Two years ago, the University of Pittsburgh Medical Center began searching for software that would automate paper-based infection control processes across its 19 hospitals. After considering four vendors, UPMC opted to do business with Salt Lake City-based TheraDoc.
But the medical center did more than obtain software licenses. UPMC wound up taking an approximately 20 percent ownership stake in the vendor, which markets a suite of clinical decision-support products. The medical center also has a seat on TheraDoc’s board. “We were so impressed with their software, we said we wanted to be part of the company’s future,” says Tami Merryman, vice president of UPMC’s Center for Quality Improvement and Innovation.
Does this deal signal a new era of hospital-IT vendor relationships? At the least, it stands in stark contrast to industry norms. Historically, many hospitals and their software suppliers have maintained what might be charitably described as a love-hate relationship. Products have been oversold, support has been lacking, promised features have failed to emerge. Any romance was fickle at best. But now some hospitals are trying to leave such contention behind with joint development deals—or in cases like UPMC, literally tying a portion of their economic future to a vendor’s destiny. Experts caution, however, that such ventures are challenging propositions—for both the hospital and the IT vendor.Aligning goals
One of the most prominent hospital-vendor development efforts belongs to Intermountain Health Care, a Salt Lake City-based hospital system, and GE Healthcare, a multi-national conglomerate known for its imaging technology. Beginning in 2006, the two began a seven-year development project, splitting roughly $200 million in cost, to build an integrated inpatient, outpatient and emergency department clinical information systems package. The new system will replace Intermountain’s homegrown legacy system, says Kevin Smith, associate vice president of information systems and general manager for the GE alliance. “We needed a partner to help us build it,” he says. “Someone with deep pockets and clinical-systems expertise.”
One goal, says Smith, is to build a comprehensive information system that can compete with such big-name vendors as Cerner, Epic and Siemens. The new system, which will be marketed by GE, will include a host of features such as order entry, results viewing, nursing and physician documentation, and a clinical data warehouse. Intermountain stands to receive royalty payments once the new system hits the market, Smith says. “The real beneficiaries of the system will be large academic medical centers, as well as 200-bed hospitals that could not build the content themselves,” he says.
Hospital-vendor development efforts are not new, observes Lewis Redd, managing partner with the provider client services group for Atlanta-based Accenture. They are often spawned when hospitals, looking to replace highly customized legacy systems, face difficult “build it or buy it” decisions, he says. “Hospitals are trying to find a way to have their cake and eat it too,” he says. “They want everything in a system, but they want it from a commercial vendor so they don’t have a custom system to support.”
But many joint development deals have filed, Redd says. “It is a risky strategy, so hospitals need to get smart about how they do it. They need to define the scope—who does what, the number of people involved, the timeline, and when the vendor will be paid. The hospital needs to make sure its goals and the vendor’s product strategy are aligned. If it is not defined, the likelihood of hitting a point where the business objectives are not aligned is pretty high.” Without clearly stated goals, a vendor might feel pressured to adapt its software to the needs of other prospective clients, Redd explains. Ownership stake
That’s one reason Intermountain insisted on a large financial and personnel stake in its GE partnership, Smith says. The 340 staff members working on the project are roughly divided between GE and Intermountain. “We felt we needed to put in significant funds, rather than having GE say, ‘Give us the requirements and we’ll engineer them.’ It’s why a number of development efforts have failed. Without sufficient skin in the game, it is difficult to drive common goals. The shared investment drives shared accountability.”
In addition, Intermountain and GE have formed a joint governance team that makes strategic decisions. Its steering committee is a 50-50 split between the two companies. Any unresolved decisions would go to the CEOs of the two organizations, Smith says. “Nobody wants to go to that level because it would look like they can’t handle it themselves,” he says.
The most controversial decisions, Smith says, revolve around how much of the new system to build from scratch and how much of it should be built around GE’s pre-existing software. Those priorities were muddied after GE acquired IDX, a hospital information systems vendor. “People tend to want to use the technology they have as opposed to reengineering the system and starting from scratch,” Smith says. In the end, however, the group will build a new product compatible with specifications set by HL7, a not-for-profit volunteer organization based in Ann Arbor, Mich., that creates exchange standards for clinical and administrative data, Smith says.
In making an investment in TheraDoc, UPMC is hoping to capitalize on a product line that already has a market presence. TheraDoc licenses its various decision-support software products to more than 100 healthcare organizations nationwide. Offering data mining capabilities, its infection control software will aggregate data from UPMC’s pharmacy and lab systems. “There are not a lot of vendors with mature products” that do this, says Merryman. She declines to name the three other vendors the medical center considered during its year-long vendor evaluation.
For his part, Redd is bearish on hospitals taking ownership positions in software companies. “A hospital’s core competence has nothing to do with the software business,” he says. Further, hospitals may be reluctant to support a vendor that has economic ties to a competitor. “The vendor has immediate conflicts,” he says.Gary Baldwin is technology editor of
HealthLeaders magazine. He can be reached at email@example.com.
Hospitalists’ DIY Approach Pays Off
When it comes to hiring commercial IT vendors, Adam Singer, M.D., is a skeptic. Singer is chief executive officer of IPC-The Hospitalist Company, a Los Angeles-based organization that provides hospitalist services nationwide. Formed in the mid-1990s, the company has grown to more than 600 physicians in 105 practices.
IPC has also become a technology developer, using its own crew of 15 IS staff to write and maintain the company’s customized practice management software. The software enables charge capture and patient tracking, providing a “virtual office” for the remote physicians to keep tabs on their practices. Backed by an infusion of venture capital, IPC began building the software in 1997. “No vendor had the technology,” Singer says.
Singer says he has attempted to work with IT vendors on various aspects of the system, but the relationships would sour. “Every time I used a vendor, I got into trouble,” he recalls. “There was always a voltage drop. We had the domain knowledge, but they had the programmers. It led to delays, increased costs and products that didn’t function as well as they should.”
IPC now leases its software, which runs on an Oracle database, to two independent hospitalist groups. There are some 15,000 hospitalists nationwide, so Singer figures his technology revenue stream can only grow. “No software project is ever finished,” he says. “So if you use an outsourced vendor, you’ll be stuck for life.” —Gary Baldwin