In the days before healthcare networks, a provider's decision to invest in facilities or services likely would have hinged on the question, "How does it affect our bottom line?"
That's still important, but senior managers of networks now have a bigger picture to consider. So they know to ask a companion question: "How will this investment help us provide a continuum of care?"
A leading contingent of integrated delivery networks has begun applying that same strategic mind-set to information systems' value, according to a new study that seeks to identify success factors for information practices within these emerging networks.
Top executives are going beyond return on investment to focus on noneconomic returns that can't be measured strictly in dollars and cents. Rather, they're measuring projects against a higher goal of network integration to see if they make sense.
Then, powered by a lean decisionmaking group of top managers, the networks are committing more than 25% of their capital budgets on average to information systems.
An intensive study of 10 integrated delivery networks, conducted by Waltham, Mass.-based First Consulting Group, combined surveys with lengthy interviews to chart the approaches IDNs are taking to information challenges.
With integrated networks becoming the dominant form of provider organization, First Consulting says the industry is "starved for comparative information on best practices for information management requirements to support these large, complex enterprises."
To address this need, it selected prominent organizations that have demonstrated advanced business integration or advanced approaches to information systems strategy, development or implementation (See chart, p. 118).
After months of exploring the organizations' track records of achievement-as well as some failings and lessons learned-a handful of pivotal activities emerged, says Peter Kilbridge, M.D., one of the study's directors.
Beyond the usual paybacks. The report on best practices suggests an approach that gets beyond preoccupation with department or facility orientation and beyond insistence on payback according to traditional financial formulas (See chart, p. 116).
The report found:
The organizations distinguish broadly between strategic and nonstrategic projects. Strategic projects support goals of network integration, such as connecting the various sites, pooling data from those sites and harnessing the networkwide potential of information.
Those projects are evaluated according to their ability to fulfill strategic goals, while projects affecting facilities or departments continue to be evaluated the traditional way-subjected to varying rigors of cost-benefit analysis.
A number of the IDNs look for wider measures of information systems' value, focusing on contributions that computerization and connection technology make to the organization. Says one chief operating officer: "We will measure not the success of the information system but the success of the process the information system supports."
Top executives of most of, but not all, the IDNs demonstrate their commitment to information systems strategy beyond lip service. They have given chief information officers the budget, authority and latitude to execute a strategy.
"Strong executive support permits the CIO to exert more control over (information) system governance and decisionmaking, work from a position of strength in relationships with vendors, and focus*.*.*.*on initiatives for systemwide integration rather than on different constituents' needs of the moment," the study concludes.
The green light. Executive support and a clear strategic focus enabled a fledgling network in northern Ohio to get most of the key information systems in place for systemwide integration in less than three years.
Toledo-based ProMedica Health Systems had added facilities during a stretch of a half-dozen years. They include Toledo Hospital near the downtown district, adjacent Children's Hospital of Northwest Ohio and Flower Hospital eight miles west in Sylvania.
The leaders of ProMedica "figured out early the importance (of integration) and immediately went with a guy who supported an aggressive approach," Kilbridge says.
Realizing it had to tie the medical campuses together and get physicians into the system, ProMedica hired David Selman as top information officer in 1995. Reporting directly to the chief executive officer, Selman "was given the green light and underwritten for the necessary investments," Kilbridge says.
The first thing Selman did was author a strategic plan identifying how some of the key integration components needed to occur and in what sequence.
The plan went straight to the basics of information transfer-speed and volume. That called for using the fastest and highest-capacity communications technology available to create local-area networks serving the two primary campuses and 11 other sites, primarily medical office buildings.
Selman says he is two years into a three-year process of wiring the sites and connecting them with a wide-area network, using a technical foundation (called asynchronous transfer mode) that can whisk images as well as data files at speeds of more than 100 miles per hour. That compares with three or four miles per hour for standard phone lines.
Early on, the ProMedica board also approved funding for the end game of multiple-site information integration, a clinical data repository.
Launched in August 1995 at Toledo Hospital, Selman put development on a fast track. Six months into the project, 750,000 patient records had been entered in the database. And a month later, in March 1996, laboratory results began making their way into the repository, Selman says.
The Toledo Hospital repository, developed by 3M Information Systems, is complete and in use in several nursing units, an emergency department, a family practice and a specialist practice. ProMedica is about two months away from including the second hospital campus, Selman says.
Records from the home health and skilled-nursing operations are next on the implementation schedule, and physician office records are set to be included by next year, he says.
Despite the technical sophistication and compact time frame, the strategic plan hasn't broken the bank.
The total three-year expense of laying the network infrastructure was $3.5 million. The networkwide repository has cost between $3 million and $3.5 million, including consulting fees, hardware and software but without capitalizing the costs of in-house labor, Selman says.
Do what needs doing. Strategic integration projects aren't that difficult to distinguish from tactical projects benefiting a particular location or constituency, Kilbridge says. "Is it essential to the support of the business plan?. . . . Most people find more things are obvious yeses than they have the money for," he says.
That calls for setting priorities within the strategic realm. If a provider organization wants to operate multiple sites of care, "you need to have them connected," he says. From that assumption, a system to identify one patient from another across sites is "an absolute strategic investment."
Integration decisions are equally necessary whether a network is relatively compact, as in Toledo, or spread out for 75 miles and still expanding. The latter case can be found in Tupelo, Miss., population 30,000 and home to what's called the largest hospital in rural America.
North Mississippi Health Services, a regional provider network operating in 26 cities, includes 699-bed North Mississippi Medical Center in Tupelo, four other outlying hospitals and 30 physician practices. About 100 buildings make up the network, half of them outside the Tupelo city limits, says Tommy Bozeman, the network's director of information systems.
Unlike the typical emerging regional network, the medical center's top management has been working on wide-area integration for nearly 20 years, partly because of the demands of running a large referral hospital as the hub of a broad rural service area.
Even with the luxury of an established strategic process, the network was signing up physician practices and building clinics at such a clip during the past four years that the information systems crew was straining to keep up with the additional automation demands, Bozeman says.
But North Mississippi has kept to a plan focused on putting all clinicians on a common system of electronic spokes leading out from the medical center hub.
The network in 1994 committed to spending $45 million over five years on new software, new people to operate the stepped-up computer network and new dedicated transmission lines connecting small hospitals more than an hour's drive away, Bozeman says.
Economies of scale. The operation of North Mississippi depends heavily on using primary-care and rural hospital sites as feeder systems for specialists at the main hospital. With distance a key factor, specialists rely on receiving test results and other valuable clinical information from those primary-care sites, Bozeman says. The alternative is to repeat costly procedures unnecessarily.
The provider organization is about 75% of the way through a project to connect all sites through a wide-area communications network, making any clinical information captured electronically at a site available to anyone else in the network, Bozeman says. Family medical clinics represent the final 25%, and the target date for getting them on line is Sept. 30.
All physician sites are being put on the same practice management and billing system, operated out of a central billing office over the electronic network, instead of allowing each practice to implement any number of different software systems locally.
Other systems in place or in progress:
A central clinical repository called the TDS Permanent Patient Record, marketed by Eclipsys Corp., is receiving information from the wired-up sites. Included are 12 home healthcare offices logging 500,000 visits a year.
Each patient or insured enrollee has a common identification number used by all hospitals, physicians and home-care providers in the network. The identification numbers are imprinted on magnetic-stripe cards issued to all 300,000 enrollees that can be read on personal computers in registration centers. The card automatically calls up the person's record, grabbing the right files even in cases of name or address changes.
Internal medicine sites are automating transcriptions of clinical notes included in the clinical repository as part of a patient's running electronic record. The IDN is six to nine months from expanding that system to family medical clinics, using a central transcription pool to feed notes into the repository.
A clinical information steering committee, including physicians considered critical to the network, oversees the networkwide strategy and has been aggressive from the start toward spending the money needed to execute the grand plan, Bozeman says.
That has included the commitment to bring every referring medical clinic into the wide-area network whether officially part of the North Mississippi system or not. Also, all outlying sites are connected by dedicated phone lines directly to the main medical center.
Different measures. Not all the surveyed IDNs embrace that strategic approach. One still demands an across-the-board 15% return on investment to justify funding. First Consulting did not identify the IDN.
"This policy has resulted in delayed development of elements of system infrastructure in some areas," the study reports, "especially those relating to support for delivery of patient care across the continuum of settings." Kilbridge says clinicians are frustrated by what they consider to be "primitive" networkwide operation of the clinical systems.
Executives interviewed for the First Consulting study say they devised alternative measures of success centered on improvement of work processes. Instead of making financial return the key focus, they cite:
Reduced average lengths of stays and reduced member days per month.
Improved contracting power through ready availability of cost and length-of-stay data.
Reduced variation in care, a central goal of clinical quality improvement efforts and a primary target of investment in clinical information systems.
n Improved clinical outcomes through better availability of clinical information at the point of care.
John Glaser, vice president and CIO of Boston-based Partners HealthCare System, tells surveyors, "Return on investment does not work in the service sector. Successful investments translate better to attainment of service goals and quality of service than to cost savings."
He added that a project must deliver on promises of enhanced care through the use of computer features and functions. And it has to adhere to a projected time line and come in at the budgeted amount.
Not all clinical improvements are big deals, but they are still meaningful to clinicians. At North Mississippi, where nurses do all charting on an electronic system, an oncology head nurse can update eight charts in 30 minutes, 15 minutes less than it took with paper charts.
There's a little something for everyone. For the billing office, the accuracy afforded by electronic charts also has resulted in an average of $8 a day extra in medication revenues per patient, Bozeman says.
"Nurses aren't concerned about that (revenue result), but they are about keeping their license, which means documenting everything they do," he says.
Some cutting-edge projects are evaluated on benefits that can be measured hypothetically using formulas reported in healthcare literature, Kilbridge says.
At ProMedica, Selman says he took documentation published by Intermountain Health Care, from which 3M had borrowed ideas to build its data repository, and computed a savings of $3 million to $4 million retrospectively from 1993 network data. Savings were based on the comparative gains made at Intermountain after instituting the computer innovations.
ProMedica plans to validate the savings projections by studying how it fares before and after implementing the repository, Selman says. That data should be available starting in 1998.
The information systems department, in partnership with the clinicians who use the system, aims to identify improvements in outcomes management, clinical documentation and other areas that stand to benefit from integrating separate sources of data at many locations, he says.
CIOs as chefs. IDNs usually aren't lacking in the core information systems that capture, report and receive data. But integration is a separate project, Kilbridge emphasizes.
Developing an integration strategy calls for investing in people and in intelligent use of the systems already in place, he says, adding that "95% of this is the cook, not the ingredients." The IDNs that are having trouble "are not pulling it together properly."
Some of the ingredients Bozeman is working with at North Mississippi might be considered too old to use. The patient accounting and general ledger system is nearly 20 years old.
But it still handles a payroll that has surpassed 5,000 people. Computer pros continue to enhance and improve it. "We've taken a venerable old system and used the heck out of it," says Bozeman, who's been at the medical center 21 years.
It didn't hurt that when the system was purchased in 1979 it already could handle multiple sites, enabling programmers to run Medicare data from separate sites on one system.
Good "cooks" watch for other ways to stretch a recipe.
When Selman was shopping for a clinical data repository, he looked for products that allowed clinical information sharing across care sites and a central strategy for coalescing far-flung records. But he selected 3M because he thought it could do more.
Two years ago, it already could process and trigger clinical alerts on such things as drug interactions, a feature that's since increased in visibility after rising publicity about medical complications that such systems could prevent.
The network has started activating alerts for critical lab values, and it's on schedule to start pharmacy alerts in midsummer, Selman says.
Bozeman got more integration for the buck by reaching agreements with vendors on expanding installed systems to additional locations for little or no extra cost.
An "unlimited" license allows North Mississippi to activate new capacity and support it at a central location instead of buying new licenses and new maintenance and support contracts every time the network gets a little bigger, he says.
Not only does that save money but it also puts the vendor on the spot to demonstrate it can "scale up" the capacity and performance of its software as needed.
"If the vendor cannot do that, that would lead us to believe they're not the kind of vendor we want," Bozeman says.