinical performance.”
At Inspira, the newly implemented CDS technology has helped to consolidate a single point of view between separate medical staffs. In doing so, it has improved interoperability in terms of reading data from disparate systems and yielding a common answer.
Moreover, the process of convening the medical staffs from separate facilities has fostered consensus on the design of clinical alerts — in terms of both practicality and relevance of medical evidence.
In form and function, the tool has helped to build bridges between medical staffs as they consider additional CDS modules for future implementation.
Lead image credit: Released to Public Domain