While looking to bring interoperability into its institution, the University of Pittsburgh Medical Center partnered with a technology company willing to work on solutions over the long term.
Despite all the recent talk in health care circles about interoperability, it is nearly impossible to point to an existing information-sharing network or vendor solution and say, “They’ve done it. This is a true example of system interoperability in which physicians have all the relevant information they need to make treatment decisions at the time they need it.”
There are many reasons for this. Some health care organizations have yet to invest significantly in technology because they are waiting for it to further mature. On the other hand, organizations that have invested in electronic medical records and clinical information systems are uneasy about making additional, sizable investments to facilitate interoperability.
The fact is, in the absence of a true interoperability model, few organizations have the incentive to pull the trigger on an initiative that may end up serving as a case study for how not to proceed.
Model for the Future
Against this backdrop, an alliance was created between the University of Pittsburgh Medical Center (UPMC) and dbMotion, a provider of health care information integration software that facilitates interoperability and health care information exchange. The alliance’s aim is to achieve true, real-time interoperability and improve the quality of patient care for UPMC’s delivery network and region.
In addition, the clinical solutions created can serve as a model for peers--helping them move more aggressively toward the universal sharing of patient information among physicians, regardless of the health care settings in which they practice.
UPMC is an integrated health care enterprise composed of 19 hospitals, nearly 5,000 employed and affiliated physicians, and more than 400 outpatient sites, doctors’ offices and other care facilities. UPMC is committed to providing physicians with the most relevant and recent patient care information available. As one of its physicians is fond of saying, “You can keep only so much information in your head at one time.” With transplant and other complex patient cases, technology is an essential tool in delivering the best care possible.
To support the institution’s commitment to patient care and research, it has invested heavily in EMRs and other clinical systems and allowed physicians and executive leadership to seek out best-of-breed technology. Along the way, this approach has yielded a patchwork of different systems across the extensive network. Like other organizations, it has struggled to find ways to make these systems talk with one another and allow physicians to access the information they need.
This challenge led to various efforts to connect the disparate systems. A single-point login was implemented that enables physicians to access the system they are accustomed to using and then “tab” their way to the relevant patient information they may need from other systems. Also implemented was a message routing system so physicians can share their notes and treatment decisions, irrespective of where they practice. As a result, more than 600 different system interfaces were created to facilitate patient data-sharing among physicians.
Filling in Holes
In more recent years, however, the organization recognized that it still fell short of a truly interoperable system that would allow physicians to access the complete patient care information they needed at the time they needed it. So a search began to find an interoperability solution. Along the way, several important lessons were learned:
- A single-vendor solution was unworkable. UPMC’s disparate network--and the investment already made in obtaining best-of-breed systems--was too large and complex for a single vendor to manage all the interfaces and data integration needed to achieve interoperability.
Instead, a best-of-breed integration vendor allowed each of the information system providers to continue to excel at what it does best. This solution does not replace the electronic medical record solutions already in place. Rather, it uses service-oriented architecture, which loosely couples application components, to enable EMR vendors to easily access data from the network without complex point-to-point interfaces and the added expense these connections require. In addition, the usability of these systems will be enhanced and give physicians greater access to the information they need in a format to which they are accustomed.
- An integration architecture was needed that would enable shared clinical documentation and decision support. The solution uses an HL7 foundation and a data repository model that fit with goals to ensure universal, real-time access to patient information across the continuum of care. Its architecture also gives the flexibility to add new systems that will help further improve patient care and safety.
In addition, the solution will help the institution comply with care standards, which are constantly evolving. For example, as pay-for-performance efforts continue to expand, health care organizations will need to prove they have complied with care standards to receive full payment for their services. That is difficult to do today because much of the information resides in disparate clinical and administrative systems.
- UPMC wanted a vendor partner rather than a mere supplier. The health care IT marketplace does not yet offer products that meet the need for systems in areas such as oncology. A vendor was found that shared an interest in developing specific technologies and systems that, in time, can be commercialized and marketed to peers.
The initial goals for the alliance include achieving interoperability within UPMC’s integrated delivery network. This work can potentially serve as a model for a workable, real-life interoperability strategy and spur other interoperability initiatives. In fact, discussions are already under way to form a regional health information organization in western Pennsylvania, and the alliance’s work should speed its development.
Product Development
In addition, the alliance is an opportunity to develop technology products that will have commercial appeal for the industry. Goals for the creation of new products include:
- Clinical decision support tools to improve patient care quality and compliance with changing care standards.
- Tools to identify specific patient cohorts for clinical trials. Clinical researchers will be able to more easily find patients whose conditions and drug regimens fit the criteria to be included in specific trials.
- Tools that will close the gaps in the continuum of care between acute, post-acute, ambulatory and other care settings. The solution will enable interfaces with a wide array of outpatient clinical systems to facilitate the sharing of patient data--a critical need at any delivery network.
All in all, the alliance will create an industry model demonstrating that physicians can get the patient information they need, in the format they prefer, with only minimal disruption to their workflow. Most importantly, it will bring the vision of true interoperability to life for the benefit of patients.
Duane Falk is information services division vice president in charge of the interoperability project, and G. Daniel Martich, M.D., is vice president of eRecord and associate chief medical officer, both for UPMC. Jay Srini, vice president, emerging technologies, UPMC, and Peter van der Grinten, general manager, U.S. and Canada, dbMotion Inc., also contributed to this article.
This article is also available on-line at H&HN's Most Wired.