Catchall definitions of data repositories in the marketplace may impede advances in healthcare information systems, a new study suggested.
In addition, clinical professionals and chief information officers have differing notions of what clinical data repositories should do and how they differ from other types of databases.
That disparity complicates planning for and purchasing a complex system that's key to integrating diverse healthcare organizations, according to the Clinical Data Repository Survey by the Center for Healthcare Information Management.
The Ann Arbor, Mich.-based association represents more than 100 information systems vendors and consulting firms. It conducted the survey earlier this year to test whether the marketplace differentiated labels that are sometimes used interchangeably, such as clinical data repositories, electronic medical records and computerized patient records.
The study also sought to understand why so few systems were purchased and why even fewer were fully implemented.
Part of the problem is financial, the survey showed. Insufficient money is available to commit to such a project, and providers' cost projections are low compared with their expectations.
Another problem is confusion between what providers think they are asking for and what vendors think they are delivering.
"Going to the vendor community and stating `I need a database' will bring a deluge of possible products and solutions," the study said. "There are many vendors offering CDRs, and as many offering data warehouses and their associated components. It can be very confusing for a new buyer."
Depending on whether healthcare organizations are trying to support daily clinical operations or strategic and research objectives, two very different types of databases are required. Clinical data repositories are designed for immediacy, whereas data warehouses are retrospective (See chart, page 58).
CHIM received 317 responses from executives who are members of the American Association of Nurse Executives, the College of Healthcare Information Management Executives and the American College of Physician Executives.
A preliminary look at results in June found consensus among CIOs and clinical executives about the most important features of CDRs, such as central storage of clinical data, 24-hour access to current patient data and integration of test results and observations (July 6, p. 35).
But the two groups disagree fundamentally on the purpose of such repositories. CIOs want to integrate clinical results, and clinical executives want to measure quality.
"CIOs are looking toward a CDR solution to solve different problems than the ones providers are seeking to solve," the study said. "Vendors will need to listen to both the (information systems) and clinical `wish lists' to integrate a solution that satisfies both constituencies."
Unlike most information systems purchased for administrative and financial reasons, CDRs are valued for their clinical usefulness. However, the study suggested that physicians and nurses are not being brought into the strategic process and educated about the repositories as much as they should be.
Overall, 59% of the respondents said their organizations were implementing or planning CDRs. But within the three responding groups, 81% of CIOs said CDRs were in the works, whereas only 47% of nurses and 56% of physicians acknowledged such projects.
"It is significant that such a disparity exists between the (information systems) and clinical leadership on whether or not a CDR is part of the strategic plan," the study said. "If the CDR is a clinical tool-which one might safely assume it is in some form-why would the clinical leadership be less aware of it (than information systems staff are)?"
Nearly 70% of CIOs said there was a distinction between a CDR and a data warehouse, compared with 40% of physician executives and 43% of nurse executives. More than half the clinical executives said they didn't know.
"It cannot be emphasized too strongly that it is imperative to involve the widest and deepest levels possible of clinical staff (in the CDR planning and strategic process)," the study said.
The study also found a gap between what organizations expect to spend on a CDR and what their expectations are.
A significant majority of all three groups of respondents said their organization's spending on a CDR during the next two years will be $5 million or less.
Among CIOs, 31% projected expenses of less than $1 million, and 43% anticipated $1 million to $5 million.
Those ranges seemed low to the study's analysts, who expected to see budgets more in the range of $5 million to $15 million or higher.
"Providers seem to have high expectations for quality, quantity, access and storage in their CDR, yet the budgeted figures seem incongruous with those expectations," the study said. "This could indicate that healthcare organizations severely underestimate the cost and complexity of implementing a CDR."
But it could also indicate that buyers want to spend incrementally on the clinical database-and the related integration of systems that feed information into it-as money is available.
"Few organizations may be able to commit to the `whole enchilada' in one contract," the study concluded. "It may be that organizations will want to gain entry to a CDR solution with a smaller investment and be able to prove value before committing further."
Instead of trying to provide and sell the entire range of features and functions, vendors may have to make new database capabilities fit the information systems components of several vendors.
Sometimes vendors may have to wait for customers to replace one or more existing systems-such as the pharmacy, laboratory or billing system-to update computer capability and enable systems to feed information into a repository.
"It could be that the successful vendors will be those that specialize in a particular aspect of the CDR . . . created in an environment that fits with other specialized solutions," the study said. "This may be the chance for many Davids to thrive in the land of a few Goliaths."