Drug Firms, Government Seek It Standards for Clinical Data

Wyeth worked with several consulting firms to set up its Internet-based clinical data management system, says Robert T. Maguire, chief of operations for clinical development. One firm was hired four years ago to determine whether there was such a thing as a commercially available clinical trial management system. There wasn't, Maguire says, and there still isn't.

"There is nobody offering that single solution set, so we've had to build it ourselves," he says.

That set would include a database and a range of tools that give workers throughout clinical management--from doctors in the field to number crunchers at headquarters--real-time access to data. While Oracle, the database computing giant, says it is close to offering a turnkey clinical management system, there are still no finalized industry data standards for its operation. This is a big problem, given that the ultimate data exchange in this industry is between private drug companies and federal regulators.

The emergence of a clearly defined clinical data management system is following the typical route for new information technology (IT), where users convene standards organizations that force vendors to develop interoperable products. The process is complicated, however, by the essential conservatism of the pharmaceutical industry regarding changes to IT systems, especially where filing information with federal agencies is involved.

"The pharmaceutical industry is underinvested in IT and hasn't benefited from it as have other industries," says Mark A. Goldberg, president of Perceptive Informatics, a clinical data management software subsidiary of Parexel, a clinical research organization. "There is almost an irrational level of fear in regard to changing practices and installing technology"--ironic, Goldberg says, considering that the Food & Drug Administration is issuing guidance on how to make such changes.

Stephen I. Katz, director of the National Institute of Arthritis & Musculoskeletal & Skin Diseases advisory council at the National Institutes of Health, also sees clinical management IT as lagging. "Clinical research informatics is clearly behind," Katz says. "The needs were not realized early enough, perhaps."

Katz is interested in seeing NIH improve its own IT infrastructure and its ability to share data with drug companies and other government agencies. "NIH would like to better utilize the massive information that we already have for musculoskeletal, pulmonary, and metabolic systems," he says. "There is a tremendous amount that is lost without having data interoperability."

Standards for data management may be even more important at FDA. "Without a standard, people file data in different arrangements. We have to relearn the data sets each time," says Randy Levin, associate director for medical informatics at the agency's Center for Drug Evaluation & Research. "Once data is in a standardized structure, we can preprogram software to run a macro script," so that data coming from different sources will automatically format to conform.

There are two important standards bodies working on clinical data management: Health Level Seven (HL7), an International Organization for Standardization-affiliated group studying standards for health care data management, and the Clinical Data Interchange Standards Consortium (CDISC), a user and vendor group. The organizations have several linking points, including the Regulated Clinical Research Information Management working group at HL7, which is jointly run by HL7, CDISC, and FDA.

While final standards are still pending, the industry is at least starting to receive some guidance. Rebecca D. Kush, president of CDISC, says the group issued model operating standards last year that conform to how FDA would like to have data filed electronically. She adds, however, that it generally takes a federal agency two to five years to issue a final rule instituting such a standard.