CDISC and HL7: aBRIDGed

In October 2003, former FDA Commissioner Mark McClellan told an audience in Bethesda, Maryland: "The kinds of information bottlenecks created by poorly accessible paper records is dangerous, inefficient, and expensive, whether it's our own evaluation of a new drug application or a doctor's evaluation of her patient."

McClellan's criticism came during the keynote presentation he gave at Interchange 2003, the inaugural industry-wide meeting of the Clinical Data Interchange Standards Consortium (CDISC), which develops and supports global, platform-independent data standards that enable information system interoperability in research and related areas of health care. Since that keynote address, CDISC has worked hard to make the paper-based bottlenecks McClellan referred to a thing of the past by not only making it easier to transition from paper to computers, but also making it possible to integrate clinical research data and health care data. Their partner in this Herculean harmonization effort is Ann Arbor, MI-based Health Level 7 (HL7), an ANSI-accredited standards developing organization for health care that's been around for three decades. In February 2008 the two volunteer-driven groups, who first met when CDISC accepted an invitation to speak to HL7's board about their work, celebrated the second renewal of their 2001 associate charter agreement—the official initiation of their collaboration.

"Our first step was to try to integrate our XML standard (ODM or Operational Data Model) with the HL7 RIM (Reference Information Model)," said Rebecca Kush, PhD, president and chief executive officer of CDISC, "[but] this was a complete failure; the CDISC and HL7 representatives were not speaking the same language." After spending much time figuring out the basic disconnect between its data structure and HL7's and harmonizing its own standards (e.g., ODM and SDTM), domain experts from CDISC created a model to represent clinical research within HL7 RIM called BRIDG (or the Biomedical Research Integrated Domain Group Model). And it's this domain analysis model, which NCI and FDA helped develop, that's setting the stage for the seamless integration of patient care and clinical research, according to Charles Jaffe, MD, PhD, chief executive officer of HL7. Jaffe, who served six years on CDISC's board before joining HL7, says such integration will allow researchers "seamless access to patient level information whether by diagnosis or medication and [to] identify those criteria that accurately reflect inclusion/exclusion criteria." Technicalities aside and erring on the side of simplicity, the resulting benefits would include major improvements in the subject recruitment process (such as lower costs and improved efficiency and speed) and improvements in the real-time reporting and surveillance of adverse events so that postapproval tragedies like Vioxx could be avoided.

In order for such benefits to be realized, electronic health records (EHRs)—a technology reportedly used by less than 15% of U.S. physicians today—would have to become the norm in hospitals and doctors offices and they would have to use CDISC and HL7 standards with the appropriate processes. "That's the whole idea behind the standards movement," explained Kush, "to get adoption." But while other countries like Canada and the UK pour millions into upgrading their health care systems to include EHRs, the U.S. government has left the decision to the private sector, which is dragging its feet in the face of out-of-pocket costs and interruptions in work flow. Policy issues, including privacy, are also impeding the transition. "Before we can be successful with our most important opportunity to make fundamental use of data, we need to address those policies," said Jaffe. As for how to address them, neither HL7, CDISC nor FDA appear certain. In the meantime, both standards organizations continue to forge ahead with their pursuit of a collaboration on a global scale, most recently including the development of joint educational programs around the world as part of their renewed charter agreement. They also continue to boast the backing of FDA. "The [HL7/CDISC] project is very important in getting us to that single, harmonized, standardized approach to the structured scientific data that we receive," Armando Oliva, MD, CAPT, deputy director for bioinformatics, Office of Critical Path Programs, told Applied Clinical Trials.